1

INTENSIVE PATENT VALUATION

Karl F. JordaDavid Rines Professor of Intellectual Property Law & Industrial Innovation

Director, Kenneth J. Germeshausen Center for the Law of Innovation & Entrepreneurship

Franklin Pierce Law CenterTwo White Street, Concord, NH 03301 USA

Intellectual Property Finance & Valuation SeminarIncreMental Advantage

New York CitySeptember 18, 2007

2

INTENSIVE PATENT VALUATIONPART I

A. IntroductionLive in “Golden Age” for IPRs• Patent filings and issuances are skyrocketing• Talk of patent “revolution,” “explosion,” “frenzy”• Anything under the sun that is made by man” is patentable• Courts, Congress, Justice Department — pro IPRs• Corporations built on patented technologies• Motto: Innovate or perish• Value of IPRs for securing exclusivity — simply invaluable• Royalties for licensing IPRs in 2002: $150 billion• Over $1 billion for some companies• Universities jumped on bandwagon• Getting patents, concluding licenses, collecting royalties

Ronald Myrick, formerly of General Electric, put it this way: “The attraction of IP is simple; it’s at the forefront of the technology that’s driving the world and IP is one of the unique entities in the law where you’re actually creating assets.”

3

B. The Primary Objective of IP Protection

• Providers of IP consultancy services focus in particular on– “IP value extraction,”– “IP monetization,”– “Maximizing royalties.”

• However, this overlooks that much greater gains can be realized from protection of, and exclusivity for, a company’s products and processes.

Self exploiting via manufacturing and selling can be much more lucrative than licensing-out.

In an exam paper, a student of mine put it this way: “Licensing is not where the big bucks are. Patentees can most often get the best value out of their patents by commercializing and marketing the technology themselves. Licenses only happen when patentees for whatever reason cannot fully exploit patents themselves. Also, when you license technology you often create a competitor.”

4

The Primary Objective of IP Protection (continued)

• Market exclusivity under IP protection is the primary objective for all but a few of the biggest corporations.

• Entrepreneurs, start-ups, small and middle-sized companies would not last absent IP protection and market exclusivity.

• Such companies are completely dependent on IPRs for their technologies.

• Licensing their IPRs would set up competitors —good reason behind the general reluctance to license-out.

• And pharmaceutical and biotech companies need IPRs and market exclusivity to protect their enormous R&D investments.

5

The Primary Objective of IP Protection (continued)

• As is well known, licensing normally carries little risk but also little reward.

• Royalty income at prevailing rates amount to at best a small percentage of net sales of licensed product, while markups on products sold under IP protection could be much higher, by multiples, and may reach a 1000% or more.

• This is another reason for the innate reluctance to license-out IPRs.

• 97% of all patents are not licensed for this reason or because the technology they cover is not useful, feasible or marketable (Emmett Murtha, an ex-IBM and former LES President).

6

The Primary Objective of IP Protection (continued)

• Marshall Phelps, Microsoft’s new Corporate Vice President for Intellectual Property (ex-IBM), had this to say on the subject:

• Our emphasis is first and foremost about the quality of innovation and then the subsequent and logical protection of that innovation. We will be investing some $6.9 billion in R&D annually. It would be foolish if we did not do everything we could to protect the output of such a large investment…. This type of investment is going to generate a healthy stream of intellectual property. As with others in the IT industry, our most important IP strategy is to protect our innovations and our substantial investment in the area of R&D, through IP laws and, in some instances (!) to seek compensation for this investment through licensing to third parties or engaging in technology transfers with other innovators. (Emphasis added.)

• Joe Siino, IP VP of Yahoo, and David Simon, Chief IP Counsel of Intel chimed in by stating respectively:

• While it is true that IP strategy should be tied to the business strategy, it’s risky to treat IP as another profit center. A company’s most valuable IP (on its core business) will never be licensed.

• Our head is at being a successful business rather than using the IP department to make money. I’d rather have us see a lot more product — which will contribute a lot more money to the bottom line — than to maximize my assets trying to get people to take a license.

7

The Primary Objective of IP Protection (continued)

• My former employer, CIBA-GEIGY Corp. realized $3.5 billion in profits — yes, profits — from producing and selling Atrazine, a corn herbicide, over a 17-year period. This period coincided with the patent life, inasmuch as EPA permission to sell and patent issuance occurred in the same year. Had CIBA-GEIGY licensed the Atrazine patent, which they refused to do for obvious reasons, the royalty income would at best have been merely a small fraction of the profit that was garnered.

8

The Primary Objective of IP Protection (continued)

• Interestingly, from a chart of descending willingness to license out, it is clear that licensing for royalties is last. Willingness goes down from licensing a subsidiary, an associated company, in a joint venture, for cross licensing or for royalties.

• Money consideration comes last and other quid pro quos, e.g. cross-licenses under licensee’s patents covering products that can be made or sold profitably are preferred or insisted upon.

• In fact, obtaining such cross-licenses as quid pro quos rather than accepting mere royalties is one of the significant recent trends in licensing/technology transfer.

9

C. Royalty-Free Licenses

• Would IP “value extraction” and “monetization” advocates ever contemplate or recommend royalty-free licensing? Very doubtful!

• But there is significant royalty-free licensing. Makes eminent business sense.• There is indeed great virtue in royalty-free licensing in terms of good will and

good relationships, bringing about increased sales of goods and supplies and hence larger market share.

• Examples:

At one point in my career at CIBA-GEIGY Corp. (now Novartis), I prepared over 20 royalty-free non-exclusive licenses to carpet manufacturers under patents I had obtained in the U.S. and Canada on an important improvement in tufting carpets. CIBA-GEIGY was not in the business of manufacturing and selling carpets but dyestuffs. CIBA-GEIGY had no intention to practice this tufting method itself. Licensing was the best alternative. Rather than doing it for royalties, we did it for free with the expectation that this would induce grateful carpet manufacturers to buy more dyestuffs from CIBA-GEIGY. Carpet manufacturers were pleased to be licensed for free to practice an important new technique for tufting carpets.

10

Royalty-Free Licenses (continued)

A more recent example is the royalty-free licensing by Iridian Technologies of iris-scan patents.

Iridian owns a broad patent and another two dozen patents on iris-recognition software, which is able to accurately identify people at airport security or automated teller machines.

They licensed these patents also on a royalty-free basis after deciding that the “upside of software sales was greater than the downside of collecting royalties.”

They won contracts with Schiphol Airport and the UAE government and expected other big government contracts.

Iridian will “end up getting a lot of business” per US Today of August 15, 2005.

This case also shows that giving away valuable patent rights for free can be a savvy business move.

11

Royalty-Free Licenses (continued)

In the field of licensing law and practice there are other instances of, or occasions for, granting free licenses.

• Interference settlement agreements.• Grant-back provisions in license agreements often are royalty-free.• Releases of patent rights to employees, where a corporation or university

has no interest in the employee’s invention.• Hybrid patent/trade secret licenses with royalty based on the trade secrets.• Corporations owning patents that would be infringed by university research

grant the university a royalty-free license.• In standard setting situations, assurances by patentees to license on

royalty-free terms.

The conclusion is inescapable that royalty-free licensing of valuable IP rights in preference to royalty-bearing licenses, is savvy business strategy conducive to creating good will and establishing or cementing good relationships, with attendant increases in market share.

12

D. General Value Considerations

• As regards IP valuation and royalty settings in licensing, many considerations and factors play a significant role and cannot be ignored.

• Vastly different values may reside in broad pioneering or basic patents versus narrow improvement or picture patents, that it is easy to design around.

• For competitive reasons, patent applications are filed very early after conception and reduction to practice and hence have little experimental support and cover technology in a mere embryonic stage. That is entirely different from a patent that covers a successful commercial product or process.

• There is a significant difference in value between a patent that is strong and enforceable and a patent that is weak and of questionable enforceability. Also, a patent that has been upheld in court as valid, will significantly gain in value.

• And of course values may vary widely from industry to industry.• Also, in most patent transactions a package of patents (issued patents,

pending applications, rights to apply for patents) is the merchandise, but the purchase price or royalty is not cumulative.

• Due diligence is indispensable in IP transactions which may take weeks or months and without which one may “buy a lawsuit” rather than an asset.

13

General Value Considerations (continued)

• Contrary to common assumptions, it is not true that– licensors can charge what the traffic will bear– licensors can recoup their R&D expenses, the cost of the development of a

technology is a big factor,– there are royalty standards within each industry to go by, etc.

• Indeed, there is a limit to what a licensor can charge and most often it is the licensee’s economics, not the licensor’s, that controls the royalty determination (Gordon Smith).

• And isn’t there a 25/75% rule? Isn’t licensee entitled to the lion’s share because of the greater risk he/she carries, especially with less-than-fully developed technology?

• Above all, when it comes to royalties less is more and greed never pays off.• In my corporate experience, several agreements went South because the royalties

were too high, the profitability was not there and the deals could not be sustained in the end.

• On several other occasions, agreements had to be renegotiated for lower royalties for the same reasons.

In other words, they were not viable win/win license agreements to begin with.

So much for maximizing the “royalty stream!”

14

General Value Considerations (continued)

• Actually, the cost to licensor of the development of the technology is not a factor at all. These R&D costs are sunken expenses expended by the patentee/licensor whether or not it is licensed and, therefore, should not be considered in arriving at a suitable royalty. That is to say, the public’s interest in buying a product is essentially unrelated to the cost of developing it (Tom Arnold, Martin Landis, Gordon Smith).

• Anent royalty standards in industry and the figures given as industry averages, John Romary (Finnegan Henderson) called such average royalty rates “folklore” and “suspect as a royalty-rate guide.”

• He also states that these figures are based on the net sales price of a non-exclusive license and that a “20 to 50 per cent premium” and “as much as a 300 per cent premium…in the pharmaceutical field” may be a reasonable average for an exclusive license.

15

E. 100 Royalty Setting Factors

According to Tom Arnold there are 100 factors to be considered by both sides in licensing negotiations. They are tabulated and discussed in Appendix C, 1998 Licensing Law Handbook, Clark Boardman, p.295.

• This tabulation is a handy checklist• Not all factors play a role in a given technology license• They are grouped under the rubrics of

– Intrinsic Quality– Protections and Threats of Protection– Market Considerations– Competitive Considerations– Values Brought to the Table by the Licensee– Financial Considerations– Particular Risk Considerations– Patent Portfolios– Legal Considerations– Government Regulatory Considerations

16

100 Royalty Setting Factors (continued)

• Among the most important factors are– the state of development of the subject technology (embryonic and untested v.

tested and commercial),– the strength of the IP rights — “be it patent, trade secret, or both” — (solid v.

weak, easy to design around vel non),– the degree of exclusivity (exclusive v. sole, semi- or co-exclusive v. non-

exclusive) and– geographic scope.

Note reference to “trade secrets”

The amount of, and value added by, trade secrets covering essential collateral know-how is indeed important:

“Trade secrets are a component of almost every technology license…(and) can increase the value of a license up to 3 to 10

times the value of the deal if no trade secrets are involved.” (Melvin Jager).

17

100 Royalty Setting Factors (continued)

According to Martin Landis (of AT&T)“The patent royalty negotiated by the parties is determined largely by the strength of the patent itself and only secondarily, by the value of the technology. For example, a U.S. patent on a commercially significant technology may only command a low royalty rate because the most pertinent prior art, an obscure dissertation gathering dust on a library shelf in a small town, teaches the thrust of the invention, yet was never considered by the Patent Office which issues the patent.” (Journal of Proprietary Rights, August 1991)

18

F. 20 Steps For Pricing A Patent

IP monetization folks start talking about market, cost and income approaches almost from the outset on the stated or unstated assumption that patents are presumed to be valid, meaning that a patent is a patent is a patent and by definition a “Rembrandt in the Attic.”

Nothing could be further from the truth!

According to a 2004 publication of the AICPA (American Institute of Certified Public Accountants), entitled “20 Steps for Pricing a Patent,” choosing one of the traditional valuation approaches comes only as step 18, with step 19 then being the choice of the income method.

Steps 1-14 deal with a close reading of the patent and an investigation of the patent situation.

19

20 Steps for Pricing a Patent (continued)

The author of this article (Timothy Cromley) is a CPA, professional engineer and U.S. patent attorney.

This article is silent on trade secrets, but trade secrets must also be considered.

Preface: To value an invention you have to understand it.

20

20 Steps for Pricing a Patent (continued)

a. Steps dealing with patent considerations:– 1. Check whether the patent is in force.– 2. Identify the context.– 3. Gather information.– 4. Assemble a valuation team.– 5. Read the patent.– 6. Investigate the patent’s scope.– 7. Talk with a patent attorney.– 8. Inquire about the patent’s validity.– 9. Inquire into blocking patents.– 10.Consider synergies among patents.– 11.Investigate foreign patent protection.– 12.Consider the remaining life of the patent.– 13.Analyze any prior royalties paid for the patent.– 14.Inquire into any actual or threatened litigation involving the

patent.

21

20 Steps for Pricing a Patent (continued)

b. Steps dealing with commercial considerations:

– 15.Identify the next-best alternative technologies.– 16.Estimate a demand curve for the patented item.– 17.Determine the patented product’s point of profit

maximization.

c. Steps on valuation approaches:– 18.Consider the applicability of traditional valuation approaches.– 19.Do an income-approach valuation.

d. Final step:– 20.Write the patent valuation report.

22

20 Steps for Pricing a Patent (continued)

e. Missing essential steps re trade secrets: • Inquire about the existence of trade secrets on

discrete aspects as well as on collateral know-how for patented inventions

• Analyze security measures that are in place to maintain trade secrecy

• Consider the efforts and investments involved in developing the proprietary know-how

23

G. ConclusionIn IP valuation exercises one cannot ignore the fundamentals of IPRs in terms of integration strategies for dual or multiple protection of innovation, nor the relevant fundamentals of IP licensing law and practice. If this is true and if it is also true that “business decisions (should not) end up being made by patent attorneys who may not understand the long-term commercial ramifications,” as has been stated by a noted representative of the IP value extraction school, then symbiotic collaboration and teamwork between the two practices is the answer to best serve clients

8.27.07

24

INTENSIVE PATENT VALUATIONPART II

A. Introduction

• In our knowledge-based high-tech era it is important to exploit the overlap between IP categories for dual or multiple protection.

• This is true especially between patents and trade secrets.

• Patents and trade secrets are not incompatible but dovetail: the latter can protect volumes of collateral know-how.

• This results in synergistic integration and secures invulnerable exclusivity.

• Most technology licenses are hybrid licenses covering patents and trade secrets.

• Licenses under patents without access to collateral know-how are insufficient for commercial use of patented technology.

25

B. Integration Strategy for IPRs

From former fragmentation by specialties, IPRs are now a “seamless web,” due to progress in technology and commerce, per Professor Jay Dratler.

Professor Dratler was the first one to “tie all the fields of IP together.” “Integrative treatment.” (Intellectual Property Law: Commercial, Creative, and Industrial – 1991)

In 1997 the authors of “Intellectual Property in the New Technological Age” (Professors Merges, Merrell, Lemly & Jorde) also• avoid the fragmented coverage• approach IP as a unified whole; and• concentrate on the interaction between different types of IPRs.

Thus we now have a unified theory in the IP world, a single field of law with subsets and significant overlap between IP fields. Several IPRs are available for the same IP or different aspects of the same IP for dual or multiple protection. Not taking advantage of the overlap misses opportunities or, worse, amounts to malpractice.

26

Integration Strategy for IPRs (continued)

One IPR category is center of gravity and more important—most often patents

Other IPR categories are then supplementary but very valuable to

• cover additional subject matter

• strengthen exclusivity

• invoke additional remedies

• standup if primary IPR becomes invalid

and thus provide synergy and optimize legal protection.

Multiple forms of protection are especially important in the fields of biotechnology and computer.

The most important strategy is exploiting the overlap between patents and trade secrets.

Illustrative examples are:

• GE’s industrial diamond process technology

• Wyeth’s Premarin process

• Pizza Hut decision

27

Multiple Protection in Biotech

Protection for a diagnostic kit involving monoclonal antibodies:

• Product patent on the test kit• Process patent on the preparation of the

antibodies• Copyright for test kit’s instructions• Trademark• Trade secrecy for production know-how• Trade secrecy for collateral know-how

28

Multiple Protection for ComputersA data processing system can involve:

• patented hardware and software

• patented computer architecture on circuit designs

• patented business methods

• trade secret production processes

• trade secrecy for collateral know-how

• copyrighted microcode

• copyrighted operating system

• copyrighted instruction manual

• semiconductor chips protected as mask works

• consoles or keyboards protected by design patents

• or as trade dress under trademark principles

• trademark registration

29

IP Integration ConceptsEXPLOIT THE OVERLAP

DEVELOP A FALL BACK POSITION

CREATE A WEB OF RIGHTS

BUILD AN IP ESTATE

BUILD A WALL

BUILD A RINGFENCE (India)

OVERPROTECT

LAY A MINEFIELD

for

SYNERGISTIC EFFECT

via

DUAL OR MULTIPLE PROTECTION

30

C. The Importance of Trade SecretsTrade secrets are the “crown jewels” of corporations — not the “cesspool of the patent system.”

Mark Halligan: “Trade secrets are the IP of the new millennium and can no longer be treated as a stepchild.”

James Pooley: “Forget patents, trademarks and copyrights…trade secrets could be your company’s most important and valuable assets.”

Trade secret misappropriation recently cost Walt Disney $240 million, Cargill $300 million, and Toshiba over $400 million.

88% of responses in an IPO Survey indicate trade secrets to be the really important intellectual assets because patents have limits: patentability requirements, publication and invent-around feasibility.

31

The Importance of Trade Secrets (continued)

Trade secret protection operates without delay and undue cost against the world — unlike patents which are territorial and so expensive to obtain and maintain that only very selective foreign filing is done.

Patents are tips of icebergs in an ocean of trade secrets• Trade secrets cover over 90% of new technology• Over 80% of technology licenses cover trade secrets or are hybrid licenses

Trade Secrets are the “workhorse of tech transfer.” (Bob Sherwood).

32

D. Patent/Trade Secret Interface

As a practical matter, licenses under patents without access to associated, collateral know-how are often not enough, because patents rarely disclose the ultimate scaled-up commercial embodiments of products and processes.

“In many cases, particularly in chemical technology, the know-how is the most important part of a technology transfer agreement.” (Homer Blair).

“It is common practice in industry to seek and obtain patents on that part of a technology that is amenable to patent protection, while maintaining related technological data and other information in confidence. Some regard a patent as little more than an advertisement for the sale of accompanying know-how.” (Peter Rosenberg).

In technology licensing “(r)elated patent rights generally are mentioned late in the discussion and are perceived to have ‘insignificant’ value relative to the know-how.” (Michael Ward, Honeywell VP Licensing).

33

Patent/Trade Secret Interface (continued)

“Trade secrets are a component of almost every technology license…(and) can increase the value of a license up to 3 to 10 times the value of the deal if no trade secrets are involved.” (Melvin Jager).

“One potential shortcoming of focusing on patents as a measure of innovation, besides the fact that it ignores the other types of intellectual property, is that patents are often valueless absent the ‘know-how’ that translates protected intellectual property into viable products.” (Gavin Clarkson, Harvard).

“A company with one or more patents for its technology will usually have substantial valuable technical and business information related to, but outside the direct coverage or disclosure obligations of, its patents. The company can maintain vigorous efforts in both areas of legal protection. (Jerry Cohen, Perkins, Smith & Cohen).

“It is frequently stated that know-how is the most valuable element of technology transfer. This is consistent with the writer’s own experience.” (Robert Goldscheider).

Failed Brazilian tactic — translation of foreign patents

CIBA-GEIGY examples: Eastman Kodak & DuPont patent licenses were useless.

34

Patent/Trade Secret Interface (continued)

• In the past- and even today – the question always was phrased in the alternative

• E.G., titles of articles discussing the matter read “Trade Secret v. Patent Protection,” “To patent or not to patent?” “ Trade Secret or Patent?” “To Patent or to Padlock?,” etc.

• Anent this choice, the respective advantages and disadvantages, e.g. in terms of duration and scope of protection, were considered controlling.

• On scrutiny the perceived differences are not there.• The patent life may be more or less than twenty years from filing and a

garden-variety type of trade secret, far from being indefinite, may last but a few years.

• Nor is there a difference as regards to the scope of protection with “everything under the sun made by man” being patentable.

• And while a patent does, and a trade secret does not, protect against independent discovery, a patent leads to efforts to design or invent around and a trade secret, properly guarded and secured, may withstand attempts to crack it.

35

E. Patent/Trade Secret Complementariness

• Supreme Court (Kewanee Oil, 1974): perfectly viable alternatives.

• Not mutually exclusive but mutually reinforcing — dovetail, in harmony

• “Coexistence is well-established.” (Don Chisum).

• Inextricably intertwined: Most R&D data and collateral know-how cannot and need not be included in patent applications — grist for trade secrets.

• Trade secrets precede, accompany and follow patents.

• Tom Arnold: it’s “flat wrong” to assume that “because the patent law requires a best mode requirement, patents necessarily disclose or preempt all the trade secrets that are useful in the practice of the invention.”

36

Patent/Trade Secret Complementariness (continued)

All patents are born as trade secrets• In the critical R&D state and before any patents issue, trade secret law

“dovetails” with patent law.

• Assuming that a development has been enabled and the best mode described, all collateral know-how not disclosed, whether or not inventive, can be retained as a trade secret.

• All R&D data, including data pertaining to better modes, developed after filing, again whether or not inventive, can also be protected as trade secrets.

• With respect to technologically complex developments consisting of many patentable inventions and volumes of associated know-how, complementary patenting and secreting is tantamount to having the best of both worlds

The question is not whether to patent or to padlock but ratherwhat to patent and what to keep a trade secret.

37

Patent/Trade Secret Complementariness (continued)

• Best policy and strategy is to patent as well as to padlock.• Initially file a broad or several patent application(s) simultaneously

or sequentially per the time honored maxim” file early, file often”• Pending applications are secret• This keeps options open and permits to defer a decision to keep

invention secret if application is not allowed.• Even if allowed, application can be abandoned and invention is kept

secret• Continue filing on improvements and additional patentable aspects

throughout R&D stage and beyond — offensively and defensively — “evergreening.”

• A la IBM, CIBA-GEIGY and other corporations, going for “big numbers.” Pitney Bowes obtained over 100 patents on their Paragon Mail Processor, a “simple machine.”

• Concurrently always keep innovations of disallowed applications and the huge volumes of collateral know-how as trade secrets.

38

F. The Best Mode RequirementConventional wisdom: coexistence is impossible because of the “best mode”

requirement. The “best mode” requirement applies • only to the knowledge of the inventor, • only at the time of filing and • only to the claimed inventionHence “best mode” requirement is no impediment, because —

1. Patent applications are filed early in the R&D stage to get the earliest possible filing or priority date.

2. The specification normally describes in but a few pages only rudimentary lab experiments or prototypes.

3. The best mode for commercial manufacture and use remains to be developed later.

4. Patent claims tend to be narrow for distance from the prior art.

5. As shown by case law, manufacturing process details are, even if available, not a part of the statutorily-required “best mode” disclosure of a patent.

39

G. Exemplary Trade Secret Cases

1. GE’s exclusive industrial diamond process technology

• Holds patents (some expired) and trade secrets

• Refused to grant licenses

• Fast-track GE scientists stole trade secrets for Far Eastern interests for million dollar payments

• In the end got caught, tried, jailed

2. Wyeth’s exclusive Premarin manufacturing process

• Has market exclusivity since 1942

• Patents expired decades ago

• Closely guards its trade secrets

• Natural Biologics stole these trade secrets

• Wyeth sued, got sweeping injunction

40

Exemplary Trade Secret Cases (continued)

3. Pizza Hut case

• Pizza Hut supplier, C&F Packing, invented and patented a manufacturing process for pizza sausage toppings and kept improvements secret

• Pizza Hut misappropriated trade secrets and got sued

• Court decision:

1. patents are invalid on on-sale bar grounds (on Summary Judgment)

2. trade secrets are enforceable and Pizza Hut had to pay $10.9 million (after trial)

41

H. Conclusion

• These and other cases are examples of how trade secrets serve as a fall-back position when patents fail

• Patents can be at risk due to three dozens of invalidity and unenforceability reasons and many attrition factors such as– Narrow claims granted by IP offices– Enforcing patents being a daunting and expensive task– Only very limited or no coverage in foreign countries– Only about 5% of a large patent portfolio have commercial value

(per Emmett Murtha, ex. IBM and former LES President) and– The effective economic life of a patent being only about five

years (Murtha)– As well as others.

42

Conclusion (continued)

• Trade secrets are indeed a viable mode of protection.• They can be used in lieu of patents but, more

importantly, they can and should be relied upon at the same time and side by side with patents to protect any given invention as well as the volumes of collateral know-how.

• Hence, it is patents and (not “or”) trade secrets. A happy marriage!

8.27.07