What Litigation Finance is Really About

By Joshua Hunt 

Professor Anthony Sebok

September 1, 2016 The New YorkerTwice a year, Y Combinator, a startup accelerator based in Mountain View, California, holds what it calls a Demo Day, which is a showcase of the latest batch of new companies it has nurtured. In the past, those have included Dropbox, Airbnb, and Reddit (as well as a number of companies no one remembers). Last week, at the most recent Demo Day, Eva Shang took to the stage to talk about her startup, Legalist.

Dressed in jeans and a black T-shirt emblazoned with Legalist’s logo, Shang, who dropped out of Harvard to focus on her startup with a classmate, spent her allotted two and a half minutes telling the Demo Day audience about Legalist’s business model. Using an algorithm to calculate the likelihood that a lawsuit will succeed, the company then invests in cases it deems promising. If a plaintiff it has funded prevails, Legalist takes a percentage of the winnings—usually between twenty-five and thirty per cent. The algorithm considers, among other things, the size of the presiding judge’s caseload, which is an indicator of how long a trial will take, and the judge’s past decisions, which is an indicator of how he or she might rule.

Read more in The New Yorker. 

 

 

 

 

Before the Supreme Court’s decision in that case, which prevented Virginia from blocking the N.A.A.C.P.’s efforts to help local victims of racial discrimination, laws against third-party funding of lawsuits were selectively applied to civil-rights groups, Sebok said.