united states of america - etracomvferc.com shaun ledgerwood.pdf · united states of america before...

UNITED STATES OF AMERICA

BEFORE THE

FEDERAL ENERGY REGULATORY COMMISSION

AFFIDAVIT

OF

SHAUN D. LEDGERWOOD

ON BEHALF OF

ETRACOM LLC AND

MICHAEL ROSENBERG

February 16, 2016

20160216-5346 FERC PDF (Unofficial) 2/16/2016 4:13:50 PM

i

Table of Contents

I. Introduction and Background ..................................................................................................... 1

II. The Market Characteristics of CAISO Interties ......................................................................... 5

A. Convergence Bidding at the CAISO Interties and at New Melones ............................... 5

1. Description of the CAISO Interties ......................................................................... 5

2. Description of Intertie Congestion .......................................................................... 6

3. Description of Convergence Bidding ....................................................................... 8

4. Problems with Convergence Bidding on the CAISO Interties ............................ 10

5. Additional Problems with Convergence Bidding at New Melones ..................... 11

a. The Effects of New Melones’ Fully Encumbered Status ............................. 11

b. The Effects of Software Errors at New Melones .......................................... 15

B. Intertie Congestion and the Role of Congestion Revenue Rights (“CRRs”) ................. 16

III. A Framework for the Analysis of Market Manipulation ......................................................... 18

A. The Trigger, Nexus and Target ........................................................................................ 18

B. The Proof of Intent .......................................................................................................... 19

C. The Importance of Intent to the Proof of Causation ...................................................... 20

IV. Does the Logic of the Alleged Behavior Fit This Framework for Analyzing Market

Manipulation? ............................................................................................................................ 21

V. An Economic Analysis of ETRACOM’s Behavior Relative to the Framework’s

Three Elements .......................................................................................................................... 22

A. The Trigger: Staff’s Allegations Fail to Demonstrate that ETRACOM’s Trades

at New Melones Were Intentionally Uneconomic ........................................................ 22

1. It Was Economically Rational for ETRACOM to Believe During the

Relevant Period that Its Convergence Bidding Strategy at New

Melones Could Be Profitable on a Stand-Alone Basis .......................................... 22


ii

a. Staff’s Ex-Post View of the Profitability of ETRACOM’s

Hydroelectric Trading Strategy Ignores the Market Presented at

New Melones ................................................................................................. 23

b. Data Confirms ETRACOM’s Expectations Concerning the

Profitability of Its Hydroelectric Trading Strategy at New

Melones .......................................................................................................... 28

c. Staff Mischaracterizes ETRACOM’s Tracking of Its Virtual Losses

during the Relevant Period ........................................................................... 33

d. ETRACOM’s Virtual Trading at New Melones during the

Relevant Period Was Based on a Rational Strategy Unconnected

to its CRRs ...................................................................................................... 36

2. ETRACOM Was Incentivized By the Software Pricing/Modeling

Errors to Bid at Zero or Negative Prices to Clear Its Virtual Supply

Offers at New Melones ........................................................................................... 38

3. Staff’s Allegations that ETRACOM’s Behavior at New Melones Was

Intentionally Uneconomic, Unusual and Anomalous Are Incorrect and

Not Supported ......................................................................................................... 42

a. ETRACOM’s Behavior at New Melones Was Incentivized By the

Prices Presented There and Was Not Intentionally Uneconomic .............. 42

b. Staff Errs in Alleging that ETRACOM Sought to Purposely Lower

Prices at New Melones .................................................................................. 45

c. A Review of Actual Trading Data Proves That ETRACOM’s

Trading at New Melones Was Neither Unusual nor Anomalous ............... 46

d. ETRACOM’s Virtual Bidding Behavior at New Melones Was

Consistent with the Expansion of its Portfolio as a New

Participant in the CAISO .............................................................................. 55

e. The Potential Profitability of Other Locations Does not Confirm

ETRACOM's Manipulative Intent ................................................................ 57

B. The Nexus: ETRACOM Reasonably Could Not Have Known the Strength of

the Nexus Linking Its Virtuals to the Value of Its CRRs ............................................... 58

C. The Target: ETRACOM’s Profits in Its CRR Portfolio in Late May 2011

Were Similar to Those Experienced in Early April ........................................................ 60


iii

VI. Staff’s and the DMM’s Allegations Concerning ETRACOM’s Intent Suffer from

Numerous Flaws and Rely on No-Win Logic .......................................................................... 62

A. ETRACOM Intended Its Hydroelectric Trading Strategy to Be Profitable

from the Start ................................................................................................................... 62

B. ETRACOM Intended for Its May 14-15 Off-Peak Virtual Supply Offers to Be

Profitable .......................................................................................................................... 63

C. ETRACOM Intended for Its May 16-31 Trades to Be Profitable .................................. 64

D. ETRACOM’s Behavior in June Confirms that It Never Intended to Affect the

Value of its CRRs .............................................................................................................. 67

E. Staff Attempts to Prove ETRACOM’s Intent through the Use of “No Win”

Scenarios ........................................................................................................................... 68

F. Summary: Staff Has Failed to Prove ETRACOM’s Manipulative Intent ...................... 69

VII. Staff’s Estimate of Unjust Profits and Harm Overstates the Impact of ETRACOM’s

Virtual Trading .......................................................................................................................... 70

VIII. Conclusion ................................................................................................................................. 75

Appendix A: CV of Shaun D. Ledgerwood

Appendix B: List of Exhibits


1

Affidavit of Shaun D. Ledgerwood

In the Matter of ETRACOM LLC and Michael Rosenberg

Federal Energy Regulatory Commission

I. Introduction and Background

1. I am a Principal with The Brattle Group (“Brattle”), an economic and management

consulting firm with offices in Cambridge, MA; New York, NY; San Francisco, CA; Washington,

D.C.; London, England; Madrid, Spain; Rome, Italy; and Toronto, Canada. My business address is

1850 M Street, NW, Suite 1200, Washington, DC 20036. I hold a Ph.D., M.A. and B.A. in

Economics from the University of Oklahoma and a J.D. from the University of Texas.

2. During my 25 year career, I have focused on issues related to regulation and competition,

particularly in electricity and other energy markets. Since joining Brattle in January 2011, my

practice has centered on the analysis of market manipulation within and across commodity and

financial markets. Prior to joining Brattle, I worked as an economist and attorney for the Office

of Enforcement (“Enforcement”) of the Federal Energy Regulatory Commission (“FERC” or “the

Commission”), where I evaluated and assisted the prosecution of manipulative behavior within

and across wholesale electricity and natural gas markets, and as affiliated faculty of Georgetown

University. Before joining the Commission in June 2008, I worked as a consulting economist,

focusing on competitive and regulatory issues related to electricity and natural gas, including

issues related to ratemaking and the function and analysis of tariffs. I also worked as an adjunct

professor of law and economics for the University of Oklahoma and as an attorney. I have

testified as an expert witness before state utility commissions and in federal court. My full

qualifications and relevant experience are provided in my CV attached as Appendix A.

3. Based upon my experience in manipulation cases and prior experience with competition

matters generally, I developed an economic framework to describe, detect and analyze alleged

manipulative behavior. This approach allows for consistent analysis of potential manipulation

through the appropriate separation of cause and effect, providing a comparable analytic approach

across products, cases, agencies, and statutes that prohibit fraud or the creation of an “artificial”

price. The framework is the cornerstone of a recent book three other Brattle co-authors and I

published in 2015, has been published in economic, law and trade journals, and has been

presented at conferences and other settings to traders, regulators, attorneys and academics in the

U.S., Canada and the EU. It is currently being used by market monitors and industry compliance

personnel for conducting trade surveillance. This Commission and its Office of Enforcement

have adopted the logic of the framework, evidenced by Chairman Norman Bay’s explanation of

the agency’s approach to analyzing manipulation cases in connection with hearings before the


2

United States Senate1 and by staff’s discussion of market manipulation analysis contained in its

Energy Primer: A Handbook of Energy Market Basics.2 A list of relevant publications, projects

and presentations is also provided in Appendix A.

4. I have been retained by Morrison & Foerster, LLP on behalf of ETRACOM LLC and its

trader Michael Rosenberg (collectively, “ETRACOM”) to examine and comment on the allegedly

manipulative behavior described by the Commission’s Order to Show Cause and Notice of Proposed Penalty and attached Enforcement Staff Report and Recommendation, dated December

16, 2015 (“OSC” and “Staff Report,” respectively). Specifically, I have been asked to analyze

ETRACOM’s convergence bidding behavior used to execute its hydroelectric trading strategy at

the New Melones intertie (“New Melones”) of the California ISO (“CAISO”) from May 14, 2011

to May 31, 2011 (the “Relevant Period”) using the analytic framework described above.

5. In the remainder of this affidavit, I: (a) describe the characteristics of the CAISO

convergence bidding market before, during, and after the behavior at issue in this case, as well as

the characteristics of the trading instruments involved; (b) discuss the problems associated with

convergence bidding on the CAISO interties generally and at New Melones given the market

design flaws and software errors presented there; (c) discuss the framework for the analysis of

market manipulation and its use for evaluating fraudulent intent and causation; (d) determine

whether ETRACOM’s behavior in this market as described in the Staff Report fits the general

logic of the manipulation framework; (e) if so, discuss whether ETRACOM’s behavior satisfies

each of the three specified elements required to prove the cause and effect of a manipulation; (f)

if so, consider what conclusions should be drawn therefrom given the market flaws and software

errors presented at New Melones and other economic and documentary evidence of ETRACOM’s

intent; (f) assuming that a manipulation did occur, estimate the unjust profits and harm caused

by the behavior; and (g) discuss the ramifications of this case on future market efficiency.

6. As background for my analysis, I reviewed several documents and trading data related to

this investigation, including: the OSC and Staff Report; staff’s Preliminary Findings contained in

Re: Etracom, LLC and Michael Rosenberg, dated July 17, 2014 (“Preliminary Findings”); the

CAISO Division of Market Monitoring’s referral of ETRACOM, dated July 29, 2011 (“DMM

Referral”); and various other documents relevant to assessing ETRACOM’s behavior and intent.

1 Testimony of Norman C. Bay, Director, Office of Enforcement, Federal Energy Regulatory

Commission Before the Committee on Banking Financial Institutions and Consumer Protection Subcommittee, United States Senate (January 15, 2014), pp. 4-6. Mr. Bay referred to the

manipulation’s trigger as the “tool,” its nexus as the “target” and its target as the “benefitting position.”

See also Mr. Bay’s responses to the Committee on Energy and Natural Resources’ Questions for the

Record (June 4, 2014), response to Senator Debbie Stabenow’s Question 4(b).

2 Available at http://www.ferc.gov/market-oversight/guide/energy-primer.pdf. Specifically, see the

discussion of “Manipulative Trading Techniques and Cross-Product Manipulations” on pp. 118-119.


3

7. The market at New Melones was profoundly impacted by two key flaws that were not

known to ETRACOM or other market participants until after the Relevant Period. First, New

Melones was a “fully-encumbered” intertie, meaning that only one entity could schedule

physical power flows over the line. As a result, a transmission line with a 384 MW import limit

and 15 MW export limit effectively functioned as a market with no available capacity in either

direction, causing cleared (or even uncleared) virtual bids and offers placed there to set the set

market price and create import or export congestion. Second, errors in the CAISO’s software

that determined day-ahead prices at New Melones caused prices to resolve there at $0/MWh in

29.6% of hours in the thirty days prior to the Relevant Period. Combined, these flaws created an

unpredictable and dysfunctional market that induced ETRACOM’s behavior and produced the

market outcomes of which staff now complains.

8. A presumption of transactional legitimacy—i.e., innocence—must follow all open market

trades in a free economic system.3 Unlike some prior cases that this Commission has considered

involving the alleged exploitation of known flaws in market rules or market design,4 staff has

acknowledged that ETRACOM did not know of any of the flaws or software errors present at

New Melones in advance of or during the execution of the hydroelectric trading strategy that is

the subject of this proceeding. To the contrary, these flaws and software errors resulted in price

formation contrary to the process described in the CAISO’s tariff5 and created unpredictable,

arbitrary, and unforeseeable prices and congestion. These flaws and errors also disrupted

effective price discovery and incentivized ETRACOM to bid at zero or negative prices to ensure

that its virtual supply offers would clear, a fact admitted in the Staff Report.6 Staff has used the

resulting price outcomes to allege a trading pattern consistent with manipulation, and has

3 In the absence of this presumption, every market trade placed would be subject to endless regulatory

scrutiny, thus discouraging significant volumes of legitimate trading and undermining the purpose for

and benefits derived from free trade within competitive wholesale electricity markets.

4 See, for example, In Re Make-Whole Payments and Related Bidding Strategies, Order Approving Stipulation and Consent Agreement,144 FERC ¶ 61,068 (2013) (alleged exploitation of market rules in

the CAISO and Midwest ISO to garner out-of-market payments); Deutsche Bank Energy Trading,

LLC, Order to Show Cause and Notice of Proposed Penalty, 140 FERC ¶ 61,178 (2012) (alleged use of

uneconomic or fraudulent physical flows to benefit the value of CRRs); and MISO Virtual and FTR

Trading, Order Approving Stipulation and Consent Agreement, 146 FERC ¶ 61,072 (2014) (alleged use

of virtual bids to benefit the value of financial transmission rights on a radial tie, consistent with the

trader’s dissertation thesis).

5 For example, CAISO’s tariff stipulates the methodology by which it should have calculated LMPs for

New Melones during the relevant period. CAISO Fifth Replacement Tariff at § 27.1.1 (defining LMP

as “the marginal cost of serving the next increment of Demand” at a node) and Appendix C (detailing

calculations required for marginal congestion and other components of LMP).

6 See, for example, Staff Report, p. 33 (“this argument [concerning software errors] could only explain

why ETRACOM’s offers were zero or negative”); and p. 36 (“[s]taff does not dispute that ETRACOM

may not have known the line was fully encumbered”).


4

evaluated other economic and trade-based evidence presented in this case through a lens that

presumes ETRACOM’s guilt. Given staff’s approach, the outcome is preordained—and incorrect.

9. Based on an economic analysis of the evidence presented in this case when viewed

through an ex ante presumption that ETRACOM did not—and reasonably could not—know of

these anomalous outcomes, I find that: (a) ETRACOM did not know and could not have known

that market flaws and software errors set prices at New Melones based on cleared (or even

uncleared) convergence bids, despite the fact that scheduled physical and virtual flows never

exceeded the transmission line’s capacity; (b) these flaws and errors frustrated effective price

discovery at New Melones and resulted in market outcomes in violation of the CAISO’s tariff,

preventing ETRACOM from understanding the characteristics of the intertie and incentivizing it

to offer virtual supply at zero or negative prices to assure that those offers would clear; (c)

ETRACOM’s convergence bidding activity was economically rational given day-ahead price

signals, published reports of an imminent significant hydro event, and its correct (though mis-

timed) pursuit of stand-alone profits consistent with its hydroelectric trading strategy; (d) the

appearance of HASP congestion in July 2011, as well as in prior and later years, confirms both

the legitimacy of ETRACOM’s hydroelectric strategy and its expansion into peak and off-peak

hours; (e) ETRACOM’s trading strategy at New Melones shared several attributes with other

strategies executed by the company elsewhere, contrary to staff’s assertions in the Staff Report

and Preliminary Findings; (f) ETRACOM did not recognize or intentionally exploit a nexus

between its virtual supply and CRR positions before, during or after the Relevant Period, as

confirmed through documentary evidence that was mischaracterized or ignored by staff; (g)

given the unique flaws and software errors presented at New Melones, it would have been

reasonable given ETRACOM’s experience in the CAISO not to perceive its CRR gains at New

Melones as unusual or related to its convergence bidding; (h) ETRACOM would not have viewed

the increased value of its CRRs during the Relevant Period as unusual, either when compared

against its profits there earlier in the month or its total CRR profits the month before; and (h)

consistent with its presumption of guilt, the logic used by staff and the CAISO Market Monitor

in an attempt to prove ETRACOM’s manipulative intent in this case is couched in “no win”

scenarios where a finding of manipulation is predetermined.

10. Based on these findings, there is an insufficient basis to conclude that ETRACOM acted

with fraudulent intent in executing its hydroelectric trading strategy at New Melones; thus, staff

has not demonstrated that ETRACOM engaged in manipulation. As I will discuss, ETRACOM

followed the price signals produced from this deeply flawed market to offer virtual supply at zero

or negative prices in pursuit of stand-alone profits, trades that staff now alleges were

manipulative despite ample documentary and economic evidence showing otherwise. If this

Commission punishes ETRACOM for its pursuit of a legitimate, risk-taking strategy that was

incentivized and ultimately rendered unprofitable by an unknowable and irreparably

dysfunctional market design, it will impede efficient future trading for fear that legitimate risk-

taking strategies that lose money on a stand-alone basis may be identified by and prosecuted by

Enforcement, irrespective of documentary evidence that proves the trader’s legitimate intent.


5

11. The “harm” experienced at New Melones during the Relevant Period resulted from CRR

revenue adequacy issues created by phantom congestion that was produced by multiple market

flaws, culminating in a dysfunctional market that this Commission ultimately determined should

never have existed. The physical flows over the intertie during this time never approached its

import or export transmission limits, meaning that the day-ahead congestion price at New

Melones generally should have remained at zero irrespective of ETRACOM’s trading. If the

problems associated with the encumbrance and software issues presented at New Melones are

corrected, the market impact of ETRACOM’s trading should have been zero—as would be any

resulting calculation of harm. Nevertheless, if the Commission finds manipulative intent gave

rise to harm in this case, staff’s calculations of disgorgement and harm should be reduced to at

most $121,426 and $388,007, respectively, to account for the export congestion that would have

incentivized ETRACOM’s offers in off-peak hours and that the prices that were set by other

market participants in 43.5% of the hours at issue. These figures represent upper bounds of harm

and disgorgement based on staff’s perfunctory methods, as necessitated by the CAISO’s inability

to rerun the market given the multiple flaws and software errors presented at New Melones

12. I have organized the remainder of this affidavit as follows. Section II briefly discusses the

characteristics of, and fundamental problems with, convergence bidding at the CAISO’s interties

generally and at New Melones specifically during the Relevant Period, and describes the trading

instruments associated with the behavior at issue. Section III briefly describes the framework for

analyzing suspected market manipulation. Section IV addresses whether the behavior that staff

describes as manipulative in its Staff Report fits the characteristics of a manipulation as described

by the framework. Section V evaluates the cause and effect of ETRACOM’s behavior using the

three elements of the framework. Section VI discusses what conclusions should be drawn

therefrom given the evidence presented concerning ETRACOM’s intent, as viewed through the

lens of the decisions it made based on information available ex ante the convergence trades at

issue and without knowledge of the market flaws. Assuming that ETRACOM’s behavior is found

to be manipulative, Section VII discusses the accuracy of staff’s estimate of alleged unjust profits

and market harm. Finally, Section VIII summarizes my conclusions and further discusses the

potential ramifications of this case on future market liquidity and efficiency.

II. The Market Characteristics of CAISO Interties

A. CONVERGENCE BIDDING AT THE CAISO INTERTIES AND AT NEW MELONES

1. Description of the CAISO Interties

13. The CAISO is physically connected to the rest of the Western Interconnection by a set of

transmission lines, collectively referred to as the interties, which allow for the import of power

from the Western Electricity Coordinating Council (“WECC”) into the CAISO and the export of

power from the CAISO to buyers outside of the state. The New Melones intertie at issue in this

case is a 384 MW transmission line connecting hydroelectric generation from the control region

operated by the Sacramento Municipal Utility District and Western Area Power Administration


6

(“WAPA”) to the CAISO.7 Importantly, WAPA is responsible for the marketing and scheduling

of power from these hydroelectric generation assets.8

14. While interconnected, the power markets of the WECC and the CAISO were not fully

harmonized during the Relevant Period due to incongruities in how the markets balanced loads

with generation.9 Specifically, nodes internal to the CAISO were balanced every 5 minutes and

were settled on a 10-minute basis using the average of two 5-minute prices, whereas external

nodes (for the interties) were balanced on a 15-minute basis and settled on an hourly basis by

averaging four 15-minute prices. These differences in the “real-time” settlement process used for

internal nodes and the Hour-Ahead Scheduling Process (“HASP” or “hour-ahead” market) used

for the interties resulted in persistent differences in the hourly locational marginal prices

(“LMPs” or “prices”) used for financially settling real-time internal and hour ahead intertie

transactions.10 By comparison, internal node and intertie transactions in the day-ahead market

cleared together at a common LMP.

2. Description of Intertie Congestion

15. Hourly LMPs consist of three components: an energy component, equal to the marginal

cost of the last generator needed to supply the market’s energy requirements; a loss component,

capturing the cost of energy lost by transmission; and a congestion component.11 Congestion

occurs on a transmission path if the demand for flowing power over the path approaches or

exceeds the path’s flow capabilities. For example, if the transmission capacity going from Point A

(the ‘‘source’’) to Point B (the ‘‘sink’’) is 500 MW, but the Regional Transmission Organization

(“RTO”) seeks to send 600 MW of power from Point A to Point B when calling on the least-cost

generators to serve load, congestion occurs on the path. This causes the congestion component

of the price at the source to decline and the congestion component of the price at the sink to

increase, raising the cost of serving Point B from Point A to a level at which the RTO can

economically limit power flows to the available 500 MW of transmission capacity.12 In the

context of each of the CAISO’s interties, congestion can arise in either an import or export

7 DMM Referral Attachment 1, p. 1.

8 See WAPA website at: https://www.wapa.gov/newsroom/FactSheets/Pages/products.aspx.

9 The CAISO’s dual real-time clearing model was replaced with a Fifteen Minute Market effective May

1, 2014. See Integration of Variable Energy Resources, Order No. 764, 77 FR 41,482 (2012), FERC

Stats. & Regs. ¶ 31,331, order on reh’g, Order No. 764-A, 141 FERC ¶ 61,232 (2012), order on reh’g,

Order No. 764-B, 144 FERC ¶ 61,222 (2013); and Cal. Indep. Sys. Operator Corp., 146 FERC ¶ 61,204

(2014), P 1. However, problems remain in completing this integration. See Cal. Indep. Sys. Operator

Corp., Order on Compliance Filing, 152 FERC ¶ 61,063 (2015).

10 See Cal. Indep. Sys. Operator Corp., Order on Tariff Revisions, 146 FERC ¶ 61,204 (2014), P 2.

11 See CAISO Fifth Replacement Tariff at Appendix C.

12 Shaun Ledgerwood and Johannes Pfeifenberger, “Using Virtual Bids to Manipulate the Value of

Financial Transmission Rights,” The Electricity Journal, vol. 26, issue 9, (November 2013), p. 15.


7

direction when the amount of power attempting to flow over the tie approaches or exceeds its

capacity. A theoretical representation of how congestion should have behaved at the New

Melones intertie under the CAISO Tariff is shown below in Figure 1.

Figure 1

The Behavior of Congestion on a Normal Intertie

16. Assuming the day-ahead congestion market at New Melones was functioning properly,

the green region of Figure 1 represents that no congestion should have occurred there within a

wide band of physical flows in both the import and export directions, including at 0 MW when

no power was flowing over the line. If imports into the CAISO were to approach (yellow region)

and ultimately exceed (red region) the intertie’s physical limit of 384 MW, the “shadow cost” of

trying to flow each additional MW of power would increase, such that more costly resources

inside of the CAISO would become more likely to be economically dispatched to deliver that

same MW to the market. Likewise, if exports from the CAISO toward New Melones approached

(yellow) or exceeded (red) the intertie’s 15 MW limit, export congestion would arise.

17. These examples expose a flaw in staff’s reasoning. If the CAISO sought to export 60 MW

of power over the New Melones intertie, a 1 MW, 5.03 MW or even 44 MW physical or virtual

import transaction should not eliminate the export congestion there. Staff erroneously states:

The line was congested before ETRACOM placed its virtual supply offers, so

ETRACOM knew that the line was at its limit (regardless of what the limit was),

and, consequently, that small virtual transactions would have an effect on

pricing[….] While ETRACOM’s supply offers between 1-5 MW are a small

portion of the import OTC, they are up to one-third of the export OTC. Because

ETRACOM was aware of the small position limits in the export direction at New

Melones, it was clear that it would only take a small MW amount of imports to

counter export congestion.13

13 Staff Report, p. 36.


8

In this example, the size of ETRACOM’s virtual bids relative to the size of the export constraint

at New Melones is irrelevant. What matters was size of the excess demand that was presented

there, which under normal circumstances would result from any combination of physical exports

and virtual demand that exceed the line’s 15 MW rating. Because ETRACOM had no

transparency into the trades that occurred at New Melones until 90 days after they were placed,

it had no idea whether the export congestion was being caused by a net flow of 16 MW, 21 MW

or 100 MW attempting to flow over the line’s 15 MW maximum export capacity. Therefore,

staff’s presumption that ETRACOM knew that its 1-5.03 MW virtual supply offers would affect

congestion at New Melones is incorrect.

3. Description of Convergence Bidding

18. Convergence bidding (also known as “virtual” bidding) allows market participants to

profit from expected differences between day-ahead and real-time (or, in the case of the interties,

hour-ahead) prices. Convergence bids simultaneously clear with bilateral and other physical

trades in the day-ahead market and thus will tend to impact day-ahead and (at times) real-time

LMPs. The distinguishing characteristic of a convergence trade is that the quantity of megawatts

(MW) bought or sold by the trader in the day-ahead market is exactly offset by a sale or purchase

of an identical quantity in the real-time market, such that there is no net effect on the physical

market quantity traded.

19. For example, if in a given hour a trader expects the day-ahead LMP at a location to

exceed the real-time LMP, it would be economically rational for the trader to place virtual

supply offers at that point for that hour. If the offers clear, the trader receives the day-ahead

price on all MW sold and pays the real-time price for the same number of MW purchased,

netting a profit or loss equal to the difference of the day-ahead and real-time prices multiplied by

the quantity of MW cleared. In essence, the trader uses the day-ahead market to sell to itself in

the future real-time market such that the net effect is that of a purely financial transaction.14

The process of clearing a virtual supply offer is illustrated below in Figure 2.

14 Ledgerwood and Pfeifenberger, p. 11.


9

Figure 2

The Process of Clearing Virtual Supply Bids

20. The trader offering virtual supply into the market expects the transaction to operate as a

financial swap. Specifically, the trader offers to sell the swap for a price “PVS”” and expects it to

clear the market if the day-ahead price “PDA”” exceeds or equals its offer (i.e., PDA PVS). If

cleared, the swap then pays the trader the difference of the day-ahead price (not its offer price)

and the real-time price “PRT”” multiplied times the number of MW cleared. The trade is

profitable if the day-ahead price is greater than the real-time price or loses money if the opposite

is true. Thus, a rational trader should be willing to “sell the swap” for any price greater than the

expected real-time price, potentially profiting from either a higher day-ahead price or a lower

than expected real-time price. Importantly, it follows that the trader could be willing to pay a

premium for this virtual supply swap—i.e., sell it for a negative price—if the expected real-time

price is negative. This is consistent with how ETRACOM characterized its view of the virtual

supply offers at New Melones.15

21. If the market at the location where the virtual supply offer is placed is well-functioning

and competitive, the clearing of a small quantity of virtual supply at the price PVS should have

little effect on the day-ahead price PDA. The consequence of placing such offers at negative prices

should then serve only to make those offers more likely to clear, effectively making them “price

taking” to whatever day-ahead market price results. However, every sale tends to lower prices,

while every purchase tends to raise them. Profitable virtual trades thus improve market

efficiency by converging day-ahead and real-time LMPs, indicative of the better alignment of

day-ahead awards with real-time market conditions. In addition to their role in improving price

15 See Rosenberg Deposition, p. 62:17-24, 181:7-10, and 351:9-19. As applied to New Melones, the “real-

time” market of Figure 2 is the HASP.


10

discovery and mitigating market power, this efficiency gain underlies the Commission’s directive

to the CAISO to incorporate convergence bidding into its market design.16

4. Problems with Convergence Bidding on the CAISO Interties

22. The CAISO allowed convergence bidding on internal nodes and at the interties beginning

on February 1, 2011.17 While all convergence bids cleared in the day-ahead market, bids placed

on internal nodes settled against the real-time LMP while bids placed on the interties cleared

against the hour-ahead LMP set by the HASP. The CAISO also proposed (and FERC approved)

temporary position limits designed “to ensure that no single market participant can exercise

market power at an individual node.”18 These were to be phased out over twelve months for

internal nodes and sixteen months for intertie scheduling points.19

23. However, the CAISO identified “adverse market impacts” resulting from convergence

bidding on the interties,20 prompting it in August 2011 to request that increases in the related

position limits be postponed21 and in September 2011 to request that such bids be eliminated

altogether.22 The Commission suspended (but did not eliminate) convergence bidding on the

interties effective November 28, 2011, with possible reinstatement pending the outcome of a

technical conference.23 The CAISO indicated that a redesign of its markets due to FERC Order

76424 would facilitate the eventual reintroduction of convergence bidding on the interties,25 but

this process later was quashed by the Commission’s determination that the proposed changes to

the market design were unjust and unreasonable given continuing liquidity concerns.26

16 Cal. Indep. Sys. Operator Corp., 116 FERC ¶ 61,274 (2006), PP 450-451.

17 Cal. Indep. Sys. Operator Corp., 133 FERC ¶ 61,039 (2010), P 1.


19 Id.

20 Id., P 5. Specifically, “CAISO observed market inefficiencies related to an increase in price divergence

and higher than expected real-time imbalance energy offset charges, { } as well as price inconsistencies

between the bid price offered and the market clearing locational marginal price.” Id., P 6.

21 See Cal. Indep. Sys. Operator Corp., 136 FERC ¶ 61,214 (2011).


23 Id., PP 38-39.

24 FERC Stats. & Regs. ¶ 31,331 (2012).

25 Cal. Indep. Sys. Operator Corp., 146 FERC ¶ 61,204 (2014), P 8. The CAISO later delayed the return

of convergence bidding on the CAISO interties beyond the May 1, 2015 start date due to perceived

lack of liquidity in the new 15-minute EIM. Cal. Indep. Sys. Operator Corp., 151 FERC ¶ 61,074

(2015), P 8.

26 Cal. Indep. Sys. Operator Corp., 152 FERC ¶ 61,234 (2015), PP 41-45.


11

5. Additional Problems with Convergence Bidding at New Melones

24. The market settlement mechanism at New Melones was negatively affected by two other

key flaws at the time when convergence bidding commenced that were not made known to—

nor reasonably could have been known by—market participants until after the Relevant Period.

First, New Melones was a “fully encumbered” intertie, meaning that the entire transmission

capability of the tie was always fully subscribed. Second, multiple software errors were present

at New Melones which frustrated price discovery and incentivized market participants to bid

virtual supply at zero or negative prices to ensure that they would clear. Understanding these

flaws is essential to understanding this case, as they incentivized ETRACOM’s behavior and were

responsible for the alleged market harm for which staff brings this case.

a. The Effects of New Melones’ Fully Encumbered Status

25. New Melones’ “fully encumbered” status meant that the entity with scheduling rights, in

this case WAPA,27 fully subscribed the entire transmission capability of the tie. This prevented

the placement of physical bids and offers at New Melones by any other market participant but

WAPA. However, convergence bids by any market participant were allowed pursuant to the

position limits set by the CAISO and ostensibly put in effect to protect against market power

concerns.28 This flaw, which was not made known to ETRACOM or other market participants

until June 2011,29 allowed small convergence bids to disproportionately impact the intertie’s day-

ahead congestion price and thus affect the day-ahead LMP. This is illustrated in Figure 3, below.

27 DMM Referral Attachment 1, p. 1.

28 See Staff Report, p. 15, n. 81.

29 Rosenberg Deposition, 99:19-100:5.


12

Figure 3

The Effect of New Melones’ “Fully Encumbered” Status

26. The upper portion of Figure 3 depicts how the New Melones intertie should have worked

in the day-ahead market if it functioned properly pursuant to the CAISO tariff. Export

congestion should occur only if the net volume of physical exports and virtual demand bids

approached or exceeded 15 MW, and import congestion should occur only if the net volume of

physical imports and virtual supply offers approached or exceeded 384 MW.30 Relief of export

congestion with less than a near-400 MW31 net swing in the import direction would tend to

cause congestion to fall to zero, not necessarily cause import congestion. By comparison, because

of New Melones’ fully encumbered status, the limits of the tie in either the import or export

direction fell to zero MW because the CAISO set the day-ahead transmission limit at the physical

schedule of WAPA. According to the DMM, and as shown in the lower portion of Figure 3, this

meant that a single MW of cleared (or even uncleared) virtual trades could create or reverse

congestion at New Melones.32

30 The yellow regions shown in Figure 2 denote that congestion tends to appear on a transmission path

before the path’s physical limit is reached.

31 Assuming that 1 MW is sufficient to clear the export congestion, import congestion should not arise

without nullification of the 15 MW export limit and 384 MW import limit, thus totaling to 400 MW.

32 DMM Referral Attachment 1, pp. 1-2. Based on our review of trading data in 2011, the presence of

virtual trades (whether cleared or not) seemingly did cause congestion in most hours. However, there

are several instances where virtual trades did not create any congestion—another outcome that staff

and the CAISO DMM have yet to explain.


13

27. The effect of the intertie’s fully-encumbered status is illustrated below in Table 1, which

shows the top four quantities of physical imports flowing over the New Melones intertie into the

CAISO from February through August 2011. The four hours with the largest imports (370 MW)

are shaded, showing that the highest physical flow rates of this period resulted in no congestion whatsoever. This is consistent with the upper portion of Figure 3, for imports well below the

line’s 384 MW maximum capacity should not create congestion. However, the remaining hours

shown in Table 1 all show (negative) congestion in the import direction, even though the size of

the physical flows (340-360 MW) are smaller. The common element amongst all of these hours

is the presence of virtual trades, most of which also set the market price.33

Table 1

The Effect of New Melones’ “Fully Encumbered” Status34

Sources & Notes:

[1]: Congestion expected given physical flows below the line’s 384 MW import capacity.

[2]: Day-Ahead Hourly Congestion, taken from Velocity Suite, ABB Inc.

[3]: Day-Ahead Hourly LMP, taken from Velocity Suite, ABB Inc.

[4]: Day-Ahead Scheduled Import at New Melones, taken from CAISO OASIS.

[5]: Flags if ETRACOM's Virtual Supply Bid equals LMP in that hour.

[6]: Flags if Virtual Supply Offer set the LMP in that hour.

[7]: Flags if Virtual Supply was bid in that hour.

33 Note that HE 20 on May 15, 2011 shows that a virtual supply bid was placed into the market and did

not set the market price, yet the market total LMP resolved to $0/MWh. This is indicative of a flaw in

the software at New Melones that inhibited price discovery. I discuss this flaw later in this affidavit.

34 See Excel workbook titled “Transmission Congestion Analysis.xlsx,” tab “Table 1.”

Date HE

Expected DA

Congestion

per Tariff

($/MWh)

Actual DA

Congestion

($/MWh)

Actual DA

LMP

($/MWh)

DA Physical

Import

(MW)

Did

ETRACOM's

Bid Set the

LMP?

Did Virtual

Supply Offer

Set the

LMP?

Was Virtual

Supply

Offered in

the Hour?

[1] [2] [3] [4] [5] [6] [7]

5/15/2011 19 $0.00 ($28.82) ($10.32) 340 No No Yes

5/15/2011 20 $0.00 ($26.88) $0.00 340 No No Yes

5/15/2011 21 $0.00 ($32.57) $17.72 340 No No Yes

5/15/2011 22 $0.00 ($14.83) $14.88 340 No No Yes

6/4/2011 21 $0.00 ($46.21) $5.00 350 No No Yes

6/4/2011 22 $0.00 ($42.94) $5.00 350 No No Yes

6/5/2011 21 $0.00 ($7.83) $42.68 360 No No Yes

6/5/2011 22 $0.00 ($1.76) $39.38 360 No No Yes

6/12/2011 21 $0.00 ($31.45) $8.65 350 No No Yes

6/12/2011 22 $0.00 ($24.92) $7.94 350 No No Yes

8/2/2011 18 $0.00 $0.00 $55.27 370 No No No

8/2/2011 19 $0.00 $0.00 $49.10 370 No No No

8/2/2011 20 $0.00 $0.00 $41.92 370 No No No

8/2/2011 21 $0.00 $0.00 $38.22 370 No No No


14

28. Juxtaposed to the platform for a well-functioning market shown in the upper portion of

Figure 3 (above), this dysfunction led to the eventual prohibition of convergence bidding on all

such ties on August 29, 2011.35 Staff has mischaracterized the root cause of the decision to end

convergence bidding at New Melones and other fully-encumbered interties as a sole function of

revenue inadequacy, with the insinuation that this inadequacy was disconnected in some way

from the market dysfunction that incentivized ETRACOM’s actions:

CAISO’s decision to discontinue offering CRR positions and virtual trading at

New Melones occurred after ETRACOM’s conduct in May 2011 and is irrelevant

to ETRACOM’s conduct. As ETRACOM admits in its response, the substantive

concern underlying CAISO’s decision to stop offering CRR positions at New

Melones was based on revenue inadequacy.[] Due to the fully encumbered nature

of the line (physical flows were limited to one entity and perfectly hedged) there

simply were not enough funds from physical transactions to pay the congestion

fees to CRR holders. ETRACOM fails to argue why revenue inadequacy justifies

its conduct. Similarly, CAISO terminated virtual trading at New Melones in

August 2011 due to inefficiencies related to the fully encumbered nature of the

line.36 (Emphasis added).

29. The sole reason for CRR revenue inadequacy at New Melones was because there was no

linkage between physical power flows over the tie—none of which approached its physical limits

in either the import or export direction during the period when convergence bidding was

allowed—and the appearance of congestion that resulted from cleared (or sometimes uncleared)

virtual trades. Because the net quantity of physical and virtual megawatts actually cleared was

insufficient to pay the virtually-created congestion tab for the much larger set of CRR positions,

revenue inadequacy resulted. The causes of the revenue inadequacy were therefore the same

market design flaws that caused the market outcomes at issue here, and affected ETRACOM’s

ability to engage in effective price discovery and transact normally as it did at other well-

functioning interties. Staff’s argument that “ETRACOM fails to argue why revenue inadequacy

justifies its conduct” is a red herring that only serves to distract from ETRACOM’s legitimate

defenses relevant to supporting the presumption of transactional legitimacy it should be afforded

in this case.

30. Coincidentally, it is essential to understand the significance of Figure 3 to this case and to

the break in the logic of staff’s allegations. For ETRACOM to have known that virtual supply

offers of between 1-5.03 MW would significantly diminish the export congestion at New

Melones presumes that it knew and fully understood that the day-ahead market functioned as described in the lower portion of Figure 3. Otherwise, such small virtual trades (a) should have

35 CAISO Market Notice, Virtual Bidding Not Allowed at Fully Encumbered Interties (August 22, 2011).



15

been price-taking and thus have had no effect on the day-ahead LMP, and (b) would not have

been expected to reverse export congestion given the disclosed physical limits of the intertie.

31. The Staff Report states “[s]taff does not dispute that ETRACOM may not have known the

line was fully encumbered. However, the unique characteristic of the line is irrelevant.”37 To the

contrary, if ETRACOM did not understand this critical market design flaw, it would have

assumed that the market at New Melones worked as shown in the upper portion of Figure 3 and

thus never have considered that its small virtual supply offers would have had any significant

market impact. As I will discuss later herein, ETRACOM’s subsequent actions support this latter

interpretation, including (a) its successful clearing of 1-5.03 MW price-taking virtual supply

(offered at prices of -$20/MWh) at other interties from May 25-31, which did not set the prices

at those locations, and (b) its willingness to place virtual demand bids at New Melones in June

2011 that would have injured the value of its CRRs at New Melones given this flaw.

b. The Effects of Software Errors at New Melones

32. The second defect presented at New Melones involved software errors that miscalculated

the day-ahead price as zero in a given hour whenever only virtuals supply offers were placed in

that hour (i.e., a day-ahead import into the CAISO over the tie) and the lowest-priced virtual

supply offer was positive.38 These software errors impeded the ability of convergence bids placed

at New Melones accurately to serve as a price discovery tool, and erroneously suggested to

market participants that only zero or negatively-priced virtual supply offers (such as those used

by ETRACOM during the Relevant Period) would clear. Day-ahead LMPs reflected the $0/MWh

software errors at New Melones in 213 out of 720 hours (29.6%) from April 14, 2011 through

May 13, 2011, the 30 days prior to the Relevant Period.39 Although staff’s allegations of

intentional, uneconomic trading rely heavily on ETRACOM’s placement of virtual supply offers

at zero or negative prices during May 2011,40 the Staff Report explicitly states that “this argument

[concerning software errors] could only explain why ETRACOM’s offers were zero or


38 DMM Referral Attachment 1, p. 2, n. 2. While the DMM’s description of the software error cited only

cases where the virtual supply offer was priced below the system marginal energy plus loss component

of LMP at New Melones, the same type of zero LMP error also occurred in several other hours where

the virtual supply offer was priced above the system marginal energy plus loss component, such as for

HE 20 on May 7-10 and May 12-15. See Excel workbook titled “Zero-Neg LMP Bid Analysis.xlsx,” tab

“Software Error Breakdown.” These software errors would have had the same effect of suppressing

efficient price discovery and suggesting that only zero or negatively-priced virtual supply offers would

be guaranteed to clear.

39 See Excel workbook titled “Zero-Neg LMP Bid Analysis.xlsx,” tab “Full Summary.”

40 OSC, P 3; Staff Report, pp. 1, 22-23, 29, 33, 34-35, and 42.


16

negative.”41 That staff’s own findings so clearly contradict its conclusions suggests that staff did

not provide ETRACOM the required presumption of transactional legitimacy in this case.

33. Combined with the problems associated with New Melones’ fully-encumbered status, the

software errors caused several of the attributes for which ETRACOM is now blamed, including

the perceived need to bid virtual supply at zero or negative prices to assure its offers would clear,

the ability of small virtual trades to set the LMP in an unpredictable manner, and the creation

and reversal of phantom day-ahead congestion when no actual congestion was present. The

flaws presented at New Melones frustrated effective price discovery and resulted in congestion

prices and LMPs that were capricious and arbitrary, resulting in unwarranted wealth transfers to,

from and between unsuspecting market participants who transacted there in good faith based on

the presumption that the market followed the terms of the CAISO tariff as approved by the

Commission. The resulting outcomes were dysfunctional and were inconsistent with a well-

functioning market, in violation of the CAISO’s tariff’s strict requirements for price formation.

Despite this issue, the CAISO DMM singled out ETRACOM as a bad actor and shifted the blame

for the CAISO’s own failures to execute the market’s intended design. The flaws played a major

role in ETRACOM’s economic decision to trade at New Melones and the outcomes that resulted.

Critically, had there been no flaws, ETRACOM’s virtual trades would have had a de minimis

effect on the value of CRRs valued there. Given that ETRACOM did not know of these flaws

prior to or during the execution of its hydroelectric trading strategy, the cause of any harm

arising at New Melones should be attributed to the CAISO, not ETRACOM.

B. INTERTIE CONGESTION AND THE ROLE OF CONGESTION REVENUE RIGHTS (“CRRS”)

34. Because competition can raise the cost of serving consumers over a transmission path,

load-serving entities are given or may purchase CRRs as a hedge against congestion in the day-

ahead market. CRRs pay their holder the difference in the day-ahead congestion prices at the

path’s source and sink. For example, assume a generator (the source) needs to flow 25 MW of

power over a constrained path to serve its customer’s load (the sink). If this flow would increase

the day-ahead congestion price at the sink by $10/MWh, the generator’s customer would have to

pay an additional $250 to complete the transfer. However, if the load is hedged by a 20 MW

CRR over the path, it would receive the $10/MWh congestion payment times its 20 MW CRR

position, thus partially hedging $200 of the $250 cost it incurred due to the congestion.

35. In addition, CRRs can be purchased bilaterally or through competitive CAISO auctions

by any creditworthy entity that seeks their financial attributes as a hedge or a speculative

investment. Robust participation by speculators serves to maximize payments to utilities and

customers who wish to sell their CRR positions for a lump-sum payment, as opposed to retaining

the rights to the payments for actual congestion over time. However, as I discuss later, the

ability of market participants to acquire speculative positions raises the possibility that those



17

positions could be used as the target of manipulation. This ability is not unique to financial

participants; for example, if the generator in the prior example chose to “over-hedge” its 25 MW

physical obligation with a 100 MW CRR, every $1/MWh paid in increased congestion charges

would result in a $4/MWh gain from its CRR position. The concern, of course, is that the

generator is then incentivized to intentionally worsen congestion by executing unprofitable

trades over the path so as to increase the (net profitable) payments to its CRR position.

36. Several FERC enforcement actions have been based on this logic, each involving the

alleged use of physical and/or virtual trades to manipulate the value of congestion contracts.42

One such case involved the alleged manipulation of a CAISO intertie that was uni-directionally

constrained.43 Importantly, however, every one of these cases alleged that the market participant

who executed the scheme was expressly aware of its ability to exploit the inherent market flaws

to capitalize on the “over-hedges” presented. By comparison, ETRACOM had no reason to

understand the flaws that were present at New Melones; indeed, those flaws were responsible for

incentivizing ETRACOM to offer virtual supply at zero or negative prices and for the ability of

small virtual trades to set prices and reverse congestion during the Relevant Period.44 Likewise,

ETRACOM would have had no reason to know about the existence of the flaws raised by these

other cases until December 15, 2011, at the earliest.45

37. Notwithstanding these concerns regarding causation and ETRACOM’s intent, the logic of

an “over-hedged” CRR position is the foundation of the allegations made against it: that

uneconomic virtual supply offers were intentionally placed at the New Melones intertie to

exacerbate import congestion (or mitigate against export congestion) to benefit ETRACOM’s

CRR positions. These allegations rely entirely on hindsight-based determinations presented in

staff’s Report that misinterpret, ignore or do not discuss salient facts relevant to ETRACOM’s

intent at the time the behavior occurred and the harm caused by the market flaws. To fully

assess the validity of staff’s allegations, it is necessary to understand the impact that these flaws

had on ETRACOM’s view of market conditions and the basis for its trading activity at New

Melones during the Relevant Period, as well as to assess the likelihood that it understood the

cause and effect of its behavior given the flawed market at New Melones presented at that time.

42 See Constellation Energy Commodities Group, Inc., Order Approving Stipulation and Consent

Agreement, 138 FERC ¶ 61,168 (Mar. 9, 2012); and MISO Virtual and FTR Trading, Order Approving

Stipulation and Consent Agreement, 146 FERC ¶ 61,072 (Feb. 7, 2014).

43 Deutsche Bank Energy Trading, LLC, Order Approving Stipulation and Consent Agreement, 142

FERC ¶ 61,056 (Jan. 22, 2013).

44 Staff Report, pp. 33 and 36.

45 This date corresponds with the issuance of a Staff Notice of Alleged Violations (NAV) in the Deutsche

Bank case. See http://www.ferc.gov/enforcement/alleged-violation/notices/2011/deutsche.pdf. NAVs

were issued for Constellation on January 30, 2012 and for LDES on January 6, 2014.


18

III. A Framework for the Analysis of Market Manipulation

A. THE TRIGGER, NEXUS AND TARGET

38. The following framework describes a market manipulation based on three elements: a

trigger, a nexus and a target. The trigger begins the manipulation with an act intended to bias a

market outcome to cause the manipulation to occur. This biased outcome, such as a distortion in

a market price or output, is the nexus that links the manipulation's cause and effect. This effect

alters the worth of the target, which then produces the manipulation’s profits. Proof of a

manipulation therefore requires proof that the manipulator intentionally acted in a manner

designed to cause (trigger) a change in some market mechanism (nexus) to alter the value of one

or more positions (target) that benefit from the change. The anti-manipulation enforcement

actions that have been brought by the FERC can be analyzed using this framework, as they

ultimately follow a similar logic of cause and effect. For simplicity, I provide an illustrative

graphic of the logic of the framework below in Figure 4.

Figure 4

A Framework for Analyzing Market Manipulation46

46 Discussions of the framework have been published in economic, law and trade journals and recently

in a book. For example, see Taylor, Ledgerwood, Broehm and Fox-Penner, Market Power and Market Manipulation in Energy Markets: From the California Crisis to the Present, PUR Inc. (April 2015);

Ledgerwood, “Uneconomic trading and market manipulation,” Energy Risk Magazine, p. 33 (July

2013); Ledgerwood and Carpenter, “A Framework for Analyzing Market Manipulation,” Review of Law & Economics, v. 8, No. 1, pp. 253, 290, (September 2012); and Ledgerwood and Harris, “A

Comparison of Anti-Manipulation Rules in U.S. and EU Electricity and Natural Gas Markets: A

Proposal for a Common Standard,” Energy Law Journal, v. 33, No. 1, pp. 1, 33 (April 2012).


19

39. Triggers to a manipulation can include (a) outright fraud, such as filing a fraudulent

report or failing to divulge a material fact; (b) the exercise of market power, typically executed

through an act of withholding; or (c) intentional uneconomic trading, such as by trading

excessive quantities in the market to deliberately incur a loss that biases a market outcome.

Behavior that does not fall into one of these three categories generally serves a stand-alone

legitimate business purpose and thus cannot be considered manipulative. The manipulation’s

nexus can be any market-related mechanism that is subject to bias by the manipulation trigger,

such as a market price or quantity traded. The manipulation’s target is then one or more

positions designed to benefit from the bias created, such as physical or financial derivatives tied

to the market price or other market or out-of-market payments linked to the bias created.

40. In conjunction with the development of this framework, I developed an economic model

that explains the principles that underlie manipulations triggered by uneconomic trading—i.e.,

where the trader intentionally incurs a financial loss to trigger the manipulation. The revenues

derived from the manipulation must exceed this loss for the manipulation to be profitable

overall, an “over-hedged” condition which I refer to as financial leverage. In Figure 4, financial

leverage exists if the revenues produced from the manipulation’s target are greater than the losses

incurred in the trigger, such that the manipulation is profitable overall. In a 2013 publication, I

applied this model to “Day 2” RTOs using an example where virtual bids are used to manipulate

the value of financial transmission rights.47 This is functionally similar to the assertions made in

this case, where convergence bids are allegedly used to manipulate the value of CRRs.

B. THE PROOF OF INTENT

41. Because a presumption of transactional legitimacy follows all open-market trades in a free

economic system, demonstration of the assemblage and movement of the framework’s three

components is a necessary, but not sufficient, requirement for proving a market manipulation.

Specifically, proof that a market participant controlled and operated the three framework

components to its benefit is necessary to demonstrate the manipulation’s cause and effect, but is

not dispositive proof of the actor’s manipulative intent. Under a fraud-based statute and rule,

such as the Commission’s Rule 1c, proof of intent requires a showing of additional evidence to

demonstrate that the alleged manipulator acted with requisite scienter—i.e., fraudulent intent.

42. In cases such as this one where uneconomic trading is the alleged manipulation trigger,

proof of a market participant’s fraudulent intent is problematic because losses are a normal

market event endemic to legitimate risk-taking. To overcome the presumption of transactional

legitimacy, parties with the burden of proof typically provide a combination of witness testimony

and/or contemporaneous documentary evidence—emails, voice recordings or other information

that provide objective evidence of the trader’s intent—and economic evidence of repeated

behavior that demonstrate that the trades of concern were not intended to be profitable on a

47 Ledgerwood and Pfeifenberger, pp. 9-25.


20

stand-alone basis. In the absence of documentary proof, it follows that the amount of economic

evidence needed to prove intent must be sufficient to overcome possible legitimate motivations

for the behavior, with the presumption of transactional legitimacy remaining with the market

participant until the requisite evidentiary standard of proof is met. Conversely, where the

available documentary evidence supports the legitimacy of the trading behavior under scrutiny,

the economic evidence should be discounted accordingly in favor of a presumption that the

losses incurred were benign.

C. THE IMPORTANCE OF INTENT TO THE PROOF OF CAUSATION

43. At its core, the framework described above is a model for proving the cause and effect of

a manipulation, such that the suspected manipulator is shown to intentionally engage in some

fraudulent act (e.g., an intentionally-uneconomic, price-making trade) to cause a change in some

market mechanism (e.g., a day-ahead congestion price) with the intent of effecting a benefit to a

position that is tied to the mechanism (e.g., a CRR). However, the critical element in this process

is not the simple tying together a suspected trigger, nexus and target, but the proof that the actor

intended to operate these three pieces in furtherance of a manipulative scheme. Interactions of

trading instruments within a portfolio are commonplace, and repeated interactions certainly can

draw a regulator’s interest. However, without proof that the interactions among the instruments

were known to the portfolio holder, without proof that the holder then intentionally executed

trades to cause those interactions to produce some effect, without proof that the holder had in

place instruments that would then benefit from that effect, an allegation of manipulation is

nothing more than supposition. This cannot overcome the presumption of transactional

legitimacy that must attach to open market trades, for to do so would mistake correlation for

causation, as staff has done here in its allegations against ETRACOM.

44. There are two key aspects of this case that are necessary for staff to prove to establish the

alleged manipulation’s cause and effect. First, although most of ETRACOM’s virtual transactions

placed at New Melones from May 14-31 did indeed lose money, staff must also prove that those

losses were intentionally incurred and were not the result of legitimate risk-taking in pursuit of

stand-alone profits. Second, staff must prove that ETRACOM understood the strength of the

manipulation’s nexus—i.e., that its virtual supply offers would set the price and/or cause

congestion to benefit the value of its CRRs at New Melones—and intentionally exploited it. As I

will discuss, staff does not and cannot meet its evidentiary burden in this case due to a lack of

documentary proof of ETRACOM’s manipulative intent and substantial economic and

documentary evidence showing that ETRACOM neither perceived the strength of nor exploited

this nexus, thus supporting the stand-alone, legitimate business purpose of ETRACOM’s trading

strategy when all available evidence is considered properly.48

48 Whereas staff’s case is based on a selective use of facts discovered before and after the Relevant Period,

ETRACOM did not have the benefit of hindsight when making decisions to place the trades at issue in

this case. As I will discuss, an assessment of the information available to ETRACOM ex ante will Continued on next page


21

IV. Does the Logic of the Alleged Behavior Fit This Framework for

Analyzing Market Manipulation?

45. Staff’s Report alleges that ETRACOM engaged in two periods of intentional uneconomic

trading at New Melones in May 2011. With respect to the first period, described by staff as the

“test period,” Enforcement alleges that ETRACOM intentionally placed “$0 virtual supply offers

in mostly off-peak hours” on May 14-15 to test whether uneconomic virtual supply offers could

reverse the adverse effect export congestion had on the value of ETRACOM’s CRR positions

from May 8-13.49 In the second period, commencing on May 16 and running through the end of

the month, ETRACOM allegedly “expanded its virtual trading strategy to all hours of the day and

began making virtual supply offers below $0.” and “was willing to sell at least a portion of its

MWs between -$28 and -$30 (the offer floor).”50 Without addressing their merits, these

allegations assert behavior that could serve as a potential manipulation trigger due to the impact

that uneconomic convergence bids could have on day-ahead congestion prices at New Melones

when traded in sufficiently large quantities.

46. The nexus of the alleged manipulation described in the Staff Report is the day-ahead

congestion price at New Melones during the Relevant Period. Specifically, staff alleges that

ETRACOM submitted uneconomic virtual supply offers to remove export congestion and create

import congestion at New Melones to improve its CRR positions that were short to its day-ahead

congestion price (i.e., that benefitted from a lower price).51 As I will discuss, the significant flaws

of the market at New Melones and other economic and documentary evidence raise serious

doubts as to the extent to which ETRACOM tracked the link between its virtual trades and the

value of its CRR positions. However, without considering the merits of staff’s claims, the

potential linkage of convergence bids to positions tied to day-ahead congestion prices at New

Melones could allow those prices to serve as a nexus for manipulation.

47. The Staff Report alleges that ETRACOM’s actions to reduce congestion prices at New

Melones were designed to impact its CRR positions sourced at the intertie.52 Without

considering the merits of staff’s allegations, these CRR positions could feasibly have served as a

manipulation target because the CRRs were directionally aligned to benefit from lower day-

ahead congestion prices and the size of those CRRs (34.668 MW on-peak and 25.326 MW off-

peak) was larger than the size of the 1-5.03 MW virtual supply offers alleged to have triggered

the manipulation, thus providing financial leverage.

Continued from previous page

demonstrate the reasonableness of its hydroelectric trading strategy, as confirmed by economic and

documentary evidence that staff failed to consider both before and after the Relevant Period.


50 Id., pp. 17-18.

51 Id., p. 18.

52 Id., pp. 10-11.


22

48. Staff makes many and varied allegations concerning ETRACOM’s behavior that it asserts

are indicative of fraudulent intent.53 Although these allegations are refutable, and indeed many

are incorrect, staff essentially asserts that they are collectively sufficient to meet its ability to

prove that ETRACOM’s hydroelectric trading strategy was manipulative.

49. Thus, taking the allegations made in the Staff Report at face value, staff alleges a

manipulation that follows the general pattern of the framework. However, as I will discuss in

the next sections of this affidavit, staff’s arguments are based on flawed economic reasoning that

is channeled by a presumption of ETRACOM’s guilt, and several of its resulting conclusions lack

documentary support. If staff’s assertions with respect to any of the framework’s three elements

or intent are refuted, they fail to overcome the presumption of transactional legitimacy that must

otherwise be afforded to ETRACOM’s open market trades, and thus fail to prove a manipulation.

V. An Economic Analysis of ETRACOM’s Behavior Relative to the

Framework’s Three Elements

A. THE TRIGGER: STAFF’S ALLEGATIONS FAIL TO DEMONSTRATE THAT ETRACOM’S

TRADES AT NEW MELONES WERE INTENTIONALLY UNECONOMIC

50. There are numerous logical and factual problems with staff’s allegations concerning

ETRACOM’s hydroelectric trading strategy at New Melones in May 2011. Some of these

problems relate to the mechanics of ETRACOM’s bidding behavior, while others misstate the

likely motives underlying its trading strategies. Further, staff’s assertions rely on a revisionist

historical perspective suggesting that ETRACOM had knowledge at the time it placed the virtual

supply offers of factors that were not known, nor could have been reasonably known, to it or to

other market participants, including the ability to accurately assess the market at New Melones

given the significant market flaws and software errors that existed at that intertie. This is true

both for convergence bids placed during the alleged “test period” of May 14-15, 2011, as well as

the convergence bids placed later during the remainder of that month.

1. It Was Economically Rational for ETRACOM to Believe During the Relevant

Period that Its Convergence Bidding Strategy at New Melones Could Be

Profitable on a Stand-Alone Basis

51. In response to staff’s investigation of its convergence bidding behavior at New Melones,

ETRACOM explained that it intended to profit from virtual supply offers placed there in May

2011 due to expectations of high hour-ahead import congestion over that intertie given an

expected El Niño weather pattern in 2011 and the fact that other import constraints into the

CAISO were already binding.54 ETRACOM explained that this strategy was designed to

53 Staff Report, pp. 23-24.

54 ETRACOM Narrative Response to FERC Data Request 3, p. 2.


23

capitalize on anticipated downward spikes in hour-ahead congestion prices at New Melones

relative to day-ahead congestion prices, such that ETRACOM could profit from virtual supply

offers until the day-ahead market could adjust to the actual congestion experienced.55

52. Staff’s Report alleges that ETRACOM’s explanation is “unpersuasive and unsupported”

given historical data available for New Melones.56 Staff asserts ETRACOM’s daily losses on its

virtual supply offers ranged from $871 to $5,851 per day, ultimately totaling $42,481 from

May 14-31.57 While stepping back from an erroneous statement in its Preliminary Findings that

ETRACOM “needed a -$1,000/MWh HASP price to be present in a majority of hours” for the

strategy to be profitable,58 staff now states “[h]istorical data available in May 2011 shows that

only 0.21% of hours had HASP prices lower than ETRACOM’s -$30/MWh supply offers”59 and

that “prices during the July 8-22 period isolated by ETRACOM also do not indicate an

unprecedented event.”60 Staff further contends that ETRACOM’s choice to continue placing

virtual supply offers at New Melones throughout May despite evidence of mounting losses

demonstrates that its motives were never to make a profit on these offers, but only to manipulate

the value of its CRR positions.61

a. Staff’s Ex-Post View of the Profitability of ETRACOM’s Hydroelectric

Trading Strategy Ignores the Market Presented at New Melones

53. ETRACOM anticipated a unique hydroelectric event unlike that experienced in the

region since 199762—the existence of which the CAISO DMM explicitly acknowledged63 and the

potential profitability of which staff previously had admitted was possible64 but now asserts is

55 See Rosenberg Deposition, pp. 312:21-314:14.


57 Id., pp. 10, 20.

58 Preliminary Findings, p. 18.


60 Id., p. 30.

61 Id., p. 32.

62 The phenomenon expected at New Melones was compared to the 1997 “Pineapple Express,” an event

in which high snowpack, followed by warm rains, caused significant congestion over the Pacific

Intertie and NOB. See Rosenberg Deposition, pp. 140:20-142:10.

63 The CAISO DMM’s document titled “Etracom at New Melones: Follow Up” dated December 9, 2013

(“DMM Follow Up”) states “[h]ydroelectric output was expected to be abnormally high in 2011. This

period historically persists at least into June[...] One would expect longer persistence in an abnormally

high hydro year such as 2011[...] Low system and regional HASP prices related to over supply from

high hydroelectric output historically have persisted well into June.” DMM Follow Up, p. 2.



24

“implausible.”65 Staff’s negativity does not comport with its recognition that “snow pack and

water levels at New Melones in 1997 and 2011 were similar,”66 suggesting that an infrequently

available opportunity was at play. This is validated by actual weather data which indicates that

snowpack and water levels were much higher in 2011 than in 1997, as is shown in Figure 5.

Figure 5

Sierra Nevada Snowpack and Water Content (1971-2015)67

Source: California Data Exchange Center; Snow data collected at Black Springs (Stanislaus R).

54. Figure 5 shows the monthly snow depth in inches (dark blue line) and water content

thereof (light blue line) at the Black Springs weather station upriver of the New Melones

reservoir in the Sierra Nevada Mountains from 1971 to 2015. The dotted lines show these values

for March 29, 2011, the highest levels of that year, just before ETRACOM had discussed the

65 Staff Report, pp. 28, 30.

66 Id., p. 28.

67 See Excel workbook titled “Sierra Nevada Analysis.xlsx,” tab “Figure 5.”


25

possibility of capitalizing on likely HASP congestion generally using virtual trades on April 25,68

and at or near New Melones as early as May 1.69 As the graph shows, maximum annual

snowpack and water content levels exceeded those in 2011 only once in this history (in 1983).

Staff now contends (with the comfort of perfect hindsight) that “there is no evidence that an

event like the Pineapple Express, and the associated accelerated snow melt, could reasonably be

expected” at the time. Staff is wrong, for the key ingredient needed for such an event—wet,

abundant snow—was present at level in excess of that which gave rise to the 1997 “Pineapple

Express” event. Since ETRACOM did not have the benefit of hindsight when formulating its

strategy, staff’s presumption of ETRACOM’s guilt relies upon the failure of the event to

materialize, while ignoring the highly increased likelihood that it could.

55. ETRACOM developed the hydroelectric trading strategy based on several fundamental

and technical factors, including these weather conditions and the congestion perceived at other

locations near New Melones.70 ETRACOM chose New Melones because the expected HASP

congestion had yet to arise as of May.71 Staff contends that there was nothing unique about the

hydro conditions at New Melones, and in fact that “ETRACOM’s virtual strategy would have

been potentially more profitable at other similar locations where the day-ahead LMP was

typically positive.”72 Aside from the fact that the exhibit cited provides no analysis to support

this statement, staff once again relies on the benefit of hindsight to portend that ETRACOM

failed to execute trades that there is no evidence it knew of or considered at the time.

56. More critically, staff has apparently missed that ETRACOM placed and cleared 1-5 MW of virtual supply bids at seventeen interties other than New Melones from May 25, 2011 through

May 31, 2011 in off-peak hours.73 Moreover, every single one of these virtual supply offers was made at the price of -$20/MWh,74 meaning that the ETRACOM used the same type of

purportedly “uneconomic” virtual trades at these other locations that it used at New Melones.

These trades were net profitable at all but one location and were placed at seven locations where

ETRACOM also held CRRs. The key difference between ETRACOM’s price-taking virtual trades

made at these other interties and those made at New Melones was the fully encumbered status of

latter, which staff has admitted ETRACOM would have known nothing about at the time the

trades at issue were placed. The price-taking nature of the (likely) well-functioning markets at

68 See ETRACOM Instant Message, 4/25/2011 10:41:54 AM – 10:44:07 AM, Bates No. ETR01450.

69 See ETRACOM Instant Message, 5/1/2011 2:20:51 PM – 2:21:26 PM, Bates No. ETR01459.

70 Rosenberg Deposition, p. 105:1-19.

71 Id., p. 302:6-303:14.


73 See Excel files titled “VT Bids May 2011.xlsx” and “May 2011 VT Awards.xlsx.”

74 See Excel file titled “VT Bids May 2011.xlsx.”


26

these other locations is confirmed by the fact that every one of ETRACOM’s -$20/MWh virtual supply offers cleared and yet not one of them set the day-ahead LMP.75

57. A deeper analysis of the interactions between ETRACOM’s virtual supply offers placed on

May 25-31 and its portfolio of CRRs provides further insights into the fallacies that underlie

staff’s ex post analysis of ETRACOM’s behavior at New Melones. Specifically, this analysis shows

that random interactions amongst the instruments within ETRACOM’s portfolio could be used to

support an erroneous manipulation claim based only on the “false positives” provided by

economic data, as staff has done here. This is shown in Table 2, below.

Table 2

Interaction between ETRACOM’s Virtual Trades and CRRs

(May 25-31, 2011)76

Sources:

[B]: "VT Bids May 2011.xlsx".

[C]-[D]: Velocity Suite, ABB Inc.

[F]-[I]: "FERC Request 20111216 Q7.csv".

Notes:

[C]-[D]: Averages prices during hours ETRACOM cleared.

[E]: [C] - [D].

[F]: Average volume cleared across time period.

[H]: [E]*[F]*[G].

[I]: Sink CRR Position - Source CRR Position.

[J]: - [I] / [F].

75 See Excel file titled “VT Bids May 2011_All Nodes.xlsx,” tab “Table 2.”

76 See Excel workbook titled “VT Bids May 2011_All Nodes.xlsx,” tab “Table 2.”

Intertie

Average

Offer Price

Average

Day Ahead

LMP

Average

Hour

Ahead LMP

Profit of

1 MW

Average

Hourly

Volume

# of Hours

Offered and

Cleared Total Profit

CRR Net

Position

($/MWh) ($/MWh) ($/MWh) ($/h) (MW) Off Peak ($) ($/MWh)

[A] [B] [C] [D] [E] [F] [G] [H] [I] [J]

CAPTJACK_5_N505 $20.00 $4.49 $0.98 $3.51 5.00 43 $754.39 $3.30 0.66

CAPTJACK_5_N512 $20.00 $4.49 $0.63 $5.12 1.00 43 $220.15 $0.00 0.00

CRAGVIEW_1_GN001 $20.00 $3.22 $0.94 $4.16 1.00 43 $178.72 $5.47 5.47

MALIN_5_N101 $20.00 $0.46 $2.85 $2.39 5.00 43 $513.29 $3.30 0.66

MEADS_2_N101 $20.00 $4.77 $1.75 $3.02 5.00 43 $648.62 $0.00 0.00

MERCHANT_2_N101 $20.00 $4.77 $1.75 $3.02 5.00 43 $648.45 $0.00 0.00

MOENKOPI_5_N101 $20.00 $4.73 $1.74 $2.99 5.00 43 $643.31 $15.82 3.16

NGILA1_5_N001 $20.00 $4.82 $1.78 $3.04 5.00 43 $653.80 $7.54 1.51

PALOVRDE_ASR APND $20.00 $4.72 $1.74 $2.99 5.00 43 $642.40 $7.54 1.51

PARKER_2_N101 $20.00 $4.74 $1.74 $3.00 5.00 43 $645.54 $0.00 0.00

ROA 230_2_N101 $20.00 $4.87 $1.80 $3.07 5.00 43 $660.25 $0.00 0.00

SMD5_ASR APND $20.00 $4.67 $1.65 $3.01 5.00 43 $648.14 $0.00 0.00

SMDH_ASR APND $20.00 $4.67 $0.55 $5.21 1.00 43 $224.11 $0.00 0.00

SYLMARDC_2_N501 $20.00 $0.70 $0.85 $1.55 5.00 43 $333.74 $2.78 0.56

SYLMARLA_2_N501 $20.00 $4.83 $1.77 $3.05 5.00 43 $656.57 $0.00 0.00

TJI 230_2_N101 $20.00 $4.92 $2.32 $2.60 5.00 43 $558.10 $0.00 0.00

VICTORVL_5_N101 $20.00 $4.82 $1.77 $3.05 5.00 43 $655.34 $0.00 0.00

Financial

Leverage


27

58. Table 2 provides a natural basis for comparison because ETRACOM placed and cleared

the same number of virtual supply offers for the same price at all seventeen locations over the

same period of time. For each market across all hours cleared, the table reports the average day-

ahead congestion price, day-ahead LMP and hour-ahead LMP associated with ETRACOM’s

offers and calculates the average profitability of the associated virtuals (potential triggers). The

table next calculates ETRACOM’s net CRR position in place at each intertie (potential targets)

and calculates the resulting financial leverage presented. Leverage greater than one77 suggests a

concern based on the framework, because losses on the virtuals could be used to benefit the

corresponding CRR position such that the net result is profitable overall. Based on this overly-

simplistic viewpoint, three markets (shaded in green) theoretically look problematic, one

particularly so due to the relatively large amount of import congestion shown (potential nexus)

and the fact that the virtual supply offers cleared there lost money (potential trigger).

59. Without any documentary evidence of intent—much like the case at issue here—this

behavior could be viewed as suspicious when viewed in isolation. However, correlation is not

causation. The virtual supply offers at this location, like ETRACOM’s trades at New Melones, are

examples of “false positives,” proven by the fact similar trades were placed at seventeen locations

with varying results. However, unlike the (likely) well-functioning markets at these seventeen

other locations, ETRACOM could not anticipate how the flawed, dysfunctional behavior of New

Melones would treat its zero or negatively-priced price-taking virtual supply offers. Staff’s

allegation that ETRACOM’s offers at New Melones were intentionally uneconomic or anomalous

ignores the causal impact that these flaws had on (a) incentivizing the virtual supply offers it

placed there and (b) driving the losses on those virtual supply offers that cleared. Staff thus

denies ETRACOM the presumption of legitimacy that should be afforded to the other losing

trades shown in Table 2. Indeed, random interactions are expected in any portfolio, some

theoretically helping, others injuring the values of the related instruments. This is verified by

other locations in Table 2 (shaded in blue) characterized by negative financial leverage, meaning

that the virtual trades tended to injure the value of the associated CRRs at those locations.

60. It is noteworthy that the trading strategies at sixteen of these interties were profitable,

while one lost $4.16/MWh—about $178.72. Staff complains that if ETRACOM had wanted to

make money using a hydroelectric trading strategy, “there were many other potentially more

profitable locations ETRACOM could have chosen for such a strategy.”78 Indeed, staff’s list

includes three locations where ETRACOM cleared virtual supply from May 25-31, a fact that it

77 The CRR position is expressed as a positive number if the net of all positions is a sink, and a negative

number if it is a source. Financial leverage is calculated as –[CRR MW/Virtual MW], which will be

positive if the net CRR position is a source. Leverage then exists if the size of the CRR source in MW

exceeds the size of the average virtual supply offers in MW, shown by a value greater than one.



28

either does not recognize or does not acknowledge.79 As at New Melones, ETRACOM tried to

capitalize on potential HASP congestion at other interties, but with better success. Whereas staff

expounds at some length as to why ETRACOM’s trades at New Melones should be distinguished

from its other trades in May,80 no mention is made of this key and critical distinction:

ETRACOM’s price-taking virtual supply offers on the interties were generally profitable in May

2011 when the markets functioned properly. But for the flaws in the CAISO’s market design,

ETRACOM’s offers at New Melones would likewise have been price-taking and had no effect on

prices—and the congestion generally present in the region could have proven profitable there as

well. As the success of ETRACOM’s strategy at these other interties confirms, ETRACOM’s

losses at New Melones were caused by these flaws, not by its negatively-priced offers.

b. Data Confirms ETRACOM’s Expectations Concerning the Profitability of

Its Hydroelectric Trading Strategy at New Melones

61. Staff maintains that the potential profitability of ETRACOM’s hydroelectric trading

strategy at New Melones was “implausible” “[g]iven the difficulty in predicting the timing of an

event, the uncertain payout, and the fact that a significant hydro event was not likely to occur at

all[.]”81 Staff’s presumption of what constitutes a “significant hydro event” aside, the fact is that

ETRACOM’s sixteen-day trading strategy conducted across all hours—commensurate with staff’s

“fourth phase” of the alleged manipulation—would have been profitable at New Melones if: (a)

the strategy was placed in July 2011 instead of May; and (b) the market there had functioned

normally, such that virtual supply bids placed by ETRACOM would have been price-taking.

Staff believes that this second assumption leads to results that are “misleading” because the flaws

and software errors at New Melones caused ETRACOM’s negatively-priced virtual supply bids to

set the day-ahead market price there in several hours.82 However, because ETRACOM did not

know of these flaws, its ex ante understanding would be that prices were formed in accordance

with the CAISO tariff—and under the tariff, ETRACOM’s offers would not have set the price or

caused congestion. Indeed, staff ignores that ETRACOM also offered virtual supply at prices

of -$20/MWh at other interties from May 25-31, none of which set the prices at those locations.

It is therefore reasonable to consider ETRACOM’s offers as price-taking for this analysis.

62. Figure 6, below, shows a time series of the daily average spread between the day-ahead

and hour-ahead LMPs experienced at New Melones during 2011, which would equal the average

79 Hourly_Charts_Hydronodes.pdf. The three locations are CAPTJACK_5_N505, CAPTJACK_5_N512,

and MALIN_5_N101.

80 Staff Report, pp. 12, 23, 24, and 38.


82 Id., pp. 30-31. Were the 384 MW New Melones intertie a well-functioning and competitive market,

the negatively-priced 1 MW to 5.03 MW virtual supply bids placed by ETRACOM would have been

price-taking to day-ahead and HASP prices. In reality, ETRACOM’s bids were infra-marginal in

43.5% of the hours from May 16-31, despite the flaws presented at New Melones during that time.


29

hourly profit or loss made on a price-taking 1 MW cleared virtual supply offer. Prices above the

horizontal axis show profits driven by either higher day-ahead prices or lower hour-ahead prices.

The red box represents the sixteen-day period from May 16-31 in which ETRACOM executed its

trading strategy. As the losses indicate, the expected downward spike in hour-ahead congestion

prices never materialized during this period, thus contributing to the losses from ETRACOM’s

virtual supply offers. However, had these virtual supply offers been placed in July 2011 instead

of May, the potential profit could have been much greater.

Figure 6

The Potential Profitability of Virtual Supply at New Melones83

Source: Velocity Suite, ABB Inc.

63. For example, a 1 MW price-taking virtual supply bid placed every hour over the 16-day

period from July 8, 2011 through July 24, 2011 (shown by the green box in Figure 2) would have

made a net profit of about $5,132, $1,317 during off-peak hours and $3,815 during peak hours.84

83 See Excel workbook titled “Virtual Profits.xlsx,” tab “Figure 6.”

84 See Excel workbook titled “Virtual Profits.xlsx,” tab “Table 3.”


30

A 5 MW price-taking virtual supply position placed over this period would thus have netted

$25,661, $6,587 in off-peak hours and $19,074 in peak hours. In addition, had ETRACOM placed

a 5 MW price-taking virtual supply bid every hour of July 14, 2011, the strategy would have

earned $8,763 ($249 in off-peak hours, $8,514 in peak hours), due to hour-ahead LMPs that fell

as low as -$389.90/MWh.85 Although mistimed, ETRACOM’s strategy was based on reasonable

view of market conditions that ultimately came to fruition two months later in the 2011

hydroelectric season.

64. The profit potential of ETRACOM’s strategy was not limited to 2011. Table 3 shows the

maximum potential profit that could be earned from a 1 MW or 5 MW price-taking virtual

supply offer cleared in every hour over a sixteen-day period at New Melones if started between

the months of April and September during any year from 2009 to 2015. As the table shows, this

strategy could have made money in every year, with maximal profits in two of the seven years

derived by starting the strategy before the May 16 start date used by ETRACOM in 2011.

85 See Excel workbook titled “LMP and Congestion Prices.xlsx,” tab “Daily Profit.”


31

Table 3

The Potential Profitability of the Hydroelectric Trading Strategy

at New Melones (2009-2015)86

Source: Output from R code, "Maximum Summer Rolling Profits.csv".

Notes:

[1]: Maximum summer profits from following the 16-day hydroelectric strategy.

[2]: Cumulative on peak profits from that period.

[3]: Cumulative off peak profits from that period.

[4]: Starting HE of the 16-day strategy.

[5]: Ending HE of the 16-day strategy.

65. Because of the difficult timing needed to obtain these maximal profits, I calculated the

percentage of days in which a 1 MW price-taking virtual supply offer cleared in every hour over

a sixteen-day span would be profitable during each given year, assuming the strategy began


Year Total Profits

On Peak

Profits

Off Peak

Profits Start Date HE End Date HE

[1] [2] [3] [4] [5]

2009 $6,986 $2,987 $3,999 4/10/2009 9 4/26/2009 8

2010 $4,857 $1,226 $3,631 6/19/2010 2 7/5/2010 1

2011 $5,132 $3,815 $1,317 7/8/2011 9 7/24/2011 8

2012 $7,812 $6,023 $1,789 7/8/2012 2 7/24/2012 1

2013 $4,522 $1,301 $3,221 8/31/2013 3 9/16/2013 2

2014 $7,293 $2,957 $4,337 5/11/2014 11 5/27/2014 10

2015 $6,609 $8,442 $1,833 6/25/2015 14 7/11/2015 13

Year Total Profits

On Peak

Profits

Off Peak

Profits Start Date HE End Date HE

[1] [2] [3] [4] [5]

2009 $34,932 $14,936 $19,995 4/10/2009 9 4/26/2009 8

2010 $24,286 $6,129 $18,157 6/19/2010 2 7/5/2010 1

2011 $25,661 $19,074 $6,587 7/8/2011 9 7/24/2011 8

2012 $39,059 $30,115 $8,944 7/8/2012 2 7/24/2012 1

2013 $22,608 $6,505 $16,103 8/31/2013 3 9/16/2013 2

2014 $36,467 $14,784 $21,683 5/11/2014 11 5/27/2014 10

2015 $33,044 $42,209 $9,164 6/25/2015 14 7/11/2015 13

5 MW Virtual Supply Bid


32

between April 1 and August 31 of that year. These results are shown for the period 2009-2015 in

Table 4, below.

Table 4

Days When a Sixteen Day Virtual Supply Bid Would Be Profitable

at New Melones (1st Trade Starts April 1 – August 31)87

Source:

Output from R code, "Percentage of Profitable Days with Trades Starting on HE 1".

Notes:

[1]: Summer months are defined as periods starting in April through August.

[2]: Trading is assumed to start on HE 1 and continue at all hours for 16

consecutive days.

[3]: [2] / 153.

66. The hydroelectric trading strategy would have been profitable on average in almost 72%

of all hours over this span, and would have been profitable almost always in the two years prior

to 2011. Indeed, had convergence bidding been allowed since the inception of the CAISO latest

market design on April 1, 2009, a 1 MW virtual supply bid placed in every hour at New Melones

prior to May 1, 2011 would have on average netted $3.22/MWh.88 This shows conclusively that

ETRACOM’s actions were rational even without the extraordinary hydro event expected,

assuming the market at New Melones had functioned properly. As it was, ETRACOM may have

thought that higher profits—or lower premiums for executing the strategy—would have been

possible given the minimal losses it incurred during staff’s “test period,” the small gain it made on

May 15, and the fact that it did not set the price in almost half of the hours in which it cleared

virtual supply at New Melones from May 14-31.


88 See Excel workbook titled “LMP and Congestion Prices.xlsx,” tab “LMP and Congestion Data.”

Year Profitable Days % Profitable

[1] [2] [3]

2009 142 92.8%

2010 153 100.0%

2011 72 47.1%

2012 129 84.3%

2013 72 47.1%

2014 125 81.7%

2015 76 49.7%

Average 110 71.8%


33

c. Staff Mischaracterizes ETRACOM’s Tracking of Its Virtual Losses during

the Relevant Period

67. Staff lists several electronic communications that it alleges “show ETRACOM’s

disproportionate interest in New Melones; its employees discussed ETRACOM’s performance at

New Melones almost daily[.]”89 Interestingly, the first document listed by staff is dated May 1,

2011 and obliquely discusses virtual trading with a reference to New Melones a full week before

the WAPA flows that motivated the alleged manipulation had occurred.90 Staff also fails to

include in its list nine other documents in its Administrative Record that show ETRACOM’s

conversations regarding New Melones that took place before May.91 Of the documents that were

listed by staff, only two specifically mention losses on the virtual supply offers at New Melones.

In the first, a contractor to ETRACOM reported toward the end of an exchange on May 16, 2011,

“[w]e lost $800 on Melon[e]s but made back $200 on some evening trades[.]”92 At no point in the

cited instant message is this loss linked to gains in the ETRACOM’s CRRs at New Melones.

68. The second document cited twice by staff is a May 20, 2011 instant message in which the

same contractor reports “[y]esterday Melon[e]s cost us about $2K[.]”93 Although in one of its

references staff adds the rest of this sentence of the message—“- continue with it?”94—staff

mischaracterizes the broader intent connoted by this message. The full exchange follows:

5/20/2011 7:33:20 AM Mike Davis Yesterday Melons cost us about $2K-

continue with it?

89 See Staff Report, p. 19 and n. 102, pp. 19-20.

90 At 2:15:04 PM, Arik Kapulkin wrote “Q: what is the location of the document that summarizes (and

includes plans for) the state of affairs of the VT trading?.” After responses from Michael Rosenberg, at

2:20:51 PM Kapulkin wrote “slightly OT comment about 33936_MELNS JB_115_33951_ VL YHM: it

is NOT directly related to the NEWMELON lTC, its cause is massive production of Hydro power

(same as the cause of NEMELON lTC binding events), hence it is generally related (= similar).” Less

than a minute later at 2:21:26 PM, Kapulkin wrote “Q re: VT risk analysis.” See ETRACOM Instant

Message, 5/1/2011 2:15:04 PM – 2:21:26 PM, Bates No. ETR01459.

91 See ETRACOM Instant Message, 3/10/2011 12:23:18 PM – 12:35:21 PM, Bates No. ETR01387

ETRACOM Instant Message, 3/13/2011 1:19:50 PM, Bates No. ETR01398, ETRACOM Instant

Message, 3/22/2011 11:33:10 AM – 11:36:0 AM, Bates No. ETR01411, ETRACOM Instant Message,

3/29/2011 11:18:16 AM – 11:24:037 AM, Bates No. ETR01431, ETRACOM Instant Message, 3/30/2011

8:02:49 AM – 8:03:10 AM, Bates No. ETR01435, ETRACOM Instant Message, 3/30/2011 3:26:13 PM,

Bates No. ETR01438, ETRACOM Instant Message, 4/8/2011 4:12:46 PM, Bates No. ETR01654,

ETRACOM Instant Message, 4/16/2011 1:07:49 PM – 1:07:55 PM, Bates No. ETR01443, and

ETRACOM Instant Message, 4/25/2011 10:55:36 AM, Bates No. ETR01450.

92 Staff Report, p. 10, citing ETRACOM Instant Message, 5/16/2011, 9:47:36 PM, Bates No. ETR01508.

93 Id., pp. 10 and 20, citing ETRACOM Instant Message, 5/20/2011 7:33:20 AM, ETR01509.

94 Id., p. 10.


34

5/20/2011 7:33:58 AM Michael Rosenberg not sure, i am thinking we

should stop putting positions on melon until the

auction end

5/20/2011 7:34:48 AM Mike Davis Or just put them in off peak.

5/20/2011 7:35:11 AM Michael Rosenberg neither95 (emphasis added).

This exchange proves that ETRACOM did indeed evaluate the profitability of the hydroelectric

trading strategy on a stand-alone basis, shown by the fact that Rosenberg was willing to end it on

May 20, 2011 based on the prior days’ losses and his need to focus on the pending CAISO auction

for June CRRs.

69. Far more importantly, ETRACOM’s consideration of stopping the hydroelectric trading

strategy due to a roughly $2,000 loss observed on May 19, 2011 demonstrates that it was not

linking the profitability of these virtual trades with the profitability of its CRRs sourced at New

Melones at the time. This is demonstrated in Table 5, below.

95 ETRACOM Instant Message, 5/20/2011 7:33:20 AM – 7:35:11 AM, Bates No. ETR01509.


35

Table 5

ETRACOM Did not Compare Virtual Losses to CRR Gains96

Sources: Velocity Suite, ABB Inc. and client data.

Notes:

[1]: Profits and losses calculated from DA and HA LMPs.

[2]: CRR profits calculated based on shadow prices at New Melones.

[3]: "Unjust" profits calculated using FERC's method.

[4]: Rolling sum of [1].



[7]: ETRACOM instant messages ETR01508 and ETR01509.

70. Table 5 shows that if ETRACOM had been tracking the profitability of its CRRs relative

to the $1,997 loss on its virtual trades on May 19, 2011—a must for staff’s allegations to hold—it

would have observed that the value of its CRRs at New Melones increased $38,195 that day. If

ETRACOM tracked its cumulative gains and losses through May 19, it would have known that

the $7,199 in losses on its virtual trades corresponded with a $176,943 gain in its CRRs. Indeed,

if ETRACOM had been tracking only the allegedly “unjust” profits derived from its CRRs as staff

alleges, it would have observed gains of $24,127 on May 19 and $93,354 from May 14-19.97

Given the financial leverage of about 19:1 on May 19 and almost 25:1 for the period May 14-19

96 See Excel workbook titled “May Virtual and CRR Daily Profits.xlsx,” tab “Table 5.”

97 For the calculations listed in this paragraph, see Table 5 above or Excel workbook titled “Shadow

Prices Profits Screen and Financial Leverage Analysis.xlsx,” tab “Financial Leverage Summary.”

Date

Daily Virtual Bid

Profits Daily CRR Profits

Alleged Daily

Unjust CRR Profits

Cumulative

Virtual Bid Profits

Cumulative CRR

Profits

Cumulative

Alleged Daily

Unjust CRR Profits

[1] [2] [3] [4] [5] [6]

5/14/2011 ($59) $17,870 $2,366 ($59) $17,870 $2,366

5/15/2011 $6 $10,189 $719 ($52) $28,059 $3,085

5/16/2011 ($871) $39,278 $22,866 ($923) $67,338 $25,951

5/17/2011 ($2,252) $33,242 $19,174 ($3,175) $100,580 $45,125

5/18/2011 ($2,027) $38,169 $24,101 ($5,202) $138,749 $69,226

5/19/2011 ($1,997) $38,195 $24,127 ($7,199) $176,943 $93,354

5/20/2011 ($2,225) $37,654 $23,586 ($9,424) $214,598 $116,940

5/21/2011 ($1,165) $24,636 $10,568 ($10,589) $239,234 $127,508

5/22/2011 ($988) $16,827 $19,446 ($11,577) $256,061 $146,954

5/23/2011 ($2,326) $32,101 $18,033 ($13,903) $288,162 $164,987

5/24/2011 ($3,334) $33,379 $19,311 ($17,237) $321,540 $184,298

5/25/2011 ($4,080) $33,021 $18,953 ($21,317) $354,561 $203,251

5/26/2011 ($2,772) $33,977 $19,909 ($24,089) $388,538 $223,160

5/27/2011 ($3,445) $34,499 $20,432 ($27,534) $423,037 $243,591

5/28/2011 ($4,652) $28,963 $14,896 ($32,186) $452,000 $258,487

5/29/2011 ($2,128) $14,709 $17,327 ($34,314) $466,709 $275,814

5/30/2011 ($2,316) $16,989 $19,608 ($36,630) $483,699 $295,422

5/31/2011 ($5,851) $33,718 $19,650 ($42,481) $517,417 $315,072


36

created by the alleged scheme,98 ETRACOM as a rational economic actor would never have

considered ending the hydroelectric strategy unless it was guided by a stand-alone, profit-seeking

motivation on those virtual trades. Having been made aware of this issue in ETRACOM’s

Response and my Affidavit filed in response to its Preliminary Findings,99 staff’s failure to present

the rest of the relevant instant message in its report filed before this Commission and made

available to the public is unacceptable.

d. ETRACOM’s Virtual Trading at New Melones during the Relevant Period

Was Based on a Rational Strategy Unconnected to its CRRs

71. Staff’s Report presumes that ETRACOM’s virtual supply offers were intended to negate

the export congestion presented at New Melones as of May 14, 2011, the first day of the alleged

“test period.” Yet staff agrees that ETRACOM would have understood neither the cause of the

congestion nor the market prices at New Melones given the intertie’s fully-encumbered nature

and software errors as of that date. Offering virtual supply for $0/MWh in off-peak hours of that

day, ETRACOM cleared and set the price in seven of those eight hours for a total loss of about

$59, but zero or negative LMPs appeared in thirteen of the remaining seventeen hours of that

day.100 ETRACOM followed this on May 15 by clearing four of nine hours at $0/MWh for about

a $6 profit, but negative LMPs appeared in fourteen of the remaining twenty hours of that day. Whether ETARCOM considered these outcomes as a “loss” is debatable, especially given that it

viewed the payment as a premium that would give it exposure to the potential HASP congestion.

Staff nevertheless alleges that these results confirm ETRACOM’s intent to place uneconomic

transactions to benefit its CRRs at New Melones, purportedly confirmed later by its willingness

to absorb losses on its virtual supply bids throughout the rest of the month.

72. However, the May 20 instant message discussed above proves that ETRACOM was not

tracking a link between its virtual trading and CRRs as of May 20, 2011, six days after the

strategy had begun. If ETRACOM did not perceive the linkage between its virtual trading and

CRRs as of May 20, then at a minimum it would not have been attempting to manipulate the

market at any time prior to that. As such, the only possible explanation for ETRACOM’s virtual supply offers prior to May 20 was the rational, profit-seeking and legitimate pursuit of its

hydroelectric trading strategy. If staff cannot prove ETRACOM’s guilt from the outset of its Test

Period, its allegations for the entire Relevant Period must fail given the rational economic

decision making displayed in the documentary evidence presented.

98 Financial leverage is measured as the gains from the alleged target divided by the losses in the alleged

trigger, equaling $38,195/$1,997 = 19.13 for May 19 and $176,943/$7199 = 24.58 for May 14-19.

99 See Affidavit of Shaun D. Ledgerwood on Behalf of ETRACOM LLC and Michael Rosenberg, filed

September 30, 2015, P 68.

100 See Table 5 and Excel workbook titled “May Virtual Bids vs. LMPs.xlsx,” tab “Summary.”


37

73. Even if ETRACOM closely tracked its daily and cumulative losses on its virtual supply

offers, it could be rational for it to continue implementing the hydroelectric trading strategy on a

day-by-day basis given the continuing expectation of an unprecedented congestion event. A

rational economic market participant would base its choice to place a pending day’s trades on the

expected profitability of those trades, not the sunk costs incurred on prior trades. Thus, if

ETRACOM believed its virtual supply offers for a given day would be profitable given the

information available to it on an ex ante basis at the time those offers were placed, it acted in an

economically-rational manner.

74. Ultimately, ETRACOM chose to continue this strategy despite each prior day’s loss. Its

explanation for doing so is consistent with the economically-rational perspective of evaluating

the relative costs and benefits of a transaction on a day-to-day basis given the trader’s forward-

looking expectations of market conditions that were ultimately confirmed, though mistimed, as

Rosenberg confirmed:

There is a lot of temptation after a few days of trying and seeing the losses to jump

out and just write off the loss. But that needs to be resisted, and it's a bit of a

trading psychology. Because when you put a long position in anticipation of real-

time or HASP event, you need to have staying power in the market.101

Given snow conditions upriver of the New Melones Reservoir that could spawn a hydroelectric

event even more dramatic than the conditions in 1997 produced, ETRACOM was rational to

continue to pursue this strategy on a day-to-day basis.

75. In sum, staff relies on the benefit of selective hindsight to allege that ETRACOM chose to

uneconomically sacrifice profits on “an average of about $2,600 per day”102 up to a lump sum of

$42,481 in furtherance of a strategy designed to manipulate the value of its CRRs, without

considering the market flaws that incentivized ETRACOM’s virtual supply offers and were

responsible for its losses. ETRACOM did not have the benefit of this hindsight when it placed

those trades day-by-day, nor could it have known before its trading that the hydroelectric

conditions needed to make such a strategy profitable were still two months away. Based upon

historical data from 2009 and 2010, however, the strategy should have been profitable if the day-

ahead market presented at New Melones had been well-functioning. Further, as I will discuss

later when refuting staff’s assertions that ETRACOM’s trades at New Melones were anomalous,

ETRACOM mimicked this strategy starting mid-June using virtual supply offers placed upriver of

the New Melones reservoir, generally at prices of either $0/MWh or -$20/MWh. ETRACOM’s

trading activity during the Relevant Period at New Melones was based on its economically-

rational view of market conditions and the publicly available information at the time.

101 Rosenberg Deposition, p. 225:18-24.



38

2. ETRACOM Was Incentivized By the Software Pricing/Modeling Errors to Bid

at Zero or Negative Prices to Clear Its Virtual Supply Offers at New Melones

76. The Staff Report alleges that ETRACOM intentionally offered virtual supply at New

Melons at zero or negative prices during the Relevant Period that was “designed to lower day-

ahead LMP [sic] at New Melones.”103 This attributes manipulative intent to ETRACOM when in

fact the numerous flaws in the CAISO’s market design and errors in its software at New Melones

incentivized its behavior. For example, staff contends that on May 14 and 15 “ETRACOM placed

$0 virtual supply offers in mostly off-peak hours, essentially offering free energy from New

Melones into CAISO.”104 Staff admits that a key software error erroneously resulted in zero

prices at New Melones, stating “[i]n some hours, the software set the LMP at $0 when it should

have been a positive amount.”105 An example of this problem is shown below in Table 6.

103 Id., p. 17.

104 Id., pp. 17-18.

105 Id., p. 33. See also DMM Referral Attachment 1, p. 2, n. 2. Note the price was also erroneously set to

zero in other circumstances not described by DMM. See supra n. 38.


39

Table 6

The Effect of Virtuals on Day-ahead Prices at New Melones

April 28, 2011106

Sources & Notes:

[1]: Congestion expected given physical flows below the line’s 384 MW import capacity.

[2]: Day-Ahead Hourly Congestion, taken from Velocity Suite, ABB Inc.

[3]: Day-Ahead Hourly LMP, taken from Velocity Suite, ABB Inc.

[4]: Day-Ahead Scheduled Export at New Melones, CAISO OASIS (transmission interface usage option).

[5]: Day-Ahead Scheduled Import at New Melones, CAISO OASIS (transmission interface usage option).

[6]: Flags if ETRACOM's Virtual Supply Bid equals LMP in that hour.

[7]: Flags if Virtual Supply Offer set the LMP in that hour.

[8]: Flags if Virtual Supply was bid in that hour.

77. Table 6 shows the hourly day-ahead congestion and total LMPs at New Melones on April

28, 2011. Virtual supply offers were placed by another market participant in all (shaded) on-

peak hours. In HE 7 and 18, these virtual supply bids set the market price with negative LMPs,

meaning that this other market participant bid virtual supply into the market at negative prices.

For all other hours, the market price cleared at $0/MWh, reflecting the software error admitted

by staff and that shaped ETRACOM’s price discovery prior to May 2011.

78. Staff does not mention the frequency of this error, which affected 29.6% of the day-ahead

LMPs at New Melones from April 14, 2011 through May 13, 2011.107 Further, ETRACOM had

106 See Excel workbook titled “Transmission Congestion Analysis.xlsx,” tab “Table 6.”

Date HE

Expected DA

Congestion per

Tariff ($/MWh)

Actual DA

Congestion

($/MWh)

Actual DA

LMP

($/MWh)

DA Physical

Export (MW)

DA Physical

Import (MW)

Did ETRACOM's

Bid Set the LMP?

Did Virtual

Supply Offer

Set the LMP?

Was Virtual

Supply

Offered in

the Hour?

[1] [2] [3] [4] [5] [6] [7] [8]

4/28/2011 1 $0.00 $0.00 $20.20 0 0 No No No

4/28/2011 2 $0.00 $0.00 $15.20 0 0 No No No

4/28/2011 3 $0.00 $0.00 $1.07 0 0 No No No

4/28/2011 4 $0.00 $0.00 $0.35 0 0 No No No

4/28/2011 5 $0.00 $0.00 $8.55 0 0 No No No

4/28/2011 6 $0.00 $0.00 $19.89 0 0 No No No

4/28/2011 7 $0.00 ($37.07) ($6.08) 0 0 No Yes Yes

4/28/2011 8 $0.00 ($34.76) $0.00 0 0 No No Yes

4/28/2011 9 $0.00 ($37.48) $0.00 0 0 No No Yes

4/28/2011 10 $0.00 ($34.33) $0.00 0 0 No No Yes

4/28/2011 11 $0.00 ($34.43) $0.00 0 0 No No Yes

4/28/2011 12 $0.00 ($37.36) $0.00 0 0 No No Yes

4/28/2011 13 $0.00 ($38.50) $0.00 0 0 No No Yes

4/28/2011 14 $0.00 ($35.74) $0.00 0 0 No No Yes

4/28/2011 15 $0.00 ($35.34) $0.00 0 0 No No Yes

4/28/2011 16 $0.00 ($36.06) $0.00 0 0 No No Yes

4/28/2011 17 $0.00 ($33.91) $0.00 0 0 No No Yes

4/28/2011 18 $0.00 ($34.19) ($1.54) 0 0 No Yes Yes

4/28/2011 19 $0.00 ($26.59) $0.00 0 0 No No Yes

4/28/2011 20 $0.00 ($32.20) $0.00 0 0 No No Yes

4/28/2011 21 $0.00 ($42.41) $0.00 0 0 No No Yes

4/28/2011 22 $0.00 ($32.84) $0.00 0 0 No No Yes

4/28/2011 23 $0.00 $0.00 $26.54 0 0 No No No

4/28/2011 24 $0.00 $0.00 $18.36 0 0 No No No


40

placed positively-priced virtual supply offers at New Melones on March 11-15, 2011, days when

$0 day-ahead LMPs appeared.108 Given this experience, it is reasonable that ETRACOM would

have perceived the need to bid its virtual supply at zero (or lower) prices—i.e., as a price-taker to

the market—to assure that it would clear in pursuit of its hydroelectric trading strategy.

79. Staff states this software error might have motivated ETRACOM’s behavior, stating “this

[software error] argument could only explain why ETRACOM’s offers were zero or negative.”109

This directly undercuts a key element of the framework that is necessary for staff to prove a

manipulation: that ETRACOM offered intentionally-uneconomic virtual supply into the day-

ahead market by offering at $0 prices on May 14-15 or negative prices on May 16-31. Because

the hydroelectric trading strategy required the virtual supply bids to clear the day-ahead market,

and because the software error and (as discussed below) other price signals indicated to

ETRACOM that $0 or negatively-priced offers were necessary to clear, those $0 or negatively

priced bids were not intentionally uneconomic, but rather served a stand-alone, legitimate

business purpose consistent with ETRACOM’s hydroelectric trading strategy. The assertions

made throughout the Staff Report that ascribe fraudulent or manipulative intent to ETRACOM’s

zero or negatively-priced virtual supply offers are invalidated by staff’s own observation.

80. In addition to the 213 hours of day-ahead LMPs that cleared at $0/MWh at New Melones

from April 14, 2011 through May 13, 2011 due to the $0/MWh software error, another 126 hours

cleared at a negative LMP during this period, 114 of which were set by convergence bidding and

95 of which were set by negative virtual demand bids—the lowest of which set a price

of -$26.80/MWh.110 Given no knowledge of the software pricing error, fully-encumbered status,

or composition of trades that set the market price at New Melones, ex ante ETRACOM’s only

known information relevant to the placement of its virtual supply offers was that 339 out of 720

hours (47%) had cleared at prices between $0/MWh and -$26.80/MWh.

81. Staff’s allegations also ignore the role that virtual demand played in this market while

ETRACOM engaged in its hydroelectric trading strategy. For example, of the seventeen hours of

“free energy” offered by ETRACOM on May 14-15, over one-third of those supply offers failed to clear given day-ahead market clearing prices of zero or less set by other market participants’

Continued from previous page

107 Specifically, 213 out of the 720 hours during this period cleared at a $0 day-ahead LMP that did not

coincide with a $0 virtual bid and when only positively-priced virtual supply bids were bid into the

market. See Excel workbook titled “Zero LMP Bid Analysis.xlsx,” tab “Full Summary.”

108 See Excel workbook “Zero-Neg LMP Bid Analysis.xlsx,” tab “March 2011 Summary.”

109 Staff Report, p. 33. Note the price was also erroneously set to zero in other circumstances not

described by DMM. See supra, n. 38.

110 See Excel workbook titled “Zero LMP Bid Analysis.xlsx,” tab “Neg Bid Summary” And tab “Full

Summary.”


41

negative demand bids.111 Given that market clearing prices were negative in 26 of the 48 hours

(54%) over this two day period, including a price of -$15.51 in HE 24 on May 15 when

ETRACOM’s $0/MWh virtual supply offer failed to clear,112 it was reasonable for ETRACOM to

believe that it needed to offer its virtual supply offers for the rest of the month at negative prices

to assure that they would clear. Staff mentions none of this in its report, as it confirms that a

presumption of transactional legitimacy must attach to ETRACOM’s trades.

82. Starting on May 13, a market participant other than ETRACOM began placing virtual

demand bids at positive and negative prices; four other parties placed virtual demand bids later in

the month, three of whom offered at negative prices.113 Due to the fully-encumbered nature of

the tie, ETRACOM’s virtual supply offers incorrectly set the day-ahead price in 56.5% of the

hours between May 16-31; however, the negatively-priced demand bids in the market, the

existence of which would not have been made available to ETRACOM and other market

participants until 90 days after the fact,114 set the price in the other 43.5% of those hours.

ETRACOM could rationally perceive this scenario as a competitive response in confirmation of

its hydroelectric trading strategy,115 thus necessitating the continued use of offers at negative

prices to assure they would clear. More importantly, ETRACOM would have viewed the 43.5%

of hours in which it was not marginal as evidence that its virtual supply offers were price-taking,

as would be expected of 1-5.03 MW trades at an ostensibly competitive intertie with a 399 MW

range (384 MW import plus 15 MW export). Such results would have been expected given

observations of negative prices in 126 of all hours between April 14, 2011 through May 13 and 26

of the 48 hours on May 14-15.116

83. It is therefore rational that ETRACOM perceived the need to set its virtual supply offers

from May 14, 2011 through May 31, 2011 at zero or negative prices to assure they would clear.

Had New Melones functioned normally, ETRACOM’s virtual supply offers would have been

price-taking, as they were at the seventeen other interties where ETRACOM made virtual supply

offers for -$20/MWh from May 25-31. Staff presumes that ETRACOM understood and exploited

the software errors and other flaws of this fully encumbered intertie’s market well before the

CAISO CRR Group disclosed the first of such problems to CRR market participants in June 2011.

111 See ETRACOM’s response to FERC Data Request, Question 7, an Excel workbook titled “FERC

Request 20111216 Q7.csv” and Excel workbook “May Virtual Bids vs. LMPs.xlsx,” tab “Summary.”

112 See Excel workbook titled “May Virtual Bids vs. LMPs.xlsx,” tab “Summary.” Note that the lowest

price on May 15 registered at -$19.88 for HE 10. Id.

113 See Excel workbook titled “May Virtual Bids vs. LMPs.xlsx,” tab “Bidder Table.”

114 CAISO posts masked virtual bids and offers after a 90 day lag. See Wholesale Competition in Regions

with Organized Electric Markets, Order No. 719, FERC Stats. & Regs. ¶ 31,281, (2008), P. 424.

115 See Excel workbook titled “May Virtual Bids vs. LMPs.xlsx,” tab “Summary.”

116 See Excel workbook titled “Zero-Neg LMP Bid Analysis.xlsx,” tab “Full Summary.”


42

Staff’s presumption is unreasonable, contrary to its admission that ETRACOM did not know of

the market design and implementation flaws, and must be rejected.

3. Staff’s Allegations that ETRACOM’s Behavior at New Melones Was

Intentionally Uneconomic, Unusual and Anomalous Are Incorrect and Not

Supported

84. Staff identifies aspects of ETRACOM’s behavior at New Melones in May 2011 as showing

its purported intent to place uneconomic offers in furtherance of a manipulative scheme. Some

of these allegations concerning ETRACOM’s trading activity are incorrect, while others are based

on supposition or speculation. For example, staff alleges that ETRACOM’s selective placement of

virtual supply offers on May 14-15 (the so-called “test period”) was both uneconomic—“[t]he

company suffered a net loss of $52”117—and intended to test its ability to thwart export

congestion perceived the week before.118 Staff then compares ETRACOM’s bidding for the

remainder of May to its virtual bids placed elsewhere in the CAISO in May 2011, stating

“ETRACOM’s virtual trading at New Melones was the only strategy that began mid-month and

encompassed all hours for an extended period.”119

a. ETRACOM’s Behavior at New Melones Was Incentivized By the Prices

Presented There and Was Not Intentionally Uneconomic

85. Staff’s contention that a $52 net loss accrued across two trading days would be sufficiently

unprofitable to deter ETRACOM from expanding a profit-seeking strategy potentially worth

thousands of dollars is not reasonable. To the contrary, this near-costless outcome suggested that

the “premium” needed for ETRACOM to sustain the hydroelectric trading strategy across a

broader set of hours over time would be relatively inexpensive. Staff cites a May 15

communication stating “were [sic] in a good shape in CA – check HPT, new strategies on VT in

CA” as evidence that ETRACOM was aware that these virtual trades were uneconomic, yet chose

to expand its strategy.120 Notably, on the 15th ETRACOM’s four cleared virtual offers made

about a $6 profit. If, in fact, this message reflected any recognition on the part of ETRACOM as

to the small premium it would pay given its anticipation of a highly profitable congestion-related

event, said awareness should have incentivized the placement of more such trades, not less.

86. Staff also alleges ETRACOM’s trading during the “test period” strategically confirmed its

ability to impact the day-ahead congestion price at New Melones by selectively choosing not to

bid in hour-ending (“HE”) 7 on May 14, then including HE 7 in its next day’s trades on May 15,


118 Id., pp. 9-10.

119 Id., p. 12.

120 Id., p. 20, and n. 106 referencing ETRACOM Instant Message, 5/15/2011 11:07:48 AM, Bates No.

ETR01499.


43

which was a Sunday.121 What is not mentioned is that the trades placed on May 14 included all (and only) off-peak hours, for which HE 7 would not have been of interest. The inclusion of

HE 7 in ETRACOM’s May 15 offers would have logically served as a next step to assess the costs

and benefits of expanding the strategy to other hours. The $26.53 day-ahead LMP on May 14

likewise suggested that the premium needed to extend the strategy to this additional hour would

be low. Importantly, however, the HE 7 virtual supply bid made a profit of $10.15/MWh122 and

made the day’s trading profitable overall. This trading was also consistent with ETRACOM’s

testimony cited in the Staff Report, which stated that is was good business practice to expand a

strategy “from a position of limited scope to the target scope.”123

87. Staff’s characterization as to the timing of ETRACOM’s choice to begin placing virtual

supply at New Melones is also problematic. The Staff Report observes that “[t]he strategy was

initiated only a few days after ETRACOM discovered that the profitability of its CRR positions

was being adversely impacted by export congestion.”124 However, this congestion created a

positive spread between the day-ahead and hour-ahead prices. During the period from May 8-

13, the average day-ahead LMP at New Melones was $31.48 when WAPA flowed power.125 By

comparison, the average day-ahead price for all of April was $10.57 ($6.58 in the off-peak) and

from May 1-7 was $15.23 ($15.53 in the off-peak).

88. The high day-ahead prices would incentivize a rational market participant to place virtual

supply in pursuit of stand-alone profits in hours when WAPA exported power from the CAISO.

Thus, irrespective of its CRR position, it was rational for ETRACOM to place virtual supply offers

in the hours when export congestion had appeared in prior days in the hopes of profiting on such

offers. The virtual supply offers placed in off-peak hours in May therefore reflected legitimate

profit-seeking behavior. Notably, and contrary to staff’s claim that this explanation was

“introduced after the fact by ETRACOM’s expert economist,”126 Rosenberg extensively discussed

this concept at his deposition.127

89. Staff disagrees with the observation that high day-ahead prices would have incentivized

ETRACOM to place virtual supply offers into the market, stating:

ETRACOM maintained throughout the investigation that it placed its virtual

supply offers to be profitable in a hydro event that would be reflected in future

121 Id., p. 24.


123 Staff Report, p 17, n. 93, citing Rosenberg Deposition at 348:9-11 et seq.

124 Id., p. 24.

125 The calculations in this paragraph are provided in the Excel workbook titled “LMP and Congestion

Prices.xlsx,” tab “LMP and Congestion Data.”


127 Rosenberg Deposition at 231:24-233:10; 234:1-20; 235:17-21; 340:7-18; and 341:6-20.


44

prices. It cannot at the same time argue that it placed the virtual supply offers in

response to current price signals.128

Staff’s position erroneously ignores ETRACOM’s testimony and misunderstands how a financial

trading firm perceives a virtual supply offer—as a swap.129 As I explained above, ETRACOM

would have viewed its zero or negatively-priced virtual supply offers placed at New Melones no

differently than the -$20/MWh offers it made at seventeen other interties from May 25-31: as

price-taking financial contracts that would be paid the day-ahead LMP and would pay the hour-

ahead LMP. Contrary to staff’s claims, higher day-ahead prices caused by the export congestion

that began at New Melones beginning on May 8 would have signaled to ETRACOM an increase

in the expected payout to the virtual supply swap, especially when observed relative to the many

hours of zero and negative day-ahead prices that arose the prior month. This profit potential was

described by ETRACOM in testimony as being relevant to reducing the “premium” that it would

ultimately pay for its broader hydroelectric strategy.130

90. Staff has not provided a single piece of documentary evidence to support its allegation

that ETRACOM ever linked the effect of its purportedly uneconomic virtual supply offers on

congestion prices at New Melones to its CRRs, much less that it intentionally attempted to

exploit that link. ETRACOM could not have known at the time that the export congestion was

enabled by a 1 MW physical flow by WAPA during off-peak hours beginning on May 8,131 nor

that a 1 MW virtual supply offer would be sufficient to reverse the direction of that

congestion.132 Because New Melones was a 384 MW import and 15 MW export intertie open to

convergence bidding, ETRACOM would not have known if the net export congestion at New

Melones was caused by an increase in physical exports, a decrease in physical imports, an

increase in virtual demand bids or decrease in virtual supply offers, nor would it have known the

volumes of the transactions that potentially could have been placed by market participants.

Without knowing of the tie’s fully-encumbered status, seeing persistent import and export

congestion at a location with a 384 MW transmission import capacity and 15 MW export limit,

and without having the benefit to see other convergence bids placed into the market until 90

days later, ETRACOM would logically conclude that its small virtual supply offers would

compete in a much larger market than actually existed at New Melones, such that its

hydroelectric trading strategy would win or lose based upon its stand-alone merits.


129 See supra, PP 20-21.

130 See Rosenberg Deposition, p. 62:17-24, 181:7-10, and 351:9-19.


132 This is an especially odd outcome given that the export capacity of New Melones was rated at 15 MW.

See DMM Referral Attachment 1, p. 2. As discussed previously in relation to Figure 3, the fact that a 1

MW physical export by WAPA created congestion on a 15 MW intertie is yet further evidence of the

dysfunction inherent to this market.


45

b. Staff Errs in Alleging that ETRACOM Sought to Purposely Lower Prices at

New Melones

91. Staff alleges as evidence of its uneconomic behavior that “ETRACOM was trying to clear

more MWs because clearing more MWs, and at the lowest price possible, served to lower LMP

further, thereby increasing the benefits to ETRACOM’s CRR position.”133 Staff is yet again

incorrect. After May 14-15, when ETRACOM offered up to 5 MW of virtual supply into the

market at $0/MWh, it began to separate its offers into segments, with each additional segment

offered into the market at a higher price. This is shown below in Table 7, which reports the

average daily price at which ETRACOM offered each of these segments.

Table 7

ETRACOM Made Price-Sensitive Virtual Supply Offers134

Source: "VT Bids May 2011.xlsx"

Notes:

[1] - [4]: ETRACOM’s New Melones bid prices at segments of the bid supply curve.


134 See Excel workbook titled “VT Bids May 2011_All Nodes.xlsx,” tab “Table 7.”

Date

Number of

Hours Bid Segment 1 Segment 2 Segment 3 Segment 4

[1] [2] [3] [4]

5/14/2011 8 $0.00

5/15/2011 9 $0.00

5/16/2011 24 $30.00 $20.00 $10.00 $0.00

5/17/2011 24 $29.01 $21.05 $10.94 $0.91

5/18/2011 24 $28.65 $24.52 $18.95 $15.36

5/19/2011 24 $29.06 $25.31 $18.75 $14.27

5/20/2011 24 $29.06 $25.31 $18.75 $14.27

5/21/2011 24 $29.05 $18.88 $17.06 $14.85

5/22/2011 24 $29.13 $28.59 $18.71 $10.92

5/23/2011 24 $29.07 $28.83 $28.57 $18.81

5/24/2011 24 $29.07 $28.83 $28.57 $18.81

5/25/2011 24 $29.07 $28.83 $28.57 $18.81

5/26/2011 24 $28.89 $28.74 $18.49 $17.94

5/27/2011 24 $29.38 $29.33 $29.28 $29.23

5/28/2011 24 $29.38 $29.33 $29.28 $29.23

5/29/2011 24 $28.56 $28.54 $28.52 $28.50

5/30/2011 24 $28.56 $28.54 $28.52

5/31/2011 24 $28.56 $28.54 $28.52 $28.50


46

92. Consistent with a presumption of guilt, staff alleges that ETRACOM’s virtual supply

offers were intended to set the market price at New Melones as low as possible so as to maximize

the price effect benefitting its CRRs. However, Table 7 shows that ETRACOM was only willing

to offer higher quantities of virtual supply if they were to clear at higher prices, particularly so

prior to May 25, the day when it began offering up to 5 MW of virtual supply at other interties at

the flat price of -$20/MWh. Later “flatness” in ETRACOM’s offers also would have been a

rational response to perceived competition from other market participants, signified by the

43.5% of hours from May 16-31 when it did not set the price.135 The timing of these “flat” offers

also comports with the end of the June 2011 CAISO CRR auction, which ETRACOM testified

had diverted its full attention away from the performance of its hydroelectric trading strategy

earlier in the month.136

93. If ETRACOM’s intent was to lower the day-ahead price at New Melones, it would have

bid all segments of its supply curve at -$30/MWh to maximize the price effect such bids might

have. Instead, consistent with the principle of an upward-sloping supply curve, ETRACOM

placed price-sensitive bids that would have raised prices to the detriment of the value of its

CRRs. This demonstrates that (a) ETRACOM did not perceive that these offers would impact its

CRRs, and (b) that its virtual supply offers were profit-seeking on a stand-alone basis, thus

serving a legitimate business purpose supportive of a presumption of transactional legitimacy.

c. A Review of Actual Trading Data Proves That ETRACOM’s Trading at

New Melones Was Neither Unusual nor Anomalous

94. Staff states “numerous characteristics of ETRACOM’s virtual trading strategy indicate

ETRACOM’s intent to manipulate. These include: the location (i.e., New Melones); timing (i.e.,

start date, test period hours, expansion to 24 hour trading and end date); and the distinctiveness

of the strategy compared to ETRACOM’s virtual trading at other locations.”137 Analysis of these

“characteristics” demonstrates that ETRACOM’s behavior was legitimate, not manipulative.

95. As discussed above, ETRACOM’s choice of location for its hydroelectric trading strategy

was based on market fundamentals, most notably the exceptional snowpack upriver of the New

Melones Reservoir at levels not observed since 1983. ETRACOM’s choice to place virtual supply

offers at that location over a period of days was justified by the historical price spreads that

would reward such a strategy and incentivized by high day-ahead prices that would increase the

profitability of (or reduce the premium paid for) the strategy. ETRACOM’s only error in placing

virtual supply there was its failure to recognize the dysfunction of the CAISO’s pricing

mechanism caused by pricing algorithm or software errors and the intertie’s fully-encumbered

status, none of which it would have known of in advance of or during the strategy.


136 Rosenberg Deposition at 137:11-138:10, 224:6-25 and 309:15-21.



47

i. The Timing and Structure of ETRACOM’s Trades at New Melones Was

Neither Unusual nor Anomalous

96. Staff’s allegations presume that ETRACOM understood a causal linkage (nexus) between

its virtual supply offers (trigger) and CRRs (target) at New Melones in advance of executing the

“test period” of the purported manipulative scheme. However, ETRACOM had recognized the

possibility of capitalizing on likely HASP congestion using virtual trades as early as April 25,138

and at or near New Melones as early as May 1.139 For staff to assert that ETRACOM was not

considering its hydroelectric trading strategy at New Melones and elsewhere until the May 8-13

export congestion appeared at New Melones is incorrect.

97. Staff nevertheless asserts that the timing and structure of ETRACOM’s trades during the

“test period” is a “strong indication” of its manipulative intent, particularly with respect to the

addition of an additional HE 7 virtual supply offer of May 15 that was not used on May 14:

The test period targeted precisely the eight hours that had experienced the export

congestion. The exclusion of just one of those hours (hour-ending 7) on May 14

was the control variable in the test that ETRACOM used to see the impact of its

trading strategy and gauge how successful it was at countering the export

congestion and lowering the day-ahead LMP. This is a strong indication that

ETRACOM intended its trades to counter the export congestion.140

As discussed above, the inclusion of HE-7 in ETRACOM’s May 15 offers would have logically

served as a next step to assess the costs and benefits of expanding the strategy to other hours,

especially given that the high day-ahead price on HE 7 on May 14 incentivized ETRACOM to

expand the strategy there next. Indeed, the HE 7 virtual supply offer on May 15 made a profit of

$10.15/MWh141 and the day’s trading was profitable overall. There was nothing anomalous about

this trade or others in the “test period.”

98. Contrary to staff’s allegations, ETRACOM regularly added or removed one or two hours

of virtual bids or offers from an existing trading strategy. As I show below in Table 8, there are

at least 80 instances where ETRACOM added or removed “test bids” from other strategies that

were placed before, during and after the Relevant Period. Staff’s portrayal of ETRACOM’s HE 7

bid on May 15 as anomalous is incorrect. Therefore, its allegation that this action is indicative of

ETRACOM’s manipulative intent is unsubstantiated and erroneous.

138 See ETRACOM Instant Message, 4/25/2011 10:41:54 AM – 10:44:07 AM, Bates No. ETR01450.

139 See ETRACOM Instant Message, 5/1/2011 2:20:51 PM – 2:21:26 PM, Bates No. ETR01459.




48

Table 8

Instances Where ETRACOM Added or Removed One or Two

Hours from Its Virtual Trading Strategy (April-June, 2011)142

Source: "VT Bids April 2011.xlsx," "VT Bids May 2011.xlsx," "VT Bids June 2011.xlsx."

Notes:

[1]: Denotes virtual supply (INC) or virtual demand bid (DEC) added/removed.

[2]: Date hour(s) removed/added.

[3]: Hour(s) of the day added/removed.

142 See Excel workbook titled “Add_Removal_VT_February - June 2011.xlsx,” tab “Table 8.”

Additions Removals

Node INC/DEC Date of Change Change Node INC/DEC Date of Change Change

[1] [2] [3] [1] [2] [3]

CAPTJACK_5_N505 INC 5/18/2011 HE 5 CAPTJACK_5_N505 INC 5/20/2011 HE 21 22

CAPTJACK_5_N505 INC 5/20/2011 HE 23 CAPTJACK_5_N505 INC 5/26/2011 HE 5

CAPTJACK_5_N512 INC 5/18/2011 HE 5 CAPTJACK_5_N512 INC 5/20/2011 HE 21 22

CAPTJACK_5_N512 INC 5/20/2011 HE 23 CAPTJACK_5_N512 INC 5/26/2011 HE 5

CRAGVIEW_1_GN001 INC 5/18/2011 HE 5 CRAGVIEW_1_GN001 INC 5/26/2011 HE 5

CRAGVIEW_1_GN001 INC 5/20/2011 HE 23 DLAP_PGAE APND DEC 4/23/2011 HE 23 24

DLAP_PGAE APND DEC 4/24/2011 HE 23 24 DLAP_PGAE APND DEC 5/22/2011 HE 3

DLAP_PGAE APND DEC 5/6/2011 HE 19 DLAP_SCE APND DEC 4/23/2011 HE 23 24

DLAP_PGAE APND DEC 6/17/2011 HE 3 DLAP_SCE APND DEC 5/22/2011 HE 3

DLAP_SCE APND DEC 4/24/2011 HE 23 24 MALIN_5_N101 INC 5/26/2011 HE 5

DLAP_SCE APND DEC 5/6/2011 HE 19 MEADS_2_N101 INC 5/20/2011 HE 21 22

DLAP_SCE APND DEC 6/17/2011 HE 3 MEADS_2_N101 INC 5/26/2011 HE 5

MALIN_5_N101 INC 5/18/2011 HE 5 MERCHANT_2_N101 INC 5/20/2011 HE 21 22

MALIN_5_N101 INC 5/20/2011 HE 23 MERCHANT_2_N101 INC 5/26/2011 HE 5

MEADS_2_N101 INC 5/18/2011 HE 5 MOENKOPI_5_N101 INC 5/20/2011 HE 21 22

MEADS_2_N101 INC 5/20/2011 HE 23 MOENKOPI_5_N101 INC 5/26/2011 HE 5

MERCHANT_2_N101 INC 5/18/2011 HE 5 NGILA1_5_N001 INC 5/20/2011 HE 21 22

MERCHANT_2_N101 INC 5/20/2011 HE 23 NGILA1_5_N001 INC 5/26/2011 HE 5

MOENKOPI_5_N101 INC 5/18/2011 HE 5 PALOVRDE_ASR APND INC 5/20/2011 HE 21 22

MOENKOPI_5_N101 INC 5/20/2011 HE 23 PALOVRDE_ASR APND INC 5/26/2011 HE 5

NGILA1_5_N001 INC 5/18/2011 HE 5 PARKER_2_N101 INC 5/20/2011 HE 21 22

NGILA1_5_N001 INC 5/20/2011 HE 23 PARKER_2_N101 INC 5/26/2011 HE 5

PALOVRDE_ASR APND INC 5/18/2011 HE 5 ROA 230_2_N101 INC 5/20/2011 HE 21 22

PALOVRDE_ASR APND INC 5/20/2011 HE 23 ROA 230_2_N101 INC 5/26/2011 HE 5

PARKER_2_N101 INC 5/18/2011 HE 5 SMD5_ASR APND INC 5/20/2011 HE 21 22

PARKER_2_N101 INC 5/20/2011 HE 23 SMD5_ASR APND INC 5/26/2011 HE 5

PARKER_2_N101 INC 6/13/2011 HE 5 SMDH_ASR APND INC 5/20/2011 HE 21 22

ROA 230_2_N101 INC 5/18/2011 HE 5 SMDH_ASR APND INC 5/26/2011 HE 5

ROA 230_2_N101 INC 5/20/2011 HE 23 SUMMIT_ASR APND INC 5/20/2011 HE 21 22

SMD5_ASR APND INC 5/18/2011 HE 5 SUMMIT_ASR APND INC 5/26/2011 HE 5

SMD5_ASR APND INC 5/20/2011 HE 23 SYLMARDC_2_N501 INC 5/26/2011 HE 5

SMDH_ASR APND INC 5/18/2011 HE 5 SYLMARLA_2_N501 INC 5/20/2011 HE 21 22

SMDH_ASR APND INC 5/20/2011 HE 23 SYLMARLA_2_N501 INC 5/26/2011 HE 5

SUMMIT_ASR APND INC 5/18/2011 HE 5 TJI 230_2_N101 INC 5/20/2011 HE 21 22

SUMMIT_ASR APND INC 5/20/2011 HE 23 TJI 230_2_N101 INC 5/26/2011 HE 5

SYLMARDC_2_N501 INC 5/18/2011 HE 5 VICTORVL_5_N101 INC 5/20/2011 HE 21 22

SYLMARDC_2_N501 INC 5/20/2011 HE 23 VICTORVL_5_N101 INC 5/26/2011 HE 5

SYLMARLA_2_N501 INC 5/18/2011 HE 5

SYLMARLA_2_N501 INC 5/20/2011 HE 23

TJI 230_2_N101 INC 5/18/2011 HE 5

TJI 230_2_N101 INC 5/20/2011 HE 23

VICTORVL_5_N101 INC 5/18/2011 HE 5

VICTORVL_5_N101 INC 5/20/2011 HE 23


49

99. Staff states: “[t]he expansion of ETRACOM’s strategy to 24 hours a day on May 16 (and

thereafter) demonstrates that ETRACOM viewed its strategy as successful in the test period and

worthy of expansion, even though the strategy suffered a net loss.”143 As discussed previously, a

$52 “net loss” —including a small net gain on May 15—was a small premium to pay for a swap-

based strategy that could be worth thousands if successful at levels no different than those seen

in 2009, 2010 or the summer of 2011. Further, as was shown above in Table 3, expansion of the

strategy into the peak hours would be expected to provide the greatest opportunity for profits

from the strategy. Staff’s allegations as to the anomalousness of ETRACOM’s expansion of its

trading strategy instead to peak hours ignore the rationality and legitimacy of its stand-alone,

profit-seeking motivations.

100. Staff’s final point concerning the timing of ETRACOM’s trading alleges its decision to end

the hydroelectric trading strategy at New Melones on May 31, 2011 was conspicuous because

“ETRACOM ended its trading strategy on the same day that the CRR positions that benefited

from the strategy substantially decreased.”144 Specifically, although ETRACOM held directionally

equivalent CRR positions sourcing to New Melones in June, staff asserts these smaller 7.24 MW

on-peak and 7.79 MW off-peak CRR positions reduced the manipulation target sufficiently such

that “the incentive to continue the manipulation was greatly decreased.”145 ETRACOM’s CRRs

held at New Melones in June 2011 were indeed smaller than the 34.668 MW on-peak and 25.326

MW off-peak positions held in May, but the financial leverage provided by these positions

remained such that intentionally uneconomic 1-5.03 MW virtual supply offers placed at the

levels ETRACOM allegedly used in May would have continued to have been profitable in June.

101. Notwithstanding this continuing opportunity, ETRACOM chose to end its hydroelectric

trading strategy at the end of May. Staff finds this to be further evidence of manipulative intent:

ETRACOM ceased virtual trading at New Melones abruptly on May 31.[]

ETRACOM offered no explanation for this. Moreover, […] abandoning this

strategy after two weeks is inconsistent with ETRACOM’s claim that its trades

were designed to capture congestion caused by an anticipated hydro event that,

by that time, had not yet materialized.146

Staff’s observation underscores the danger of interpreting trading behavior on an ex post basis

without affording the trader a concomitant presumption of transactional legitimacy. Because it

chose to end the strategy on May 31st, staff alleges that ETRACOM never intended to profit from

the hydroelectric trading strategy and thus is guilty of manipulation. However, had ETRACOM

instead chosen to continue the strategy, staff would then assert (assuming continued losses on the






50

virtual supply offers) that ETRACOM was continuing the manipulation into June given the

continuing leverage in its CRR positions. Such “no win” scenarios are inevitably determinative

of ETRACOM’s guilt.

ii. ETRACOM’s Trades at New Melones Were Neither Unusual nor

Anomalous Compared to Its Trades at Other Locations

102. Staff asserts “ETRACOM’s virtual trading at New Melones in May 2011 was anomalous

compared to its trading at all other locations.”147 Variance in ETRACOM’s approach to trading at

New Melones relative to other trading points would be expected because the selected

hydroelectric trading strategy was a new and unique opportunity presented by the conditions at

the time. Specifically, ETRACOM stated that New Melones was specifically chosen due to: (a)

the unusual snowpack conditions in the Sierra Nevada mountains, which was “150 percent and at

some locations over 180 percent of normal”;148 (b) that the New Melones constraint was the

“dominant one” in the region that would experience congestion when the snowpack melted;149

and (c) that the expected congestion had yet to occur there.150 All of these stated reasons would

rationally support implementing the hydroelectric strategy at New Melones, not elsewhere.

Nonetheless, ETRACOM offered and cleared virtual supply at seventeen other interties from

May 25-31 by bidding at -$20/MWh, a fact which staff ignores.

103. Staff states that “ETRACOM’s virtual trading at New Melones was the only strategy that

began mid-month and encompassed all hours for an extended period.”151 Staff is incorrect.

Analysis of ETRACOM’s convergence bidding from February 2011 through July 2011 reveals 76

instances where ETRACOM placed convergence bids in all hours of a day over a stretch of three

or more days, thirteen of which were for periods equal to or longer than New Melones. These

are shown below in Table 9. The length of time ETRACOM submitted continuous bids or offers

tended to grow longer as its experience increased and it grew more comfortable with its bidding

strategies. Many of its strategies started in the middle of the month and/or finished at or near

month’s-end. This pattern shows that staff is wrong in asserting that ETRACOM’s pattern of

convergence bidding at New Melones differed markedly from its behavior elsewhere. To the

contrary, ETRACOM’s sixteen day, all-day run of virtual supply placed at New Melones was

consistent with its pattern of convergence biddings placed over time throughout the CAISO.


148 Rosenberg Deposition, p. 307:16-308:3.

149 Id., p. 302:20-303:14.

150 Id.



51

Table 9

ETRACOM’s Virtual Trades Placed in All Hours of Three or More

Sequential Days (February 1, 2011 through July 31, 2011)152

Sources and Notes:

[1] - [2], [4] - [5]: Start and end dates of longest consecutive streak of 24 hour virtual bids or offers from processing

of CAISO OASIS convergence bidding data.

[3], [6]: End Date - Start Date + 1.

152 See Excel workbook titled “Virtual_Duration_Analysis.xlsx,” tab “Table 9.” Source data derived from

CAISO OASIS convergence bidding data.

INC DEC

Node Start Date End Date

Number of

Consecutive Days Node Start Date End Date

Number of

Consecutive Days

[1] [2] [3] [4] [5] [6]

102802 6/16/2011 7/6/2011 21 256006 6/14/2011 7/15/2011 32

124965 6/16/2011 7/6/2011 21 471137 6/17/2011 7/6/2011 20

459685 6/18/2011 7/6/2011 19 235513 6/18/2011 7/6/2011 19

149152 5/16/2011 5/31/2011 16 147188 7/17/2011 8/1/2011 16

473965 7/17/2011 8/1/2011 16 200099 7/17/2011 8/1/2011 16

774962 7/17/2011 8/1/2011 16 209028 7/17/2011 8/1/2011 16

781179 7/17/2011 8/1/2011 16 585380 7/17/2011 8/1/2011 16

211707 7/18/2011 8/1/2011 15 104860 7/18/2011 8/1/2011 15

438683 7/18/2011 8/1/2011 15 129183 7/18/2011 8/1/2011 15

566980 7/18/2011 8/1/2011 15 168094 7/18/2011 8/1/2011 15

713637 7/18/2011 8/1/2011 15 179460 7/18/2011 8/1/2011 15

230684 6/23/2011 7/6/2011 14 188962 7/18/2011 8/1/2011 15

109758 7/20/2011 8/1/2011 13 328813 7/18/2011 8/1/2011 15

268663 7/20/2011 8/1/2011 13 610614 7/18/2011 8/1/2011 15

809004 6/17/2011 6/29/2011 13 765910 7/18/2011 8/1/2011 15

400150 6/25/2011 7/6/2011 12 777894 7/18/2011 8/1/2011 15

186676 4/2/2011 4/12/2011 11 883706 7/18/2011 8/1/2011 15

105329 3/23/2011 3/31/2011 9 900424 7/18/2011 8/1/2011 15

111290 3/23/2011 3/31/2011 9 801655 6/23/2011 7/6/2011 14

334953 3/23/2011 3/31/2011 9 135026 7/20/2011 8/1/2011 13

478434 6/17/2011 6/25/2011 9 605483 7/20/2011 8/1/2011 13

166669 4/19/2011 4/26/2011 8 324560 6/25/2011 7/6/2011 12

171176 6/11/2011 6/18/2011 8 668372 6/27/2011 7/6/2011 10

682455 4/19/2011 4/26/2011 8 214336 6/11/2011 6/18/2011 8

160296 6/30/2011 7/6/2011 7 125967 6/30/2011 7/6/2011 7

210908 6/30/2011 7/6/2011 7 179287 6/30/2011 7/6/2011 7

761029 6/30/2011 7/6/2011 7 478434 6/30/2011 7/6/2011 7

176545 4/4/2011 4/9/2011 6 669526 6/30/2011 7/6/2011 7

135933 3/8/2011 3/12/2011 5 809004 6/30/2011 7/6/2011 7

139450 3/8/2011 3/12/2011 5 149152 6/1/2011 6/4/2011 4

154666 3/8/2011 3/12/2011 5 171176 7/26/2011 7/29/2011 4

170763 3/8/2011 3/12/2011 5 176260 7/4/2011 7/6/2011 3

184679 3/8/2011 3/12/2011 5 328620 3/3/2011 3/5/2011 3

213012 3/8/2011 3/12/2011 5

316542 3/8/2011 3/12/2011 5

328620 4/22/2011 4/26/2011 5

461023 3/8/2011 3/12/2011 5

801655 3/8/2011 3/12/2011 5

888461 3/8/2011 3/12/2011 5

165948 7/26/2011 7/29/2011 4

168149 7/26/2011 7/29/2011 4

669526 7/26/2011 7/29/2011 4

804513 7/4/2011 7/6/2011 3


52

104. Staff is also incorrect in its assertion that ETRACOM’s bidding during the so-called “test

period” was unusual. Examination of the long-term trades shown in Table 9 reveals 23 cases in

which ETRACOM placed bids or offers in some (but not all) hours of the three days immediately

prior to those trades. These are shown below in Table 10.

Table 10

Examples of ETRACOM “Test Periods” before Multiday,

All-Day Trades153

Sources and Notes:

[1] - [2]: Start and end dates of longest consecutive streak of 24 hour virtual bids from processing

of CAISO OASIS convergence bidding data.

[3]: [2] - [1] + 1.

[4]: Sum of hours when virtuals were placed at a given node in the previous 3 days.

[5]: [4] / 72.

153 See Excel workbook titled “Virtual_Duration_Analysis.xlsx,” tab “Table 10.” Source data derived from

CAISO OASIS convergence bidding data.

Node Bid Type Start Date End Date

Number of

Consecutive

Days

Hours of VB in

Prior Three

Days

Percentage of

Hours

[1] [2] [3] [4] [5]

135933 INC 3/8/2011 3/12/2011 5 25 35%

139450 INC 3/8/2011 3/12/2011 5 25 35%

149152 INC 5/16/2011 5/31/2011 16 17 24%

154666 INC 3/8/2011 3/12/2011 5 25 35%

170763 INC 3/8/2011 3/12/2011 5 25 35%

171176 INC 6/11/2011 6/18/2011 8 8 11%

176545 INC 4/4/2011 4/9/2011 6 45 63%

184679 INC 3/8/2011 3/12/2011 5 25 35%

213012 INC 3/8/2011 3/12/2011 5 25 35%

316542 INC 3/8/2011 3/12/2011 5 25 35%

328620 INC 4/22/2011 4/26/2011 5 70 97%

461023 INC 3/8/2011 3/12/2011 5 25 35%

473965 INC 7/17/2011 8/1/2011 16 51 71%

774962 INC 7/17/2011 8/1/2011 16 51 71%

781179 INC 7/17/2011 8/1/2011 16 51 71%

801655 INC 3/8/2011 3/12/2011 5 25 35%

888461 INC 3/8/2011 3/12/2011 5 25 35%

147188 DEC 7/17/2011 8/1/2011 16 51 71%

200099 DEC 7/17/2011 8/1/2011 16 51 71%

209028 DEC 7/17/2011 8/1/2011 16 51 71%

214336 DEC 6/11/2011 6/18/2011 8 8 11%

585380 DEC 7/17/2011 8/1/2011 16 51 71%

668372 DEC 6/27/2011 7/6/2011 10 49 68%


53

The characteristics of ETRACOM’s trading at New Melones were therefore a common feature of

ETRACOM’s efforts to evaluate the stand-alone profitability of its trading strategies. This

contradicts staff’s characterization of such trades at New Melones as anomalous.

105. In response to Tables 9 and 10, staff somewhat inexplicably asserts:

ETRACOM points to subsequent occasions, primarily after May 2011, when it

placed bids mid-month and for all hours for sequential days.[] It also points to

other strategies that had a test period.[] However, ETRACOM admittedly relies on

trading data from after May 2011. Staff’s observation is that the New Melones

trades were anomalous at the time the manipulation occurred. ETRACOM’s later

trading behavior does not refute this point.154 [Emphasis added].

Staff is incorrect, as several of these patterns occurred before ETRACOM’s New Melones trades.

But even if some aspects were first implemented at New Melones, this is hardly evidence of an

anomaly or manipulative intent. By staff’s logic, analysis of the pattern of ETRACOM’s

purportedly “anomalous” behavior over time is irrelevant if it looks at behavior after the alleged

event. This assertion is preposterous, especially because these other examples occur within two

months of the New Melones trades, and the virtual trading market in the CAISO had existed for

a little more than three months at that point. Also, by definition the behavior associated with a

new trading strategy would not be expected to resemble other strategies executed previously or

contemporaneously at the same or other locations. Under staff’s logic, any new trading strategy

that incidentally fits within the manipulation framework would be foreclosed from pointing to

subsequent use of the same strategy as proof of its legitimacy—a standard identical to that which

staff seeks to apply here. This position serves no purpose but to clarify that staff has presumed

ETRACOM’s guilt in this matter and asks that the Commission do the same.

106. Staff alleges that ETRACOM’s choice to end the hydroelectric trading strategy at New

Melones on May 31, 2011 demonstrates that the strategy was never intended to be profitable on a

stand-alone basis but rather intended to serve as a trigger for the manipulation of ETRACOM’s

CRR positions. However, staff has missed the fact that ETRACOM continued this same strategy

into June—just not directly at or in the size used at New Melones. Seeing that its hydroelectric

trading strategy was not successful at New Melones in May, in June ETRACOM moved the

strategy to a set of six nodes upriver of the New Melones Reservoir around the Donnells Dam on

the Stanislaus River, including two locations that staff identified as potentially being more

profitable than New Melones.155 As a collective strategy, the virtual supply offers placed across

these six nodes mimicked the virtual supply offers that ETRACOM had made at New Melones

the prior month, once again positioning it to benefit from the congestion in the HASP that it still


155 ERTACOM placed trades as BEARDSLY_7_B1, CURTISS_1_N101, CURTIS1_7_N001, POD COLVIL

7_PL1X2, SPRNGGP_7_B1 and DONNELLS_7_B1, the latter two cited by staff as potentially being

more profitable than New Melones. See Hourly_Charts_Hydronotes.pdf.


54

expected. This is shown below in Figure 7, which recreates the top panel of the graph shown on

p. 12 of the Staff Report for the May strategy (top panel) and June strategy (bottom panel).

Figure 7

ETRACOM’s Hydroelectric Trading Strategy: Cleared

Virtual Supply MW in May and June 2011156

Source: “FERC Request 20111216 Q7.csv”

107. Although the June strategy was started earlier in the month and cleared a more consistent

pattern of MW than the May strategy, the strategies share the common element of “test bids”

that were made before the strategies were expanded to all hours of the day. Similar to the May

strategy at New Melones, the June strategy began with virtual supply offers generally being made

156 See Excel workbook titled “Upstream and New Melones Position.xlsx,” tab “Figure 7.”

ETRACOM's Hydroelectric Trading Strategy: May 2011

ETRACOM's Hydroelectric Trading Strategy: June 2011

18

16

14

12

10

8

6

4

2

0

MW

6

5

4

3

2

1

0

MW


55

at $0/MWh, but later dropped to -$20.00/MWh or less as the month progressed.157 These trades,

which served as part of a spread, were sometimes profitable,158 but lost money overall.159 Like

the trades placed at the end of May at other locations, these trades confirm ETRACOM’s effort to

take the risk of losses in an effort to profit from its hydroelectric trading strategy. Such losses do

not negate that the strategy was intended to serve a stand-alone, legitimate business purpose.

Staff’s attempt to cast ETRACOM’s trading strategy at New Melones in May as anomalous from

these other trades in an attempt to bolster its allegations of manipulation does not hold weight.

iii. ETRACOM’s Trades at New Melones in June 2011 Demonstrated Its

Lack of Knowledge of a Manipulation Nexus

108. Notwithstanding the “opportunity” afforded by the continued financial leverage provided

by its June CRR positions which source from New Melones, ETRACOM reversed its strategy and

began bidding virtual demand in June at positive prices,160 thus actively placing bids that could

only serve to reduce the value of its CRR positions. Staff contends that “ETRACOM’s June

virtual demand trading does not validate its May virtual supply trading.”161 “Validation” aside, if

ETRACOM had truly understood the link between its virtual trading and CRRs—meaning that it

knew that the market operated as described in the lower portion of Figure 3, above—it would

not have placed positively-priced virtual demand bids that could only serve to negatively impact

the value of its CRRs. This economic evidence confirms that ETRACOM never understood how

the pricing mechanism at New Melones worked (or failed to work), and thus lacked the intent to

manipulate that mechanism for lack of appreciating the linkage between its purportedly

uneconomic virtual supply offers (triggers) and CRRs (targets).

d. ETRACOM’s Virtual Bidding Behavior at New Melones Was Consistent

with the Expansion of its Portfolio as a New Participant in the CAISO

109. Virtual bidding in the CAISO began on February 1, 2011, 102 days before staff’s Relevant

Period began. ETRACOM’s entry into this new market was cautious and measured, as

demonstrated in Figure 8 below. This graph presents a time series of the profitability of

157 See “VT Bids June 2011.xlsx”

158 Rosenberg remarked in a June 22, 2011 instant message after learning of a $610 gain “ok not bad after

losses on melones[.]” See ETRACOM Instant Message, 6/22/2011 11:46:56 AM – 11:47:15 AM, Bates

No. ETR01582.

159 The total loss on these nodes for June was -$2,436.87. See DR 7.

160 Staff Report, p. 13. ETRACOM submitted virtual demand bids from June 1-7. None cleared until June

7, when ETRACOM cleared 1 MW in seven hours (HE 1-6, and 24), resulting in losses of $53.88. In

addition, it appears that ETRACOM’s demand bids set the clearing price in HE 8-22 despite these bids

not clearing. See Excel workbook titled “LMP and Congestion Prices,” tab “Virtual Demand

Summary.” This is yet another example of the software errors that were present at the New Melones

intertie.



56

ETRACOM’s convergence bidding portfolio over the six-month period from February 1, 2011

through July 31, 2011. The portfolio is broken apart by the profitability of the virtuals over time,

with the red line representing accrued losses, the green line representing accrued profits and the

blue line representing net profits. Figure 8 shows that ETRACOM traded relatively few virtuals

in early months, gradually increasing its activity through April into May, with a substantial

uptick in volume beginning in mid-June. To assert that ETRACOM, or any trader in this nascent

market, had sufficient experience to evaluate all contingencies surrounding these trading

instruments before mid-May 2011 is questionable, at best. This is especially true at New

Melones, where the dysfunction created by software errors and other encumbrance-driven flaws

served to thwart effective price discovery.

Figure 8

The Profitability of ETRACOM’s Convergence Bidding Portfolio

at All Nodes (February 1, 2011 to July 31, 2011)162

Sources: “FERC Request 20111216 Q7.csv”; Velocity Suite, ABB Inc.

110. Figure 8 further shows that ETRACOM’s portfolio of virtuals produced modest net profits

and losses until mid-June, when its virtual activity and associated profits increased. Significantly,

while a slight drop in the portfolio’s net profitability is observable from mid-May to the end of

that month, those losses were minor if viewed from the perspective of the gains and losses of

ETRACOM’s broader virtual portfolio.

162 See Excel workbook titled “Virtual Profits.xlsx,” tab “Figure 8.”


57

e. The Potential Profitability of Other Locations Does not Confirm

ETRACOM's Manipulative Intent

111. Staff seeks to discredit the rationality of ETRACOM’s hydroelectric trading strategy at

New Melones by arguing that it could have pursued the same strategy at other similar locations,

and potentially stood to earn more profit by doing so:

Staff finds that ETRACOM’s virtual strategy would have been potentially more

profitable at other similar locations where the day-ahead LMP was typically

positive. Consequently, ETRACOM likely could have cleared positive (as opposed

to negative) supply offers. That would have lessened its losses on non-event days

because the spread between the day-ahead LMP and HASP would have been

smaller. However, at those locations ETRACOM did not hold CRR positions that

were unexpectedly declining in profitability.163

Enforcement’s arguments are entirely speculative in referencing the many alternative trading

options that ETRACOM may have rationally dismissed or never known of, yet are now theorized

in hindsight. Although the universe of such “false negatives” is undoubtedly large, it ultimately

is of little relevance to whether the hydroelectric strategy that ETRACOM executed at New

Melones was reasonable and intended to be profitable on a stand-alone basis. Staff’s hindsight is

not evidence of the opportunity costs ERTACOM recognized on an ex ante basis.

112. Contrary to staff’s assertions, ETRACOM did place virtual supply offers at other interties

where it held CRRs on May 25-31 at -$20/MWh, all of which cleared as price-taking offers. This

included three locations that Staff references as being “potentially more profitable” than New

Melones.164 As shown in Table 2, above, ETRACOM attempted to capture HASP congestion at

interties other than New Melones, and generally did so successfully because the markets at those

locations functioned properly. Staff’s comparison of ETRACOM’s purported willingness to incur

losses at New Melones relative to the profits it made on other interties ignores that the

dysfunction at New Melones incentivized ETRACOM’s negatively-priced virtual supply offers,

was responsible for the failure of its offers to be price-taking, and was the cause of the losses it

incurred on those trades. Staff’s allegations reverse cause and effect, thus ignoring evidence

supportive of a presumption of transactional legitimacy.


164 Id., pp. 23, 31; Hourly_Charts_Hydronotes.pdf. The three locations are CAPTJACK_5_N505,

CAPTJACK_5_N512, and MALIN_5_N101.


58

B. THE NEXUS: ETRACOM REASONABLY COULD NOT HAVE KNOWN THE STRENGTH OF

THE NEXUS LINKING ITS VIRTUALS TO THE VALUE OF ITS CRRS

113. The Staff Report alleges that “ETRACOM understood and intended its virtual trading to

impact its CRR positions”165 and that “it had to know that the negative LMPs at New Melones

benefited the profitability of its CRR positions: the relationship between the day-ahead price

(including congestion) and the profitability of CRR positions is fundamental to the product’s

value, and Rosenberg understood this concept.”166 Rosenberg’s background aside, staff’s

observation sidesteps the real question: Did ETRACOM understand the nexus between the losses on its virtual supply offers and the gains in its CRR portfolio? To answer this question with a

“yes” requires that ETRACOM knew of, understood and intentionally exploited the many flaws

presented by this market.

114. ETRACOM had every reason in May 2011 to believe that the New Melones intertie was a

competitive market. The CAISO position limit set for virtual supply offers at New Melones was

19.2 MW,167 significantly higher than ETRACOM’s 1 MW to 5.03 MW offers in May. In placing

such small virtual supply volumes on a 384 MW intertie, especially one that was routinely

congested, a market participant would not reasonably contemplate an ability to set the market

price or affect the flow of congestion under normal market conditions, much less reliably do

either. Indeed, had it wished to increase the likelihood that it would impact congestion,

ETRACOM would have offered virtual supply at the 19.2 MW position limit, instead of less than

a third of that limit. Lack of market participation may have been hindered further by

asymmetrically-small position limits on virtual demand bids (1 MW), the clearing of which

ETRACOM could not have learned of until the virtual data was posted some 90 days later.

ETRACOM would have been irrational to assume that the market was so dysfunctional that a

1 MW physical export bid by WAPA was sufficient to have reversed the flow of congestion on

the intertie, thus leading to increased day-ahead LMPs and incentivizing the very offers that staff

now claims to be manipulative.

115. Further, it is not reasonable to conclude that ETRACOM had sufficient data to accurately

assess whether its trading was affecting the LMP at New Melones—and even more critically, the

congestion component. As was discussed previously, ETRACOM had placed positively-priced

virtual supply offers at New Melones on March 11-15, 2011, days when $0 day-ahead LMPs

appeared due to errors in the CAISO’s pricing model.168 It then regularly observed negative

prices between April 14 and May 13 which cleared at prices as low as -$26.80/MWh without its

participation.169 Once its hydroelectric trading strategy had fully commenced, ETRACOM’s


166 Id., p. 35.

167 Id., p. 6.

168 See Excel workbook “Zero-Neg LMP Bid Analysis.xlsx,” tab “March 2011 Summary.”

169 See Excel workbook titled “Zero-Neg LMP Bid Analysis.xlsx,” tab “Neg Bid Summary.”


59

virtual supply offers set the day-ahead LMP in 56.5% of all hours between May 16-31, but other

bids set the LMP in the other 43.5% of those hours.170 ETRACOM had no way of knowing at the

time whether these other prices were set by supply or demand or by virtual or physical bids or

offers. Because these LMPs always cleared at negative values, and the congestion component of

those prices showed the logical and expected result of import congestion over the intertie given

hydroelectric conditions and its prior experience, ETRACOM could not reasonably have known

that it was responsible for setting either the day-ahead LMP at New Melones or its congestion

component. This is especially true given the dysfunction caused by market design flaws and

software errors and ETRACOM’s limited experience trading virtuals in the CAISO generally.

116. Staff admits that ETRACOM did not understand the cause of congestion at New Melones

and that the tie was fully encumbered,171 meaning that it could not have understood the ability of

virtual bids to create, much less totally reverse congestion on a line with a 399 MW span

between its import (384 MW) and export (15 MW) limits. Staff also admits the existence of the

software error,172 and further recognizes that its anomalies would explain why ETRACOM was

incentivized to make its virtual supply offers at zero or negative prices. There is therefore no

basis for staff to assert that ETRACOM would have known in advance of its trading on May 14

that the market at New Melones was anything but a normal, competitive market, or that the zero

or negative day-ahead prices clearing there were set by anything other than a robust competitive

process with active participation of both physical and virtual players. As such, staff cannot

reasonably allege that ETRACOM would have perceived its hydroelectric trading strategy as

comprising anything other than competitive, price-taking offers, the negative price of which was

fully incentivized by a flawed, broken and opaque pricing mechanism. ETRACOM would not

have understood the strength of the alleged nexus of the manipulation in advance of beginning

its hydroelectric trading strategy, a fact that seriously hampers a fundamental premise of staff’s

argument of manipulative intent.

117. This serious deficiency explains why staff originally made the hedging statement in its

Preliminary Findings that even if ETRACOM did not know of the nexus in advance of executing

its strategy, it “would have quickly seen the dramatic increase in the profitability of Etracom’s

CRR positions at New Melones and realized it was Etracom’s virtual trading behavior that was

causing that dramatic increase.”173 However, this exposes the slippery slope fallacy upon which

staff’s allegations rely: if ETRACOM did not understand the nexus at the outset of its strategy,

why then did it trade at all beginning on May 14, 2011? The answer is clear: because it was a legitimate trading strategy that was designed from the outset to be profitable on a stand-alone

basis. Once this inescapable conclusion is recognized, staff’s claims collapse completely.


171 Staff Report, p. 17, 36.




60

118. Staff’s hedge recognizes that for any portion of its manipulation claim to survive, it must

overcome the (now validated) presumption of transactional legitimacy to prove that ETRACOM

subsequently determined the strategy to be not working, but then recognized the nexus between

its virtual losses and CRR gains and thus chose to intentionally place virtual supply offers that it

knew would be uneconomic on a stand-alone basis, to the greater benefit of its CRRs. The prior

analyses and examples presented herein prove that this is not so, as is supported by all of the

documentary evidence available and the economic analyses ignored by staff. Contrary to staff’s

allegations, ETRACOM’s lack of knowledge concerning the nexus was shown conclusively by:

Market dysfunction brought about by the fully-encumbered status of the tie and software

errors, notice of which was not made known by the CAISO and would not have been

foretold by other manipulation cases at the time of the behavior;

ETRACOM’s price-sensitive offers consistent with an upward-sloping supply curve;

The fact that ETRACOM never would have considered stopping the strategy on May 20 if

it was tracking its CRR gains relative to its virtual losses; and

June trades at New Melones that were likely to have injured ETRACOM’s CRRs.

Given the lack of any documentary evidence otherwise establishing ETRACOM’s recognition of

the nexus, all of these factors support a presumption of transactional legitimacy that staff should

have afforded ETRACOM in this case.

C. THE TARGET: ETRACOM’S PROFITS IN ITS CRR PORTFOLIO IN LATE MAY 2011 WERE

SIMILAR TO THOSE EXPERIENCED IN EARLY APRIL

119. As discussed above, ETRACOM neither understood nor purposely exploited the alleged

nexus between its hydroelectric trading strategy and CRR positions. Documentary evidence and

actual CRR profits (shown above in Table 5) demonstrate that it did not monitor the profits of its

virtual supply offers at New Melons relative to the profitability of its CRR portfolio. However,

because staff alleges that ETRACOM would have observed the “dramatic increase” in the

profitability of its CRR positions and connected that increase to its virtual trading, it is useful to

examine the context surrounding how ETRACOM would have perceived the increased value of

its CRR positions before and during the Relevant Period.

120. To provide this context, I checked the profitability of ETRACOM’s CRR portfolio during

staff’s Relevant Period to determine if the large gains experienced in May 2011 were anomalous.

This inspection revealed that ETRACOM’s CRRs also made large gains in April 2011, in part due

to gains made from CRRs sourcing at New Melones. Because ETRACOM would have viewed the

gains made on its CRR portfolio in May relative to those made in April, it is useful to compare its

day-by-day gains in the two months. This is shown below in Figure 9.


61

Figure 9

The Profitability of ETRACOM’s CRR Portfolio

(April 2011 vs. May 2011)174

Sources: “FERC Request 20111216 Q7.csv”; Velocity Suite, ABB Inc.

121. Figure 9 shows the relative cumulative profits of ETRACOM’s CRR portfolio in April

(light blue line) and May (dark blue line). The light blue line shows ETRACOM made $551,367

in revenues from its CRR portfolio between April 2, 2011 through April 8, 2011,175 a gain that

was completely unrelated to its virtual trading. New Melones was the second-largest contributor

to this total. Staff asserts that “Rosenberg would have quickly seen the dramatic increase in the

profitability of ETRACOM’s CRR positions at New Melones [in May] and, because of the obvious

relationship between day-ahead price and CRR profitability, realized it was ETRACOM’s virtual

trading behavior that was causing that dramatic increase.”176 In fact, ETRACOM may have

rationally perceived the rapid gains in its CRRs in May 2011 as altogether unsurprising given its

experience in April and not associated such gains with losses in its convergence bids.

122. As the dark blue line of Figure 9 demonstrates, ETRACOM’s CRRs at New Melones were

profitable from May 1-13, before it began its hydroelectric trading strategy. Contrary to the

174 See Excel workbook titled “CRR and VB Profit.xlsx,” tab “Figure 9.”

175 See Excel workbook titled “CRR and VB Profits.xlsx,” tab “Profits (CRR).”



62

“dramatic increase in the profitability of ETRACOM’s CRR positions” touted by staff, the effect

of the alleged behavior was only to change the rate at which ETRACOM’s revenues escalated,

shown by the slightly increased slope of the line after May 13. Because its CRRs were already

profitable, it is reasonable that ETRACOM did not perceive this change in slope (if it perceived it

at all) as a causal effect of its virtual trading strategy, especially when it experienced even larger

gains over a shorter period during the prior month where it did not place convergence bids.

VI. Staff’s and the DMM’s Allegations Concerning ETRACOM’s Intent

Suffer from Numerous Flaws and Rely on No-Win Logic

123. The Staff Report states staff’s belief that the sole purpose of ETRACOM’s hydroelectric

trading strategy was never to make money, but only to manipulate the value of its CRRs

positions: “ETRACOM’s virtual trades as a whole were uneconomic, a fact known to ETRACOM

prior to initiating its trading strategy and throughout the trading period.” [Emphasis added].177

This presumption of ETRACOM’s guilt before it ever placed a single virtual supply offer guides

staff’s exposition, which then interprets every subsequent action of and statement made by

ETRACOM to support its conclusion that ETRACOM acted with the fraudulent, manipulative

intent needed to trigger liability under this Commission’s Rule 1c. An alternative narrative must

be considered, one that instead recognizes that ETRACOM’s trading behavior might have been

motivated by a stand-alone, legitimate business purpose. If this balanced perspective yields a

different perspective of ETRACOM’s behavior, the Commission should give added weight to this

evidence relative to staff’s presumptions.

A. ETRACOM INTENDED ITS HYDROELECTRIC TRADING STRATEGY TO BE PROFITABLE

FROM THE START

124. There is no evidence to support staff’s view that ETRACOM knew that its hydroelectric

trading strategy would lose money in advance of its trades placed on May 14, 2011. ETRACOM

saw historic snowfall and water-content unlike that seen before 1983. Staff asserts that the lack

of “warm rain” made ETRACOM’s strategy “implausible”178 but this belittles the fact that the key

predicate for the hydroelectric event was present. Staff’s assertion also ignores that a sixteen-day

price-taking virtual supply strategy would have been profitable every day if started from April 1,

2009 to August 31, 2009 or 92% of days from March 1, 2010 to August 31, 2010, as verified by

Table 4. Staff’s likewise admits that ETRACOM had no understanding of the fully-encumbered

status New Melones—represented by the lower portion of Figure 3 above and responsible for the

ability of virtual bids to set both prices and congestion—yet somehow ETRACOM understood,

prior to ever trading, that it could reverse congestion on an intertie with a 399 MW net span (384

MW import limit, 15 MW export limit) with virtual supply offers sized between 1-5.03 MW.

177 Id., p. 23.

178 Id., pp. 28-29.


63

125 Indeed, staff entirely ignores the effect the market flaws had on the profitability of

ETRACOM’s virtual trades. Staff notes that the software errors inherent to New Melones could

have incentivized ETRACOM’s offers at zero or negative prices—47% of all prices in the 30 days

prior to May 14 were zero or negative, the lowest at -$26.80/MWh—yet argues that those bids

were intentionally uneconomic. Had the prices been set in conformance with the CAISO tariff,

historical HASP prices suggest that ETRACOM’s offers could have been profitable but for the

flaws associated with the encumbrance. Staff’s allegations that ETRACOM knew nothing of

these flaws yet intended to manipulate the market at New Melones in advance of its actual

trading there are inconsistent and unsupported. The only possible explanation for ETRACOM’s

conception of the hydroelectric trading strategy was that it was intended it to be profitable and

economic from the start.

B. ETRACOM INTENDED FOR ITS MAY 14-15 OFF-PEAK VIRTUAL SUPPLY OFFERS TO

BE PROFITABLE

126. Staff’s intent narrative next alleges that ETRACOM’s trades on May 14-15 targeted hours

of export congestion produced from the 1 MW WAPA flows. Staff’s assertion that ETRACOM’s

$0/MWh virtual supply offers were uneconomic “free energy” contradicts its own recognition

that the software problems at New Melones incentivized such bids, and ignores that fact that

over one-third of those $0/MWh supply offers failed to clear given day-ahead prices of zero or

less in 54% of hours over the two-day period, including a price of -$15.51/MWh. Staff ignores

that high day-ahead prices corresponding with WAPA’s 1 MW export incentivized ETRACOM

by reducing the premium it would pay for the virtual supply offers it made, evidenced by the

small $52 loss it incurred over those two days and small profit it made on May 15. Staff’s need to

rely on a $52 loss incurred over two days to support its allegation that ETRACOM’s engaged in

intentionally-uneconomic trades on May 14-15 is indicative of the lack of evidence it has to

prove ETRACOM’s intent it has in this case.

127. Staff’s support for its allegation that ETRACOM’s HE 7 offer on May 15—which profited

by $10.15—was evidence of a ploy to reverse congestion is a nonspecific instant message stating

“we[‘]re in good shape in CA[.]”179 Rather than confirming manipulative intent on the part of

ETRACOM, this document confirms its legitimate expectation that the premium it paid given its

anticipation of a potentially profitable congestion-related event was minimal, thus incentivizing

the continuation and expansion of its hydroelectric trading strategy. As discussed above, staff

erroneously characterizes the “test period” strategy ETRACOM used at New Melones on May 14-

15 as anomalous, when in fact ETRACOM previously and subsequently added or removed hours

in advance of multi-day, all-day strategies (Tables 8 and 10). Staff’s belief that consideration of

ETRACOM’s addition or removal of hours in strategies executed after May is irrelevant to

evaluating its intent on May 14-15 is absurd and suggests a desire to exclude evidence relevant to

properly evaluating its conclusions and ETRACOM’s legitimate intentions.

179 Id., p. 20, referencing ETRACOM Instant Message, 5/15/2011, 11:07:48 AM, Bates No. ETR01499.


64

C. ETRACOM INTENDED FOR ITS MAY 16-31 TRADES TO BE PROFITABLE

128. Staff carries its presumption of ETRACOM’s manipulative intent through to the virtual

supply offers placed from May 16-31. Based on the foregoing analysis, it is certain that:

Day-ahead prices at New Melones cleared between $0/MWh and -$26.80/MWh in 47%

of hours between April 14 and May 13, consistent with the prices ETRACOM offered;

Negative prices cleared in day-ahead in 54% of the hours of May 14-15, causing over one-

third of ETRACOM’s bids at New Melones to fail to clear;

Software errors unknown to (and, at the time, unknowable by) ETRACOM incentivized

it to offer virtual supply at negative prices to assure those offers would clear;

ETRACOM did not set the day-ahead market LMP in 43.5% of the hours between May

16-31, which reasonably would have led it to believe that it faced competition from other

market participants at New Melones such that it needed to offer at more aggressively-low

negative prices to clear the market; and

As staff admits, ETRACOM did not know of or understand the fully-encumbered status of

New Melones, which caused its offers to intermittently set the price, was responsible for

its virtual losses, and is alleged to have provided the nexus that is the basis for staff’s case.

Despite these facts that support a presumption that ETRACOM rationally followed the economic

incentives provided to it by a market that staff and the CAISO DMM admit was deeply flawed

and dysfunctional—a presumption of transactional legitimacy—staff nevertheless concludes that

ETRACOM “had to know that the negative LMPs at New Melones benefited the profitability of

its CRR positions”180 Staff’s use of the phrase “had to know” is telling: while ETRACOM lost

money on its hydroelectric trading strategy in May and its CRRs were profitable at that same

time, staff has no evidence to prove that ETRACOM intentionally incurred losses on its virtual

trades, or that it intentionally used those transactions to manipulate the value of its CRRs.

129. Staff relies on the false premise that no rational firm would willingly risk “about $2,600

per day” over sixteen days to pursue a speculative trading strategy.181 Staff ignores that snow and

water-content conditions at New Melones were as favorable for the strategy to work as they had

been since 1983; that ETRACOM would assume that it hydroelectric trading strategy would be

price-taking given that it did not know of New Melones fully-encumbered status, and as such

could have been as profitable as it would have been in prior years (Tables 3 and 4) or would have

been in July 2011 (Table 3 and Figure 6); that a rational trader would evaluate a trading strategy

day-by-day, without necessarily viewing prior losses as anything other than a sunk cost; and that

ETRACOM may have viewed its losses at New Melones as less significant given its larger virtual

portfolio (Figure 8). Clearly, if staff perceives that a $52 loss incurred over two days was

180 Id., p. 35.

181 Id., p. 30.


65

sufficient to rule ETRACOM’s actions as intentionally uneconomic—and therefore fraudulent—the view of ETRACOM’s loss of “about $2,600 per day” was determinative.

130. That said, staff has no documentary evidence to support its allegations, and the evidence

that staff cites is in fact exculpatory. Staff alleges that two instant messages show ETRACOM’s

awareness of its virtual losses at New Melones on two occasions, the first stating “[w]e lost $800

on Melon[e]s but made back $200 on some evening trades”182 on May 16, the second stating

“[y]esterday Melon[e]s cost us about $2K”183 on May 20. As discussed above—and which must be

restated given the magnitude of its bearing on ETRACOM’s intent—staff failed to include

Rosenberg’s response to this second instant message, “I am thinking we should stop putting

positions on [at New] Melon[es] until the auction end.”184 Aside from showing that ETRACOM

tracked its virtual losses at New Melones two times over the course of sixteen days, these

documents demonstrate that ETRACOM and Rosenberg evaluated the hydroelectric trading

strategy only on the basis of its stand-alone profitability and without any consideration of its

effects on their CRR portfolio at New Melones. That ETRACOM considered ending the strategy

on May 20 given $1,997 in losses given $38,195 in concomitant gains in its CRR portfolio

(Table 5) demonstrates that staff’s allegations concerning ETRACOM’s intent, both with respect

to its losses on its virtual trades and its recognition of a nexus between its virtual losses and CRR

gains, are erroneous and biased, and therefore must be ignored.

131. Staff alleges that the choice and timing of ETRACOM’s May 16-31 trades is indicative of

manipulative intent. Staff asserts that “having seen [on May 14-15] that it could effectuate a $0

LMP at New Melones, ETRACOM expanded its virtual trading strategy to all hours of the

day[.]”185 This insinuates that in the wake of the off-peak $6 gain on May 15 and $52 net loss

over both days of the “trial period,” ETRACOM intentionally chose to expand its (presumptively

losing) virtual strategy to all hours to amplify gains to its CRRs. Aside from staff once again

pushing its unsupported allegation that ETRACOM understood and intentionally exploited the

nexus between its virtual trades and CRRs caused by New Melones’ fully-encumbered status,

staff’s position ignores that the greatest possible profits from the strategy in 2011—corresponding

with the greatest risk of losses—would derive from peak hours (Table 3). Staff’s assertion that

ETRACOM’s placement and expansion of its hydroelectric trading strategy at New Melones was

intentionally uneconomic is baseless.

132. Staff’s allegations that ETRACOM’s virtual supply offers placed May 16-31 differed from

its trades placed at other locations are similarly unfounded, shown by 76 other instances where

ETRACOM placed continuous bids or offers for 24 hours a day for more than a few days at a time

(Table 9) and by seventeen other interties where ETRACOM offered 1-5 MW of virtual supply at

182 Id., p. 10, citing ETRACOM Instant Message, 5/16/2011, 9:47:36 PM, Bates No. ETR01508.

183 Id., pp. 10 and 20, citing ETRACOM Instant Message 5/20/2011, 7:33:20 AM, Bates No. ETR01509.

184 See Id., p. 20-21, discussing ETRACOM Instant Message, 5/20/2011 7:33:58 AM, Bates No. ETR01509.

185 Id., p. 18.


66

the flat price of -$20/MWh from May 25-31 (Table 2). Likewise, staff’s erroneous allegation that

ETRACOM intended to clear its virtual supply at the lowest prices possible is easily disproven by

looking at the upward slope of its offers (Table 7), showing that its intent was not to lower the

day-ahead price at New Melones, but instead to reduce the premium it paid for the trades.

133. Staff’s allegation that the timing ETRACOM’s cessation of its hydroelectric trading

strategy at the end of May was indicative of manipulative intent is similarly groundless. Aside

from the simple fact that a month’s-end serves as a natural breakpoint for decision making,

ETRACOM retained sufficient leverage in its CRRs at New Melones in June to have profitably

incentivized the continuation of staff’s alleged manipulative strategy. That ETRACOM did not to

do so confirms both that its hydroelectric trading strategy was legitimate and that it never

recognized the strength of the nexus between its virtual trading and CRRs, thus showing the

error of staff’s assertions.

134. Finally, staff’s alleges that ETRACOM’s “gains on its CRR revenues grew dramatically as

its virtual trading increased. These gains dwarfed the losses associated with its virtual trading.”186

Staff’s graphic purports to show the “dramatic” relationship between gains in ETRACOM’s CRRs

and its losses in its virtual trades at New Melones, shown below in Figure 10.

Figure 10

Staff’s Depiction of ETRACOM’s May 2011

Virtual Losses and CRR Gains187

186 Id., p. 11.

187 Id., p. 11.


67

135. The lower part of Figure 10 shows ETRACOM’s virtual losses (red line) and CRR gains

(blue line) in May 2011. Three things should be noted about this diagram. First, contrary to

staff’s assertion that “ETRACOM’s mounting losses at New Melones, which ranged from $871

and $5,851 per day, could not be overlooked[,]”188 ETRACOM’s small daily losses and total

$42,481 loss at New Melones in May 2011 was small compared to its broader portfolio of virtual

(Figure 8) and CRR positions. Second, while ETRACOM’s CRR gains at New Melones look large

when compared to its virtual losses in May, they look unremarkable when compared to

ETRACOM’s CRR gains in April (Figure 9). Third, and most importantly, the fact that these

lines are placed on the same diagram does nothing to demonstrate that ETRACOM connected the

losses in one to gains in the other (or vice versa). In short, staff once again shows the pieces of

the framework without successfully demonstrating ETRACOM’s manipulative or fraudulent

intent to connect them or operate them in concert.

D. ETRACOM’S BEHAVIOR IN JUNE CONFIRMS THAT IT NEVER INTENDED TO AFFECT

THE VALUE OF ITS CRRS

136. Staff seeks to sidestep the fact that ETRACOM reversed its virtual trading strategy in June

2011 by bidding virtual demand at New Melones in quantities smaller than the size of its CRRs,

stating that “[s]taff does not argue that ETRACOM’s June trading was part of its manipulative

scheme.”189 Indeed, staff could not possibly make such an argument, since the act directly and

conclusively ruins its argument concerning ETRACOM’s knowledge of the alleged manipulation

nexus. A rational economic actor with such knowledge would avoid offering virtual demand at

the same location where it holds a CRR source position because the demand bid could only serve

to increase day-ahead LMPs and congestion prices, to the detriment of the larger CRR position.

That ETRACOM placed such trades confirms that it did not perceive the alleged linkage between

its virtual trades and CRRs in June, thus confirming that it held no such knowledge during the

Relevant Period in May alleged by staff.

137. ETRACOM did not exploit such nexuses generally within its portfolio. From February 1

through August 31, 2011, ETRACOM held over 1.45 million hourly positions—comprised of

virtuals and/or CRRs—across various locations in the CAISO.190 Of these, only 15,104 (1.04%)

were hours where a virtual trade was placed at the source or sink of a CRR. Without additional

documentary proof of manipulative intent, identification of these trades as anything more than

random interactions is spurious.

188 Id., p. 20.

189 Id., p. 37.

190 See Excel workbook titled “Interacting_Node_Analysis.xlsx,” tab “Summary.”


68

E. STAFF ATTEMPTS TO PROVE ETRACOM’S INTENT THROUGH THE USE OF “NO WIN”

SCENARIOS

138. Staff’s Report uses “no win” logic to support its allegations of ETRACOM’s manipulative

intent. For example, ETRACOM’s choice to place virtuals only in off-peak hours on May 14-15

was viewed as evidence of its intent to manipulate the value of its CRRs, for off-peak hours were

typically when WAPA engaged in its 1 MW physical export. When ETRACOM expanded its

strategy to all hours of the day starting on May 16, that too was considered evidence of

manipulative intent, for ETRACOM had then ostensibly forsaken all hopes of profitability—

although the events of July (and many other locations and time periods) proved otherwise. If

ETRACOM had limited its bidding from May 16-31 to off-peak hours only, staff could assert

logic identical to that it used to describe ETRACOM’s behavior during the so-called “test

period”—that the virtual supply offers were targeting its CRRs, despite day-ahead price signals

that incentivized the behavior. The implication is that ETRACOM’s guilt is presumed, with the

facts chosen and presented selectively to paint a preordained picture of manipulative intent.

139. ETRACOM’s decision to discontinue its hydroelectric trading strategy at the end of May

2011 provides another example of staff’s “no win” logic. Because ETRACOM ended its strategy

on May 31st, staff alleges that the timing of its relinquishment with the expiry of its May CRR

positions at New Melones demonstrate that it never intended to profit from the hydroelectric

trading strategy and thus is guilty of manipulation. However, had ETRACOM instead chosen to

continue the strategy into June with smaller but still financially-leveraged CRR positions, staff

would then assert (assuming continued losses on the virtual supply offers) that ETRACOM was

simply continuing the manipulation. Given that ETRACOM chose not to continue the strategy

into June, and in fact placed virtual demand trades that would only serve to hurt the value of its

CRRs, staff has instead tried to minimize the economic implications of the behavior by stating

“ETRACOM’s June virtual demand trading does not validate its May virtual supply trading.”191

This approach is inconsistent with affording ETRACOM the presumption of transactional

legitimacy that must follow its open market trades.

140. Such “no-win” logic can also be found in the CAISO DMM’s “Follow Up” document that

describes its (and, necessarily, Enforcement’s) single-minded logic in this case concerning the

proof of ETRACOM’s intent. Page 3 of this document states:

Etracom offered no more than 4 MW of VIRTUAL SUPPLY at New Melones from

May 1 through May 25. The position limit for virtuals at New Melones allowed up

to 19 MW. If Etracom's VIRTUAL SUPPLY price expectations led to expectation

of profitable VIRTUAL SUPPLY position with 4 MW then it would be even more

profitable at 19 MW.



69

Thus, DMM asserted that ETRACOM’s convergence bids were too small compared to position

limits in place at New Melones to have maximized profit from the hydroelectric trading strategy,

thus demonstrating its supposed indifference to making a profit on the trades and confirming its

purportedly manipulative intent. However, the very next paragraph of the document states:

Starting May 26, Etracom began slightly increasing its VIRTUAL SUPPLY offer

quantity at New Melones (from 4 MW to 4.03 MW and then 5.03 MW). The

purpose was to provide more VIRTUAL SUPPLY than was offered and clearing

for VIRTUAL DEMAND and thus continue to set low LMPs at New Melones and

receive higher CRR revenues.

Therefore, the DMM also accuses ETRACOM of manipulation because its convergence bids were

too large given the limited participation of virtual demand at New Melones, again leading to the

inexorable conclusion that ETRACOM acted with manipulative intent. That staff relied on the

DMM’s logic in whole or in part in drawing its conclusions in this case is revealing.

F. SUMMARY: STAFF HAS FAILED TO PROVE ETRACOM’S MANIPULATIVE INTENT

141. This case underscores the importance of documentary evidence to the proof of fraudulent

intent in market manipulation cases generally and in this case in particular. To overcome the

presumption of transactional legitimacy that must be afforded to ETRACOM’s hydroelectric

trading strategy, staff must demonstrate that ETRACOM did not intend for its convergence bids

to be profitable on a stand-alone basis and that it intentionally used those bids to benefit the

value of its financially-leveraged CRR positions. While circumstantial evidence may have

provided staff with reason to investigate ETRACOM’s behavior, the investigation led to

documentary evidence that not only fails to support a finding of manipulation, but confirms both

the legitimate business purpose of its convergence bidding strategy and its lack of awareness of

how the strategy could have or would have affected the value of its CRR positions.

142. Based on the above, I opine that: (a) ETRACOM did not know and could not have known

that flaws in the market and its implementation and software errors set prices at New Melones

based on cleared (or even uncleared) convergence bids, despite the fact that scheduled physical

and virtual flows never exceeded the transmission line’s capacity; (b) these flaws and errors

frustrated effective price discovery at New Melones and resulted in market outcomes in violation

of the CAISO’s tariff, preventing ETRACOM from understanding the characteristics of the

intertie and incentivizing it to offer virtual supply at zero or negative prices to assure that those

offers would clear; (c) ETRACOM’s convergence bidding activity was economically rational

given day-ahead price signals, published reports of an imminent significant hydro event, and its

correct (though mis-timed) pursuit of stand-alone profits consistent with its hydroelectric

trading strategy; (d) the appearance of HASP congestion in July 2011, as well as in prior and later

years, confirms both the legitimacy of ETRACOM’s hydroelectric strategy and its expansion into

peak and off-peak hours; (e) ETRACOM’s trading strategy at New Melones shared several

attributes with other strategies executed by the Company elsewhere, contrary to staff’s assertions

in the Staff Report and Preliminary Findings; (f) ETRACOM did not recognize or intentionally


70

exploit a nexus between its virtual supply and CRR positions before, during or after the Relevant

Period, as confirmed through documentary evidence that was mischaracterized or ignored by

staff; (g) given the unique flaws and software errors presented at New Melones, it would have

been reasonable given ETRACOM’s experience in the CAISO not to perceive its CRR gains at

New Melones as unusual or related to its convergence bidding; (h) ETRACOM would not have

viewed the increased value of its CRRs during the Relevant Period as unusual, either when

compared against its profits there earlier in the month or its total CRR profits the month before;

and (h) consistent with its presumption of guilt, the logic used by staff and the CAISO Market

Monitor in an attempt to prove ETRACOM’s manipulative intent in this case is couched in “no

win” scenarios where a finding of manipulation is predetermined.

143. Because ETRACOM’s convergence biding and CRR trading strategies each were intended

to be independently profitable on a stand-alone basis, and because documentary evidence

confirms that ETRACOM was not aware of and thus did not intentionally exploit a linkage

between these positions, there is an insufficient basis to conclude that there was a manipulation

trigger, target or nexus in this case; no basis for concluding that ETRACOM acted with

fraudulent intent in executing its hydroelectric trading strategy at New Melones; and, given the

many flaws present at New Melones, no basis for concluding that ETRACOM’s trading caused

the alleged market harm. As a result, staff has not demonstrated that ETRACOM engaged in

manipulation.

VII. Staff’s Estimate of Unjust Profits and Harm Overstates the Impact

of ETRACOM’s Virtual Trading

144. Staff’s Report alleges that ETRACOM made unjust profits of $315,072 as a result of its

virtual bidding strategy at New Melones from May 14, 2011 through May 31, 2011.192 For this

calculation, staff determined that ETRACOM earned $517,417 on its CRR sourced at New

Melones from May 14-31.193 Because some of this revenue (presumptively during on-peak hours)

would have been earned in the absence of any virtual activity, “staff determined that the average

profits earned between May 8 and 13 provide a reasonable measure of what profits would have

been for the rest of the month had ETRACOM not engaged in manipulation.”194 Staff calculated

these “legitimate” profits to equal $202,345, resulting in purportedly unjust profits of $315,072.195

Staff also alleges that ETRACOM and caused market harm equal to $1,514,207 over the Relevant

Period, calculated by finding that CRR source holders at New Melones were paid $2,122,947,

$608,740 of which was legitimate gain and $1,514,207 paid due to the alleged manipulation.196

192 Id., p. 39.

193 Id.

194 Id.

195 Id.

196 Id.


71

Finally, staff seeks $2.5 million in civil penalties from ETRACOM ($2.4 million from the

company and $100,000 from Rosenberg) based on its interpretation of these numbers relative to

the Commission’s Penalty Guidelines.197

145. The fully-encumbered status and software errors present at New Melones distorted price

signals, LMPs and congestion severely. Had these flaws not existed, ETRACOM would not have

been incentivized into placing the virtual supply offers associated with its hydroelectric trading

strategy at zero or negative prices; ETRACOM’s offers would have been price-taking, as they

were at the seventeen other interties where it placed -$20/MWh virtual supply offers during the

Relevant Period; the resulting LMPs would have resolved more favorably, reducing the losses

ETRACOM incurred on those trades; and, most importantly, no congestion should have occurred

on the line as physical flows never approached its physical limits in either direction. ETRACOM

knew nothing about the flaws responsible for driving this market’s dysfunction, and all

documentary and economic evidence confirms that it never did recognize a causal linkage

between its virtual activity and CRR’s performance. Had CAISO followed its tariff, ETRACOM’s

virtual trades would have had a de minimis effect on the value of all CRRs valued there, as no

phantom congestion would have occurred and no revenue inadequacy would have resulted.

Given these circumstances, staff’s recommendations as to ERTACOM’s harm and unjust profits

assessments should be ignored with ETRACOM assigned no liability for its actions.

146. I have replicated staff’s unjust profits and harm calculations on an hourly basis.198 If

arguendo, this Commission determines ETRACOM’s behavior was manipulative, staff’s

calculations overstate both ETRACOM’s unjust profits and harm because: (a) the day-ahead price

signals created by the 1 MW power export by WAPA incentivized the rational placement of

virtual supply offers at New Melones in the off-peak hours of May 14-31; and (b) ETRACOM did

not set the market price in hours where its offers were infra-marginal or did not clear.

Correction for each of these problems is presented in Table 11, below.

197 Id., p. 40 and Revised Policy Statement on Penalty Guidelines, 132 FERC ¶ 61,216 (Sep. 2010), P 2

198 See Excel workbook titled “May Virtual and CRR Daily Profits.xlsx,” tab “Table 5.”


72

Table 11

Correction of Staff’s Disgorgement and Harm Calculations199

Source and Notes:

[1]: Brattle replication of FERC's baseline for calculating harm and disgorgement.

[2]: [1] + Removing the hours in which negative demand bids = LMP.

[3]: [1] + Removing off-peak hours.

[4]: [1] + Removing the hours in which a negative demand bid = LMP and off-peak hours.

147. As discussed above, the average day-ahead price for hours when WAPA exported power

on May 8-13 cleared at $31.48, compared to an average off-peak price of $6.58 in April and an

average off-peak price of $15.53 on May 1-7. This high Day-head price incentivized ETRACOM

to execute its hydroelectric trading strategy in off-peak hours on May 14-15, which resulted in a

small net loss of $52. Given that WAPA continued to export power from the CAISO in most of

the off-peak hours throughout May, ETRACOM rationally would have continued to place virtual

supply offers at New Melones throughout this period. Because four of its nine $0/MWh failed to

clear on May 15, and because the expected gains from HASP congestion substantially exceeded

its expected losses on its virtual supply offers, ETRACOM would rationally have been

incentivized to continue offering virtual supply during the remainder of May at negative prices

in pursuit of stand-alone profits on those offers. Staff’s calculations should therefore remove all

off-peak hours from its analysis. As Table 11 shows, doing so reduces ETRACOM’s calculated

unjust profits from $315,072 to $174,336 and harm from $1,514,207 to $557,078.

148. ETRACOM did not set the price in 43.5% of hours from May 16-31, all of which were set

by negative virtual demand bids. While staff may assert that these negatively-priced virtual

demand bids were drawn into the market only as a result of ETRACOM’s negatively-priced

199 See Excel workbook titled “Shadow Prices Profits Screen and Financial Leverage Analysis.xlsx,” tab

“Table 11.”

FERC (Baseline)

Baseline Less

Negative

Demand Bids

Baseline Less Off

Peak

Hours

Baseline Less Off

Peak Hours and

Negative

Demand Bids

[1] [2] [3] [4]

Unjust Profits

ETRACOM's profits betweenMay 14 31 $517,417 $315,074 $387,175 $256,786

Legitimate profits betweenMay 14 31 $202,345 $126,929 $212,839 $135,360

Alleged unjust profits betweenMay 14 31 $315,072 $188,145 $174,336 $121,426

Harm

CRR holders' profits betweenMay 14 31 $2,122,947 $1,216,954 $1,237,189 $820,541

Legitimate profits betweenMay 14 31 $608,740 $375,195 $680,111 $432,534

Alleged harm created betweenMay 14 31 $1,514,207 $841,759 $557,078 $388,007


73

virtual supply offers,200 such conjecture would ignore that negative virtual demand bids set the

price 95 times from April 14 to May 13, with prices clearing as low as -$26.80/MWh. Further, a

competitive response to ETRACOM’s virtual supply offers would have raised the market-clearing

LMP only by as much as other rational market participants were willing to pay in accepting the

risk of potential hour-ahead price spikes given the market conditions at New Melones. Because

these outcomes would have reinforced ETRACOM’s view of itself as a price-taker, ETRACOM

would have perceived these hours as indicative of a normally-functioning market, within which

the premium of its legitimate hydroelectric trading strategy was reduced by a competitive

response. Staff’s calculations should therefore remove all hours when ETRACOM was not the

marginal bidder. Doing so reduces ETRACOM’s calculated unjust profits from $315,072 to

$188,145 and harm from $1,514,207 to $841,759, also shown in Table 11.

149. The two problems with staff’s calculations are conceptually distinct but are not mutually

exclusive, meaning that several of the hours in which ETRACOM did not set the market price

also occurred in the off-peak. To fully correct for both issues requires the prevention of double-

counting for hours excluded. Table 11 shows that this results in further reduction of the unjust

profit calculation to $121,426 and harm to $388,007. Additional downward adjustments to staff’s

disgorgement and harm calculations are needed if the Commission finds that ETRACOM lacked

cognition of the nexus between its virtual supply offers and CRRs for part of the Relevant Period

or if a methodology is devised that can compensate for the many flaws and software errors that

were present in this market during the Relevant Period.

150. Ideally, the validity of staff’s calculations would need to be verified by “rerunning” the

CAISO’s network model for New Melones assuming that the subset of ETRACOM’s convergence

bids found to be manipulative can be removed from the market. This analysis would need to

compensate for the various flaws that were endemic to the market at that time, including the

software pricing issues that inhibited valid price discovery and the fact that the intertie’s “fully

encumbered” status reduced the market’s size of available transmission capacity the from the

intertie’s capacity to effectively 0 MW. Unfortunately, as the CAISO DMM wrote in the

footnote on p. 16 of its 2011 Q2 Report, “DMM has limited ability to assess the impact of virtual

bidding on congestion due to problems with re-running the day-ahead market software…”201

Given the flaws created by the CAISO’s poor execution of its market design at New Melones

during the Relevant Period, the DMM’s difficulties are not surprising. However, because the

Commission must ultimately rely upon these calculations to evaluate the impact of ETRACOM’s

behavior on the market, any estimations of undue profits and harm are speculative as a result.


201 See http://www.caiso.com/Documents/RevisedQuarterlyReportonMarketIssuesandPerformance-

August2011.pdf.


74

151. Staff’s argument against these (or any other) reductions relies entirely upon its erroneous

and unsupported position that ETRACOM’s allegedly manipulative intent was formed in advance

of its first trades on May14:

Staff finds incorporation of these hours appropriate because ETRACOM’s trading

was not responding to price signals from WAPA’s scheduled export; its trades

were placed with an intent to lower prices to benefits its CRR positions.

ETRACOM’s intent to manipulate prices occurred in all hours it bid, whether or

not it was inframarginal or cleared.202

Without the benefit of presumptions and suppositions, there is not one shred of economic or

documentary evidence that confirms staff’s determination of ETRACOM’s allegedly manipulative

intent. If this Commission believes that ETRACOM’s hydroelectric trading strategy could have

been legitimate such that its trades on May 14 were placed in pursuit of stand-alone profits—a

possibility increased by snow conditions unprecedented since 1983, profitable trades at sixteen of

seventeen other interties later in May and even staff by staff’s own admissions in its Preliminary

Findings—a presumption of transactional legitimacy must be confirmed for those trades.

152. However, staff knows that this recognition will doom its case, for at no point after May

14 can it demonstrate that ETRACOM’s behavior “crossed the line” into manipulation, especially

given ETRACOM’s: (a) consideration of ending the strategy on May 20 despite large CRR gains;

(b) placement of negative virtual supply offers at seventeen other interties starting on May 25

after the June CRR auction was at an end; and (c) trading of virtual demand into it CRRs source

at New Melones in June, which could only serve to injure the value of those positions. Indeed,

staff’s entire case is built around nothing more a correlation between a suspected trigger, nexus

and target, but without the requisite proof that the losses in the virtual supply offers were

intentionally incurred (trigger) or that the day-ahead congestion price (nexus) was intentionally

exploited. The same presumption of transactional legitimacy must therefore successfully attach

to the rest of ETRACOM’s trades, for staff’s suppositions are insufficient to prove otherwise.

153. Considering the contribution of a deeply flawed and dysfunctional market, the substantial

documentary and economic evidence that supports the stand-alone, legitimate business purpose

of ETRACOM’s hydroelectric trading strategy, and staff’s lack of basis to support its allegations of

manipulative intent, no civil penalties are appropriate because the investigation of ETRACOM

should be closed. However, in the event that this Commission determines that some liability for

harm or unjust profits is warranted, civil penalties should be minimized so as to reduce the

likelihood that the precedent of this case will thwart future legitimate risk taking by market

participants seeking to engage in new strategies but fearful of incurring losses in instruments that

might inadvertently interact with others in their portfolios.



75

VIII. Conclusion

154. Staff’s Report seeks to prove manipulative intent based upon a revisionist ex post view of

trades placed by a market participant who ex ante traded without the benefit of the interpretive

hindsight now available. The behavior at issue was a first-time hydroelectric trading strategy

based on a unique opportunity, executed by a company with relatively little experience with

virtual trading in the CAISO. Any allegations of market distortions created by this strategy must

first consider CAISO’s flawed market design that portrayed a 384 MW import and 15 MW export

intertie as competitive when, in fact and unknown to market participants, a single MW (whether

placed by ETRACOM, WAPA or other entities) could reverse the congestion on that path.

155. Any analysis of the zero or negative pricing of ETRACOM’s virtual supply offers must

likewise control for the distortion of incentives caused by the CAISO’s multiple market flaws

pricing software errors at that location. Any examination of ETRACOM’s motivations must be

tempered by a realization that the CAISO’s market at New Melones was so severely flawed that

trading was suspended at that location on August 29, 2011 (“fully encumbered” interties) and

throughout the CAISO (interties generally) on November 28, 2011. ETRACOM, a firm with less

than six months of experience with placing virtuals generally and with less than 3.5 months of

experience in placing virtuals specifically in the CAISO, is accused of manipulating a market that

subsequent remedial measures determined should never have existed in the first place.

156. Consistent with my framework for the analysis of market manipulation, staff alleges that

ETRACOM intentionally placed uneconomic virtual supply offers (the trigger) to bias the day-

ahead congestion price at New Melones (the nexus) to benefit the value of its CRRs (the target).

However, while staff’s narrative fits the general logic of the framework, the proof of its specific

elements and of ETRACOM’s intent fails. ETRACOM’s virtual supply offers were not

intentionally uneconomic, but were executed pursuant to a trading strategy that rationally

considered potential profits on a day-to-day basis and that even staff admits could have been

profitable. ETRACOM exhibited no knowledge of a nexus between its virtual trades and CRRs,

for it placed virtual demand bids for which any gains would have been more than offset by even

greater losses to its CRR positions at New Melones immediately following the Relevant Period.

ETRACOM would not have attributed the profits garnered on its CRRs at New Melones in late

May to its virtual trading, for it had experienced even greater profits to its CRR portfolio only

one month before and without the benefit of such an alleged “scheme.”

157. An alternative narrative is more plausible: ETRACOM conceived of a hydroelectric

trading strategy designed to profit from a once-in-a-generation snowpack not experienced since

1983. While contemplating this strategy, export congestion appeared during the off-peak hours

of May 8-13, 2011, incentivizing ETRACOM to deploy virtual supply offers consistent with its

strategy on May 14-15. The losses on these offers being negligible, ETRACOM decided to

employ the virtual trading strategy across all hours beginning May 16, assessing its costs relative

to its outlook of potential profits expected to accrue due to impending hydroelectric conditions.

ETRACOM considered shutting the strategy down due to losses on May 20, completely oblivious


76

to the benefit that staff asserts the strategy was conferring to its CRRs. As the June FTR auction

was ending on May 25, ETRACOM expanded its strategy to seventeen other interties, sixteen of

which were profitable. ETRACOM ended the strategy at New Melones at the end of May, but as

of June 1 placed virtual demand without recognition that those trades would injure its CRRs still

sourcing there. Hydroelectric conditions materializing in July ultimately demonstrated the

potential profitability ETRACOM’s hydroelectric trading strategy.

158. Through the virtual offers associated with its hydroelectric trading strategy, ETRACOM

engaged in profit-seeking transactions such that its expected gains from HASP congestion were

hoped to exceed the expected cost—i.e., “premium”—associated with those offers. This strategy

was unsuccessful due to poor timing relative to the expected event and, more importantly, the

multiple, severe flaws that dogged the market at New Melones. Given the losses incurred, staff

and the CAISO DMM assert that the intent of the strategy must have been fraudulent, as no one

would rationally have lost money under the circumstances—an easy argument given the

information now available ex post of ETRACOM’s trades. However, it is ETRACOM’s intent ex ante of its trading that is relevant to a finding of market manipulation.

159. Unlike several previous cases that this Commission has considered involving the alleged

exploitation of flawed market rules or market design,203 there is no evidence that ETRACOM

knew of the flaws created by the encumbrance and software errors present at New Melones

before executing its hydroelectric trading strategy. To the contrary, these subsequently disclosed

flaws—and other yet to be disclosed and explained—disrupted effective price discovery and

incentivized ETRACOM to bid at zero or negative prices to assure that its virtual supply offers

would clear. While it is true that ETRACOM lost money on the virtual offers that comprised

this strategy, those losses were also caused by the same market flaws that ETRACOM never

knew of until after the Relevant Period.

160. This Commission has noted that “profitability is not determinative on the question of

manipulation[…].”204 Likewise, it must also be true that lack of profitability in and of itself is not

dispositive of manipulation. If this is not so, then this Commission would effectively force every

trader seeking to execute a stand-alone, profit-seeking strategy to question whether unknown

interactions within the rest of its portfolio may draw an enforcement action should it lose

203 See In Re Make-Whole Payments and Related Bidding Strategies., Order Approving Stipulation and

Consent Agreement, 144 FERC ¶ 61,068 (2013) (alleged exploitation of market rules in the CAISO and

Midwest ISO to garner out-of-market payments); Deutsche Bank Energy Trading, LLC., Order to

Show Cause and Notice of Proposed Penalty, 140 FERC ¶ 61,178 (2012) (alleged use of uneconomic or

fraudulent physical flows to benefit the value of CRRs); and MISO Virtual and FTR Trading, Order

Approving Stipulation and Consent Agreement, 146 FERC ¶ 61,072 (2014) (alleged use of virtual bids

to benefit the value of financial transmission rights on a radial tie, consistent with the trader’s

dissertation thesis).

204 Deutsche Bank Energy Trading, LLC., Order Approving Stipulation and Consent Agreement, 142

FERC ¶ 61,056 (2013), P 20.


77

money, despite its best intentions. This is why documentary evidence is so central to these cases,

for it alone serves as either confirmation or refutation of the economic evidence that otherwise

cannot distinguish between a manipulative scheme and/or an errant interaction.

161. Here, the documentary evidence available supports ETRACOM’s version of the facts, not

that presented by staff. If this Commission were to punish ETRACOM for the losses it incurred

in a failed attempt to profit from a legitimate, risk-taking strategy incentivized by an irreparably

flawed market design, the outcome would chill future trading for fear that legitimate risk-taking

strategies that lose money on a stand-alone basis might be identified by and prosecuted by this

Commission, irrespective of documentary evidence that proves the trader’s legitimate intent.

The resulting reduction of market liquidity would discourage legitimate trading, thus reducing

market efficiency and making those markets more susceptible to manipulation over time.

162. Allegations that the effects of ETRACOM’s pursuit of its hydroelectric trading strategy

created market distortions are attributable to the many severe flaws present in this market and

are not the fault of ETRACOM. Any resulting assessments of unjust profits or market harm are

unwarranted as a result. However, if such assessments are ultimately made, they should be

adjusted in a manner consistent with this Affidavit.

163. Staff’s case against ETRACOM gives this Commission with an opportunity to demonstrate

that the market participants subjected to the investigations process can receive due process

following the receipt of an Order to Show Cause. ETRACOM has been subjected to this process

since 2011, and has cooperated with the investigations process and incurred significant expense

in defense of the allegations brought by staff. The economics of the situation, as reflected in so

many of the settlements reached to date, has already confirmed the Commission’s message to

market participants as to the consequences of engaging in behavior it deems illegal. Believing it

has found the components of such behavior, staff has aggressively pursued this case, so much so

that it has ignored, misinterpreted and sometimes misrepresented evidence that is exculpatory

for ETRACOM. But without proof of fraudulent, manipulative intent, its case must be dismissed.

Only this will restore the presumption of legitimacy that should follow ETRACOM’s trades.


Appendix A

CV of Shaun D. Ledgerwood


SHAUN D. LEDGERWOOD

Principal

Washington, D.C. +1.202.419.3375 [email protected]

A-1

Dr. Shaun Ledgerwood is an expert in market competitiveness with an emphasis on the economic

analysis of market manipulation claims in energy, financial and commodities markets. He specializes in

the analyses of competitive matters within and across physical and financial markets; issues pertinent to

economic regulation, ratemaking, and resource planning; asset valuations; and analyses pursuant to

matters in tort, contracts, or involving fraud. His 25 years of experience as an attorney and an economist

allow him to provide clients with unique insight into matters involving potential litigation, especially as

they relate to meeting burdens of proof under statutory requirements.

As a former economist and attorney for the Office of Enforcement for the Federal Energy Regulatory

Commission (FERC), he evaluated manipulative behavior within and across wholesale electricity and

natural gas markets. During his tenure, he assisted in the enforcement proceedings involving Energy

Transfer Partners, L.P. and Amaranth Advisors, LLC, Constellation Energy Commodities Group, and

several other manipulation-related investigations. He led the agency’s pilot project for energy market

surveillance of wholesale natural gas and electricity markets, as well as evaluations of the impact that

proposed environmental (Waxman-Markey) and financial (Dodd-Frank) reform legislation would have

on the FERC’s jurisdiction and resources.

Dr. Ledgerwood’s experience at the FERC led him to develop a framework for detecting and analyzing

manipulative behavior. The framework separates actions that cause the manipulation from those that

benefit from its effect, simplifying the proof (or disproof) of the manipulation. This logic is particularly

useful in assisting surveillance efforts, maintaining compliance, and for supporting or defending

enforcement actions brought under the anti-manipulation rules of the FERC, Commodity Futures

Trading Commission (CFTC), Federal Trade Commission, Securities and Exchange Commission, the EU’s

anti-manipulation provisions under REMIT and MAD, and other similar statutes based on the principles

of fraud or the creation of an “artificial” price.

Dr. Ledgerwood previously worked as a consultant for C.H. Guernsey and Company in Oklahoma City.

While there, he assisted clients in complex market power issues and testified as an expert witness before

state utility commissions and in federal court. He also taught graduate level courses in microeconomic

theory, regulation, law and economics, antitrust and remedies at the University of Oklahoma

Department of Economics, College of Law, and Price College of Business. He also served as an Affiliated

Faculty member at the Georgetown University Public Policy Institute.

Dr. Ledgerwood holds a Ph.D., M.A., and B.A. in Economics from the University of Oklahoma and a J.D.

from the University of Texas (Austin).


Shaun D. Ledgerwood

A-2

AREAS OF EXPERTISE

Market Manipulation Detection, Analysis, Compliance and Litigation

Utility Regulatory Policy and Ratemaking

Analysis of Competitive Issues in Regulated Industries

Analyses in Support of Antitrust or Other Market-Related Proceedings

Analysis of Liability and Damages in Tort, Contracts and Fraud

Environmental Policy and Regulation

Valuation Analysis

EXPERIENCE

MARKET MANIPULATION DETECTION, ANALYSIS, COMPLIANCE AND LITIGATION

Analyses of Alleged Market Manipulation

Dr. Ledgerwood assists clients, regulatory agencies, and other experts on matters involving alleged

manipulative activity in commodity and financial markets. He specializes in the analysis of acts that

can trigger a manipulation (uneconomic trading, outright fraud, and the exercise of market power) and

their nexus to potential manipulation targets such as financial derivatives or physical index positions.

He has published extensively on this subject since joining Brattle in January 2011.

Successful Defense of Client under Investigation. Assisted client alleged to have engaged in

uneconomic physical transactions across RTOs to benefit financial positions. Analysis proved

that the company’s physical power flows were typically uncorrelated with the directionality

of its financial positions and that those positions which could directionality benefit lacked

sufficient financial leverage to make the manipulation profitable overall. The investigation

was closed with no action taken against the company.

Support of Agency Investigation. Assisted a regulator investigating the alleged exercise of

market power by a company that simultaneously held related derivatives contracts that could

benefit from higher prices. Analysis determined that the exercise of market power was

profitable on a stand-alone basis without the need of any additional profits garnered from the

derivatives. As such, no manipulation was found, for the behavior was neither fraudulent

nor created an “artificial” price.

Successful Defense of Client under Investigation. Assisted client alleged to have engaged in

prohibited transactions in the California ISO. Analysis proved that the company flowed

physical power when the economics dictated it could do so profitably, and likewise reversed

those transactions financially based on legitimate profit-seeking motivations. The agency

investigating the matter closed the case with no further action against the company.

Support of Client under Investigation. Assisting a trader to defend against allegations of

violations of the FERC’s anti-manipulation rule. Dr. Ledgerwood provided an expert report

explaining why the behavior alleged served a stand-alone legitimate business purpose,

specifically avoided manipulative outcomes, adhered to the results that the agency explicitly

contemplated in its prior orders, and generally served to enhance the efficiency of the market

as intended given the instruments involved.


Shaun D. Ledgerwood

A-3

Support of Client under Investigation. Assisting European firm under investigation for the

suspicious trading of energy futures contracts. Analysis provided to assist client’s response to

the regulatory agency’s inquiries and to assess potential liability under the applicable market

manipulation rules.

Support of Client under Investigation. Assisting a large electricity trader in evaluating its

potential exposure given an agency investigation into the industry’s financial trading activity.

Analysis uses the trader’s internal trading records and publically available data from the RTO

to evaluate if there is potential liability under the FERC’s anti-manipulation rule, as

evaluated using Dr. Ledgerwood’s framework. Where concerning behavior is found, other

client records are reviewed to determine whether the trading at issue served a stand-alone

legitimate business purpose, thus blunting potential assertions of manipulative intent.

Assisting Internal Investigation. Assisting a client concerned that various trades placed over

time could be questioned as “uneconomic” and thus fraudulent under the agencies’ market

manipulation rules. These transactions may have led to various financial benefits and/or out-

of-market payments that could be alleged as potential manipulation targets. The analysis

evaluates the trading data to determine whether intentionally uneconomic trades were in

fact placed into the market and, if so, whether those trades distorted market outcomes. This

analysis will provide the client and its counsel with an understanding of the agency’s

potential “best case” and the associated potential liability involved.

Support of Clients under Investigation. Assisting several electricity traders and trading firms

accused of manipulating the value of Financial Transmission Rights and/or financial swaps

using physical and/or virtual transactions claimed by the agency to be uneconomic. Analyses

center on evaluating the strengths and weaknesses of agency claims using the manipulation

framework and associated action (or inaction) relevant to intent. These analyses use trade

data, agency findings and other available data to evaluate liability and potential harm.

Construction of the agency’s “best case” using the framework allows for the isolation of the

behavior of concern, allowing for a focused analysis of the economic and behavioral evidence

relevant to disproving assertions as to the manipulation’s components and the trader’s intent.

Internal Investigation of Suspicious Trading Activity. Assisted a large physical market

participant in assessing its potential liability with respect to specific trading strategies that

potentially altered the value of related physical index and financial derivatives positions. Dr.

Ledgerwood performed this analysis in cooperation with the client’s external counsel to

provide a clear, consistent understanding of the behavior that could be of concern, as well as

meaningful perspectives as to how the client could improve compliance going forward.

Agency Self Report. Successfully assisted a major financial trading organization in preparing

and submitting a self report concerning legitimate trading behavior that might have been

misconstrued as manipulative. The analysis provided was consistent with the logic of the

analytic framework Dr. Ledgerwood has developed and extensively published. The agency

decided to take no action against the market participant in this case.


Shaun D. Ledgerwood

A-4

Support of Client under Investigation. Assisted a commodity trader’s preparation for

deposition pursuant to an investigation for alleged manipulative activity. Using market data,

the client’s trading book, documents provided by the government, and objective data

including IMs, emails, and public data surrounding the suspected trade events, Dr.

Ledgerwood evaluated the government’s likely “best case” that could be brought against the

trader, as well as potential “best responses” the trader would have in response to refute the

causal elements necessary to prove manipulative intent.

Support of Agency Analysis of a Suspected Market Manipulation. Examined the behavior of

several precious metals traders for specific trading days in which manipulative behavior was

suspected to occur. Work involved analysis of various physical and financial positions in

relation to emails and deposition transcripts provided by the agency sponsoring the work.

Analysis of Potential Carbon Market Manipulations. Assisted studies to evaluate the potential

exposure of the proposed California and EU cap and trade markets to manipulative activity.

Deutsche Bank Energy Trading. Provided analysis and opinions related to the investigation

of an alleged manipulation of Congestion Revenue Rights using physical power trades.

Constellation Energy Commodities Group. Provided the conceptual framework used for the

analysis of Constellation’s alleged uneconomic trading of physical and virtual bids and offers

to benefit the value of financial transmission rights and other related derivatives positions.

Energy Transfer Partners, L.P. (ETP). Worked with staff attorneys and analysts to assist staff

expert witnesses in developing their direct and rebuttal testimonies. Also prepared for and

participated in the depositions of expert witness for ETP.

Amaranth Advisors, LLC et al. (Brian Hunter). Worked extensively with staff’s expert

witness to develop his direct and rebuttal testimonies.

Compliance Seminars and Internal Trade Surveillance

Dr. Ledgerwood regularly works with traders and executives with compliance-related responsibilities,

using his framework as a vehicle for compliance with the market manipulation rules of the various

enforcement agencies and to devise internal surveillance systems to detect and deter manipulative

trading. He also meets with clients and their external counsel in the financial, commodities and

securities spaces to discuss manipulation-related compliance strategies. These increasingly include the

establishment of customized internal trade surveillance systems to monitor for behavior of concern.

Compliance Plans for Physical and Financial Electricity Trading Firms. Assisted in-house

and external counsel in the development of compliance plans tailored to the specific trading

operations relevant to their client’s firms. Plans are to be distributed to and acknowledged by

trading, compliance and risk management personnel to further of the creation/maintenance

of compliance programs consistent with FERC and CFTC manipulation rules.

Seminars for a Multinational Oil Trading Company. Delivered presentations to the traders,

risk managers and compliance personnel of a global energy trading firm. These presentations

focused on avoiding the types of trading behavior that have been and are likely to be found to

be manipulative under U.S. and EU manipulation and antitrust laws.


Shaun D. Ledgerwood

A-5

Compliance-Based Risk Optimization Strategy. Assisted a trading firm in wrestling with the

decision as to how to expand its trading activity by trading a more diverse set of physical and

financial products while remaining certain of compliance with agency anti-manipulation

rules. This assessment considered the client’s efforts to incorporate a trade surveillance

system as an opportunity for expanding its trading platform, such that the costs of the

increased compliance capabilities are potentially offset by the increased profitability of its

trading business.

Seminars for a Large North American Utility. Delivered a set of presentations to the traders

and executives of a large utility with a substantial North American energy trading presence.

Presentations to the traders focused on identifying the types of behavior likely to be deemed

suspicious by the FERC and CFTC. Executive presentations were interactive sessions

designed to strategically detect and deter such behavior.

Compliance Seminars for a Major International Bank. Provided four days of compliance

seminars for the traders, compliance personnel, and executives of a major banking institution

with an extensive commodities trading unit. Focus was on the recognition of behavior likely

to draw regulatory scrutiny under the fraud-based and “artificial” price anti-manipulation

rules in place in the U.S. and Europe; understanding of the potential civil and criminal

liability associated with violations of these rules; and mechanisms for establishing internal

surveillance systems to detect such behavior.

Compliance Seminars for the Trading Staff of a Major European Utility. In cooperation with

counsel from a major European law firm, delivered two presentations to the energy trading

group of a major European utility with an international trading presence. Focused on the

prosecution of uneconomic behavior under REMIT and the MAD, as well as jurisdictional

challenges presented by overlapping and potentially uneven regulatory enforcement regimes.

Compliance Seminar for the Trading Staff of a Large U.S. Electric Utility. Presentation to the

electricity trading group of a large load-serving utility with a national trading presence.

Focused on the definition of “uneconomic” behavior as described by Dr. Ledgerwood’s

framework and recent manipulation cases brought by the FERC and CFTC.

Compliance Seminar for the Trading Staff of a Large Energy and Commodities Firm.

Assembled and delivered two half-day presentations to the trading and compliance staff of a

large U.S.-based trading firm engaged with international operations. Presentation was made

in conjunction with external counsel for the client’s law firm who has significant experience

with manipulation cases before the FERC and CFTC.

Compliance Seminar for Regulators in the Russian Federation. Delivered a 3.5 hour web-

based presentation to NP Market Council and the representatives of several Russian market

participants, ministries and other regulators of the Russian electric markets.

Compliance Seminars for the Trading Staff of a Major U.S. Electric and Natural Gas Utility.

Assembled and delivered two 1.5 hour presentations to the electricity and natural gas trading

groups for a utility with a national trading presence. In addition to discussing the framework

and case precedent, these seminars focused on the need for traders’ responsibility for

compliance given recent agency willingness to prosecute individuals for manipulation.


Shaun D. Ledgerwood

A-6

Market Monitoring and Oversight

While working with the Division of Energy Market Oversight (DEMO) at the FERC, Dr. Ledgerwood

worked with market data and participants and served as a mentor and advisor to managers and staff.

This work included interaction with other offices inside the FERC, as well as contact with state,

regional, federal, and international oversight agencies. He was responsible for providing input on

Notices of Intent (NOIs) and Notices of Proposed Rulemaking (NOPRs), on topics ranging from credit

conditions within the RTOs to price volatility in the natural gas markets and optimality of the capacity

auction mechanisms used in electricity markets.

Dr. Ledgerwood assisted the Commission in reviewing actions by the CFTC which could foreclose the

FERC’s jurisdiction to oversee the markets for wholesale natural gas and electricity. He analyzed the

CFTC’s declaration of contracts traded on Exempt Commercial Exchanges as Significant Price

Discovery Contracts (SPDCs), the effect of proposed CFTC position limits on RTO-related contracts,

and the effect that the passage of proposed legislation extending CFTC’s jurisdiction to all financial

swaps would have upon FERC’s ability to ensure just and reasonable rates.

Assistance to Canadian Electricity Market Monitor. Working for the market monitoring unit

of a provincial regulator to assist it with the implementation of its anti-manipulation rules.

The project involves the application of Dr. Ledgerwood’s framework to various scenarios

cited by the market monitor as potentially in violation of the rules, as well as to develop

processes and procedures to monitor for and detect manipulative activity and enforce the

rules if violations are confirmed.

Assistance to the Italian Natural Gas Market Monitor. Retained by the Italian electricity and

natural gas market monitor to conduct a study to assist that agency in implementing its data

surveillance responsibilities under REMIT and other applicable European competition laws.

The study presented a comprehensive evaluation of the data that is and will be available to

the market monitor, the types of manipulative and/or anticompetitive behavior that the

market monitor should seek to detect, and the tools the market monitor can employ to screen

the available data for the types of behavior identified.

Assessment of Electricity Market Surveillance Methodologies. Helped to identify the

marginal benefits and costs associated with obtaining, compiling, and analyzing data from a

variety of existing and potential new sources available to the FERC. Inherent to this effort

was the development of an understanding as to what linkages may be formed between data of

varying transparencies and value. Made recommendations regarding the priority of the

surveillance efforts that the FERC should pursue.

Analyses in Support of Agency Investigations

In addition to support of the FERC’s Order to Show Cause actions, Dr. Ledgerwood was engaged by

the Commission’s Division of Investigations to assist other investigations of market manipulation

within the natural gas and electric sectors, with an emphasis on the exploration of new theories of

how manipulations can be executed and detected.


Shaun D. Ledgerwood

A-7

Investigation of Market Manipulations within Electricity Markets. Engaged in several public

and non-public investigations involving trading behavior alleged to set physical power prices

to benefit financial positions, the values of which are tied to that physical price. Developed

analytic yardsticks to support the legal standard of proof required by the FERC’s anti-

manipulation statute and established evidence showing intent to manipulate and the negation

of erroneous affirmative defenses concerning the legitimate business purpose of such trades.

Investigation of Market Manipulations within Natural Gas Markets. In investigations of

suspected market manipulations of the wholesale natural gas markets, evaluated whether a

trading entity accumulated a portfolio of financial instruments tied to one or more physical

pricing points. Analyzed how such aggregated positions provided evidence of the incentive

to manipulate, and how subsequent attempts to alter physical prices could be shown to meet

the burden of proof of manipulative intent.

UTILITY REGULATORY POLICY AND RATEMAKING

Utilizing his background in law, economics, and knowledge of the energy sector, Dr. Ledgerwood has

analyzed retail and wholesale power contracts to structure new long-term power agreements and to

assist in resource planning and procurement. His analyses have helped clients plan for future power

plant construction and transmission line improvements and upgrades.

PJM FTR Netting Rule. Assisting client in challenging effort by PJM to declare the current

rule allowing the netting of counterflow and prevailing flow FTRs given underfunding unjust

and unreasonable and replace it with a rule precluding netting and requiring counterflow

holders to repay PJM for the entire target allocation of their counterflow portfolios. Affidavit

points out an erroneous assumption made by PJM that FTR auction prices are based on the

market’s expectations of full funding; if this assumption fails to hold, then PJM’s proposed

rule will lead to an unjust and unreasonable outcome.

PJM FTR Forfeiture Rule. Assisted client in challenging effort to expand the PJM Forfeiture

Rule to other contexts. Affidavit focused on the inefficiencies presented by the existing rule

and its lack of relevance given the active market monitoring performed by the Independent

Market Monitor as a backstop to the rule.

Energy and Ancillary Services (E&AS) Offset. Assisted client in challenging market monitor

decision to calculate the E&AS offset (used to mitigate bids into the PJM capacity market)

using the lesser of bid cost and unit cost. Method penalized generation resources that would

bid into the market below cost in off-peak hours to prevent cycling.

Statistical Analyses in Support of Ratemaking & Cost of Service Studies. Performed statistical

comparisons of historic energy usage/loads to yield regression equations used to “normalize”

test year loads using a comparison of actual and normal values for the dependent variables.

These analyses were then used to evaluate how the test year should be adjusted to account for

rate differentials amongst various service classes.

25-Year Analysis of Power Resources for Generation and Transmission Cooperative.

Performed the economic analysis for a long range plan of generation resources for a large

electric utility in the eastern U.S. Compared the utility’s existing and planned future asset


Shaun D. Ledgerwood

A-8

mix (including coal, nuclear, hydroelectric combined-cycle, combustion turbine, and

purchased power) against planned load growth to determine when new resources would be

required and what type would best suit the seasonal load duration curves in each future year.

Short-Term Load Forecast and Power Supply Analysis for Generation and Transmission

Coop. Used regression analyses to determine the relationship between weather and other

dependent variables upon hourly loads for a G&T cooperative in Texas. Using Excel, hourly

data for weather and prior loads were retrieved and the regressions applied to provide hourly

forecasts up to two weeks from the date the model was run. The future loads were then

compared to available resources and projected power prices to determine whether additional

strips of energy would need to be purchased or whether the supplier would rely upon the

spot market to serve its energy requirements.

Retail Power Contract Analysis and Reconstruction. This engagement involved the analysis

of existing retail agreements between a public power supplier and industrial customers served

under direct retail contracts. Reconstruction of contractual terms and conditions allowed for

a single contract to be used in all similar future agreements.

Load Forecasting. Performed or supervised the performance of long-term economic forecasts

of electric loads for clients including generation and transmission cooperatives, public power

authorities, and associated distribution member systems. Data from sources including NOAA,

the Bureau of Labor Statistics, the Department of Energy, and other demographic data

derived from public and private sources used to forecast electrical energy usage and demands

using econometric techniques. Sensitivity analyses were performed to account for extremes

in weather conditions or in other dependent variables significant to the models forecasted.

ANALYSIS OF COMPETITIVE ISSUES IN REGULATED INDUSTRIES

Dr. Ledgerwood has performed many studies to support petitions before the FERC or other regulatory

authorities. Analyses were submitted to support bond ratings, as market penetration studies in advance

of acquisitions, or as evidence to be used in hearings before legislative bodies.

Support of Expert Testimony Involving Merger. Provided support for expert witness

testimony regarding a proposed merger of two utilities. Evaluated the merger applicants’

expert testimonies concerning the efficiencies to be derived from the merger. Assisted the

development of testimony evaluating concerning these efficiencies.

Analysis of Mitigation of Buyer-Side Capacity Market Bids. Assisted expert in reviewing

NYISO analyses and methodology used to implement its buyer-side Mitigation Exemption

Test. Assisted the development of affidavit supporting the resulting findings.

Analysis of Proposed Construction of Combined-Cycle Plant. Assisted client analysis in

support of determining the cost of new entry associated with a proposed power plant to be

located in the eastern U.S. Analysis included the consideration of market rules, legislation,

and the historic filings of various market participants with respect to the minimum offer

price rule of PJM and other rules used in eastern capacity markets.


Shaun D. Ledgerwood

A-9

Analysis of Entry Strategy into U.S. Market. Prepared a report for a major European

construction firm seeking to determine best strategies for entry into U.S. HVAC construction.

Study began with an overall examination of entry strategy (i.e., acquisition, merger, joint

venture, or new entry) and geographic location, and culminated in a comprehensive entry

strategy. Analysis involved a survey of potential acquisition targets.

Texas Electric Cooperatives Intervention into ERCOT Transition to Competitive Markets.

Submitted filings in support of TEC’s positions concerning the potential for market abuse

within ERCOT given the impending transition to a competitive imbalance energy market

using locational marginal pricing. Commented upon projected market outcomes resulting

from proposed rules by the Texas Public Utility Commission and critiqued comments made

by other market stakeholders engaged in the proceedings.

Economic Evaluation of Alternatives Concerning the Coal Generating Station. Prepared an

economic, environmental, and engineering evaluation of the potential benefits derived from

reopening the Dean H. Mitchell coal plant. This included estimating the construction and

operating costs of alternate repowering configurations to offer alternatives to rebuilding the

plant. Environmental cost estimates were then used to evaluate options which would avoid

the need for new source review. Estimated the profitability of these options using forecasted

fuel prices, weather, and financing costs, resulting in discounted cash flow analyses that were

subjected to sensitivity analyses for robustness.

Economic Impact of Avoided Environmental Costs Caused Construction of Coal Units.

Assisted the economic and environmental evaluation of the expected pollution arising from

operation of the coal-fired Holcomb Plant relative to other alternative energy sources. The

focus was on the benefits of clean-coal technology versus less efficient sources.

Intervention Regarding the Application of Southwestern Public Service for Cost-Based Rates.

This case required a market concentration analysis using the Herfindahl-Hirschman Index to

establish the existence of high market concentration within the Southwestern Public Service

(SPS) control area. Supported and contributed filings before the FERC indicating that said

market concentrations were significantly higher than those within the zonal regions of

ERCOT, wherein competition had commenced. Performed additional analysis using the

Supply Margin Assessment methodologies prescribed by the FERC.

Fuel Plan for Combined-Cycle Power Plant. This analysis examined the demand profile for

natural gas of existing units and planned future units at a combined-cycle generating facility.

The study included regional analyses of the transportation and storage capacity within the

region, analysis of gas supply, projections of the delivered cost of gas, and a discussion of fuel

management issues under the existing contracts. The resulting fuel plan was used to support

the bond rating of the plant owner.

Intervention into the Application of Southwestern Public Service for Cost-Based Rates.

Assisted analysis of market concentration in the SPS control area using HHIs. Supported

filings before the FERC comparing market concentrations in the region to those within the

congestion zones within ERCOT to demonstrate relative market power.


Shaun D. Ledgerwood

A-10

Intervention into FERC Conference on Supply Margin Assessment. Assisted in the

preparation of comments concerning the applicability of the Supply Margin Assessment

proposed by the FERC in replacement of the “hub and spoke” Delivered Price Test based

upon the 1992 U.S. Department of Justice Merger Guidelines. This analysis focused on how

utilities with high market shares operating in highly concentrated markets could pass the

proposed Supply Margin Assessment and thereby avoid mitigation.

Estimation of Future Stranded Costs in Texas. Analyzed financial data and comparable asset

sales to determine the amount of stranded costs that would be incurred by generators in

Texas upon the commencement of retail competition. This report was submitted to the Texas

Legislature and recommended total estimated costs that were ultimately used as the basis for

exit charges applied for affected utilities.

Intervention into the AEP/CSW Merger. Assisted in the analysis of the merger application of

American Electric Power and Central South West Corp. for intervention based upon market

power concerns. The analysis included a review of market concentration provided by AEP

and the production of interrogatories for associated witnesses.

Analysis of Contractual Obligations of Capacity Contracts. Evaluated a chain of contracts

related to capacity ownership of a power plant shared between a municipality and electric

cooperative. Case involved issue of which party was entitled to the revenues from surplus

energy sold available out of the plant given different definitions of excess capacity provided

by different generations of the underlying contracts.

Support of Action for Breach of Contract and Patent Infringement. Assisted witness in

preparing an expert report supporting plaintiff’s claims of contractual breach involving fuel

oil testing technology subject to patent.

Damages for Wrongful Termination. The defendant in this case was sued by a former

employee claiming wrongful termination due to age-based discrimination. Used company

and macroeconomic data to show the employee’s termination was based upon economically

justifiable grounds. Additionally, performed an analysis to determine the damages owed had

evidence of discrimination prevailed.

Damages for Wrongful Termination. In a matter involving wages lost as a result of alleged

disability-based discrimination and the termination of the employee at a major local hospital,

prepared expert testimony on behalf of the defendant regarding lost wages and the

appropriateness of punitive damages.

Damages for Water Utility Condemnation. Provided written testimony regarding the monies

owned for the condemnation of a water utility’s service territory by a municipality. The

analysis included estimating damages for customers affected by the territorial exclusion, as

well as future customers projected to be served. Damages were assessed based upon the

incremental loss of revenues from those customer’s needed to support the debt load of the

utility’s outstanding bond issues, future capital upgrades, and expansions and identifiable

annualized fixed expenses.


Shaun D. Ledgerwood

A-11

Damages for Non-Payment under Utility System Lease. Determined the value of historical

payments owed under the terms of a valid lease for a distribution system owned by a

development authority, wherein the last payments were made in 1994 and the basis for

measuring those payments (a load-measuring meter) was removed that same year. Filled the

time gap using known average monthly per-customer energy demand data and surrogate data

for the number of customers served borrowed from city sewage accounts for the development

authority’s region. An econometric analysis was then used to determine the stream of

payments owed. Client received a settlement for all payments due under the lease.

Damages for Wrongful Termination. Represented Xerox against a claim by a laid-off

employee of gender-based discrimination in her termination. Used company-specific and

macroeconomic data to demonstrate that the layoff was not gender-specific and that damages

were inappropriate. Additionally, performed an analysis to determine the actual damages

owed should evidence of gender-based discrimination prevail. Case tried as Jane Johnson v.

Xerox Corporation, Bill Collins, Rick Hall, Kerry Summers, Ja Ann Alderman and Cathy

Shuffield; Northern District of Oklahoma; Case No. 02-CV-596 H (J). Verdict for Xerox.

Damages for Wrongful Termination. Statistical analyses were used to depict the presence of

gender-based discrimination among group of faculty at a major university. Assisted another

witness in preparing expert testimony for this purpose.

Damages for Wrongful Death. This case required a determination of the future earnings

stream for a popular local television sportscaster killed in a plane crash. Used a discounted

cash flow methodology to project the future earnings stream of the decedent and to provide a

sensitivity analysis of that stream give various alternative assumptions.

Fines for Violations of Environmental Protection Agency Rules. In this engagement, the

client was a defendant to an action by the Environmental Protection Agency seeking to levy

fines for non-compliance with anti-pollution orders. The EPA’s complaint used the standard

economic benefit model (BEN) to compute damages for economic savings from delaying or

avoiding pollution control expenditures. However, the assumptions used in applying this

model were erroneous. Used an updated model to demonstrate that the damages sought by

the EPA were excessive, and furthermore that correction for errors in the model’s application

would reduce the damages to a de minimus level.

ANALYSES IN SUPPORT OF ANTITRUST OR OTHER MARKET-RELATED PROCEEDINGS

Dr. Ledgerwood has performed market analyses to support administrative or legal proceedings for a

variety of market sectors. He has also assisted the preparation of testimony for actions brought under the

Sherman and Clayton Antitrust Acts, as well as under state below-cost sales laws or franchise-related

entry restrictions. His work includes cases on behalf of plaintiffs and defendants to such proceedings.

Damages for Price Squeeze and Vertical Restraint of Trade. Hired as an expert in case

involving an upstream supplier of minerals raising input prices and disrupting the regularity

of supply to injure a downstream service provider to benefit its own downstream subsidiary.


Shaun D. Ledgerwood

A-12

Damages for Antitrust Violations in Discount Grocery Market. Assisted a local grocer facing

predation by large national chain in all phases of analysis and resulting testimony to establish

the predatory behavior and to assess damages associated therewith.

Market Analyses to Support Franchise Applications. Assisted in performing analyses and

production of testimony pertaining to the applications of clients to obtain franchises to enter

the markets for automobiles and outboard boat motors. The presiding administrative

authority approved both applications.

Damages for Antitrust Violations in Beauty Supply Market. In this case, a local beauty supply

company sued a large national chain for miscellaneous violations of the antitrust laws

pertaining to the market foreclosure of plaintiff’s generic brands. Developed the analyses and

testimony required to demonstrate the injury to competition and damages claimed.

Damages for Antitrust Injury in Automobile Transportation Market. The client was a hauler

of automobiles suing an auto auction house for foreclosing access to the auction floor.

Assisted in the study of the product and geographic dimensions of the affected market to

support plaintiff’s case.

Defense of Payphone Service Provider against Antitrust Claims. The client in this

engagement was accused of monopolizing the market for pay phone services by several

independent payphone service providers. Assisted in the creation of analysis and testimony

designed to show the complimentary nature of payphones and cell phones, thereby asserting

the defendant’s lack of market power within the broader product market.

Analysis of Antitrust Applicability to Electric Generators. Supported the preparation of a

report to Texas Electric Cooperatives concerning the existing case law and market potential

for mounting antitrust-based claims against generators perceived to abuse market power in

wholesale power markets.

ANALYSIS OF LIABILITY AND DAMAGES IN TORT, CONTRACTS AND FRAUD

Dr. Ledgerwood has performed many analyses using statistical and discounted cash flow techniques to

determine the amounts owed due to discriminatory behavior or from losses arising from wrongful

termination or wrongful death. Other studies have focused on the damages owed for administrative

penalties or for non-payment of rents owed for lease of a utility system.

Analysis of Disclosures of Material, Non-Public Information. Retained by the regulator to

examine whether non-public information disclosed by the employees of an exchange to an

external broker was material, as measured through trading subsequent to the disclosures and

its price impact on the market. Analysis will make use of event studies to evaluate causation

and measure any price impacts, if discernable.

Evaluation of Applicability of Research Tax Credit. Evaluated if contracts that involved

highly specialized and unique research were “funded” within the context of the applicable

statute. Case required the evaluation of various contractual provisions to determine whether

the risk of non-performance due to the research was borne by the buyer or seller of these

contracts. Prepared expert report for settlement negotiations.


Shaun D. Ledgerwood

A-13

Defense against Allegations of a Pump and Dump Scheme. Assisted expert witness in

preparing a report defending against allegations of client’s assemblage of an over-the-counter

stock position and subsequent “pumping” of the stock through fraudulent statements and

uneconomic trades. Allegations were proven false using trading records of the accused using

market data to demonstrate that the client’s trades did not fit the fact pattern alleged in the

plaintiff’s complaint.

Support of Action for Fraudulent Distressed Asset Sales. Assisted expert witnesses in

preparing reports supporting the IRS in pursuing taxes owed for a fraudulent tax shelter

involving the sale of a portfolio of distressed foreign assets. Assisted the preparation of

rebuttal reports in response to testimonies of expert witnesses on behalf of the defense.

ENVIRONMENTAL POLICY AND REGULATION

Several pieces of federal legislation were introduced during 2008 and 2009 that tasked the FERC with

oversight of proposed environmental allowance markets for carbon and renewable electricity credits

(RECs). Dr. Ledgerwood was asked to study the likely resource requirements that would befall the

Commission in the event the bills were to become law. He has also worked in cooperation with state and

regional efforts to implement and operate markets for renewable allowances.

Analyses Regarding the Federal Regulation of Carbon Emissions. Staff from seven offices

joined to estimate the resources required of the FERC should it be tasked with oversight of

the cash and derivatives markets for carbon emissions as envisioned by the Dingell-Boucher

Cap-and-Trade bill. Role was to assess the resource requirements of the enforcement

provisions of the bill, requiring close coordination with the four Division Directors of the

Office of Enforcement and their staffs. This effort was followed by analysis of the Waxman-

Markey bill; however, the reduced Commission role (with oversight duties limited to the

cash market) shifted focus to likely problems should the bill pass, such as frictions that could

arise if the Commission were required to share oversight duties with the Commodity Futures

Trading Commission.

Coordination with State and Regional Environmental Efforts. To the FERC’s efforts to track

efforts to implement environmental programs, met with several state programs and industry

groups, as well as with regional groups such as the Regional Greenhouse Gas Initiative.

Represented the FERC in a consortium of federal regulatory agencies that advised California

in its implementation and oversight of a carbon cap-and-trade system.

Creation of a Federal Renewable Electricity Standard Program. FERC staff met to estimate

the resources required of the FERC in the event that the Commission was tasked with the

establishment and oversight of a Renewable Electricity Standards program as specified in the

Waxman-Markey bill passed by the House of Representatives in June 2008. This project was

separated into two phases. The first phase, Dr. Ledgerwood contributed as the lead for the

Office of Enforcement, assessed the need for regulations to establish the RES program and the

markets for Renewable Electricity Credits and Electricity Savings. The second phase, for

which Dr. Ledgerwood served as team lead, estimated the resources that would be required

across the Commission to implement and oversee the federal RES program.


Shaun D. Ledgerwood

A-14

VALUATION ANALYSIS

Dr. Ledgerwood has performed analyses to determine the present value of the capital assets of utilities or

other entities on the basis of comparable market based sales, the present value of projected income

derived from the forecasted future life of the assets, and other valuations based upon the principles of

original cost and replacement cost new less depreciation. He has performed sensitivity analyses to

determine the viability of the valuations reached under a variety of scenarios. His work has focused on

matters in the electrical, natural gas, telecommunications, and water/wastewater industries.

Valuation of a Rural Wireless Telephone Company. Team performed a valuation of a 20

percent interest in a Rural Service Area to be used for debt securitization. The valuation used

discounted cash flow, comparable sales, and spectrum valuation methodologies to provide a

band of values within which the interest could sell. This analysis was updated in 2015.

Evaluation of Privatization of University Utility Systems. Led the team that produced a

feasibility study to determine the potential value of and market interest in seven utility

systems owned by a public university. Sensitivity analyses were used to produce a range of

discounted cash flow models of the forecasted revenue streams produced by each system.

These outcomes were then compared to the relative capital improvements and operating

expenditures needed to modernize and run the systems in perpetuity. Significant

consideration was given to the likely cost of capital required of potential bidders, which

ultimately constrained the likelihood of successful privatization for several of the systems

absent the payment of large subsidies by the university.

Evaluation of Procurement of the Sugar Creek Generating Station. Led a cost-benefit analysis

of the benefits derived from the purchase of the Sugar Creek power plant by Northern

Indiana Public Service Company. The analysis used recent sales of similar plants,

construction costs, historic plant characteristics, future sales into MISO and PJM, and future

difficulty in building other fossil projects.

Valuation of the Oklaunion Generating Station. Team performed a valuation of a 7.8 percent

interest in a coal-fired power plant targeted for acquisition. The valuation used a discounted

cash flow methodology supported by fuel price projections and a variety of sensitivity

adjustments to key variables.

Valuation of Indiana-American Water Company Inc. This engagement involved a rate case

intervention to determine the fair value the water and wastewater assets of Indiana-

American Water Company (Cause No. 42029, Indiana Utility Regulatory Commission).

Performed a market-based valuation of these assets to evaluate whether the client should

protest the asset valuation proposed by the company in this case.

Investigation of the Rates and Charges for Electric Service Provided by NIPSCO. This was

the first “inverse rate case” performed in the history of Indiana. It was designed to challenge

the existing rates of Northern Indiana Public Service Company (Cause No. 41746, Indiana

Utility Regulatory Commission). In direct testimony, Dr. Ledgerwood used a market-based

valuation of the company’s generation, transmission, and distribution assets to challenge

valuations asserted by the Company’s economic experts. In rebuttal, Dr. Ledgerwood directly

critiqued the company’s expert’s analyses based upon replacement and reproduction costs.


Shaun D. Ledgerwood

A-15

Valuation of the Assets of Citizen’s Utilities. This was an acquisition feasibility study of the

water, wastewater, electrical, natural gas, and telecommunications assets of Citizen’s Utilities

sited in Arizona, California, Colorado, Hawaii, Illinois, Indiana, Louisiana, Ohio

Pennsylvania, and Vermont. Assisted in the valuations performed using discounted cash flow

and comparable sales of other utility assets.

Valuation of the Assets of North Carolina Power. In a feasibility analysis for acquiring a

portion of the transmission and distribution system of North Carolina Power, assisted

valuation analysis using replacement cost less depreciation and performed an economic

analysis centered on ratepayer effects and sensitivity studies pertaining to the valuation.

ADDITIONAL PREVIOUS EXPERIENCE

ATTORNEY AT LAW (OKLAHOMA BAR ASSOCIATION NO. 14266) – 1990 TO PRESENT

Dr. Ledgerwood joined the Oklahoma Bar Association in October 1990 and has remained an active

member. He uses his legal training extensively in his economic consulting roles, working with attorneys

in a manner designed to maximize complementarities such that the economic arguments he makes are

fully consistent with the attorney’s legal strategies and arguments required to address the client’s needs.

His legal background is particularly useful in cases involving allegations of market manipulation, which

often require the interpretation of various physical and financial contracts to establish net positions and

explain the interactions that can arise among such instruments.

PROFESSORIAL DUTIES – 1985 TO PRESENT

Affiliated Faculty, Georgetown University, Georgetown Public Policy Institute, 2010-2012

Adjunct Professor, Univ. of Oklahoma Dept. of Advanced Programs, 2001-2011

Adjunct Professor, University of Oklahoma Department of Economics, 1995-2011

Adjunct Professor of Law, University of Oklahoma School of Law, 1995-1996

Graduate Teaching Assistant, University of Oklahoma Department of Economics, 1989-1995

Undergraduate Teaching Assistant, Univ. of Oklahoma Dept. of Economics, 1985-1986

PUBLICATIONS

“Enron’s California schemes haunt regulators 15 years later,” coauthored with Gary Taylor.

Energy Risk Magazine, January 2016.

“DC Circuit Clarifies That SEC Need Not Show Market Impact in Manipulation Cases,”

coauthored with Paul Hinton. Pending publication in the Securities Journal, Fall 2015.

Market Power and Market Manipulation in Energy Markets: From the California Crisis to the Present, coauthored with Gary Taylor, Peter Fox-Penner and Romkaew Broehm. PUR,

Inc., May 2015.

“The Chasm between ‘Following the Rules’ and Manipulation.” Energy Risk Magazine,

February 2015.

“An Economic Evaluation of ‘Funding’ for Research Tax Credits,” coauthored with Darrell

Chodorow, Tax Notes, v. 144, n. 13, pp. 1593-1602, September 2014.


Shaun D. Ledgerwood

A-16

“How to think about market manipulation,” coauthored with Mike Cragg, Financier

Worldwide, May 2014.

“Using Virtual Bids to Manipulate the Value of Financial Transmission Rights,” coauthored

with Hannes Pfeifenberger, The Electricity Journal , v. 26, issue 9, pp. 9-25, November 2013.

“Uneconomic trading and market manipulation.” Energy Risk Magazine, p. 32, July 2013.

“A Framework for Analyzing Market Manipulation,” coauthored with Paul Carpenter,

Review of Law & Economics. Volume 8, Issue 1, Pages 253–295, ISSN (Online) 1555-5879,

DOI: 10.1515/1555-5879.1577, September 2012.

“A Comparison of Anti-Manipulation Rules in U.S. and EU Electricity and Natural Gas

Markets: A Proposal for a Common Standard,” coauthored with Dan Harris, Energy Law

Journal v. 33, No. 1, pp. 1-40, April 2012.

“Rummaging through the Bottom of Pandora’s Box: Funding Predatory Pricing through

Contemporaneous Recoupment,” coauthored with Wesley Heath, University of Virginia

School of Law: Virginia Law and Business Review, v. 6, No. 3, pp. 509-568, April 2012.

“Completing the Puzzle: Defining Manipulation under REMIT,” coauthored with Dan Harris,

The Brattle Group, January 2012.

“Triggers and Targets: The Anatomy of Market Manipulation,” on SSRN, July 2011.

“Market Manipulation without Market Power,” published in Law360, June 2011.

“Protecting Market Integrity by Defining Market Manipulation,” coauthored with Dr. Dan

Harris, Dr. Bin Zhou, and Dr. Pinar Bagci, The Brattle Group, June 2011.

“Losing Money to Increase Profits: A Proposed Framework for Defining Market

Manipulation,” coauthored with Mr. Gary Taylor, Dr. Romkaew Broehm, and Dr. Dan

Arthur, The Brattle Group, March 2011.

“Screens for the Detection of Manipulative Intent,” available at SSRN, December 2010.

Editor, “2008 State of the Markets Report” and contributing author to Section 6: Natural Gas

Capacity Release and Electricity Transmission Reassignment, Division of Energy Market

Oversight, Office of Enforcement, Federal Energy Regulatory Commission, August 2009.

"Deregulated Electric Markets a ‘Target Rich’ Area for Antitrust Litigation," The Competitive

Edge, Vol. 1, No. 1, C. H. Guernsey & Company, February 1999.

"An Economic Model of Illegal Profit Deterrence and Corporate Punitive Damages," co-

authors: Dr. William Clark and Dr. David Huettner, 9 J. Forensic Economics 325, Fall 1996.

PRESENTATIONS

Panelist, “A Framework for Understanding Market Manipulation” EBA 2nd Annual Canadian

Forum (Calgary, Alberta), October 2015.

Panelist, “CFTC and FERC Enforcement Trends and Developments,” Reed Smith Energy and

Commodities Conference (Houston, TX), September 2015.


Shaun D. Ledgerwood

A-17

Panelist, “What Trading Organizations Can Do to Proactively Manage Swaps Manipulation

Risks,” EUCI Conference on Financial Transmission Rights (Washington, DC), July 2015.

“Market Power and Market Manipulation in Energy Markets: From the California Crisis to

the Present,” UBS Securities, LLC Conference Call (webinar), June 2015.

Panelist, “Market Manipulation in Energy Markets,” Georgetown Center for Business and

Public Policy Symposium on Market Power and Market Manipulation in Energy Markets

(Washington, DC), May 2015.

“A framework for the analysis of market manipulation,” 30th Annual EISG Conference

(Toronto, Ontario), May 2015.

Panelist, “The State of Market Manipulation: Oversight & Enforcement in U.S. Electricity

Markets,” Gulf Coast Power Association’s 29th Annual Spring Conference and Exhibition

(Houston, TX), April 2015.

“Differentiating a Legitimate Hedge from a Target for Manipulation,” EUCI Conference on

Integrating Energy Trading, Valuation, Market and Credit Risk Management (Denver, CO),

February 2015.

“The Chasm between Following Market Rules and Fraud,” Stanford University Program on

Energy and Sustainable Development (Palo Alto, CA), December 2014.

Panelist, “A Framework for Market Manipulation Compliance and Enforcement,” Dorsey &

Whitney Enforcement Conference (Washington, DC), November 2014.

“A Framework for Surveillance and Detection of Market Manipulation,” Council of European

Energy Regulators Conference on Integrity and Transparency of Wholesale Energy Markets:

WEM Task Force (Milan, Italy), November 2014.

Panelist, “Do Energy Market Speculators Manipulate Oil Prices?,” University of Oklahoma

Price College of Business Faculty and Industry Roundtable (Norman, OK), October 2014.

Panelist, “FERC Enforcement: Cases of First Impression,” Cadwalader, Wickersham & Taft

(Houston, TX), September 2014.

Panelist, “Differentiating a FTR Hedge from a Target for Manipulation,” EUCI Conference on

Financial Transmission and Congestion Revenue Rights (Washington, DC), July 2014.

Keynote address, “A Framework for Market Manipulation Compliance and Enforcement,” Schiff Hardin LLP Annual FERC Briefing (Chicago, IL), May 2014.

Panelist, “A Framework for Market Manipulation Compliance and Enforcement,” presented

at Akin Gump’s Conference on Enforcement (New York, NY), May 2014.

Chair & presenter, “Uneconomic trading as transactional fraud: EU compliance lessons from

across the pond,” a Master Class presented at the MarcusEvans Benchmarking Market

Integrity and Transparency in Energy Trading Conference (London, UK) , April 2014.

“Uneconomic trading as transactional fraud: EU compliance lessons from the US,” presented

to the Italian Association of Energy Traders (Milan, Italy), April 2014.

“Understanding Market Manipulation,” presented at EUCI’s conference on Best Practices in

Decision and Risk Analysis (Philadelphia, PA), April 2014.


Shaun D. Ledgerwood

A-18

Panelist, “Market Manipulation and Antitrust: Complements or Substitutes?” presented to the

New York City Bar Association Antitrust and Trade Regulation Committee (New York, NY),

February 2014.

“Market Manipulation and Antitrust: Complements or Substitutes?” presented to the

partnership of a major international law firm (New York, NY), February 2014.

Panelist, “A Framework for Market Manipulation Compliance and Enforcement,” presented

to the Energy Bar Association’s Enforcement Outlook (Houston, TX), January 2014.

Panelist, “A Holistic Approach to Manipulation Compliance and Enforcement,” presented at

the LSI Buying & Selling Electric Power in the West Conference (Seattle, WA), January 2014.

“A Holistic Approach to Market Manipulation Compliance,” presented to the traders and

executives of a major U.S. utility trading electricity and natural gas, October 2013.

“A Unifying Approach to Market Manipulation Compliance,” presented to the executives,

compliance officers and senior trading staff of a diversified international commodities trading

firm (Houston, TX), September 2013.

“A Holistic Approach to Market Manipulation Compliance,” presented to the energy trading

and compliance staff of a multinational financial institution (New York, NY), August 2013.

“Using Virtual and Physical Trades to Manipulate the Value of Financial Transmission

Rights,” EUCI FTR Conference (Washington, D.C.), July 2012.

“A Unifying Approach to Market Manipulation Compliance,” presented to the partners of a

national law firm interested in applications to energy, commodities and securities trading

(Washington, DC), July 2013.

“An Example of Competing Interests in a Cross-Subsidization Program: PJM Capacity

Market,” delivered at the Platts European Gas and Power Trading Conference (London, UK),

June 2013.

“A Unifying Approach to Market Manipulation Compliance,” delivered at the Platts

European Gas and Power Trading Conference (London, UK), June 2013.

Panelist, “Market Manipulation post Hunter vs. FERC: A Framework for Unified Analysis,” a

presentation delivered at the Harvard Electricity Policy Group’s 71st Plenary Session

(Calgary, AB), June 2013.

“A Perspective on Market Manipulation,” an anti-manipulation compliance seminar

presented to the traders and compliance staff of a major U.S. electricity trading company,

May 2013.

“A Holistic Approach to Market Manipulation Compliance,” presented to the partners of an

international law firm interested in applications to energy, commodities and securities

trading (Brussels, Belgium), May 2013.

“A Holistic Approach to Compliance: REMIT and the MAD,” an anti-manipulation

compliance seminar presented to the traders and compliance staff of a major European energy

trading company, April 2013.


Shaun D. Ledgerwood

A-19

A Holistic Approach to Market Manipulation Compliance,” presented to the partners of a

national law firm interested in applications to energy, commodities and securities trading

(Washington, DC), April 2013.

Panelist, “ISO/RTO Financial Products and Recent Trends in FERC Enforcement,” a

presentation sponsored by the Energy Bar Association Electricity Regulation Committee and

the Compliance and Enforcement Committee (Washington, DC), April 2013.

“Uneconomic Trading in Energy Markets: A Framework of Market Manipulation,” Penn

State Faculty Colloquy (State College, Pennsylvania), April 2013.

“Market Manipulation in Wholesale Electricity Markets,” made to NP Market Council

(regulator of wholesale electricity markets for the Russian federation), March 2013.

“An Economic Approach to FERC Anti-Manipulation Compliance,” an anti-manipulation

compliance seminar presented in separate sessions to the compliance staff and traders of a

major commercial energy trading company, March 2013.

“Market Manipulation Cases in U.S. Wholesale Electricity and Natural Gas Markets,” Bird &

Bird Seminar on Market Abuse (London, UK), February 2013.

“Market Manipulation in Wholesale Electricity and Natural Gas Markets,” presented for the

West LegalEd CLE program (webinar), January 2013.

“Using Virtual Bids to Manipulate the Value of Financial Transmission Rights,” USAEE/IAEE

Annual Conference (Austin, TX), November 2012.

Panel Moderator, “Recent Market Violations and Their Repercussions on Compliance and

Counterparty Risk,” EUCI Conference on Credit & Counterparty Risk (Baltimore, MD),

November 2012.

“An Economic Approach to FERC Anti-Manipulation Compliance,” an anti-manipulation

compliance seminar presented in separate sessions to the gas and electricity trading groups of

a major utility’s trading group, November 2012.

Panelist, “Meeting Cross-Border Requirements: Understanding Changing International

Regulations and Ensuring Global Compliance,” ACI Energy Trading Compliance and

Regulatory Enforcement Conference (Washington, DC), November 2012.

Panel Moderator, “FERC’s Anti-Market Manipulation Rule and the Impact on Physical and

Virtual Market Participants,” EUCI FTR Conference (Houston, TX), August 2012.

“An Analytical Framework for Anti-Manipulation Compliance, Monitoring, and

Enforcement,” Marcus Evans Conference on European Energy Market Integrity: REMIT &

Regulatory Compliance (London, UK), April 2012.

“Energy Trading Compliance and Enforcement Post Dodd-Frank: A Framework of Market

Manipulation,” an anti-manipulation compliance seminar for law firms (Washington, DC),

January and March 2012.

“The Search for a Uniform Definition of Market Manipulation: A Proposed Framework,” A

Conclave Concerning the REMIT (London, UK), October 2011.


Shaun D. Ledgerwood

A-20

“Manipulation of FTR Markets,” presented to EUCI’s Conference on Financial Transmission

Rights, (Houston, TX), July 2011.

“Triggers and Targets: A Framework for Analyzing Market Manipulation,” presented to the

American Gas Association (Washington, DC), July 2011.

“A Framework for Defining Energy Market Manipulation,” presented to the Bracewell &

Giuliani LLP Energy Compliance Network (Houston, TX), May 2011.

Panelist, “Market Manipulation,” presented at Skadden’s 6th Annual Enforcement and

Compliance Conference with John Estes III and Joseph Kalt (Washington, DC), February

2011.

“The Impact of FERC’s Anti-Manipulation Statute on FTRs and Related Positions,” EUCI’s

Conference on Financial Transmission Rights (Alexandria, VA), July 2010.

“The Correlation of Energy Prices to Long-Term Interest Rates,” delivered to the PACAF

Command Financial Specialists Roundtable in coordination with the University of Oklahoma

(Okinawa, Japan), August 2008.

“Avoiding Dutch Disease: Diversification of the Azeri Economy Given the Discovery of

Caspian Oil Reserves,” Training Session conducted with USAID and the Univ. of Oklahoma

and presented to member of the Oil Ministry of Azerbaijan (Baku, Azerbaijan), August 1999.

"Poland: A Case Study of Industrial Development within Eastern European Nations,"

delivered at the Southwest Economic Association Conference (New Orleans, LA), March

1993.

PROFESSIONAL AFFILIATIONS

Oklahoma Bar Association

American Bar Association

American Economic Association

American Finance Association

Federal Energy Bar Association

Phi Beta Kappa


Appendix B

List of Exhibits


List of Exhibits

Excel Workbooks

Excel: "Shadow Prices Profits Screen and Financial Leverage Analysis.xlsx"

Excel: "Transmission Congestion Analysis.xlsx"

Excel: "Sierra Snowpack Analysis.xlsx"

Excel: "VT Bids May 2011_All Nodes.xlsx"

Excel: "LMP and Congestion Prices.xlsx"

Excel: "Virtual Profits.xlsx"

Excel: "May Virtual and CRR Daily Profits.xlsx"

Excel: "Zero Neg LMP Bid Analysis.xlsx"

Excel: "May Virtual Bids vs. LMPs.xlsx"

Excel: "Virtual_Duration_Analysis.xlsx"

Excel: "CRR and VB Profits.xlsx"

Excel: "Add_Removal_VT_February June 2011.xlsx"

Excel: "Interacting_Node_Analysis.xlsx"

Excel: "Pivot Table Analysis of May CRR Positions.xlsx"

Excel: "Upstream and New Melones Position.xlsx"

STATA/R Code

STATA program: "Process CAISO Public Bids"

STATA program: "Process Shadow Prices"

STATA program: "CRR and VB Profits"

STATA program: "Virtual Duration Analysis"

STATA program: "Process LMP Data"

STATA program: "Interacting Node Analysis"

STATA program: "Process CAISO Convergence Bids Data"

STATA program: "Process Transmission Data"

R Program: "Virtual Profits"

Sources

OASIS: "[enterdate]_PB_CB_PUBLIC_BIDS_N.csv" (19 files)

OASIS: "May 2011 Shadow Prices.csv"

OASIS: Scheduled Transmission 2/2011 8/2011: "[month]_TRNS_USAGE_DAM_20160121_12_30_21_v1.csv"

OASIS: "May 2011 TX Interface Usage.csv"

EV: "CAISO_DAH_FEB_JUL.csv," "CAISO_HAH_FEB_JUL.csv," and "CAISO_RTH_FEB_JUL.csv."

EV: "Convergence Bid Data_MELONES.csv"

EV: "Convergence Bid Data_122954.csv"

EV: "End of May 2011 LMP_All.csv"

EV: "Melones LMP_All.csv"

Client data: "MP Bids.xlsx"

Client data: "VT Bids March 2011.xlsx," "VT Bids April 2011.xlsx," "VT Bids May 2011.xlsx," "VT Bids June 2011.xlsx."

Client data: "FERC Request 20111216 Q7.csv"

Client data: "May 2011 VT Awards.xlsx"

California Data Exchange Center: "Black Springs Snow Course Data.csv"

"date_hour_mapping.xlsx"

B-1
