Friday, March 10, 2006

My personal Prosper bidding philosophy

A few caveats:
a)my strategies/goals/targeted risk & returns are likely to differ from yours, even if we're both perfectly rational -- such things can be matters of heterogeneous tastes rather than right or wrong

b)hopefully, the benefit of stimulating discussion and having more efficient lenders, will outweigh the potentially increased competition from posting about this

c)In this context, I don't really care about the oodles of formal mathematical proofs of this type of auction optimality or that type of auction that economists love to bandy about -- I'm hoping to stimulate discussion of real-world/prosper-specific issues

1.Bidding early on insta-funding loan:

a)only valuable insofar as
i)you worry others will fund so much of the loan before you realize it that you get less than your desired allocation.
ii)you suspect that by bidding early, you make it more likely that the loan will be fully funded early/fully fund at all -- that is, maybe no one's going to bother bidding on a $10,000 loan that's only been funded to the tune of $100, and is expiring in 3 hours. In a truly efficient market, this wouldn't be a big issue, as everyone who likes the loans bid-snipes it at the last possible moment, easily funding it fully. (Bid sniping is the practice of automating your bid on an online auction to go in at the last possible second; typically accomplished through various websites like when one bids on ebay)
iii)you think the borrower might withdraw the loan opportunity and not relist/relist at a crappier rate, and you want to make it more likely that it closes at the current rate. (this could backfire -- say lots of people bid early on a really nice loan but not quite to completion -- maybe borrower thinks, hey, let me withdraw and relist -- clearly I listed at too hi a price)

b)bad insofar as:
i)your opportunity cost in terms of other, sweeter loans that may pop up in the interim
ii)you think that the rate could be even better if the loan closes w/o being fully funded, and the borrower relists.

c)handwavey analysis: If you're bidding on a really sweet (hi return) loan, the opportunity cost of tying up your funds early probably isn't insanely high -- how likely is it that another super-sweet loan will pop up? Even if we're talking an OK loan, as long as you have a fair amount of capital to lend, and you still have a lot of time to dole it out, probably OK to bid early and make your life easier/guarantee you catch a piece of the loan. However, if you only have a little bit of money left to lend, and the loan you're looking at is looking fairly generic in terms of attractiveness, maybe it makes more sense to delay bidding...

2.abbreviated thoughts on non-insta-funding loans
some differences include:

a)ability to have rates bid down: may make early bidding more advantageous (as you are then the last person to get bid down). However, in my book, I want the juiciest loans anyway, so any too-attractive loan that might get bid down is already not appropriate for me.

b)perhaps greater role of signalling to other bidders -- that is, if no one bids on a large loan, no one will want to be the first to bid, tying up their money w/o any guarantee that the loan will close. In practice, if the loan is nice enough, people will bid early anyway because they don't want to worry about missing it, and the opportunity cost of tying up your capital until loan closes is small in comparison. As always, the flip side to this is that bids frequently beget other bids, increasing competition/desirability of loan in eyes of others.

c)the ability of borrowers to close loans early once they're satisfied w/ rates being offered, as long as full funding occurs -- another reason to get in slightly earlier.

Other schools of thought on bidding?

Read more!

In defense of realizable, and perhaps sustainable, double digit returns

In response to nonattender'sargument on the prosper forum board:
I mean, if you're lending 10 grand to someone at 25%, you're essentially wagering that they'll be able to invest that 10 grand better than you could, and make a (cellar, you can do the math for me, it'll be a funny number) HUGE ROI on that money in order even to BREAK EVEN on the proposition..

that's a really funny (comedic) way of looking at it. there aren't many people
I know who I trust to be able to find better investments than I can... that's essentially what lending out money is! at these rates, you better hope that you're lending to some *smart* people who know where these kinds of returns can be made (and since they're HR: who you hope will be RIGHT about it THIS time, because to BE hr means they were probably wrong about it LAST time...)

ho ho ho!

I actually don't think we're fundamentally disagreeing on anything. For instance, I basically agree with him on the ROI figures -- that's why I typically am far more willing to invest in the "need to borrow $1.5k to stay afloat/pay off a higher rate card" type loans, and loath to invest in the "need $10k for some business opportunity" loans, because lending to each makes radically different assumptions about smart/shrewd borrowers are, and where my investment edge comes from.

I agree w/his critique for the most part when it comes to new business loans -- yes, I bet that i invest a lot better than the HR borrower who wants to start a new business, but there might be some valid counterarguments -- not enough to convince me to invest in them, but enough for me to refrain from saying that such investments are stupid. (see #4)

A couple of issues that make me unsure of the validity of your critique elsewhere:

1.I can easily guarantee you >50% returns -- just let me lever up an investment in the stock market, and give me a 40 year time horizon. I mean, sure, you may have swings that make the Nasdaq in 2000 look like T-bills, but hey -- perfectly doable, right? More seriously, I not quite so crazily lever up, and blindly trade index-reproducing futures in my retirement accounts, where I have tax advantage, long time horizons, expect the future value of my IRA's to likely only be a very small fraction of my savings, and almost entirely care about return and not risk.

2.Another way to look at it -- in equilibrium, high grade and high yield corporate debt should have the same risk-adjusted return, or else the "smart" money would buy one and short the other. In practice, this doesn't hold -- and i know there are 20 million arguments for this (lots of institutions can't hold junk bonds, supply demand, etc) -- the argument i'm going to pretend is relevant here, is that in practice, it is expensive to synthetically recreate a high-yield bond by borrowing lots of money to buy lots of high grade bonds.Borrowing has hi costs, you could get margin called, etc etc, so they don't trade evenly to one another. When I trade for myself, as opposed to a hedge fund, my margin rates are even more ridiculous. So, consider HR loans as a way to lever up, w/o getting the shaft from a brokerage firm's margin rates.

The minimalist case for investing: don't expect the HR asset class as a whole to be significantly higher return w/significantly less risk, but expect to extract the brokerage margin fees you'd otherwise incur if you wanted hi-return/hi-risk investments. This is the weaker effect that makes me personally pursue this strategy, but I have higher belief in it. This also roughly fails as a valid reason if I believe penny stocks, say, have similar risk, return, and correlation characteristics (I don't, but I don't have strong evidence for my belief).

3.The bigger gamble I make personally (because I bet on HR liquidity loans, but not the big ones looking to fund new businesses):
a.I opportunistically select HR, E, or D borrowers w/unusually high rates and good characteristics (for instance, I actually consider a DTI ratio just slightly above 20% to be a big positive -- there's lender stigma, and the borrower probably isn't gaming the loan). I hope I'm not also selecting the most financially irresponsible ones (because they set their own rates too high), or that their desperation (setting rates extra high) doesn't result in a huge default rate jump. To be more careful -- i hope I'm being adequately compensated by the negative signal of unusually high rates. I assume that borrowers are frequently inefficient in setting rates (separate from desperation and overall irresponsibility) and as a result, there is some edge to be harvested from such loans. Emphasis on "hope" and "gamble." This is the stronger effect that makes me personally want to try pursuing this strategy, but I have lower belief in it.

b. more generally, I'm not betting my borrowers can make 30% returns over all their cash. I'm just hoping that by facilitating their financial well being with a loan, their lower returns over their entire portfolio (future income included), can cover the interest on their loan to me.

Say there are certain trigger events/barrier levels to my borrower's financial well being -- eg, they have steady work, but had one-time unusual expenses, and alo frequently make late payments on credit cards, etc -- but all of a sudden, another big event occurs. S/he now must choose between medical treatment, and missing a mortgage payment. In scenario A, s/he can only choose one, and her finances go into a death spiral. In scenario B, a small loan tides her over, and raises her expected wealth manifold in comparison to scenario A. Unfortunately, for all I know, she's going to keep getting hit by these events, and every month, another Scenario B occurs.

Another corporate analogy: Say you have a firm w/good prospects/earnings, but poor cash flow management -- in a position to miss one interest payment. Then, they're technically in default, their rating collapses, cost of borrowing skyrockets, they go out of business. In such a situation, someone willing to step in instantly and pledge enough money to cover the one interest payment, can rationally and realistically expect to make huge ROI on the loan.

The problem w/my analysis -- in my day job, I can actually roughly ascertain if this is the case for some public company. On prosper, unless I start IM'g borrowers for copies of their credit reports w/their names crossed out, there's no way for me to do my research. Even if I had the info -- it just isn't worth the time spent on every loan.

4.In the potential defense of those who jump upon business loans, I would point out that it is fairly easy to end up with ridiculously high return figures on lots of small businesses when one fails to calculate various opportunity costs -- which frequently is the case. Call it the family-run deli effect -- if one factors in the enormous overhead of having every family member working 18 hours a day for free (caused by the huge mental/emotional costs of deciding that the family business just isn't profitable enough to stick with), then the true return one ekes out from a deli is pretty fricking low. However, if one (not I) were to lend to said deli owner, one could depend on the behavioral issues/failure to account for opportunity costs on the part of the deli owner to fund your hi-interest loan.

The flaws to this argument, of course, are also manifold -- not every entrepreneur is a deli owner -- many are Internet-bubble MBA students and will walk away from anything lacking equity, guaranteed 200k salaries, etc. Furthermore, lots of businesses are structured so that the deli effect can't occur -- even if a desperate owner worked his butt off, he'd be better off just working a different job entirely and funneling the profits to his crappy, but beloved business -- but no one will think to actually do so.

Nonetheless, there are many such niches where lending to small biz owners can have edge -- non-scaleable businesses, businesses where owners get pride of ownership (enough to throw money down the drain), deli-effect businesses, and so on.

A sample of what I do worry about, because I think I might be able to think of something:
What if my nondefaulting loans also tend to prepay, as nondefaulting borrowers are the ones who actually stay on track, improve credit grades -- and voila -- refi my loan out of existence? Hello, reinvestment risk! Even if naive default rates hold perfectly across all my loans, i'm effectively making fairly low returns -- defaulting loans are big negatives, and prepaying loans result in hi returns for the time period my money is invested, but not enough actual $'s of interst returned to offset defaults.

Read more!