francis a. longstaff and ashley w. wang august 2004 reporter: you-cheng luo

19
Electricity Forward Prices: A High-Frequency Empirical Analysis FRANCIS A. LONGSTAFF and ASHLEY W. WANG AUGUST 2004 Reporter: You-cheng Luo

Upload: rosaline-norman

Post on 25-Dec-2015

216 views

Category:

Documents


0 download

TRANSCRIPT

Electricity Forward Prices: A High-FrequencyEmpirical Analysis

FRANCIS A. LONGSTAFF and ASHLEY W. WANG

AUGUST 2004

Reporter: You-cheng Luo

OutlineIntroductionDataEmpirical TestsTime Variation in Forward PremiaVolatility AnalysisConclusion

IntroductionThese types of derivative contracts are

rapidly growing in importance as both financial risk management tools for hedgers as well as liquid investment vehicles for energy trading firms. Since electricity is not storable, the standard no-arbitrage approach to modeling forward prices cannot be applied.

Now we focus on the question of how electricity forward prices are related to expected spot prices

PJM marketWhat’s PJM market?

Data

Data

Data

Data

Forward PremiaThe forward premium can now be defined as

FPit = Et [Fit − Si,t+1]

Fit : the electricity forward price observed on day t for delivery during hour i of day t + 1

Si,t+1 :the spot price for hour i of day t + 1.

Empirical Tests

Empirical TestsBessembinder and Lemmon show that the

forward premium can be expressed in reduced form as a simple linear combination of the variance and skewness ofthe endogenous spot price

Time Variation in Forward Premianote that the realized or ex post forward

premium can be expressed as

where represents the unexpected component of the realized forward premium and is orthogonal to information at time t

Time Variation in Forward PremiaMost asset pricing models have in common the

feature that risk premia are directly related to measures of risk, typically expressed in terms of second moments

VSit :the conditional variance of unexpected price changes

VLit :the conditionalvolatility of unexpected changes in load

VRit :the conditional volatility of unexpected changes in revenue

Time Variation in Forward Premia

Time Variation in Forward Premia

Volatility AnalysisUnder the null hypothesis that the forward

premium FPit equals zeroFit = Et [Si,t+1]

Consequently, all moments of the left-hand and right-hand sides of equation should be equal

Volatility Analysis

Volatility Analysis

Conclusion1. the pricing of electricity forward contracts in the

dayahead forward market and their relation to the corresponding spot prices

2. We find that there are significant forward premia in electricity forward prices and forward premia are negatively related to price volatility and positively related to price skewness.

3. We find that each of these risk measures plays a significant role in explaining the forward premium these results demonstrate that electricity forward premia vary significantly through time.