From a financial standpoint, Option One is a familiar one. Your loans are sold to a junk debt buyer (which, in this case, would be Prosper itself) for pennies on the dollar. You get a pre-disclosed amount right away.
The downside is that, if the litigation strategy succeeds, this provides Prosper with an incentive to continue to screw up pre-default collections. The more loans go into default, the more loans Prosper buys for pennies on the dollar, and the more money it makes from litigation. (When Doug Fuller originally announced the litigation test, Option One was the only option, and a number of lenders pointed out the conflict of interest problem.)
[ETA: Ira01 and xraider have persuaded me that the italicized paragraph is probably incorrect, and that all net proceeds from the litigation test will go to the lenders who opt in.]
Option Two gives you a stake in the outcome of the litigation. You get paid if and when the law firm extracts money from the defaulted borrowers in the pool, and the amount lenders receive (amount collected less court costs and attorney fees) is likely to be more than lenders would receive under Option One. (At least, that's the theory under which JDBs operate.) But there is no guarantee, and the money would trickle in over a period of years.
The issue here, IMO, is the pooling of the loans. This means that lenders who bid on homeless unemployed borrowers with lots of cleavage get just as much money from the litigation project as lenders who carefully selected borrowers with good jobs and attachable assets. This strikes me as a bit unfair.
The reason for the pooling, I assume, is to ensure that Prosper doesn't get stuck with any out-of-pocket expenses for this project. For simplicity of math, assume that there are 100 cases, each costing $300 for service and filing. 20 lawsuits are successful, yielding $10,000 each, 25% of which goes to the law firm. The other 80 borrowers have no attachable wages or assets. (The total portfolio would be $1 million -- 100 times $10,000.) (Ordinarily, the $300 in costs would be added to the judgment and collected from the 20 who pay the judgments, but since I'm going for simplicity here, I'm ignoring that.)
Prosper fronts $30,000 in costs to cover service and filing in all 100 cases. Under the pooled system, the first four successful collections (net amount of $7500 each after attorney fees) go entirely to reimburse Prosper. Lenders don't see a penny until the fifth collection. Eventually, another $160,000 is collected, with $40,000 going to the lawyers. That leaves $120,000 for the lenders, or a 12% recovery.
If the loans were not pooled, lenders on each of the 20 loans which were succesfully collected would get $10,000 minus $2500 in fees and $300 in costs, or $7200. That would be a 72% recovery. Lenders on the uncollectible loans would get nothing, nada, zip. And Prosper would be out the $24,000 it paid for costs on the 80 lawsuits against judgment-proof borrowers.
If I were confident that I had used good judgment in selecting borrowers (at least by comparison to the rest of the poor schlubs with defaulted loans in the pool), I'd obviously prefer a non-pooled system (not that Prosper gave lenders that choice). And, since this is an experiment intended to "create awareness of consequences" and increase payment performance for all Prosper loans, I'd argue that Prosper should eat the $24,000 in costs, either as R&D expense or loan servicing expense (creating fear among all deadbeats is a reasonable loan servicing expense) which should be spread across all Prosper loans.
I don't have any loans in the New Agency Test. If I did, I'd probably tell Prosper that both options are unacceptable, and that they should use a non-pooled system where lenders get paid based on the quality (collectibility) of their loans. (Unfortunately, there's little to be gained by making a huge stink, since Prosper has the option of abandoning the litigation project and selling off the loans to an unaffiliated JDB for the same pennies that they are ofering in Option One.)