Arnold Kling  

Bernanke Could Use a Clue

Left versus Right: What's the ... Health Care Free Lunch...

Lots of big names at a Berkeley symposium on housing finance, including Ben Bernanke. Reportedly, Bernanke focused on the need to rework the government guarantee of mortgage securities.

I really believe that is wrong. Moreover, he knows less about it than I do, even though he has tremendous power and influence and I have essentially none. But we do not need mortgage securitization at all to have a mortgage market. Given how badly securitization has turned out, we would have been better off without it. Going forward, the cost of reviving that market seems to me to be pretty high. The benefits, relative to just letting banks do the lending the old-fashioned (1960's) way, are miniscule.

Symposium papers included Dwight Jaffee:

After all, subprime loans appear to represent no more than 4 percent of all US investment securities, and the aggregate loss rate on subprime loans in the end may not exceed 25 percent.9 Using these numbers, a "representative" `US investment portfolio would suffer a one-time loss of 1.0 percent (= 0.04 x 0.25) as a result of its subprime investment. This is certainly not a number that would be expected to bring the largest US financial firms to their knees

Well, sure if all the subprime mortgages were evenly spread across the entire universe of financial institutions, they would all still be solvent. The rest of the paper is equally asinine, if you ask me.

Edward Leamer says that housing is highly procyclical, so that the Fed should use housing starts as an economic indicator. When housing starts are low, you reduce interest rates, and when they are high, you increase interest rates, over and beyond what you would do based on the inflation and unemployment rates. My guess is that this has a real-world relevance on the order of rounding error. Brad DeLong's take:

That is a question. I know that Ed Leamer is very smart and very brave to be a macroeconomic time-series econometrician. But it is not clear what macroeconomic time-series econometricians have to offer here. I have no answer.

Later, in a discussion of a paper by Shiller, DeLong writes,

Greenspanism is the doctrine that the Federal Reserve must make sure that consumer-price inflation stays low and that expectations of future consumer-price inflation stay low, but that otherwise should let the macroeconomy govern itself and if exuberance leads to an episode in which the economy is saying laissez les bon temps roulez, then laissez les bon temps roulez.

No Greenspan-hatred here. Move along. I mean, come on. The absolute consensus among academics was exactly what Brad calls Greenspanism. If you had offered a petition in 2005 saying, "The Fed should raise interest rates when inflation is low, as long as there is somebody, somewhere, crying bubble," who would have signed?

He says, in effect, that everyone would sign that petition now. Well, count me out. All I can say is that after the fact, I sure wish someone had done something differently about the mortgage crisis. But at the time, nobody could connect all the dots. What if Dwight Jaffee had been right, that this should have been just a minor blip?

Christopher Mayer and R. Glenn Hubbard make the point that house prices appreciated in many markets outside the U.S. where there was direct lending. This argues against the view that securitization played a big role in the housing bubble in the U.S. They want to attribute the bubble mostly to declining interest rates, with some irrational exuberance thrown in, particularly at the end. They point out that mortgage rates now are unusually high, and they think that bringing them down by a percentage point would help moderate the decline in house prices. Perhaps, but I'm tired of policies that try to levitate house prices. And I'm tired of government trying to make sure that mortgage credit is cheap.

Shiller answers Jaffee's question of how the subprime crisis blew up so many firms by point out that nobody knows who holds what. This creates a "lemons problem" among financial institutions (which ones are lemons?) and caused a "negative bubble in valuations of firms holding subprime securities."

Shiller is one of those who thinks that the only thing wrong with derivative markets is that they are over-the-counter and therefore involve counterparty risk. He and I could have a debate about that, particularly when it comes do credit default swaps. Trading them on an organized exchange would be putting lipstick on that pig, if you ask me.

Most of the other speakers had no written materials (one paper came with a very serious "preliminary, do not quote" disclaimer, which is fine).

Comments and Sharing

COMMENTS (3 to date)
Bill Stepp writes:

In a free market for money and banking, and mortgage underwriting and financiing, I doubt securitizatiion would be a problem. It had a bad turn not becuase of problems trading securitized mortgages and various derivatives, but because of the macroeconomic instability and ABC caused by the Fed.
Under free banking the recent market decline and credit squeeze (or whatever it was) wouldn't have happened.

elfin writes:


I believe you have criticized securitization in the past.

Housing securitization by the GSEs has been around for 30 years at least and has received widespread acceptance. In a sense, Jaffee is right; the direct losses in aggregate by most investors from subprime and housing securities is pretty minimal. The total return in the last year from the Lehman Mortgage-backed index is 4.6% (and I admit that this is a high quality, not subprime index).

But you are right, too. What amazes me is how concentrated the subprime losses are: banks, hedge funds, and SIVs mostly from what I can tell. Banks liked them because they could hold a higher yielding AAA security and conserve capital reserves. Hedge funds liked leveraging nickels into dollars with the additional yield.

The flaw was the mispricing of risk via the rating agencies, egged on and accepted by all participants--investment banks who made fees putting them together, regular banks desperate for AAA debt, rating agencies that found another line of business, hedge funds looking to leverage and arbitrage. I was exposed to the magical world of structured finance in the early 1990s. It alwasy amazed me how you could take a batch of A rated debt and convert 85% of it into AAA debt. Or how an investment bank could set up a subsidiary with a AAA rating, yet not lose anything in their overall rating.

Someone said that the mortgage securitization model didn't work because after hedging out default risk and pre-payment risk, you earned less than a treasury. Nonsense! What that means is that the risk of mortgages haven't been priced properly. My guess is that the market won't make this mistake going forward and mortgage spreads are permanently 0.25% to 0.50% higher.

Lord writes:

Securitization really goes hand in hand with globalization. It is possible to do without it, but it would be much more difficult than in the 60s. It would be a nice method to strike against globalization though, if that is your goal.

Comments for this entry have been closed
Return to top