David Dayen has a piece in The New Republic about the trend toward big investment vehicles buying up single family homes to rent them out. I’m skeptical that this business is going to work out, but moderately glad that someone is giving it the old college try. But one way or another, I wanted to speak up in defense of securitization as a business practice:
Data from the Federal Reserve Bank of New York shows that securitization inevitably leads to riskier behavior. There’s no reason on earth financial institutions should be able to convince investors again that, through sophisticated alchemy, they can slice the rental revenue pools into tranches and guarantee returns no matter the vacancy rate or the economic climate. But we’ve seen this movie before, and the first one ended rather badly.
Here’s the basic issue. Say you have enough money to buy 100 houses. If all 100 of those houses are in Atlanta, you’ll have a much easier time scouting the properties for potential purchase and managing them. But you’ll also be exposed to a lot of Atlanta-specific risk. Something bad could happen in that particular metro area. But if you take those 100 houses and combine them with 100 houses in Chicago and 100 in Phoenix and 100 in Riverside County and 100 in Sacramento and 100 in Miami and 100 in Providence and 100 in Las Vegas and 100 in Houston and 100 in New York, then you have a pool of 1,000 houses that’s much less volatile than the original 100.
The problem with that is that a pool of 1,000 houses is going to be very expensive. But if you slice that pool into 1,000 pieces, then each resulting security will be as cheap as one house, only with dramatically less volatility.
The alchemy is not that sophisticated, and while you can’t “guarantee returns no matter the vacancy rate or the economic climate” you can truly reduce the average volatility without reducing the average return. It’s basic diversification that works for the same reason that holding an S&P 500 index fund is probably a better idea than rolling the dice and buying a single S&P 500 stock at random. During the great mortgage lending boom of the aughts, bankers believed—or chose to pretend to believe—that a systematic nationwide decline in housing prices was impossible. That was wrong and dumb. But it’s still true that securitization of mortgages reduced idiosyncratic regional risk, and the same is true of securitizations of rental income.
