The Precipice of Moral Hazard

The $700B bailout being discussed this week is a bitter pill. It's costly, and yet it doesn't really make sense to the average person, because it appears to bail out the wrong people.

The question being asked seems to be, "Why bailout Wall Street instead of the average home owner?"

The short answer? Because everyone wants a bailout, but not everybody will lose their house.

Bernanke has said that he favors paying a rather high price for mortgage-backed assets, which effectively means "pretend houses are worth almost as much as people paid, and sell them slowly so you get a good return."

He has to do it this way, because his goal is to stabilize the markets. Stabilizing the markets means making most of the mortgage-backed security holders whole. What Bernanke means is, buy back most of the foreclosures and pay off the bond holders. If he pays them less, the bond holders will still blow up, even though they'll blow up a little less. One of the oldest money market funds blew up this week for losing 3 percentage points, so paying 50 cents (or even 80 cents) on the dollar just isn't an option.

But this means if you could look back in ten years, sweep up all the foreclosed homes, and spread the dollars out to just those home "owners" who lost a house because they couldn't afford to pay for it, you would spend the exact same amount of money to keep people in houses they couldn't afford. This appeals to some people (the ones who can't afford their houses) and it doesn't appeal to others (the ones who pay their bills). But the "moral hazard" involved in making this offer to anybody today is just dangerous. Game theory is working overtime to do a fine job here.

If Paulson goes to all homeowners and say, "Hey, if you owe more on your home than it's worth, we'll bail you out," then everyone will sign up. One estimate from Zillow says that 40% of LA buyers in the last 5 years are underwater (owing more than their house is worth). So sign them up, at least.

The thing is that not all of these 40% would actually default and lose their homes--probably a much smaller fraction. But give them a government bailout? They would absolutely take it. Even if that bailout were an offer from their lender to reduce their loan amount. Of course they would.

The second problem with bailing out homeowners is that asset prices are actually still too high. Housing prices in Los Angeles are at 200-300% of the inflation trend line (which housing has tracked for hundreds of years). Without crazy credit and toxic loans, sustaining these prices artificially simply harms the country and its ability to use capital for productive uses for the next 10-20 years.

And even this description is inexact, and it gets back to the one of the weaknesses of the Bernanke/Paulson plan: what is the "market value" of a house, especially in the din of foreclosures we're going to see over the next 3 years? You would have to look at housing prices ten years ago to estimate a "normal" index of affordability in order to mark these assets to market. There are many regions in the country (see this housingtracker list) where "market prices" are still not back to a reasonable multiple of income (and yes, thanks, LA's at the top).

So actual flaw #1 with the plan: without an absolute "value" to mark down assets, we wind up making a guess, and the assets may fall in value further, requiring more bailouts. And you can't "test" the market as has been proposed: the "market" for a small number of units is simply NOT the long-term value. There are still sales in my area at 2006 prices, but this is not the long-term trend line. Housing prices are not set by a small number of units, they're set by an entire market of sellers and buyers matching their prices to the "most they can afford". You will find a buyer for a distressed asset, but it may not be the "usual buyer". It might be a buyer who makes 3x as much as the guy who would have bought that house 10 years ago. You only know the price when you sell. When you sell Everything.

Actual flaw #2: is there a parallel moral hazard in bailing out the banks? I haven't seen this addressed, and I don't know if some magic regulation secret sauce is enough to prevent it. Recapitalizing banks at 2006 levels is not exactly a good plan, as it would instantly lead to another credit bubble. It's extremely hard for me to understand the systemic risk that's posed by this much capital flowing into the system quickly.

I bought a seemingly over the top book last year called "Financial Armageddon", written in late 2006--and friends and family have ribbed me for owning a book with such an extreme title. I re-read a bit of it last night, and most of Chapter 3 has happened (blow-ups for Freddie, Fannie, insurance, MBSes, etc.) Sounded quite a bit different than 18 months ago. More like a checklist.

Chapter 4? That's all about the $100T in credit-default swaps and other derivatives. I didn't know, for instance, that a lot of derivatives are accounted for off the books, and so you can owe some humongous amount of money to another institution and not even report the fact to shareholders. This is how a lot of leverage got created, and it's most likely the underlying reason that a bailout has to happen. The average American taxpayer doesn't understand why you'd write highly-leveraged insurance guaranteeing things like, foreclosure rates would be less than 1% or interest rates on money would stay cheap forever. Or why you'd do it if you were a major US bank?

Anyway that's mostly off-topic, but it explains why the bailout must happen. It's not just liquidity and things "freezing up"--it's actually highly-leveraged counterparty risk that I think we just don't want to know about. Paying $1 today will probably save us $100 later.

No comments:

Post a Comment