Thursday, September 22, 2011

Oracle contradicts Google on damages, vows to fight hard for injunction against Android

A new filing by Oracle sheds light on the damages claims issue that confused lots of people yesterday and today. Besides making it clear that Oracle hasn't slashed its damages claims, it also confirms what I have been telling everyone for several months:

Damages can be substantial but they are not what Oracle is primarily trying to get out of this case. They are a potentially substantial consolation prize. The grand prize, however, is an injunction, and Oracle's filing makes this clear with the following statement, which explains why Oracle's new damages computation -- unlike the old one that included damages until the year 2025 -- relates only to past damages and apparently provides only some data points with a view to future damages, which would be rendered totally irrelevant by an injunction:

"Third, Google complains that Prof. Cockburn does not 'provide a firm calculation of future damages.' The Court directed that past and future royalties not be mixed by advancing royalties in a lump-sum payment. Prof. Cockburn complied. Oracle intends to strenuously pursue injunctive relief to resolve the key issue in this case: whether Google can use Oracle's intellectual property to create an incompatible clone of Java and thereby undermine Oracle's and many others' investments in 'write once, run anywhere.' If future royalties are applied, it is well established that they should be based on a separate, post-verdict assessment."

Compared to the above paragraph, forget all the billion-dollar figures that make headlines. Besides explaining that the new damages computation doesn't specify future damages but wants those to be subject to a "separate, post-verdict assessment" (which would be made only in the event Oracle is denied an injunction),

  • it's a clear vow to fight hard for an injunction,

  • it makes it clear that Google's design choice to deviate from the official Java standard (which Oracle referred to as fragmentation in other pleadings) is a key part of Oracle's argument for an injunction (and for why a mere damages award wouldn't be sufficient in this case), and

  • it refers to "many others' investments", presumably meaning other holders of Java-related intellectual property (note that Oracle is not even the only such right holder to be suing Google at this stage, since Gemalto is also asserting Java-related patents against Google, and there could be more than those two in the future).

In this blog post, I yesterday explained -- on the basis of a flowchart -- the possible outcomes of this dispute in terms of the ultimate licensing cost to Google. The bottom line is that a Java license can become more costly, possibly even far more costly, to Google than the proposed acquisition of Motorola Mobility if Oracle obtains an injunction that Google can't engineer around. In that case, Oracle could shut down Android and make Google a take-it-or-leave-it offer, the terms of which would not in any way be limited by whatever Oracle previously calculated as damages.

Past damages likely below $1 billion, but could be tripled

It's important to make a clear distinction between past damages (relating to when Android was launched until the presumed but uncertain trial date, which would still be Halloween at this stage) and future damages, which are compensation determined by the court in the event that Google is found to infringe but Oracle is denied an injunction, meaning that Google can continue to infringe but has to pay the amount determined by the court, which might be a per-unit royalty or a percentage of revenues.

Oracle's letter says the following about its new computation of past damages (click to enlarge the table):

So Oracle wants to add up these three components:

  1. patent damages, calculated based on the (hypothetical) license method: $176-202 million;

  2. copyright actual damages: either $102.6 million (based on the license method) or $136.2 million (based on the lost profits method); and

  3. copyright infringer's profits: there's a $823.9 million figure for "Google Gross Revenues" but Oracle's letter stresses that the copyright owner "is required to present proof only of the infringer's gross revenue, and the infringer is required to prove deductible expenses and the elements of profit attributable to factors other than the copyrighted work", so it's up to Google now to prove how much it subtracts from the $823.9 million figure for deductible expenses and other elements of Android's value than Java.

So the most conservative estimate is that this past damages amount consists of

  • $176 million (low end of patent damages range)

  • + $102.6 million (the lower of the two copyright actual damages figures)

  • + a moderate percentage of $823.9 million (Google's gross revenues or "copyright infringer's profits")

This means a very conservative estimate based on Oracle's proposed methodology would amount to roughly $500 million, while a more aggressive estimate would be closer to $1 billion. Either way, if Google is found to have infringed willfully (for which there are strong indications in the form of Google-internal emails), the past damages amount will be tripled. In that case, the amount would almost certainly be above $1 billion and might even exceed $1.5 billion or come close to $2 billion (in the event of Google failing to prove major deductions from its "infringer's profits").

But again, that wouldn't cover whatever happens post-trial. The future would depend on whether Oracle gets the injunction it "intends to strenuously pursue" according to today's filing.

Yesterday's reports of Oracle having drastically reduced its damages claims were all wrong. They were based on a starting point ($6.1 billion) that Oracle itself consistently contradicted and didn't take into consideration that Google's letter referred to $1.2 billion for 2012 alone -- and in order to compare the new calculation with the old one, one must always consider that the old one included damages from the launch of Android until the year 2025 (as opposed to 2012).

In my report on that Google letter, I started a key sentence this way: "Assuming that Google accurately represented the estimate for 2012 and Oracle wants $1.2 billion just for that year [...]" -- as opposed to just taking all those numbers at face value. If Google had been right with its assertion that Oracle wants $1.2 billion for 2012 alone, then Oracle would not have lowered but most likely raised its damages claims. Now we know that Oracle didn't do either: it just focuses on past damages now, and those are at about the same level of before. Here's what Oracle's letter says about that $1.2 billion component that accounted for more than half of the figure that previously confused people:

"Although the latter figure certainly illustrates some of the enormous profits enabled by Google’s infringement, it is not part of Prof. Cockburn's damages estimates."

As I explained before, Oracle also justifies why its new damages report doesn't contain precise numbers for future damages.

As far as past damages are concerned, Oracle appears to have used a different methodology but arrived at roughly the same figure as before.

Various analysts have speculated about where Oracle and Google might agree between the $100 million license deal Sun purportedly once proposed and the §6.1 billion figure Oracle itself firmly denied. Those who did so all missed the point. But the ones who predicted that Google may very well end up having to pay a per-unit royalty (and I saw at least two analyst statements to that effect) were right. However, the key to a really high per-unit royalty is not the damages report. It's an injunction -- and that's the very scenario against which the judge explicitly warned Google at a recent hearing.

If you'd like to be updated on the smartphone patent disputes and other intellectual property matters I cover, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents and Google+.

Share with other professionals via LinkedIn: