A long commentary on someone else’s commentary on the economy

A friend made some comments about the economy, which basically amounted to the notion that the problem was the decoupling of consequences in the mortgage market — people were “playing with monopoly money” and behaving irresponsibly due to the invention and application of mortgage-backed securities and credit default swaps. I think what he said is more or less true, but I’d make several observations:

You didn’t need the CDSes to get from “mortgages are safe” to chaos. You needed three other things, which we had:

Mortgage-backed securities, which you note above as a critical part of the problem. You’re right that decoupling the investment from the people making the mortgage broke responsibility, and encouraged people to make irresponsible loans — but the thinking of the purchasers of the mortgage-backed securities ends up being equally relevant.

The thinking there was exactly the same as what happened with junk bonds in the eighties — the notion at that time (which was mathematically demonstrated — hold that point) was that one junk bond was unsafe, because of the chance of default — but that a large, diversified portfolio of junk bonds was safe, because the chance of many simultaneous defaults was low (it had never happened before) and the return on the portfolio was, even accounting for defaults, still higher than safer, higher-rated corporate bonds alone or in portfolio.

The real problem with CDSes — although you describe them as insurance above, they were really bets — because you didn’t need to hold the securities being insured. That’s right, if I had one set of mortgage-backed securities there was no limit to the amount of insurance being written on each of those securities. The problem wasn’t that AIG was ‘insuring’ these — the problem was that they had “backed” their insurance with CDSes with other third parties. (There’s a great This American Life about this, “The Giant Pool of Money,” that explains the whole thing.) See, AIG would have been fine, and would have been able to pay everyone off — except that Lehman collapsed and was allowed to go bankrupt, so all of these hedges they set up lost one side. AIG had insured themselves to cover their costs — but then the people insuring them vanished, or possibly vanished — and with that, lots of money.

Which brings up the second problem: modern capitalism’s engine isn’t just “making money” it’s *growth*. And not just the growth in absolute dollars being made, but growth of the profit margin as a percentage of expenses. (This point goes all the way back to Adam Smith’s “The Wealth of Nations,” so it shouldn’t be news to anyone.)

This is what drove the boom in both MBSes and CDSes. Institutional investors liked the high rate of return on MBSes (because they were like junk bonds, see above), and so the financial industry needed to originate more mortgages, which is what really led to all of the risky mortgages being made — because there weren’t enough high-quality mortgages to be had. Similarly, if you hedged your CDS, it appeared to be a completely safe investment (because the insurance you made was insured by someone else, so if you had to pay out you could be reimbursed for that expense) that still made good money (because the rate you were charging for insurance was less than the rate you were paying for it). So this need for growth drove both crappy mortgages and a giant web of CDSes that nobody could untangle.

This brings us to the final problem: a failure of modeling. There were so many CDSes going on — the NPR program (or its sequel, “Another Frightening Show About the Economy”) claimed something like 40 CDSes for each mortgage-backed security being traded — that nobody could understand what would happen if any of the parties disappeared. Even now that this trade is in the midst of unwinding, nobody is still sure what anybody else owes on any of those deals, or to whom. It sounds crazy, but literally nobody can figure out what anybody owes.

Still, that’s a minor failure of modeling (even though it’s the proximate cause of AIG’s collapse in the wake of Lehman being allowed to fail) compared with several others that contributed more significantly to the economic collapse.

A more significant failure of modeling was the “junk bond” thinking — that because something had never happened before it was unlikely to happen at all, and because the risk had been measured it had been managed. (See Michael Lewis’ NYT Magazine cover story from the other week for more on risk modeling.) But the “extremely unlikely” event of course occurred and took everything out.

Here’s the thing, though — in the long term, the unlikely is sure to occur. But people’s time horizons have been dramatically compressed. A typical Wall Street career is less than two decades — that’s not really a complete business cycle. You’d need to work on the street thirty years to work all parts of a “normal” business cycle, and longer than that to apply what you learned the first time around. So virtually nobody has instincts developed during their career that apply to whatever’s happening.

It’s worse in this case, because the investments being made are themselves less than two decades old — so there’s no historical data to understand or compare against. You see graphs of these things that go back ten or fifteen years to inception, and you wonder how they could possibly make any judgments about them based on a time with only a pair of unusually shallow recessions — the 1991 and 2001 recessions — where the market in the damn things didn’t really take off until 2002 or thereabouts anyway.

Even my Grandfather, who is over 80 and still works on Wall Street, has been overly optimistic — because he’s never seen anything that terrible. The Great Depression was when he was a child, and so his “long term” view of the stock market is still little more than half a century, less than sixty years — not long enough for “long term” trends to be experienced, let alone analyzed. (Of course, the “long term” view doesn’t help make money in the short term, and is thus not particularly conducive to a successful Wall Street career.)

Of course, the same failure of modeling impacted the people justifying even the riskiest loans as safe — because none of those people had ever worked in an economy where house prices ever went down. I can’t remember if it’s Case or Schiller, but one of them went back and basically demonstrated that housing prices in the 20th century were effectively flat, once you corrected for inflation and improvements to the properties themselves. But that stopped in 1995 or so, and suddenly the market went up.

Of course, almost everyone working in housing as a mortgage broker, or real estate agent, or representing an institutional investor has started since then — or at least since the last significant real estate crashes in the late eighties. (Most people who were in the industry at that time left and never came back.) So everyone’s personal experience says that the housing market only goes up — opening the road to unlimited refinancing, or in the worst case selling and making some money. In a world where real estate only goes up, it’s hard to end up underwater on the investment.

Which brings us to the banks, which aren’t making any loans to anybody, pretty much, because they don’t know the value of the investments they’re holding. Since they can lend out only X dollars for every dollar they have in reserve without being shut down by the FDIC, the number of dollars they have in reserve is critical. But they were using the collateral on mortgages (i.e., houses and other properties) as a very significant fraction of that reserve — so as long as the reserve keeps shrinking, the amount of money that they can lend at any time keeps dropping. Until the housing market stops dropping, nobody knows how much they can safely lend — so the amount that they feel they can lend approaches zero. Bailout money is being used to slow the rate of reserve shrinkage to cover what they’ve already got outstanding in terms of loans, to avoid falling below capital requirements.

Anyway, why should banks lend any money until the value of the assets they’re lending for stop shrinking? Nobody in their right mind will pay a mortgage on which they’re substantially underwater, so loans on houses are pretty risky right about now.

Banks not lending for housing is pretty disastrous, since basically all of the job growth since the end of the 2001 recession has been in housing-related industries (mortgages, construction, furnishing, etc).

As a result, banks acting rationally is putting people out of work, which further depresses the economy and with it the housing market. Ugh. Nothing’s going to improve until housing prices stop falling, and nobody has any idea of when that will be.

So while the problem may have been kicked off by “everyone behaving like children playing with Monopoly money,” the problem now is that everyone is acting like a grown-up — only our economy was never built for people to do that…