Print

[nonattender]

Full post here:  ProsperLenders.com

Teaser:

---
Notes on Prosper Lending: Interest Rate & Bidding Guidance

Prosper has recently done something *wonderful* for current and future Prosper Lenders. But I bet you don't know what it is - or, if you do, why it's so unbelievably important.

Since it's garnered very little attention in the P2P Lending Community thus far, I wanted to shine a little bit of light on it here, because, quite frankly, it's a monumental step - both for Prosper Marketplace *and* for the evolution of the entire Peer to Peer Lending space. Ok, alright already... enough pre-amble.

Let's hear it:

Since Prosper's launch in February 2006, lenders were presented with Experian default projections as baseline guidance for predicting default rates for each credit grade. Judging by the bidding behavior of many lenders, these default projections were trusted, to a large degree, as being both accurate and applicable...

---

Full post here:  ProsperLenders.com

-t

[Nontyper]

Awesome post... Well documented..

[caliostro]

blah

[Nontyper]

And I bet a snickers that at the end of the day, performance of Prosper loans will not improve significantly in the months to come, even with this guidance in place.

I may take that bet.... How long is the window and what is significant???

[Mark12547]

While the bidding strategy may have to be taken with a grain of salt¹, I think it is a step in the right direction; it is helping many lenders to set bidding interest rates more in line with the risks identified so far.

Note 1: The grain of salt includes these concerns:

• The after-default returns are over-simplified: you can't just subtract the default rate from the interest rate. It may be close enough for small default rates, but as the default rates go up, the error escalates far more. A discussion about their over-simplistic calculation was on the old .com forums. (No, I wasn't one of the literate ones contributing to that thread, but I realized it's implication: the simplified calculations had way too optimistic estimated return rates for the lower credit grades.)
• There is still need for more data. Prosper opened its "doors" in February 2006, which is less than 2 years ago. (Loans before February 2006 are between coworkers, families, friends, but not between strangers over the Internet, so the demographics changed radically on February 2006.) As much as I like horror movies, eh, Pninen's charts of 1+mo late & worse, we still don't know how the loans will perform during the third year.
• Even if we did have three years of history, we don't yet have a history of performance through economic cycles. Generally, during economic recessions, lower-quality loans default more. Since we know Prosper loans default much more than Experian's "average annualized default rate" for all card products, add the stress of a sour economy (sub-prime melt-down, anyone?), and even the best estimates are out the window with this systemic risk.
• Knowledgeable bidders may end up cherry picking some of the loans that are "outliers" within some of the slices, which could cause some distortion in the numbers. This is probably more likely in the new Portfolios and result in higher default in the portfolios as planned as some of the better loans get bid below the portfolio's bidding rates. In any case, in a fluid market, as perturbations are introduced, distortions occur, so it is often a guess on how significant the distortion will be and sometimes there are "unintended consequences."
• In addition to the distortions mentioned above, Prosper has been tweaking its site, so we have even less history on the newer additions to the "additional credit data" and other changes.
• It is said that you have to pick your bait to catch the type of fish you want. The effects of advertising, if any, may affect the demographics of future borrowers, which may affect where within the various slices the new borrowers will fit in. (Remember: a slice tries to group similar groups of listings together, but they don't have identical numbers, so there are variations within each slice.)
• There is also the risk that Prosper may go under. (Witness the 2-month slide to a 6-month plateau as seen on RateLadder on the Prosper Loan Growth by Month chart, and that plateau is about one-fourth of what Chris Larson said would be needed to make Prosper profitable). If Prosper Marketplace, Inc. goes through bankruptcy, contracts could be rewritten by the courts that might include increasing servicing fees paid by lenders (and thus decrease lenders' profits). I don't know if it will happen, but it is a distinct possibility.

At least they are using Prosper data, which is more realistic than Experian's "aveage annualized default rates" they were showing before.

[mjerryfirst]

A lender on Prosper should bid with an expectation of a 4-5% default rate.   

[Greebo]

A lender on Prosper should bid with an expectation of a 4-5% default rate.   

This is true only for a given value of true.

The rate to expect varies widely based on credit rating, dq's, pr's, etc.

[bamalucky]

A lender on Prosper should bid with an expectation of a 4-5% default rate.   

you left out annually

[BrassKnuckles]

I checked out the new feature and it does make you stop and think a bit before bidding.  If it helps newer lenders slow down and start bidding reasonable (read: higher) rates then I like it.  I am the first to admit I bid too low for loans when I first started (sort of like when you bid to high for something stupid on eBay).

I will not be using the new guidance in my own bidding decisions, as I have my own "formulas" now, but overall I think it is a plus to the platform, since it will help newer lenders smarten up a bit.

[Mtnchick]

A lender on Prosper should bid with an expectation of a 4-5% default rate.   

My default rate has been much higher. It's closer to 10-20%

[ira01]

A lender on Prosper should bid with an expectation of a 4-5% default rate.   

My default rate has been much higher. It's closer to 10-20%

Don't worry, mjerryfirst will explain to you what you did wrong. 

[nonattender]

• There is still need for more data. Prosper opened its "doors" in February 2006, which is less than 2 years ago. (Loans before February 2006 are between coworkers, families, friends, but not between strangers over the Internet, so the demographics changed radically on February 2006.) As much as I like horror movies, eh, Pninen's charts of 1+mo late & worse, we still don't know how the loans will perform during the third year.

Absolutely, but I'd imagine that chance of default in year 3 would be at LEAST equal or lesser than in the first 24 months, don't you?

• Even if we did have three years of history, we don't yet have a history of performance through economic cycles. Generally, during economic recessions, lower-quality loans default more. Since we know Prosper loans default much more than Experian's "average annualized default rate" for all card products, add the stress of a sour economy (sub-prime melt-down, anyone?), and even the best estimates are out the window with this systemic risk.

The same could be said about any asset class that has a limited performance history and/or exists in the present economy.  I'm not going to argue re: recession, though I disagree.

• Knowledgeable bidders may end up cherry picking some of the loans that are "outliers" within some of the slices, which could cause some distortion in the numbers. This is probably more likely in the new Portfolios and result in higher default in the portfolios as planned as some of the better loans get bid below the portfolio's bidding rates. In any case, in a fluid market, as perturbations are introduced, distortions occur, so it is often a guess on how significant the distortion will be and sometimes there are "unintended consequences."

Knowledgeable bidders = who?    I know what you mean, but... outliers being cherry picked just means that "cherrypickers" will get "better" loans at lower rates, and the Portfolio people will get "less better" loans at slightly higher rates.  In aggregate, I don't think it's a big deal.  Might be a wash performance wise - and/or might lead to more originations (which would go to your point below re: growth).

• In addition to the distortions mentioned above, Prosper has been tweaking its site, so we have even less history on the newer additions to the "additional credit data" and other changes.

Yes, but, if anything, the addition of more credit data can only have a long-term positive influence.  There may be short-term flux, but I'm happy to trade that for long-term gain..

• It is said that you have to pick your bait to catch the type of fish you want. The effects of advertising, if any, may affect the demographics of future borrowers, which may affect where within the various slices the new borrowers will fit in. (Remember: a slice tries to group similar groups of listings together, but they don't have identical numbers, so there are variations within each slice.)

Certainly - but, given the types of (free/internet) advertising employed so far, I don't see that any active/traditional marketing could make things any worse, borrower-pool wise.

• There is also the risk that Prosper may go under. (Witness the 2-month slide to a 6-month plateau as seen on RateLadder on the Prosper Loan Growth by Month chart, and that plateau is about one-fourth of what Chris Larson said would be needed to make Prosper profitable). If Prosper Marketplace, Inc. goes through bankruptcy, contracts could be rewritten by the courts that might include increasing servicing fees paid by lenders (and thus decrease lenders' profits). I don't know if it will happen, but it is a distinct possibility.

Going-concern issues are always "possible".  How "distinct" such a possibility might be is subjective.  Growth rates are absolutely something to keep an eye on - though I'm not too concerned, this early.  Am somewhat confident in Prosper's ability to secure VC - don't think anyone went into this thinking it was a 2 year experiment to enter a multi-multi-billion market - esp. given the ability of all of its recent competitors to do same - with more and more on the way every day...

[/list]
At least they are using Prosper data, which is more realistic than Experian's "aveage annualized default rates" they were showing before.

Indeed.

-t