Patent Valuation Comes Down to Litigation

Two days ago, I attended the Theory and Practice of Patent Valuation conference held at the Bancroft Hotel in Berkeley, California. The conference was hosted by the Berkeley Center for Law and Technology.

Although the conference was interesting, at first blush, it seemed to be a mish-mash of vaguely-related ideas. This post attempts to rationalize that conference, revealing a tightly coherent thread running through it.

At the conference, the speakers and the topics could be categorized according to three distinctly different approaches to patent valuation:

  1. Litigation
  2. Manual Transaction
  3. Data Analytics

[caption id="" align="alignright" width="76"]Vince
OBrien Vince O'Brien[/caption]

In a typical patent Litigation, one or a few patents is being asserted against one or more defendants. In that scenario, the legal system provides various witness-driven methodologies for valuing patents. The first business speaker at the conference, Vince O'Brien, spoke to this first approach to patent valuation.

Then, in the middle of the conference, two different panels addressed the second approach: Manual Transaction. In this second scenario, as with Litigation, one or a few patents is at issue. But instead of being litigated, the patents in this scenario are being traded (i.e. bought, sold, and/or licensed).

The groups that spoke to this scenario were Brokers and Corporate Counsel. Basically, the latter serve as buyers in the market for patents; the former sell to the latter, and also buy for them.

[caption id="" align="alignright" width="98"]Ron
Laurie{width="98" height="95"} Ron Laurie[/caption]

There was close agreement between these two groups as to the best methodology for valuing patents in this second scenario. Broker Ron Laurie of Inflexion Point Strategy said that any sales package in this scenario needs claim charts. John Amster of RPX Corporation concurred, saying that, of all the factors important in this scenario, showing infringement is the most important. (Claim charts show a case for infringement.) Then, among the Corporate Counsel, Doug Luftman of Palm, Inc., addressed what he wanted to see from anyone trying to sell his company a patent. Consistent with what Messrs. Laurie and Amster had said, Mr. Luftman exclaimed: "Show me the claim charts!".

[caption id="" align="alignright" width="100"]John
Amster John Amster[/caption]

Clearly, there is a connection between the court-mandated patent valuation methodologies used in Litigation, and the claim charts used in Manual Transactions. Indeed, the claim chart is a format taken straight from the Litigation scenario. So what's going on in the Manual Transaction scenario is that claim charts are serving as a proxy for Litigation. Specifically, claim charts take a reasoned stab at the who, what, and why of a potential, future litigation.

But although there is this tight nexus between the first two approaches to patent valuation, the connection to the third approach — Data Analytics — is more mysterious. In this third scenario, a great many patents (e.g. 1000) are being transacted. Here, there are just too many patents to read individually (as is done in Litigation and Manual Transaction). So valuation methodologies used in this third scenario must be based on something else.

At the conference, toward the end of the day, a number of speakers explained their own particular method for performing Data Analytics on a portfolio of patents. This is the practice of culling certain common attributes from the portfolio, and applying algorithms to those attributes, in an attempt to reveal value.

The most striking thing to me about the various approaches to Data Analytics described at the conference is the apparent incoherence. None of the approaches seemed to have much of anything to do with any of the other approaches.

[caption id="" align="alignright" width="92"]Ron
Epstein{width="92" height="106"} Ron Epstein[/caption]

Evidently I wasn't the only one in the audience to register this incoherence. At the end of the conference, in reference to one of these approaches, Mr. Laurie said that the approach reminded him of collateralized debt obligations. Similarly, on other occasions, I have heard Ron Epstein of leading broker IPotential, LLC, express his puzzlement as to what exactly some of these Data Analytics approaches are useful for.

Stepping back, and looking at the conference as a whole, it occurred to me that if the best approach to Manual Transaction (claim charts) is a proxy for Litigation, then best method for Data Analytics ought to be one too.

Having posed the problem in this way, one of the Data Analytics methodologies stood out to me. This is the method of studying forward citations of a patent. Basically, a patent with a significant number of forward citations suggests that multiple parties have products in the same space as the patent. This, in turn, suggests that claim charts could be generated for that patent. This, in turn, suggests that the patent has Litigation value. Indeed, lots of people have found that forward citations correlate with market value.

This analysis suggests a simple litmus test for deciding whether a particular Data Analytics approach is sound: Does the approach ultimately tie back to Litigation?

And so, with the foregoing thoughts, a light appeared, to me at least, illuminating the otherwise foggy nine hours I spent in Berkeley on Friday.