Fall 2011

Greetings!

Invotex® Group is pleased to share insights about current trends and issues of interest to litigators and counsel, particularly those with which we have recent experience. We hope you find this information informative, and we welcome your feedback.

In this Issue

1. Intellectual Property
   Prove Your IP’s Value…Please!
2. Economics
   Regression Analysis and Credibility in Legal Proceedings: Part III – R-Squared
3. Life Sciences
   IP Licensing and Anti-kickback Considerations for Life Science Companies
4. Trademark
   Monetary Damages in Trademark Disputes
5. Continuing Professional Education
   Upcoming Events

Intellectual Property

Prove Your IP’s Value…Please!
by Michele M. Riley and Lynton L. Markham

The importance of intellectual property is being discussed in the mainstream media as Google, Microsoft, Apple and others continue to battle over patented technology in the mobile phone market. It is unclear how each party has justified its actions, including the price for some of these very large transactions; however, one thing is clear: IP is important to the big technology companies in today’s economy.

These transactions raise some very interesting questions about the value of IP and its significance in company strategy. To determine the answer to these questions, one must understand and prove what the IP is worth – to the marketplace, competitors, customers and any other interested parties. And sometimes, these other interested parties are the members of a jury in a patent infringement litigation matter. As Ronald Reagan famously said, “Trust, but verify.” Too often IP owners assert the importance or value of a particular patent or trade secret when there is nothing, anywhere, to link that particular IP to any economic activity or benefit.

In a non-litigation setting, the stakeholders in an entity’s IP are varied and numerous. These stakeholders expect returns from the IP just as they would from any other asset. Therefore, stakeholders need to be convinced of the IP’s value for at least the following reasons:

• Shareholders do not want to invest for zero or negative returns.
• The expected value of the investment in IP must be ascertained to validate any investment decision that favors the IP over other desirable investments.
• There is no public market for intangible assets that offers observable market prices. An occasional IP auction, such as the Nortel patent sale discussed previously, can provide a data point. But, for the outsider, it is difficult to understand all the details impacting the transaction.
• Intangible assets are difficult for shareholders to understand and analyze because they cannot be seen or touched as tangible assets can. “For many reasons, it is easier to conceptualize how to safeguard and maximize the value of tangible assets: they exist in a controlled environment and may be easily observed, doors can be locked to protect them, and their value is fairly easily determined. IP is not so easily confined.”^[1]

Therefore, it is very important for a company to document how it values its IP and use such documents to show shareholders how the IP is being used to generate both current and future returns. This understanding of how IP is being used and valued will allow the company to effectively articulate to shareholders the benefits of such an investment.

Take, for example, the valuation of a trademark. Assume your company has been selling products featuring your long-used trademark for years. You use this trademark on all your products, feature it prominently in all your product literature and always display it in the same color scheme and font and with the same picture. In short, you would not sell a product that did not feature this trademark presented in this particular way. So, how much is this trademark worth? Is it worth something to you only because you have used it for so many years, just as an old hat or coat has value to a person that has worn it for a long time? Or, does it have meaning to customers in the marketplace who have come to associate the trademark with a quality product? Hopefully the answer is the latter rather than the former, but, without proper study and documentation, the latter cannot be proven.

So how do you prove it? If we were to perform a valuation of the trademark as, for example, part of an effort to obtain additional funding from venture capitalists (or other sources), the following questions must be asked:

• What is included in the valuation?
• What is the premise of value?
• What products do we sell that feature this trademark?
  • How much revenue and profits have we earned on these products?
• What products do we plan to sell that feature this trademark?
  • How much revenue and profits do we anticipate we will earn on these products?
• What benefits do we receive from featuring this trademark on all our products?
  • What form do these benefits take, and can they be quantified? If so, what is the quantification?
• What benefits do we forego by not putting this trademark on all our products?
• Why do customers like this trademark?
  • How does it make customers feel?
• How does this trademark contribute to your overall brand?

Documenting the answers to these questions, regardless of whether or not a valuation is being performed, can assist you in making smarter business decisions in the future as you try to monetize your IP. This exercise puts your trademark in an important context, which is the context of value. By moving the trademark from the vague and amorphous world to the quantified results world, you are moving your understanding of the trademark’s value from “unknown” to “worth something.”

Once in litigation, this information generated in the ordinary course of business becomes the factual evidence in establishing the value of the IP. A good example of this importance is purchase price allocations. When performed for financial statement reporting, these allocations are an important documentation resource in litigation matters. These reports can contain information on a variety of intangible asset classes, including patents, copyrights, trademarks, trade secrets and license agreements. In addition, the reports may include the following: 1) market, income and cost approach valuations, 2) revenue projections and expectations, 3) royalty rate estimations, 4) market comparables analysis and 5) relief from royalty valuations.

Such reports should be considered in litigation relating to IP purchased by the patent holder. “The amount paid by a plaintiff in acquiring a company with desired patents is unquestionably relevant in the calculation of damages should those patents later be asserted in litigation. Litigants have long stressed the importance of balancing the consideration paid for declared assets against damages claimed for their subsequent infringement.”^[2]

Another important example of information captured in the ordinary course of business is customer testimonials. What do your customers, the people who buy your products, say about your products? If you have a product that embodies several patents, find out from your customers what they like about the patented features. These testimonials will be useful in a royalty calculation that invokes the entire market value rule.

A customer survey may be the only way to prove that the patented feature was the reason for a customer’s purchase. A judge in the Southern District of New York recently highlighted just this, stating “[w]hile one can assume that those who wanted the feature found it desirable and were willing to pay for it…, there is no way of knowing (short of interviewing the customers, which was never done) whether this feature alone drove the decision to purchase from Otis – or put otherwise, whether the presence or absence of the allegedly infringing feature ‘was of such paramount importance that it substantially created the value of the component parts.’”^[3]

There are many sources of useful documentation that can provide an understanding of the economic value of a company’s IP, whether in the setting of a license negotiation, royalty compliance, litigation or a valuation. It is important for all IP stakeholders to be aware of these differing types of documentation, understand their uses and, thereby, have the ability to substantiate their IP’s worth.

Managing Director Michele Riley will discuss this topic in greater detail later this month when she participates in a panel discussion titled Understanding the Real Value of Your Patent Portfolio at the AIPLA Annual Meeting on October 20, 2011.

For more information, contact Michele M. Riley.
______

^[1]Stewart, Debora R. and Judy A. Byrd, Underpaid Royalties Are Pervasive, Yet Proper License Management Maximizes Profits, 2011.
^[2]Pursel, Brad L. and Mike R. Annis, The Role of FAS 141 Valuations in IP Litigation, Business Valuation Update 14 (March 2011).
^[3]Rulings on Motions in Limine, Inventio AG v Otis Elevator Co., Civil Action No. 06 -05377 *6 (SDNY filed June 23, 2011).

Back to top.

Economics

Regression Analysis and Credibility in Legal Proceedings: Part III – R-Squared
by Michael J. Petron and Emraan S. Khan

In the previous installment of this series, we continued a discussion of potential traps when regression analysis is offered as evidence before a court. The second topic, data mining/over-fitting, described the need to avoid statistical “fishing expeditions,” as we called them. A good economist or statistician is trained to walk the fine line between “mining” the universe of data for potentially relevant information and avoiding over experimentation so as to lose objectivity. To avoid the problem of “data mining” while still being thorough in its approach, a regression should involve: (1) a robust economic intuition (i.e. does this make sense?) and (2) variables and methods that add to the statistical significance of the regression.

In this installment, we will explore a related and often misunderstood aspect of regression analysis. Similar to adding variables to make a model fit the data better (data mining/over-fitting), is the tendency to judge a model purely on how well it fits the historical data. The simplest measure of this, the R-squared, is the most basic capture of the “fit” of the model. The R-squared is a measure of the extent to which the total variation of the dependant variable is explained by the independent variables of the regression. For example, if one had the regression equation:

y=c+ßx+αz+ e

Where “y” is the price of lumber, “x” is the import tariffs on lumber, and “z” is the humidity levels of climates where lumber is harvested.

If we were to run this regression, assume the resulting R-squared is .23. While only approximate, this R-squared roughly means that only one quarter of the variability of the price of lumber is explained by movements in input tariffs and humidity levels. An R-squared of .23 is on the low end of the R-squared range of 0 to 1. A low R-squared means that there are potentially many more variables besides tariffs and humidity levels that affect the price of lumber. It is intuitive to believe there are other variables besides these that would increase the R-squared. This also does not mean that our current model is wrong or that our parameter estimates “ß” and “α” are not statistically insightful for the court.

Oftentimes, the R-squared is improperly used as the “holy grail” of measuring the quality of a regression analysis. In fact, many good models may be incorrectly ignored due to a low R-squared, and many bad models may be incorrectly relied upon simply because of a high R-squared. MIT economist Franklin Fisher describes the dangers of such an approach:

“Since it is often possible to fit the data well with models that do not reflect any structural characteristics of the phenomenon being investigated, the danger here is that the laypersons involved will be impressed by poor work to the detriment of better models that do not fit or predict quite so well but are in fact informative about the phenomena being investigated.”^[1]

A model with a large number of variables and a higher R-squared does not necessarily provide additional understanding of the relationship between the key variables of interest and the dependant variable. Estimated coefficients and their standard errors are more important for this purpose. As discussed in Part I of this series, parsimony is a desired criterion in model creation, and one should resist the temptation to increase the number of independent variables solely to increase the model’s R-squared. Well grounded theory, not simply a high R-squared, should be used to drive a model and the best R-squared can vary greatly depending on the subject matter being examined. As David Anderson and Dennis Sweeny of the University of Cincinnati note:

“As a practical matter, for typical data found in the social sciences, values of r-squared as low as .25 are often considered useful. For data in the physical and life sciences, r-squared values of .60 or greater are often found; in fact, in some cases, r-squared values greater than .90 can be found. In business applications, r-squared values vary greatly, depending on the unique characteristics of each application.”^[2]

There are other reasons why simply setting a high R-squared as the sole object can be problematic. By the very nature of how it is calculated (we will not go into the details here), adding a variable to a regression can never cause the R-squared to go down. This means that the price of lumber regressed on variables x and z, can never have an R-squared lower than the price of lumber regressed only x. Even in cases where z is not statistically significant, the R-squared will be higher, but not necessarily producing a better model. A potential problem may arise if one assumes the model with x and z is better simply because of the higher R-squared.

This is why many regression packages often report an adjusted R-squared. The adjusted R-squared revises the original R-squared for the inclusion of additional variables. Essentially, it gives the economist a measure that gauges the relative added value of the additional variable when comparing against a model without the additional variable.

Ultimately, however, while both the R-squared and adjusted R-squared serve as a good starting point for economists and statisticians, they should be understood in their proper context.

For more information, contact Michael J. Petron.
______

^[1]Fisher, Franklin M. "Statisticians, Econometricians, and Adversary Proceedings." Journal of the American Statistical Association 81.394 (1986), p. 279.
^[2]Anderson, David Ray and Dennis J. Sweeney. Statistics for Business and Economics. Australia: South-Western Cengage Learning (2011), p. 580.


Back to top.

Life Sciences

IP Licensing and Anti-kickback Considerations for Life Science Companies
by Edward A. Gold, Scott J. Williams and Marylee P. Robinson

Life science companies hire surgeons, physicians and other healthcare professionals (“HCPs”) to provide a wide range of consulting services. Because these consultants may be in a position to purchase, prescribe or influence the use of the life science company’s products, the compensation for the consultant’s time must be at a fair market level. Anti-kickback regulations such as 42 CFR 411.351 require that:

Fair market value means the value in arm’s-length transactions, consistent with the general market value. ‘‘General market value’’ means the price that an asset would bring as the result of bona fide bargaining between well-informed buyers and sellers who are not otherwise in a position to generate business for the other party, or the compensation that would be included in a service agreement as the result of bona fide bargaining between well-informed parties to the agreement who are not otherwise in a position to generate business for the other party, on the date of acquisition of the asset or at the time of the service agreement.

The HCPs are engaged for a wide range of needs including scientific presentations, training programs and occasionally working with a life science company’s R&D team on product development projects. In many cases, the healthcare consultants are compensated via an hourly or lump sum payment for the time actually provided. However, when the consultant is working with a product development team, there often is an expectation that intellectual property may be developed by the HCP and transferred to the company. Therefore, the compensation to the HCP for these product development efforts often are based on royalty payments associated with the future sales of products. Although complicated in their own right, the issues involved in estimating fair market hourly rates have been discussed in many articles and presentations. This article will focus on the basic issues involved in assessing fair market compensation for royalty-based consulting engagements by healthcare professionals.

The first issue worth noting is that part of the fair market value (“FMV”) definition is the price at which a property would change hands. This part of the definition means that FMV is expressed as a dollar amount. In IP licensing, a royalty rate is often defined as a percentage of a predefined sales value. As such, in the normal context of FMV, a royalty rate does not in and of itself represent FMV. However, in the anti-kickback context, an appropriate royalty rate is understood to be an important element of the FMV of an HCP-related transaction. Deriving that appropriate market royalty rate involves using elements of the cost, market or income approaches in much the same way that one would assess a royalty rate in a non-anti-kickback context.

FMV is also another way of saying the value that will be transacted in “the market” – that is a market made up of typical buyers and sellers. To achieve this result, it is important to understand that FMV does not include the value created by synergies. Synergies are the potential ability of specific parties to be more successful or productive when working together than the typical pairing of market participants. Synergies come in various forms, including the ability to increase revenues, reduce risk or eliminate costs. Therefore, synergies generally result in additional value over and above FMV.

Because FMV ignores all synergies that may exist between the specific parties to a transaction, it is important that synergies not be allowed to affect the determination of the FMV royalty rate. As an example, if an HCP brings a technology to ABC Life Science Company and ABC Life Science Company can implement that technology better than any other competitor, then the FMV should be based on what the typical market participant would pay and not the greater amount that ABC Life Science Company could uniquely afford to pay. In contrast, if an HCP has specialized knowledge that is unique among doctors, this knowledge should be considered part of the IP being acquired (and that presumably can be acted upon by multiple life science companies) rather than as a synergy that should be disallowed.

In the anti-kickback context, the two most common types of “IP” that are likely outputs from an HCP’s efforts on behalf of a life sciences company are patented inventions and know-how. Despite its lack of the standard legal protections for IP, know-how is important in this context for two reasons. First, the AdvaMed code of ethics that is relevant to medical technology companies and their interaction with HCPs states in part,

“Provisions on Payment of Royalties. … Health Care Professionals, acting individually or as part of a group in which they are an active participant, often make valuable contributions that improve products or Medical Technologies. They may develop intellectual property, for example, patents, trade secrets, or know-how, under a product or technology development or intellectual property licensing agreement. A Company should enter into a royalty arrangement with a Health Care Professional only where the Health Care Professional is expected to make or has made a novel, significant, or innovative contribution to, for example, the development of a product, technology, process, or method. A significant contribution by an individual or group, if it is the basis for compensation, should be appropriately documented.”

Thus, the AdvaMed Code specifically calls attention to know-how as a potential basis for royalty payments, and it is common in the industry to find royalty payments when know-how is the only contribution made by the HCP. Second, the HCP working with a company’s product development teams are usually required to maintain confidentiality and thus rarely work with other companies on similar projects. This confidentiality, combined with the head start that valuable know-how can provide, creates a close parallel between the information transfer of know-how and that of the typical IP licensing situation.

The reader also should note that the AdvaMed code’s discussion of royalty payments refers to situations where the HCP is “expected” to make contributions to the development effort. Although the other situation, “has made,” is the common IP licensing situation in other industries, in these life science development efforts, the HCP is signed up as a consultant and often compensated based on the circumstances surrounding the HCP’s expected contributions.

There are many other factors that influence the agreed-upon royalty rate in a transaction between two sophisticated parties. Many of the factors found in the 15 Georgia Pacific factors are raised by parties to an IP negotiation and can be considered in an evaluation of FMV for an HCP. Examples include: exclusivity versus non-exclusivity, the stage of the life cycle, the scarcity of capable HCPs, and the risk of the product development effort. These factors, as well as many others, can be used to assess risks and benefits resulting in an equitable split of the expected benefits and thus the appropriate royalty rate. However care should be taken not to incorporate the Georgia Pacific factors that focus on synergistic elements and that are more appropriate for a determination of investment value rather than fair market value.

To summarize, FMV in the anti-kickback context for IP licensing is very similar to assessing a fair market value in other contexts. The key differences include the potential importance of know-how on its own, the need to identify a fair market royalty rate, and the emphasis on expected contributions by an HCP.

For more information, contact Edward A. Gold or Scott J. Williams.

Back to top.

Trademark

Monetary Damages in Trademark Disputes
by Marylee P. Robinson

In trademark infringement cases, the focus is often on the granting of an injunction; however, monetary damages are another important aspect of enforcement. In the December/January 2010 edition of The World Trademark Review, Invotex Group’s Michele Riley and Marylee Robinson, authored an article, Damage Limitations, that discussed groundbreaking trademark damages awards in the most active federal circuits in the United States as well as the positions adopted by appellate courts on the prerequisites for damages awards. Since then, several other cases have concluded that continue to support the findings illustrated in the original article.

To recap, courts generally impose one of two requirements before the plaintiff can recover monetary relief in trademark disputes: actual confusion and/or wilfulness. An analysis of these requirements as well as relevant case law yields the conclusion that circuit courts remain divided as to whether a showing of actual wilfulness and/or confusion is required to obtain an award of monetary damages.

In reviewing trademark litigation activity by circuit, it is clear that certain circuits have more trademark litigation activity than others. As The World Trademark Review article discussed, between 2003 and 2008 roughly 55% of all cases filed were filed in the Ninth, Second, and Eleventh Circuits. A similar trend continued in 2009 and 2010 with 54% and 53% of cases filed in these same circuits. The total number of cases also increased in 2009 and 2010. 2009 saw an increase of 3.2% over 2008 with 3,398 cases filed and 2010 saw an increase of 7.6% from 2009 with 3,657 cases filed.

The 2010 article analyzed several cases, including adidas Am., Inc. v. Payless Shoesource, Inc. (546 F. Supp. 2d (D. Or. 2008)), filed in the Ninth Circuit, in which the jury awarded damages of $304 million; the Second Circuit’s Gucci America Inc v Duty Free Apparel Ltd (315 F Supp 2d (SDNY 2004)), which resulted in maximum statutory damages of $2 million; and the Eleventh Circuit matter: Nike Inc v Variety Wholesalers Inc (274 F Supp 2d (SD Ga 2003)) with a damage award of $1.3 million.

Since the original publication, more cases of interest have concluded:

• Second Circuit: Burberry Ltd. v. Designers Imps., Inc., 2010 U.S. Dist. LEXIS 3605 (S.D.N.Y. Jan. 19, 2010)
  • The Court found that the Defendant acted wilfully in selling 12 counterfeit Burberry-branded items of merchandise and stated, “Actual knowledge is not necessary for wilful trademark infringement liability; rather ‘[i]nfringement is wilful when the defendant had knowledge that [his] conduct represented infringement or perhaps recklessly disregarded the possibility.’”^[1] However, the Court also found that certain factors mitigated the degree of the Defendant’s wilfulness. These factors included introduction of some evidence of compliance with the terms of a prior settlement agreement and of some level of diligence.
  • In order to find the Defendant liable for trademark counterfeiting and infringement, Burberry had the burden to prove: (1) Burberry had valid registered marks entitled to protection under the Lanham Act; and (2) Defendant used a similar mark in commerce in a way that would likely cause confusion among the relevant consuming public.^[2] The court found that burden was met and stated, “To determine whether confusion is likely to arise, a court need only determine that the items at issue are counterfeit and that Defendant distributed, offered for sale, and sold the items.”^[3]
  • Burberry claimed statutory damages of $6.5 million. The Court awarded statutory damages of $1.5 million which equated to $100,000 per mark per types of goods sold.
• Eleventh Circuit: Abercrombie & Fitch Trading Co. v. Importrade USA, Inc., 2009 U.S. Dist. LEXIS 127132 (S.D. Fla. Aug. 18, 2009)
  • The Court found: “(1) there is a strong likelihood of confusion that arises as a result of the Importrade Defendants' use of the Abercrombie trademarks (see Order at 8-18) and (2) the Importrade Defendants used the Abercrombie trademarks knowing they were counterfeit.” In the Eleventh Circuit this knowledge equates to wilfulness. Wilful infringement in that circuit has been defined “as when the infringer acted with knowledge or reckless disregard for whether its conduct infringed up on the plaintiff’s copyright.”^[4]
  • Abercrombie suggested and the Court agreed with statutory damages of $2.1 million. The figure was calculated using a baseline statutory award of $8,822.50 which was then trebled to reflect the Defendants’ wilfulness and then doubled for deterrence purposes. This resulted in an award of $52,935.00 per registered Abercrombie mark counterfeited per type of goods sold. There were ten marks per type of goods sold and four types of goods sold. Thus the total damage awarded was the result of 40 times $52,935.00.

For more information, contact Michele M. Riley or Marylee P. Robinson.
______

^[1]Hermes Int'l v. Kiernan, 2008 U.S. Dist. LEXIS 70506, 2008 WL 4163208, at *3 (E.D.N.Y. Aug. 28, 2008). 
^[2]Cartier Int'l B.V. v. Ben-Menachem, 2007 U.S. Dist. LEXIS 95366, 2008 WL 64005, at *10 (S.D.N.Y. Jan. 3, 2008).
^[3]Gucci Am., Inc. v. Duty Free Apparel, Ltd., 286 F.Supp.2d 284, 287 (S.D.N.Y. 2003).
^[4]PetMed Express, Inc. v. MedPets.Com, Inc., 336 F. Supp. 2d 1213, 1220 (S.D. Fla. 2004) (quoting Arista Records, Inc. v. Beker Enters., 298 F. Supp. 2d 1310, 1312 (S.D. Fla. 2003)).

Back to top.

Upcoming Events

Back to top.

Disclaimer: The opinions expressed in this newsletter are the opinions of the individual author(s) and may not reflect the opinions of the firm or any other individual associated with the firm.