As I mentioned in a prior post, I did some very basic research into what would happen if one were to keep the loan balance low (to the absolute minimum of $1,000). The following is a summary.
From November 2005 through June 15th 2008, there were a total of 893 total loans of $1,000. Of these, 582 are active (65.2%). Of the active loans, 560 are current (95.7%; 3 are in payoff), 308 have been paid off (34.5%), and only 25 are late or worse (2.8% of total, 4.3% of active loans). Note that the criteria for the above loans is as follows: AA to D, $1,000 or less, any state. No verification on homeowner or auto funding.
Based on this most basic of analysis, one would conclude that it is possible to earn decent returns from Prosper, with very low risk (again very low is subjective). Oh, the ‘decent return’ in this case is about 13.05% over the period for active loans (a bit less at 12% for all loans).
As I went through the study I noticed some more flaws in this strategy: the limited number of loans makes it impossible to scale progressively. By scale progressively I mean to gradually increase the loan amounts per bid, starting from the minimum ($50). This is because by increasing the bid size, the distribution is skewed, and the money weighted returns would be affected by any eventual defaults on the higher limit bids.
Thus, if one started with $50 bid and won every single loan outstanding, there would seem to have been a limit of about $40,000 to $50,000 in the account. Granted, $50K @ 13% is still not bad. Note also that while progressive scaling is not advised, scaling up front also has a limit. Given that the loan amount is only $1,000, one is already 5% of the loan at $50. By increasing to $100 bid, one would be at 10%. For the purposes of diversification, 10% of the loan is probably the absolute max you want to be at, before it starts to defeat the purpose of the exercise.
One could argue that being a large proportion of the loan is not an issue, since it has no bearing on whether or not the loan will eventually default. While I think this is mathematically true per se, I believe there to be some level of correlation among the loans so that you lose some diversification benefits as the bid amount rises.
So what is the bottom line? This strategy would currently likely be acceptable up to about $100K ($100 bid). Anything higher and I personally would not be comfortable. But then that’s just me. I haven’t looked at the $1,500 to $2,000 loan requests, but I suspect the non-currency rate to rise. By how much I don’t know. Certainly, as the size of the loan rises, the number of loan requests will also rise. So, within reason, one could extend this strategy and simply increase the credit worthiness requirement (i.e., drop the ‘D’ ratings when moving to $1,500, and drop the ‘C’ ratings when moving to $2,000). However, as the years go on one could just stick to the original strategy and move beyond the $50K limit that currently exists simply by more loans being originated.
Are there flaws in this study? Probably. Tell me what you think.