Wednesday, July 25, 2007

Borrowing money to lend on Prosper: Wise or Foolish?

Frequently you will see listings on Prosper where someone plans to take out a loan for the purpose of reinvesting the money back into Prosper at a higher interest rate. The idea is that you borrow money at a low interest rate and then reinvest it at a higher rate to make money on the carry, or price differential.

The concept is simple, and it is how banks earn much of their money. If they can pay less than 1% on money deposited in a checking account and then lend out the money at 8% then they are making money on the difference. This is also the idea behind the Yen carry trade, a popular investment technique where money is borrowed in Japan at very low interest rates and then reinvested in developing markets that have much higher interest rates. The risk, of course, is that currency fluctuations could wipe out any gains that you make.

At Prosper there is also a significant risk to this borrow-to-lend strategy. Before we look at the risk though, lets look at a wildly optimistic scenario:

Suppose you were a AA borrower with perfect credit. You take out a $10,000 loan at 8% (this is a below average interest rate for a AA loan of this size, but we are being optimistic). Then, lets assume you get really lucky and avoid defaults and late loans while lending at an average of 19% interest. Now you are probably thinking, wow I just made 11% on the carry. A cool $1100 of free money! Lets take a closer look to see if that is the case.

When you initially take out the loan you have to pay 1% in loan closing fees ($100 in this case). Then, if you are lending at 19%, you are in the B-HR credit grades on your loans so you will have to pay a 1% servicing fee on those loans (there goes another $100). Also, you have to wait for the money to transfer out and back in to Prosper. Then, once you bid on loans you have to wait for them to close. During this time, which will probably be about 30 days, you are not earning any interest on your money. That reduces the rate of return for the first year from 19% to 17.4% (11 months at 19% and one month at 0%).

So, instead of making 11% you are now at 7.4% in earned interest (we subtracted 1% for the loan closing fee, 1% for lender loan servicing fees, and 1.6% for idle money). Well, that's not too bad you think, it is still $740 in free money, right? Not so fast...

There is one more thing to consider: taxes. Prosper is not a tax friendly investment. Interest paid on money borrowed from Prosper is not tax deductible. To make things worse, money earned on Prosper is taxable as ordinary income. Lets assume you are in a 25% federal bracket and 7% state bracket. The federal government sees that you earned $1740 on Prosper (they are looking at just the lender side). In this case, your taxes eat up $556.80 of your earnings. Now you are down to $183.20 in first year earnings, and remember this is the optimistic scenario.

Now, lets look at a more likely scenario using the same methods and example as above. Only this time we will assume that you have a 12% default rate (this is a better than average default rate for loans at a 19% interest rate). Now, instead of earning money you post a loss of a few hundred to as much as $1000 in your first year of investing depending on when the defaults occur.

The worst case scenario is that you don't diversify, overweight in HR loans, or are particularly unlucky on your default rates. Under this scenario it would be easy to see losses well in excess of $1000 during your first year.

Looking at the numbers, the potential gain is small, and the risks are too great to make Prosper a viable place for making money on the spread in interest rates by borrowing to re-lend the money. It is possible to earn a reasonable rate of return at Prosper, but borrowing at 8% to re-lend makes it unlikely to be a successful investment.

What really makes me cringe is when I see postings from people with poor credit scores who are trying to borrow at 12% or 14% to re-invest the money back into Prosper. They are almost sure to lose money on their investment.

I have loaned money to one person who was trying the borrow-to-lend reinvestment strategy on Prosper. I loaned him money at 8%, which he reinvested in loans that have an average interest rate of 16.28%. He re-invested the full amount of the loan ($3000), and a nearly a year has passed since he made the investment. He now has 4 late loans, and according to LendingStats his estimated ROI adjusted for late loans will be 7.30%. That means he will be lucky if he breaks even on his investment.

17 comments:

Unknown said...

While I agree that borrow-to-lend is a poor strategy, your tax comments are incorrect. There is no reason why one would not be able to deduct the interest cost on their prosper loan against interest income from prosper investments. The interest would be listed as "investment interest expense" on IRS Schedule A, line 13.

See IRS Publication 550:
http://www.irs.gov/publications/p550/ch03.html#d0e7288

Anonymous said...

I would just like to agree with the tax comment. I actually took this problem to my tax advisor, and he set up a few scenarios. While I agree it's a thin-margin game, if you can borrow and invest in something quality, it saves you the need to save up $25k to invest in nice fixed-income vehicle.

Anonymous said...

I also think there is a problem that you can't deduct the losses from a default, yet you still are going to be paying interest on income in the form of interest. Do you have any idea if you can deduct defaults?

Jesse said...

If it is more than a couple of hundred dollars you are probably best off setting up prosper as a business. Prosper should deduct the losses in bad loans from your return, so there may be a delay of about 5 months(they sell loans that are 4 months late).

Chrisfs said...

You can certainly deduct the losses from a default. Prosper issues the necessary tax forms.
However it is a very thin margin and keep in mind that investment interest expense is a Schedule A item, which means that it's worthless unless you have more deductions than the standard deduction to begin with.

Nick said...

Does anyone know if there is a calculator that calculates reinvested interest for loans that you give out on prosper. Like if I give out $10,000 for example. Then reloan payments made to me. And continually reloan as payments come in. How much money I will have in 20 years for example?

Anonymous said...

It amuses me that arbitrage is so often billed as a "fool proof" way to make money. Arbitrage works well on paper, but in the real world its hard to deal with such razor thin margins.

When you're not playing with your own money, things can get ugly real fast.

Anonymous said...

Randy, I agree. The potential return is very small under the borrow-to-lend arbitrage plans. The risk, however, is high.

Anonymous said...

You can deduct your losses as a capital loss on Schedule D to offset ordinary income up to 3,000 per year on any loan that goes bad. You can then carry any unused losses forward or possible deduct it entirely if you have capital gains.
If you borrow money to lend. The interest from borrowing the money is deductible on Sch. A as investment interest expenses.

Anonymous said...

This is an old thread, but it's interesting to note the outcome of the two borrow-to-lenders you link to at the end of your article.

Both have paid their loans in full earlier this year. But Eric's site shows the C who borrowed at 14.66% with an ROI of -8.80%, and the AA you loaned to at 8% has an ROI of 0.28%.

Didn't work out for either one, but worked out fine for the lenders who supported their futile effort.

Anonymous said...

Good point! I think that borrowing to lend is a poor investment strategy. However, investing in "blenders" can be safe. Borrowers who lend often pay back there loan even if they are losing money in Prosper. I have lent money to several blenders.

Anonymous said...

Just an FYI for the interested: The loan carry scheme is NOT actually how banks make money. Banks create money when they loan it to you, and they make interest as "free" gravy on this money. The principal, similarly, winks out of existence when you pay it back to the bank. The bank is limited in the number of loans they can make by regulations that limit the reserve fraction - that is, to loan $100 they must usually have about $10 on deposit. The upshot is...even real, successful banks can't really make money this way. They typically make money by charging interest on money that exists temporarily, by charging fees for services, and by increasing the number of loans they can make by attracting depositors. There's no reason to think you're being some sort of hotshot, professional banker by playing the carry trade. It's more akin to trading options - something with very high risk of loss and even liability relative to the possible return.

Anonymous said...

Has anyone run any analysis on the loan data? Like how the size of the loan affects the default rate? Based on my research I came to the very obvious conclusion that higher balance = higher default rate. By using this very obvious conclusion, could one not then generate a slate of loans which are theoretically very low default probability?

I think yes, but one cannot borrow to lend with this strategy, since as the blog points out the margin is too thin (I am seeing about 13% return on 3% default rate over 700 loans). Oh and this is inside a self-directed Roth IRA (no taxes, not even UBIT, muhahahahaha).

Anonymous said...

Finance guy - you may want to check out this article about risk and Prosper returns.

Anonymous said...

Finance guy - I just checked out your website. You do reduce your risk by loaning to C's and D's that request the min loan amount $1,000. I look forward to reading your research.

Anonymous said...

I finished the basic analysis. The average is about 13% return over the period, with about 3% default over total loans, or about 5% over the currently active loans. Sample size = 893 total, 585 active, so there have been enough loans to satisfy the t-stat.

There is a problem: it doesn't scale progressively. However, up front scaling is OK: if you have a big chunk now you can use it all at once, but if you want to increase the bids as you go, it won't work.

Anonymous said...

Financeguy - thanks for letting us know about your results. I took a quick look at loans of $1,00 vs other small loans over $1,000. (1K to 2.5K) There was a huge difference in the default rates. Perhaps many of those borrowing $1K actually need less than $1K?

A Great New Idea in Online Investing