In early September 2011, a trial court ordered a new trial involving a jury verdict in which Oracle obtained a $1.3 billion verdict from SAP. By all accounts, Oracle’s verdict is the largest ever awarded for copyright infringement. In lieu of a new damages trial, Oracle can accept a remitter of $272 million – an amount admitted by Oracle’s own damage expert witness.
The defendant, SAP, is a German-based software competitor to Oracle for programs that automate accounting, payroll, and human resource functions. Because SAP admitted copyright infringement by illegally downloading a massive volume of files that included software, source code, and support documentation, the trial was limited to a damages determination.
The Court summarized copyright damages law as follows:
“The Copyright Act allows recovery of either statutory damages or ‘actual damages suffered by [the copyright owner] as a result of the infringement’ plus ‘any profits of the infringer that are attributable to the infringement and are not taken into account in computing the actual damages’ …Actual damages are generally determined by the loss in the fair market value of the copyright measured by the profits lost due to the infringement of the value of the use of the copyrighted work to the infringer. …
The Ninth Circuit has endorsed a retroactive license fee as measure of the loss in fair market value of the copyright. In situations where the infringer could have bargained with the copyright owner to purchase the right to use the work, actual damages are what a willing buyer would have reasonably required to pay a willing seller for the plaintiff’s work.” [Citations omitted]
All parties acknowledged that SAP obtained limited success from its infringement. Consequently, disgorgement of the defendant’s profits is not a large number relative to the large amount that Oracle won from the jury.
The trial Court concluded:
“The court fines that the $1.3 billion verdict is contrary to the weight of the evidence. Rather than providing evidence of SAP’s actual use of the copyrighted works, and objectively verifiable number of customers lost as a result, Oracle presented evidence of the purported value of the intellectual property as a whole, elicited self-serving testimony from its executives regarding the price they claim they would have demanded in an admittedly fictional negotiation, and proffered the speculative opinion of its damages expert which was based on little more than guesses about the parties expectations. At the same time, Oracle urged the jury to disregard evidence of Oracle’s actual customer losses resulting from infringement. Thus, the verdict grossly exceeded the actual harm to Oracle in the form of lost customers.”
Here are practical lessons that can be learned from this damages reversal:
Borrowing from patent law, a reasonable royalty is intended to be a lower limit to damages when lost profits cannot be proven. Economically, this is sensible because a licensor does not incur the efforts and risks of manufacturing, selling and financing the intellectual property exploitation. The licensee, who accepts all such efforts and risks, needs to be compensated for this economic contribution. Yet, in the Oracle case, claimed reasonable royalty damages were higher than lost profits. This should have been an immediate warning that the reasonable royalty claim was incorrect.
Plaintiff’s damage witness made a mistake by presenting a damage range of $881 million to $2.69 billion. The fact that Oracle’s damage witness could not present a conclusion as to a more narrow range of outcomes is a strong indication that the damage range includes impermissible speculation. According to the Court:
Rather than offering objective evidence to assist the jury in determining a fair market value for the license that even Oracle admitted its expert could not quantify, Oracle’s counsel invited the jury in his closing argument to engage in guesswork and simply pick a number between $1.66 billion and $3 billion.”
Throwing out a wide damage range to the jury occurs regularly. The underlying rationale is that the jury should determine the proper damages. However, when an extremely wide damage range is provided, the plaintiff usually does not like the realistic lower portion of the range, and so invites the jury to speculate about larger (but unrealistic) amounts. This wishful thinking is understandable, but does not meet the requirements for non-speculative damages or related expert testimony.
Reasonable royalty calculations in a copyright case are best limited to those plaintiffs that wish to license their copyrighted materials. In this case, Oracle admitted that it never gave any entity a license to copy Oracle’s application software and support materials. If the materials at issue have not been licensed, damages will need to be calculated as either (i) lost profits from the plaintiff’s lost exploitation of the copyright, (ii) disgorgement of the defendant’s actual profits, or (iii) statutory damages. Because of these alternate monetary remedies, copyright damages are different than what is applicable in patent cases. If statutory damages become applicable, the relevant court cases suggest an economic analysis that bears resemblance to what would be considered in a lost profits or reasonable royalty case, but with a lower standard of proof.
Experts should be paid well for their analysis, but the damages experts on both sides of this case were paid extraordinary and counter-productive amounts. SAP’s defense expert received $14 million. The Oracle’s expert witness (whose conclusion was harshly criticized by the trial court) was paid around $4 million.
One must imagine the jury would be prone to disregard an expert’s independence after being paid these extraordinary amounts. Having performed copyright infringement damage analysis in upward of a hundred cases, I cannot imagine performing any responsive analysis that would cost anything more than a small fraction of the smaller of these two expert fees.
As a monetary remedy, reasonable royalties have recently experienced a string of limiting court cases. In each of these cases, capable and independent experts could have done better. These significant cases include:
a. The Federal Circuit ruling in Uniloc USA, Inc. vs. Microsoft Corp. (January 2011), in which the Court concluded that the widely-used but fatally-flawed 25 percent rule can no longer be part of reasonable royalty damage calculation;
b. The Federal Circuit ruling in ResQNet.com, Inc. vs. Lansa, Inc., (February. 2010), in which the Court remanded a damages determination back to the trial court because of flawed practices that were previously often seen in reasonable royalty damage calculation; and
c. The Federal Circuit ruling in Lucent vs. Microsoft (September 2009) in which the Court criticized a failed application of the total market rule in a reasonable royalty damage calculation.
Capable damages experts who understand the need to remain credible could have prevented each of the above case damage award reversals, as well as the verdict reversal in the Oracle vs. SAP case.