Economically Rational Patent Trolls

In The Prior Art blog, Joe Mullin wrote yesterday about GOOD Magazine’s profile of Erich Spangenberg, a highly successful “patent troll.” Mullin’s main beef with the article is GOOD’s comparison of Spangenberg, and patent trolls more generally, to Robin Hood. Mullin argues that in the case of patents, there is no poor person to whom wealth is being transferred. This is a fair point. GOOD overreaches by characterizing patent trolls, or non-practicing entities, as being akin to Robin Hood.

Is It Economically Rational For Patent Trolls To Be Shotgun Litigators?

Without going into a long discussion about what a patent troll is or is not, or whether they create any economic value, I offer a few thoughts on the business strategy of patent trolls.

1. Litigating indiscriminately against many targets at once is bad strategy.

Patent trolls should not adopt the business strategy of suing tens or hundreds of targets at once.  This is not an economically rational legal strategy, since it is inefficient to file suits where the costs of litigation are greater than the expected value of recovery. In many situations, targeting infringement lawsuits against a small number of potential infringers makes more sense. From a legal perspective, each suit will be more likely to proceed to trial or result in a settlement.

As the probability of going to trial or settling increases, and the value of the attached jury award or settlement award increases, the expected value of that suit increases. The calculation is:

EV = (P(S) x [Jury Award/Settlement Award]) – LC,

where EV is the expected value of the litigation, P(S) is the probability of winning a jury award or settlement, and LC is the patent troll’s litigation costs.

A troll can calculate the point where the sum of the expected value of X number of suits is greater than the LC of bringing those suits. The relevant calculations are:

EV(X) = EV(1) + EV(2) +…+ EV(X)

LC(X) = LC(1) + LC(2) +…+ LC(X)

EV(X) > LC(X)

This last equation should guide the troll’s decision of which and how many infringement suits to bring. The calculation requires careful analysis of the strength and potential monetary value and costs of each potential suit within the set of X lawsuits. I should emphasize that the specific identity of each suit within the set X is critical to the litigation strategy.

2. The difficulty of determining litigation risk makes a licensing strategy attractive.

The problem with the EV calculation is the difficulty of assigning a value to P(S). Patent litigation is risky and costly.

For this reason, trolls can rationally look to a licensing agreement as the optimal outcome. Where a licensing agreement is worth more than the EV of the particular litigation, the troll should focus its efforts on negotiating a license. As noted above, the EV of a suit is reduced by the litigation costs. But a business strategy that emphasizes licensing will greatly reduce litigation costs. Thus, the expected value of the license can be understood as:

EV(LA) = [P(LA) x (License Value)] – NC,

where EV(LA) is the expected value of the license agreement, P(LA) is the probability of reaching a license agreement, and NC is the negotiation costs.

Compared to the litigation scenario, NC will certainly be lower than LC. At the same time, the License Value may also be lower than the Jury Award/Settlement Value. But the P(LA) may be higher than P(S) and easier to estimate with confidence. These are the tradeoffs that the troll, and the troll’s target, must consider in formulating their legal strategy.

For these reasons, patent trolls can be understood as economically rational actors — or at least at rational as any party trying to create and capture value from their patents.