Free Updates

Exclusive content to your inbox for FREE!

Previous Borrowers on Prosper are a Great Investment

by Peter Renton on May 13, 2011

On Prosper there is one selection criteria that I had all but ignored until recently: previous borrowers. These are people who are on to their second (or third) loan with Prosper. When I dug into the numbers on Lendstats I could see they may be the very best group of borrowers for investors.

Now, I am not just talking about all previous borrowers. I am talking about those that have consistently made on time payments for many months or years. It stands to reason that someone who has always made on time payments would be a good bet for a new loan. The numbers confirm that assumption.

Filtering the Prosper Loan Database

Take a look at this table. I took all the loans from Prosper 2.0 (post quiet period: July ’09 to present) and added several filters to see the impact on the ROI.

All loansInq - 1, DQ - 1C-HR LoansC-HR PB Loans
Est. ROI9.8%10.1%15.0%22.7%
Total Loans1045674774130562
Avg Interest Rate19.1%17.7%26.6%26.5%
Avg Loan Age9.39.38.78.7
Defaults3412031515
Default Rate3.3%2.7%3.7%0.9%

Let’s take you through each column. The first column is for all loans originated from July 1, 2009 through May 2011. As you can see there has been a total of 10,456 loans giving an estimated ROI of 9.8%. Then in the next column I added two filters, a maximum number of inquiries of 1 and a maximum number of current delinquencies of 1. We want to filter out those people who may be out shopping for a lot of credit and also those who may be starting to get into financial trouble.

Then to bump up the interest rate and ROI I took that same group and just looked at all the loans for borrowers with a Prosper Rating of C and below. You can see that the estimated ROI jumped from 10.1% to 15% but defaults went up as well.

Default Rates Under 1%

Now, here is where it got interesting. When I included only good previous borrowers (PB) you can see the ROI jumped dramatically. By good borrowers I mean those people who have had at least 24 months of payment history on their previous loan with a maximum of one late payment. Here is a link on Lendstats to that query. What surprised me most was that the default rate dropped from  3.7% down to 0.9%. Before someone says that these loans are still very young and defaults will likely rise, I do realize that. What interests me most is how much better good previous borrowers are relative to the loan pool as a whole. This is why I believe they are good loans to invest in.

After doing this analysis I have setup a new automated plan on Prosper to find every loan that meets these criteria. I am so confident that this is a good way to go I have bumped up my investment per loan to $50. But you don’t need to take my word for it. Explore the numbers for yourself. Ken from Lendstats has several new filters (PB12, PB25, PB36) based on length of payment history of previous borrowers. Or you can make up your own filters like I did.

Unfortunately Lending Club does not offer this as a selection, so we can’t chose loans based on previous borrower history there. Although I am guessing they will add that at some point. I certainly hope they do.

Anyway, I will be curious to hear what others think. Have you been investing in previous borrowers on Prosper? If not, do you think it is a good idea? Let me know in the comments.

Like what you read?
Then please join over 1,000 people who receive exclusive weekly p2p lending tips, and get a FREE COPY of my ebook, Understanding Peer to Peer Lending. Just enter your name and email below:
C. Jensen May 13, 2011 at 12:30 pm

Wow, very interesing. I had never thought about this before now. It’s encouraging to see that there are a lot of these notes out there to choose from. Time to set up a new filter. Thank you.

Peter Renton May 13, 2011 at 12:39 pm

@Carl, You’re welcome. I have discovered these loans are always snapped up quickly so an automated plan is a good way to go.

Bilgefisher May 13, 2011 at 12:42 pm

That’s my primary criteria. I ran the numbers via lendstats on those with 10, 20, and 30 payments. The numbers get better with each one. I then further filter it based on loan quality. I found E loans with 30 plus payments and 0 late payments greater than 31 days to be the best. I also do not count out folks that have been late a few times. I only rule out those that are really late. Lowering that default rate increases returns substantially.

Hard to say how well this will work. Time will tell. I find many people with crappy credit scores and low loan quality have low credit score over something stupid like a missed cell phone payment or hospital bill. Those same people pay like clockwork on almost everything else.

One caution though, I am leery of those who made great payments on a 2k loan and are now asking for a 15k loan though. The payment difference can break some people.

I’m glad you pointed these returns out, I just hope that the good loans don’t get snapped up by incoming institutional investors before I can bid on them.

Jason

Peter Renton May 13, 2011 at 2:10 pm

@Jason, Good point on the small loans, that is another thing to look out for. The trouble with E loans with 30+ payments is that it vastly reduces the pool of available loans. Don’t you agree? My expanded selection criteria only pulled up 10 new loans on the platform (out of 570) yesterday.

The other point I neglected to mention in the article was that I added one more additional selection to my automated plans. I chose people with a credit score that is at least equal to the their credit score on the previous loan. There is no lendstats data to back this up (Ken, do you want to add this?) but I feel that it makes sense.

While there is no way to tell how this will work in the long run, I am confident of a real world return of 15% or more based on this strategy.

KenL May 14, 2011 at 8:56 am

Good idea Peter, I should be able to add a change in credit score criteria. Don’t expect anything right away though.

BTW nice post, I like the way you relatively compared the default rates.

@ Bligefisher, I’ll have to check out those criteria of yours, they sound effective.

Peter Renton May 14, 2011 at 3:24 pm

@Ken, Thanks. I really appreciate the work you have done here. I think this new filter for previous borrowers may prove to be the most lucrative one of all for investors.

Dan B May 15, 2011 at 8:09 pm

@Ken L……..How often are the numbers on Lendstats updated?

Bilgefisher May 16, 2011 at 8:53 am

@Peter. You are very right, it does limit the number of available loans. I usually only see 15-20 each week that fit my criteria and only 4-5 that are really good. But that gets back to our differences in the number of loans. I’m okay with a smaller number of loans. That may not fit your investment strategy. I think you mentioned upping your investment on those small number of loans while still investing in other loans. That’s a great way to get the best of both worlds.

I would expect to see a higher number of these loans as prosper 2.0 ages.

Peter Renton May 16, 2011 at 9:55 am

@Dan, I believe that Lendstats data is updated daily – it pulls in a feed from both Lending Club and Prosper automatically.

@Bilgefisher, It you are not adding any new money then I think 4-5 new loans a week is certainly manageable. The problem as I see it is when you are putting new money to work. If you want to add another $10,000 then it would take many months to invest at the rate even at $100 per loan.

Bilgefisher May 16, 2011 at 12:38 pm

@Peter. You make a good point. Other than investing $500 each in those previous borrower loans and the rest on the loans the fit your other criteria, there is not much you can do to take advantage of those few loans.

KenL May 17, 2011 at 11:16 am

@Dan, LendStats gets updates daily, but it’s usually a day or sometimes 2 behind. Also once in a while the export files don’t get updated from the Prosper side, then I’m at their mercy. That hasn’t happened however for a while.

If you’re asking because you have a loan that shows as current in your Prosper account but Late(<15day) at lendstats and you're wondering when it will show current again at LendStats. I've got 2 answers for you. I THINK Prosper makes those loans current as soon as a payment starts processing. Then, if the payment fails it will go back to late. I think however at LendStats that late loan will remain late until the payment is accepted. Also, at LendStats the <15 day late loans sometimes remain at that status until the next billing cycle payment is made. This is just the way the data is in the export file. I could write some complicated code to try and account for that, but actually I don't mind if the <15day late status lingers because there is a heightened chance that those loans will go late again anyway. And actually I would rather slightly under predict ROI than over predict it because I do not want to feel responsible for any ill-advised investing based on over predictions of ROI. If you want to see what your ROI is with that loan current, you can set the loss factor for <15day late loans to zero. Then your ROI will make a nice jump up. If you do have such a loan :) , and it is now current I would recommend selling it at a 5% discount.

Peter Renton May 17, 2011 at 2:13 pm

@Bilgefisher, I think $500 a loan is fine on a portfolio of $100K or more but anything less than that and I think you are risking too much.

@Ken, Thanks for clearing all that up. For my own data I have found Lendstats to be spot on. But I have been lucky this far and I have only had two notes go <15 days late and both have gone back to current now.

Dan B May 17, 2011 at 9:10 pm

Thanks Ken for the info. It was more of a curiosity question as I noticed that there appeared to be a “lag” between your numbers & the official numbers.

Moe May 18, 2011 at 2:14 pm

Nice, I figured this one out about a year ago :) but I left Prosper because they don’t allow selling late notes, not that I have any late notes in my account, I just need the control.

Peter Renton May 18, 2011 at 2:29 pm

@Moe, Good for you. Prosper seems to be much more restrictive when it comes to the secondary market. Not sure why, but Lending Club allows more states and more flexibility when using it.

{ 2 trackbacks }

Previous post:

Next post: