As regular readers know, I've been messing around with peer to peer lending. I first opened an account with Lending Club back in October. In February, I opened an account with Prosper.com. I haven't put much money there yet, only $500, which at the $25 per loan minimum, only gives me about 20 loans.
The main attraction of Prosper over Lending Club is better returns, at least with Prosper 2.0. Prosper was the first of the two big peer to peer lending services, but the initial investments were a disaster. They shut down for a quiet period, revamped the system, and started lending again in July 2009. Since then, returns have been much more promising.
Using lendstats.com, you can see that for loan originated between July 2009 and December 2009, the average Prosper return has been 8.4%. That's several percent better than Lending Club from the same time period.
My goal with these peer to peer investments is not to ensure safety of principal. I know a significant percentage of these loans are going to default. This is not money I need for decades, so I can afford to take some risk. This is also an incredibly illiquid investment. It would be a major pain to get my money back anytime soon. But it is also an investment from which I expect high returns, and very low correlation with the rest of my portfolio. Not only do I want these 8% returns, I'm hoping that by using back-tested data from lendstats, that I can actually improve on that by 2-4% or more and get double digit returns. Here is the method I plan to use to do that.
1) The Repeat Borrower
If the future resembles the past, then I can increase my return by 3% simply by only lending to people with previous Prosper loans. 9% for everyone. 12% for repeat borrowers. That's pretty low hanging fruit. The problem is that there aren't all that many of them. There might only be 2 or 3 repeat borrower loans to fund per day. So if you want to invest $5000, at $25 a piece, that's 200 loans. It might take you three months to build that portfolio. That's a lot of time, effort, and money sitting there earning you $0 until it's invested.
2) Riskier Loans
I don't even touch the AA and A loans for the repeat borrowers. The average return for an AA repeat borrower for loans originated since July 2009 has been 4.6%. That jumps to 11.8% for a B borrower, and as high as 17% for an HR borrower.
The curve is a bit more complex for the first-time borrower. Returns for AA and A borrowers are around 6%. B-D borrowers have had returns of 8-10%. But at the far end of the curve, with the E and HR loans, it drops down to around 7%. So you're looking for the middle of the range.
3) Longer Loans
Prosper offers 1 year, 3 year, and 5 year loans with corresponding returns of 4%, 9%, and 12%. That's a pretty hefty liquidity premium, and I plan to take advantage of it. Like the repeat borrower loans, the 5 year loans are also a bit rare, so it is hard to fill a portfolio with them.
4) Homeowners
Just like with Lending Club, loaning to those who own their home seems to be a better bet. That may just be because they're less likely to declare bankruptcy, but it at least shows that they can get their finances together enough to buy a home in the first place. Homeowners 9.6% on average, renters have an average return of 8.4%.
5) Have a Job
Seems like a no-brainer I know, but apparently some people looking for loans on Prosper don't have a job. Some are retired, others are unemployed. It turns out that the highest returns are from those who are self-employed (9.8%) and employed (9.2%.) The unemployed and “other” folks have returns around 4%. It's even better if they make more than $50K. You don't want someone choosing between putting food on the table and paying you back.
6) Avoid Business Loans
Just like with Lending Club, you don't want to lend to people starting a business. Businesses fail, and when they do, so does the income of the person starting it. No income= no loan payments= loss of your investment. Returns for business loans are a couple percentage points lower than loans for other purposes (around 7%).
7) Reasonable Credit History
While the credit score doesn't seem to matter much, and delinquencies and bankruptcies don't make much of a difference, you don't want to loan to people that haven't borrowed much. Look for at least 5 open credit lines, 10 total credit lines, and a 2 year history.
This strategy is all from back-tested data. The future may not resemble the past, especially with Prosper continually tweaking their models. But investing without looking at the evidence seems stupid when the evidence is so easily found. I”ll keep you updated as I go along. If you'd like to tinker as well, click on one of the links on this page. It doesn't cost you anything and I get a small commission when you open an account.
Good introduction for new Prosper investors. As you say, people should really do some data analysis before investing their hard earned money. I would also add that while Prosper does indeed offer potentially higher rewards than Lending Club, the risk profile of the average borrower is also higher risk.
“and very low correlation with the rest of my portfolio.”
Doc, this statement is wrong and your article even proves it. When did Prosper collapsed last time? 2009. I.e. during the recession and downturn (at least till March) of the stock market.
When would people have the hardest time to repay their loans? During recession.
So, you should have very high risk correlation between your stock & Prosper portfolios. BTW, the same applies to high-yield bonds.
HSV_Climber-
Several comments-
First, you may be right. I probably overstated my position when I said very low correlation. The truth is there isn’t enough data to say whether long-term low correlation should be expected. These companies haven’t even been around a decade, and in their current version, have only really been around for 3 years.
Second, the theoretical correlation, in my opinion, should be quite low. These are very different assets from loans to governments and corporations and shares of corporations. Just as real estate acts differently and thus often (but not always) has low correlation with stocks and bonds, so should consumer loans have low correlation.
Third, using the example of “Prosper 1.0” to show high correlation in economically troubled times is not necessarily appropriate. I would argue that the poor performance of initial prosper loans was a flaw in their risk control models rather than a result of the relatively poor economy. My best guess (but again, limited data) is that this flaw has been corrected in Prosper 2.0 and Lending Club. Certainly returns are much more promising. I cannot, however, disprove your theory that this is simply a result of an improving economy. Although, if you think about it, there was plenty of unemployment and economic slowing in 2009 and 2010, yet the performance of loans taken out in that time period is quite good. It seems clear to me that the stock market is a leading indicator of the economy, but that performance of consumer loans would be a lagging indicator. Corporate profits fall, stocks crater, after a while people lose jobs, then they default on their consumer loans. If your theory of high correlation with the economy in general and thus the stock market is correct, then I would think we should see poor performance of loans taken out during 2009-2011 when we were dealing with quite high unemployment and sluggish growth. There is no evidence of that.
But like I said, I could be wrong, and you could be right. You pays your money and takes your chances.
Doc,
Good points, but you are missing survivorship bias. People who were able to qualify for Prosper / Lending Club loans had the jobs in 2009-2011. The others (i.e. the unemployed) were simply filtered out by Prosper 2.0 and Lending Club filters.
So, we should not see poor performance from loans taken out during 2009-2011, since these loans were giving to people who were qualified to receive them and the economy has been improving over these years.
On the other hand, you should see poor performance from the loans taken out before 2009 and from the loans that people will take out before the next financial crisis whenever it will happen.
So, don’t look for defaults on loans taken right after the crisis and during the recovery, but look out for the loans that would be taken at the top of the cycle and right before the crash. These loans will probably show the same behavior as the stock market, just like the loans taken in 2009 – 2011 show the stock market behavior as well.
That said, there is nothing wrong with playing around with Prosper / Lending Club and I might try to start playing with it as a bit as well. But we just need to understand the risks and very high correlation to the stock market / US economy.
That hypothesis is consistent with the data. Even if you dismiss the whole Prosper 1.0/2.0 issue and just look at Lending Club (who, as I understand it, didn’t have quite as dramatic a change in 2009), you see a return on loans taken out in 2007-2008 of 1.3% and on loans taken out in 2009 of 4.8%.
But the question remains, is it because of the economy, or inadequate risk control? I don’t think it can be proven with available data.
And keep in mind, that just because performance was relatively poor (compared to 2010 loans for example), doesn’t mean it was poor compared to other assets. Think about what your stocks and corporate bonds did in 2008? A return of 1.3% sounds pretty good in comparison, no?
I lost through Prosper. I enjoyed the process of loaning to people but some of them claimed bankruptcy, which I did not enjoy.
What years were you investing?
Excellent post but I was wondering if you could write a
litte more on this subject? I’d be very thankful if you could elaborate a little bit further. Thanks!
Hello, I think your website might be having browser compatibility
issues. When I look at your website in Safari, it looks fine but when
opening in Internet Explorer, it has some overlapping. I just wanted to
give you a quick heads up! Other then that, very good blog!
Not sure if this is spam or not, but it looks fine on multiple versions of internet explorer to me. Monitor width also has a significant effect on overlap. Glad it’s looking good on Safari. That’s a lot harder to deal with than IE.
I’m not sure where you’re getting your info, but good topic.
I needs to spend some time learning much more or understanding more.
Thanks for fantastic information I was looking for this
information for my mission.
I read only couple of posts so far and I am impressed! I will keep reading more! – One suggestion: Although the site has a beautiful appearance, if you are able to change the color combination of the background (currently beige) and font (black) to make the text stand out, it will become easier to read. Thank you!
Thanks for the tip. You’re not the first to mention it, and I’ve made a change today that should help. The ads will look a little funny, but the only solution seems to be going back and editing 300 posts separately.
I think that it’s unbelievable that people can approved for a loan without a job, it’s awful hard to pay back five grand with no income.