Regruntled.com

There is a new academic study of Prosper, the peer-to-peer loan platform, by economists Seth Freedman and Ginger Zhe Lin of the University of Maryland. Among the findings:

1. The average Prosper returns since inception were estimated at 6.05% with a 5.72% standard deviation and were trending up as credit quality continues to improve (see table 9 and figure 8.3).
2. The highest returns were in Near Prime loans (grades B-D) at an average of 8.27%, followed by Prime loans (grades AA & A) at 6.78% and sub-prime loans at 1.73%.

My own experience is that I am getting about 12.5% on AA and A loans. Mentally I discount my return to 10% to allow room for defaults, and attribute the extra 2.5% to luck. I suspect that it is not so difficult to beat the averages because the averages are lowered by clueless newbies bidding on loans that they will see as hopeless mistakes after they get a bit of experience.

Even though I think I could earn 9 or 10% on Prosper, I have not put in additional funds, mainly due to a lack of confidence that Prosper is growing fast enough to be viable.

See also:

• Prosper

On Tuesday, the day of the Montana and South Dakota primaries, I backed up the truck and shorted as much Hillary as I could. The pundits on TV were talking, not about what would happen, but how it would be staged. Obama’s superdelegates were timing their announcements so that “the voters” would put Obama over the magic number. And yet, Intrade gave Hillary a 6% chance at getting the nomination.

On Wednesday, I noticed that Hillary was down to a 4.1% chance of getting the nomination, but she still had a 6.1% chance of being President. Hello!? Do the math along with me, folks: Hillary has a 2% chance of becoming President without getting the Democratic nomination. Is this a market inefficiency, or Hillary going to run as an Independent? Can you say “arbitrage”? AR bih trahzh. I knew you could.

So the obvious trade is to buy back the short position on the nomination, and short the Presidency, for a net gain of 2%. The Presidency contract runs until November instead of expiring at the end of August, but still, 2% in months is 12% annualized.

I looked into the mechanics of the trade. Intrade has the peculiar practice of charging commissions on market orders but not limit orders. I’ve always used limit orders on Intrade, but this trade would require market orders, so I had to allow for commissions. I still had my Al Gore short position, which was linked to the Hillary short position for the purposes of calculating my margin requirement. After covering Al Gore and paying three commissions, the 2% would just about disappear, so I didn’t bother. Instead, I ran a single contract through the trade to verify that I understood the commission structure.

Intrade is an inefficient market. The anomalies are real, the opportunities are there. The mechanics are a bit complicated, but nothing a computer nerd can’t handle. Mostly I can’t believe how easy it is. Nice hobby.

Intrade gives Hillary a 10% chance at winning the Democratic nomination. Tim Russert gives her a 0% chance. Is there an opportunity here?

Hillary has two paths to the nomination. One is for Obama to self-destruct, and it has to be something a lot worse than Bittergate or Jeremiah Wright. The other is to persuade the super-delegates to take the nomination away from the first African-American to get within striking distance of the Presidency. The Democrats couldn’t possibly be that stupid. Or could they?

I’m looking for slam dunks with my Intrade account, and I’m not comfortable betting on the absence of stupidity. A few days ago I sold Al Gore short, of all people. Apparently people thought there was a chance that the Democrats would take the nomination away from both Hillary and Obama and hand it to Gore. I don’t think so.

Today I shorted a Hillary contract. Hillary and Gore can’t both get the nomination, so one or the other bet has to win. If Hillary somehow weasels the nomination away from Obama, my gains on the Gore contracts offset the losses on the Hillary contract and I break even. If Gore gets the nomination, I lose money, since I’ll lose more on Gore than I win on Hillary. If Obama wins, I make money on both bets, and more than if I had bet against only one candidate.

We’ll see what happens. As events unfold, and as the numbers change, I may adjust my Hillary/Gore ratio.

The New Hampshire primary is over, and the Ron Paul contract at Intrade is in free fall. Before Iowa, Ron Paul traded as high as 9, meaning that he had a 9% chance of getting the Republican nomination. After Iowa but before New Hampshire, he traded at around 5. Now he’s at 2.5.

I still say he should be trading at zero, but here we have to face a structural bias at Intrade. Intrade requires traders to put up capital equal to their maximum possible loss. Unfortunately, Intrade does not pay interest on small accounts, so there is an opportunity cost to every trade.

The Ron Paul contract runs until August 31st, about 8 months from now. At 2.5, someone on the long side puts up $2.50, which is the most he can lose if the contract goes to zero. The person on the short side, me for example, has to put up $97.50, which is the most I can lose if the contract goes to 100 (Ron Paul gets the nomination). The most I can make is $2.50, which is less than I could make by putting the $97.50 in a money market account for 8 months. When I shorted Ron Paul at 7.7, it was an attractive trade. At 2.5, not so much. Time to take the money and run.

Alternatively, I can wait for Super Tuesday and maybe cover at 1.5 instead of 2.5. I have no problem with waiting a month to pick up another point, but my whole thesis is that the Paulistas are behaving irrationally. Do I really want to bet that Super Tuesday will make them less irrational?

See also:

• Selling Ron Paul short 3
• Selling Ron Paul short 2
• Selling Ron Paul short

Ron Paul is down to 4.7 at Intrade (I shorted at 7.7). I didn’t expect this to happen until late January. OK, so I was wrong, but at least I’m wrong in a good way. My bet is paying off faster than expected. What happened?

Mike Huckabee happened. The only plausible scenario for a Paul nomination is a surge in the polls followed by wins in the early primaries. The come-from-behind, underdog story is there alright, but it features Mike Huckabee, not Ron Paul.

Paul’s supporters are already urging him to forget about the Republican nomination and run as an independent (Intrade gives the odds at 25%). His campaign has raised a lot of money, and money has its own momentum. An independent campaign would push the day of reckoning from January to November, allowing the true believers to remain delusional for a few more months.

Intrade also has a contract for Ron Paul winning the general election. I didn’t short that one because it was priced lower, presumably because of the possibility that Ron Paul would win the Republican nomination only to lose to Dennis Kucinich in November. If he switches to an independent campaign, I may be able short Ron Paul all over again with the other contract.

Speaking of delusional:

Zopa is a new social lending site. Zopa UK has been around for a while, and Zopa US is about to launch. (I’m on their email list, and got to do a bit of beta testing.)

I’m a lender at Prosper, which is working out OK so far, but I don’t have much faith in management. At Prosper, the problem for lenders is estimating the likelihood of default, so the emphasis is on diversifying, choosing borrowers wisely, and getting compensated adequately for risk.

At Zopa, the lender buys a CD, insured by the NCUA. Zopa then lends the money out to borrowers. Zopa handles the diversification, and (with its credit union partners) assumes the risk.

Zopa requires that borrowers have a FICO score of at least 640, which drops out the bottom 25% of the population. The current CD rate is 5.10% for 1 year, which is not bad. I have a CD at GMAC Bank, which is currently offering 4.50% for 1 year.

Zopa borrowers can post their sob stories, and Zopa lenders can “help” them by diverting part of their interest to the borrowers’ payments. This is the social part of social lending. For example, a lender with a $1000 5.10% CD could reduce his interest to 4.10% and allocate the extra $10 per year to help specific borrowers. It is possible that a borrower with a good story could get a free loan.

It’s nice to see some competition for Prosper. Zopa has a completely different business model, and will attract different borrowers and lenders. There’s no reason they can’t coexist. Personally, I can’t get too excited about Zopa, but the next time I’m in the market for a CD, I’ll consider them.

By the way, ZOPA is an acronym for “zone of possible agreement”. If a borrower is willing to pay as much as 10%, and a lender is willing to accept as little as 9%, then the range from 9% to 10% is the zone of possible agreement. If there’s no ZOPA, there’s no deal.

The Predatory Lending Association is a spoof web site (or IS it?). Their motto is: “Helping payday lenders extract maximum profit from the working poor”.

And extract they do! Advance America offers loans with APRs as high as 495 percent. My experience as a lender on Prosper, and as an owner of junk bonds, gives me a certain amount of sympathy towards lenders. 495 percent sounds pretty good. Can I get some of that action? As a lender, not as a borrower.

Here is a quick comparison between a payday lender and a regular bank:

I’d have to say that payday lending, from an investor’s point of view, is comparable to banking. The outrageous fees are offset by lower volume, higher expenses, and higher default rates. Of these two, I’d rather own some BAC because I figure AEA has more regulatory risk.

Finally! After three and a half weeks, my account at Intrade is funded, and I can place an order to sell Ron Paul short.

This has been quite an adventure. My credit card company refused to process a charge from an offshore gambling operation. I could have sent a bank wire, but the transaction costs were prohibitive. I had to airmail a $250 check to Dublin, where Intrade presumably put a hold on it until it cleared. (Intrade could have handled this better. An email when they processed the check and started the hold would have been nice. I had pretty much given up on the whole thing.)

Well! While I’m composing this post, I get an email from Intrade confirming that I’ve sold 25 Ron Paul short at 7.7. So now they’re responsive?

The contract expires on August 31, 2008. If Ron Paul IS NOT the nominee, I make $19.25, which, over 10 months, is about 9% annualized. If Ron Paul IS the nominee, I lose $230.75, which is horrible.

Real Clear Politics, one the few sources where I can find a poll that lists Paul separately, shows him at 3.3%. Keep in mind that someone with 3.3% of the votes gets 0% of the nomination.

Meanwhile, Ron Paul has been very successful at raising money. I watched the last Republican debate, and Paul was the only candidate that was booed by the Republican audience. At the end of the show, he won the instant text messaging poll. I can only conclude that Paul’s people spammed the text message number to create the illusion of support. The idea must be that if they can convince enough people that Paul IS popular, then Paul will BE popular. Sort of a high-tech, magical thinking, bandwagon strategy. Very Boomerish.

I suspect that something similar is going on at Intrade. Here is an article showing Ron Paul’s Intrade chart and arguing that Intrade is a better predictor of the nomination than the polls. Nonsense. If Intrade is being manipulated, and the polls are not, then the polls are better predictors.

I got spammed with a solicitation for True Wealth, one of the top five financial newsletters in the world. Yeah, right. The email touts the “Pinchot Plan”,

“…an unusual investment most Americans have never heard of… which was made possible by a former U.S. Presidential Advisor and Governor of Pennsylvania.

This investment is safer and more profitable than stocks, bonds, real estate, mutual funds, gold… and just about every other asset on the planet.”

Gifford Pinchot was the first Chief of the US Forest Service, and advocated scientific forest management. Very Progressive, very Teddy Roosevelt. He has a National Forest named after him. The “Pinchot Plan” is investing in publicly traded timber companies. Minus the huge steaming pile of marketing, that’s all it is. No need to subscribe to the newsletter. Anyone can do this on his own.

I mentioned Plum Creek Timber (PCL) in an earlier post, but here are a couple of other possibilities: Rayonier (RYN) owns 2.5 million acres and has a 4.27% dividend. Potlatch (PCH) owns 1.5 million acres and has a 4.25% dividend. All three companies are REITs. I own some PCL right now, but both RYN and PCH are plausible alternatives.

Virtual timber is a decent investment. Buy it when it’s on sale. Better than stocks? Better than gold? Well, no, not better, just different. It’s a good diversifier.

MicroPlace, an eBay company, is a new online micro-lending platform. The idea of micro-lending can be traced back to Lysander Spooner, but the most well-known implementation was by Muhammad Yunus and Grameen Bank (2006 Nobel Peace Prize).

Foreign aid involves transferring relatively large amounts of money from government to government and is subject the usual inefficiency and corruption. Micro-lending places small amounts of money directly into the hands of specific poor entrepreneurs, for example $500 to a woman in Peru who wants to expand her pig farm. (Is a Peruvian pig an Oinca?)

The internet, of course, can help match people who have money with people who need money and there are several models available. Kiva matches lenders directly with the borrowers, but Kiva loans pay no interest. Kiva lenders diversify by spreading their money over several borrowers. MicroPlace matches lenders with organizations in specific countries. The organizatons spread the money out over many borrowers and pay the lender 2 or 3% interest. Is this a better model? Will below-market monetary interest attract more capital than individual stories with human interest?

Are loans better than grants? At Save the Children, one can sponsor a child for $28 a month. On the other hand, at current money-market rates, $28 in foregone interest works out to about $7000 in Kiva loans or $19000 in MicroPlace loans (to an organization that pays 3%).

We can compare micro-lending to peer-to-peer lending platforms like Prosper. At Prosper, the borrowers are not third-world entrepreneurs, but fellow Americans. The interest rates are much higher and the default rates are much higher. It is possible to earn above-market returns at Prosper, but in practice many lenders end up with below-market returns.

I say let many flowers bloom. Some of these models may not work, but there’s no reason that several different models can’t thrive at the same time.