Earlier this week, I defended the Prosper’s business viability. I spent two posts rebutting the idea that it might not be a good investment. Just yesterday, I discussed the Prosper mistakes I’ve made. Today, I will share with you what I’ve learned. Hopefully 5 minutes from now you’ll be on your way to 14% return after adjusting for risk…
Step #1: Know Where You Stand
It sounds so basic, but it can be extremely difficult. As I looked back on the mistakes I made yesterday, the biggest one was thinking that Eric’s Credit Community’s Expected Return was a good gauge. So now that we know that it doesn’t work, what does work? I’ve found that RateLadder’s IRR Excel Spreadsheet is the best way to gauge how things are going. (Note: If you download the worksheet be sure that you install the XIRR function from your Microsoft Office CDs. It’s not in the standard install.) It’s worth following his IRR game with other lenders. If nothing else, you’ll be able to track my progress. My current average IRR is 12.76% which is remarkably better than the 9.91% it was just over a month ago on Feb 17th. I think this may be more luck than skill at this point. For too long, I was making E loans at 29% and maybe it hasn’t caught up with me yet. Either way, it’s a much, much better gauge than what I had before.
Step #2: Find Better Loans in No Time at All
Familiarize yourself with the Prosper’s Standing Order. After doing some reading, I’ve realized that the best way to get in the better loans is to have a bid in before the loan is even created. I only put in standing orders for automatic funding loans. Otherwise, you could be tying up money in a loan that could get bid down – or even find yourself outbid. Here are a few good loans I’ve had.
Step #3: Use the Tools at ProProsper.com
I don’t think the Prosper community is really aware of ProProsper.com. I had the great pleasure of being a sounding board throughout the initial development. It’s an entirely free site full of great time saving tools. You’ll have to sign up, but it’s worth it. Here are some of the tools I use the most:
- My Cashflow – This tool is essentially the same as the IRR game discussed above. I prefer to do it in an spreadsheet because it’s currently faster than the web version.
- Listings – The listings on the left navigation doesn’t jump out at you, but click on it and then click on the “Views” drop down and you’ll find some great options… One of them is “active loans where they offer a rate greater than the average.” These views compare currently active loans to loans of similar grade, DTI, and loan amount requested over the past 120 days. The more standard deviations (STDEV) above the better.
- Loan Rate Analyzer – This is related to the listings tool we just mentioned. Fill in the information in this form and you can find out the average interest rate of that loan over the last 120 days. I like to bid in loans that will give me a better rate than average.
- Standing Order Analyzer – This is just like the Loan Rate Analyzer, except that it will give you information for multiple credit grades.
- Query Analyzer – This is my favorite of all tools. You can query the database in free form. If you did a few of the “Listing” views above you’ll note that ProProsper gives the query that’s performing. Here are my two favorite queries to run…
- Query #1 – This query will give you loans that are (a) active loans (b) close in 24 hours (c) are not HR grade (d) sorted by the difference of the current lender rate and rate of similar loans over the last 120 days:
SELECT [listing].[listingid] AS ‘listingid’, ([listing].[startdate]+[listing].[duration]) AS ‘scheduledenddate’, [listing].[amountrequested] AS ‘amountrequested’, [listing].[creditgrade] AS ‘creditgrade’, [listing].[debttoincomeratio] AS ‘debttoincomeratio’, [listing].[lenderrate] AS ‘lenderrate’, [listing].[fundingoption] AS ‘fundingoption’, [listing].[listingnumber] AS ‘listingnumber’, rtrim(ltrim(SUBSTRING([listing].[title],1,80))) AS ‘title’, [listing].[percentfunded] AS ‘percentfunded’, [listing].[status] AS ‘status’, [listing].[borrowerrate] AS ‘borrowerrate’, [listing].[likewtavg] AS ‘likewtavg’, [listing].[likewtavg1stdev] AS ‘likewtavg1stdev’, lenderrate-likewtavg as goodValue FROM [listing] [listing] WHERE ([listing].[startdate]+[listing].[duration])>getdate() AND ([listing].[startdate]+[listing].[duration])<dateAdd(Hour,24,getdate()) AND [listing].[lenderrate]>[listing].[likewtavg] AND [listing].[likenumloans]>=10 AND [listing].[status]=’Active’ AND [listing].[creditgrade]<>’HR’ ORDER BY ‘goodValue’ desc, ‘scheduledenddate’ - Query #2 – This query will give you loans that are (a) auto-funding (b) are not HR grade (c) sorted by the difference of the current lender rate and rate of similar loans over the last 120 days: SELECT [listing].[listingid] AS ‘listingid’, ([listing].[startdate]+[listing].[duration]) AS ‘scheduledenddate’, [listing].[amountrequested] AS ‘amountrequested’, [listing].[creditgrade] AS ‘creditgrade’, [listing].[debttoincomeratio] AS ‘debttoincomeratio’, [listing].[lenderrate] AS ‘lenderrate’, [listing].[fundingoption] AS ‘fundingoption’, [listing].[listingnumber] AS ‘listingnumber’, rtrim(ltrim(SUBSTRING([listing].[title],1,80))) AS ‘title’, [listing].[percentfunded] AS ‘percentfunded’, [listing].[status] AS ‘status’, [listing].[borrowerrate] AS ‘borrowerrate’, [listing].[likewtavg] AS ‘likewtavg’, [listing].[likewtavg1stdev] AS ‘likewtavg1stdev’, lenderrate-likewtavg as goodValue FROM [listing] [listing] WHERE [listing].[fundingoption]=’Close When Funded’ AND ([listing].[startdate]+[listing].[duration])>’2007-03-12 20:57:32.421′ AND [listing].[lenderrate]>[listing].[likewtavg] AND [listing].[likenumloans]>=10 AND [listing].[status]=’Active’ AND [listing].[creditgrade]<>’HR’ ORDER BY ‘goodValue’ desc, ‘likewtavg1stdev’, ‘lenderrate’
- Query #1 – This query will give you loans that are (a) active loans (b) close in 24 hours (c) are not HR grade (d) sorted by the difference of the current lender rate and rate of similar loans over the last 120 days:
- With these two queries you can’t simply take the top loans and bid on them. You’ll want to look at the DTI and make sure it is reasonable. If you are the type who like to read what people read, you can do that as well. However, these queries should filter out most of the undesirable loans and reduce your time looking for Prosper loans to just a couple of minutes a day (if that).
In the end, by using these loan analysis tools, you can make sure that you always get in loans that are significantly above the average loan. Earlier this week, we determined that the average A-D loan returned nearly 10%. I believe that I’m now able to get in loans that are enough above average to earn up to 15%. If you want lend money on Prosper.com now.
Hey Lazy —
What do you think about turning your queries into RSS feeds? I have 5 already at ProProsper: (Under Appreciated, Homeowners, 1 Sigma, 2 Sigma, and 3 Sigma)…
You would have naming rights and page explanation (and link) editorship.
I had just recently posted some thoughts about Prosper on another site, but can not seem to find them. Anyway, my two cents about Prosper:
The GIGANTIC WINNER is…Prosper
– no risk
– all reward
The RUNNER UP is…the Borrower
– pool of available lenders where more traditional methods have failed
– will always get a better deal than they could have gotten otherwise
– no pre-payment penalty
– variable underwriting process (many loans completed with minimal information). For example, did you know the debt/income ratio is based upon income entered by the BORROWER? As I understand it, there are no screens to check this (unless you asked for a W-2) Someone correct me if I’m wrong on this.
The BOOBY prize goes to…the Lender
– I believe there is a bias in the Prosper borrowers market and that historical default rates do not apply
– Variable underwriting process (less documentation / verification can only benefit the borrower, not the lender)
– Immature market and business – Prosper is continuing to improve transparency in its reporting and tools available, but the whole concept / model is in it’s early phases
– Very difficult to truly diversify – both because of the biased market I mentioned above and that people do not usually have the funds to do so
I was also intrigued by the concept of Prosper when I first learned of it. I spent some time in the forums and came away with the strong impression that it’s not really a means to make money…but rather “fun”…like GAMBLING.
In looking at other blogs who have had some experience with Prosper, my overall perception has not changed (small monies invested, nothing they would mind “losing”, it’s fun, etc)
If you like the to “play Prosper” on the chance you might “win” (again, just like gambling), then I don’t really see the harm in it. If you think that at this stage of the process it’s a reasonable means to create superior returns on your investment (verus other alternatives)…I disagree (risk / return equation not worth it).
May “Lady Luck” be on your side!
Great info.. I have been looking for ways to evaluate risk with Prosper. Thanks for sharing what you’ve learned on it.
I took a look at the “picks” available through the various screens at ProProsper and found many of them to be quite scary – but, to be fair, no more scary (and not much more functional) than the screens you can already run via Prosper.
The main “scariness” in selecting loans to invest “by the numbers” (ie, through some third party (numeric) screening process or through Prosper’s standing orders) is that neither can take into account the “duh” factor. Automated agents simply do NOT read narrative text in the listings.
If you’re interested in dumping >$10k a month into Prosper, the time involved in finding good loans may eat up some of your return – sure. But this assumes you’re following the “diversify at all costs” line touted by Prosper. Many lenders seem to be finding that putting larger amounts of money into a smaller number of much higher quality loans offers a better return than dumping little chunks into a wide variety of loans via SO in order to “diversify”. Diversification into a large pool of loans of varying quality is… not the greatest idea ever.
Until someone invents an agent that can read listings (and filter out listings like “No job, but God will provide! Thank you for bidding!” or “Great business opportunity! I’ve run the numbers and have decided that opening an online poker site will make me rich by next Tuesday! I just need $5k to get the ball rolling!”), there is definite edge to be had in reviewing listings manually for financial innumeracy, illiteracy, dishonesty, crackpot-ism, pipe dreams, plain old bad ideas, and other things that would make a loan officer blush – and politely show you the door.
-t
Pfodyssey has some good points. I’ll address them in a future post and come back and link to it when it’s done.
On March 30, 2007, Prosper locked down discussion of the largest Prosper group, and banned several users from their forums for criticizing the leader of that group, claiming ‘This thread has been locked because it violates the forum policies of “Baiting other members into a fight or displaying other aggressive behavior” and “Being mean-spirited, unruly, or rude.”‘. Discussions of other prominent groups were not locked, leading many forum users to speculate that the lockdown was in response to complaints from the leader of that group.
By running screens you are essentially building a quantitative index portfolio of loans. The more diversified you are the closer you will be to market returns with similar risk factors. Since prosper borrowers in many cases were unable to get the loans from banks, this means that one by definition gets below market returns. In other words, if there was a risk-adjusted economic profit, the banks would have picked it up.
I think the prosper idea, which is beautiful in and of it self, went awry along the way. I believe the original idea behind prosper was that borrowers formed a group with their close friends and the entailed support and moral responsibility. E.g. if a borrower started being delinquent it would reflect badly on the group and the group would step in and get borrower on the right track. This “moral support” would be the thing that could be turned into economic profit for the lenders.
Instead I see a lot of “single” borrowers with no moral backing who see it as a way to get an easier loan. These borrowers won’t think twice about defaulting on a loan on people they don’t know even if those people are not big banks.
I also see lot of groups where the point is simply for the group leader to score an easy 0.25% e.g. they’ll list anyone who asks. I have lent to some groups where the group leader actually took time to verify W2’s etc. These have generally been more profitable. I think the message with prosper is that you don’t get something for nothing. Risk adjusted, prosper loans have the same return as junk bond ETFs (check out lending stats) e.g around 8% APY, except prosper is less liquid and requires more time. If you don’t risk adjust portfolios, you will, not surprisingly, see that some portfolios are doing really well, whereas others are doing really bad.
Using automatic funding will get you the loan faster.