Saturday, August 22, 2009

Houses and Finance

You might have thought that my mad rants on finance were totally unrelated to a housing blog. turns out that housing and finance do intersect, in some ways that have turned out to be pretty unfortunate for home-owners.

There are a few financial bloggers who I really like - Mike Konczal is one of them. He wrote a series of guest posts on the The baseline Scenario (another really good finance blog) that are of interest to home-owners who might be asking the question: "how did Wall-Street-style price-insanity infect my local neighborhood?"

The Limits of Arbitrage. This is an idea that has also been explored for stocks, but the basic idea is that if enough people lose their minds and start buying up the price of an asset, it may be impossible for sane, well-informed people to bet against them, making money while forcing the price down. There can be two reasons why people can't "bet against" the bubble:
  1. There is no good way to make the bet. For example, if I believe houses are going up, I buy a house. But if I think houses are going to crash, it's a lot harder. If I don't already own a home, I can't sell a house.

    Since I can buy 2, 3, or 4 houses if I want, but I can only sell what I already own (1 house for most home owners) it's a lot easier to make the bet that houses are going up than going down.

  2. If you borrow money to bet against an asset, you can go very, very broke. The problem is this: if you bet against asset X, you lose money every time asset X gains money. Your lender isn't going to lend you money once you've lost more than your own share. Put it together and a large enough rise in asset X means you're broke.

    For example: Google is at $100 and I think it's going down. I have $1000 and I borrow another $1000 and short (bet against the stock). I now am short 20 shares. If Google doubles in price to $200 then my initial $1000 only covers half of those shares. At about this time the bank is going to ask me to unwind the trade and give them back their money, because if things get any more out of hands, I won't be able to pay the bank back at all.

    Of course, this sucks for me - if Google later crashes down to $10, I'll never see my money - the bank forced me to sell at the worst time because I ran out of credit! Why does this matter? It matters because people can use credit to bet that an asset will go up. If they can't use credit to bet that an asset will go down (without risking the bank pulling the rug out from under them) then the amount of money betting on a rise (amplified by borrowing) will dwarf the bets on a fall, and the people betting on the price going down won't be able to "hold down" the price.

Phew. So what does this mean for houses and stocks? It means that even if smart, clever people could have predicted the bubble (some did, it remain arguable if enough did to make a difference) those smart people might not be able to keep prices sane because of structural problems with the markets. The only thing you can do is get out of the asset class entirely while the bubble goes by and wait for things to calm down. This is possible for stocks but more difficult for houses.

Prepayment. Mike is a financial engineer, so his articles can get technical, but I think prepayment is worth looking at. Cynics like me might accuse the banks of speculating on rising housing prices when they wrote all of these lame loans, but how do you make the link?

In The Giant Pool of Money NPR financial reporters explore the idea that with low interest rates and a lot of international capital, there was a huge desire for mortgage debt. In Infectious Greed Partnoy argues that often derivatives were sold to clients who didn't know what the hell they were doing.

So you can put together an argument and say "oh - the banks made subprime loans because hedge funds, insurance companies, pension funds, and other pools of money wanted the derivatives that come from subprime loans, and were too stupid to see the risk." But this doesn't explain a strange market inversion, where sub-prime mortgages were in higher demand than prime mortgages, creating incentives for brokers to push them. (A broker might make a higher fee on a subprime loan, but only because the originator set things up that way - e.g. the broker gets paid more for subprime if the originating bank wants that loan...why does the bank want a subprime mortgage?)

Here Mike suggests an answer: the prepayment penalty in a sub-prime mortage changes the very nature of what a mortgage is. One of the weird features of prime (normal) mortgages is prepayment. Owners pre-pay when interest rates fall (making a re-fi a good idea). This really sucks for the lender. If I offered you a CD that has a 5 year term but will drop to a 1 year term only if interest rates fall, would you accept it? Heck no!

(If I understand my financial engineering, mortgages have "negative convexity". Basically the longer a loan, the more its value changes with interest rates. But a mortgage's time span goes down when interest rates fall due to prepayment. So a mortgage is a loan whose duration is longer when interest rates are going up. Since a loan's value (to the lender) is inversely proportional to interest rates, this is about as bad as it gets: a loan that changes its value a lot when its value falls but only a little bit when its value goes up. If this seems totally one-sided, it is...this one-sided lender-is-hosed pricing comes directly out of the one-sided nature of pre-payment. Prepayment is heads-I-win-tails-you-lose, with the winnings for the borrower and losses to the lender.)

But if you can soak your borrower with a 4% penalty every time they prepay and encourage the borrower not to run the mortgage to its full 30 year duration (by having horribly high floating interest rates after the first few years) we have something. We have a mortgage without this "negative convexity".

There's only one problem: it's a subprime mortgage - we might not get our money back. And now we can connect the dots. To the lender, a subprime mortgage has less interest rate risk but higher credit risk. And the credit risk of a mortgage has everything to do with housing prices. Ergo: the subprime mortgage as a speculative instrument exposing banks to rising real-estate prices.

(Of course we can see the problem now: as a way to speculate on housing increases, a subprime mortgage is a bit like selling a put option - you make a little money when you win but lose a lot when you lose. In the case of the mortgage, the most you make is the higher interest and prepayment penalties. The most you lose is a huge chunk of the mortgage. Which is exactly what has happened.)

No comments: