Sunday, 28 February 2010

Gary Gorton's Somewhat Flawed Take on Shadow Banking

When I saw this Feb. 20 .pdf on shadow banking, prepared for the U.S. Financial Crisis Inquiry Commission, I got kind of excited. Reading material for my one-hour stationary bike workout! And shadow banking no less, my latest pet obsession! Ah, Sunday morning nirvana for a finance wonk.

Then I started perusing the piece and, after turning over the bike pedals oh about 1,100 times (I try to spin at 90 rpm, and when I get involved in my reading, I'll hit the mid-90s), I grew a bit disenchanted.

Read it for yourself of course. If you want a better treatment of the material, Gorton is essentially cribbing off himself from an earlier paper -- "Slapped In the Face by the Invisible Hand" from May 9, 2009 -- which is a bit smarter, dares to be prescriptive and explores the subject in more depth. My nose de-wrinkled a little after reading "Slapped."

But please start with the crisis inquiry presentation because the writing is more accessible for people who aren't finance nerds. He does a few things that I really liked:

1. History: He presents some of the historical sweep of bank panics over the last couple of centuries.

2. Prime mover: He zeroes in on the real nexus where the financial system failed in our current crisis. In case you weren't aware, basically there was a bank run -- but it wasn't retail (i.e., little investors like you and me and Grandma Jones). It was wholesale, the big institutions that make up the "deposit" base of the shadow banking system.

3. Wild haircuts: he shows nicely how the "banking run" of 2008 took place, through the haircuts on repo'ed collateral that, by increasing from negligible levels, effectively "withdrew" money from the system. I know that line is a bit abstruse, so read his explanation on page 12.

But I think Gorton misses enough stuff that his paper ultimately disappoints. Here are four points he makes that left me shaking my head:
Holding loans on the balance sheets of banks is not profitable. This is a fundamental point. This is why the parallel or shadow banking system developed.
He's a bit intellectually sloppy here -- too busy to show his work or cite the evidence? If you read what he says about securitization (shadow banking's backbone and sine qua non) in his earlier "Slapped," you'll see that he is a bit more revealing, and subtle, there.
Why did securitization arise? We do not know for sure. One possibility, discussed further below, is that it was a response to bank capital requirements, which created a cost without a countervailing benefit. Banks, being private institutions, can exit the industry if it is not profitable. Another possibility is that the demand for collateral made securitization profitable, and this could not be accomplished on-balance sheet because deposit insurance was limited.
So there was a huge need for collateral (that $600 trillion worldwide derivatives market perches atop a mound of supposedly rock-solid collateral that, though relatively tiny by comparison, has grown enormously over the last two decades). And banks were chafing under capital rules.

Okay, I'll buy that. But that's a long way from a convincing argument that hold-the-loan-to-maturity banking is inherently unprofitable in the modern age. He rather disingenuously overlooks a big point: it's partly not profitable because of shadow banking which has (1) no capital requirements (2) no FDIC insurance to pay (3) no other regulatory burdens. So the shadow banks can operate more cheaply and pay more for funds, squeezing banks cleaving to a traditional model.

Point two where we don't quite see eye to eye:
In securitization, the bank is still at risk because the bank keeps the residual or equity portion of the securitized loans and earns fees for servicing these loans. Moreover, banks support their securitizations when there are problems. No one has produced evidence of any problems with securitizations generally ...
My reaction to this: ??????? No problems with securitization generally? That seems rather boldly dismissive. If banks support their securitizations when there are problems, why are they allowed to push the risk off balance sheet? That seems to be a legitimate issue. And the way that risk is imperfectly understood and shuffled down the line via securitizations -- a loan originator puts crap in the soon-to-be-securitized ground beef for the RMBS that's leaving his hands in a week and ultimately winds up in a hamburger patty served up to an investor in Singapore who can't make heads or tails of what he's buying and just relies on a AAA stamp bestowed by a team at Moody's that couldn't really understand what the hell they were rating either -- well, some would call that a legitimate issue too.

Another point where I think he's off base:
Why would dealer banks be growing their balance sheets if there was not some profitable reason for this? My answer is that the new depository business using repo was also growing ... Now, of course there is the alternative hypothesis, that the broker-dealer banks were just irresponsible risk-takers.
I'm not really going into this one -- this blog entry is getting too long already -- only to say this appears to be what they call a "false dichotomy" in debating circles. After all, dealer banks could have been pursuing irresponsible risk-taking that had a profitable reason behind it. That's pretty much self-evident if you read the entire section here.

Now, on to one last point, where he contends that AAA rated securities were marked too far down because of fire sales and cites, as proof, the fact that AA rated corporate bonds at certain times paid more than AAA corporate bonds. That's because so many AAA's were being dumped that the spread flipped, he says.

Okay, first, I'll confess I don't have the data set on that graphic he provides. But I'd sure like to explore it in more depth. Because, honestly, it makes no sense.

Quick bond primer: One piece of the yield attached to any bond represents credit risk. Credit risk is the chance that say IBM goes belly up and you're standing in line with other creditors, trying to get back money that you "lent" to IBM by buying its bond. If you assume more credit risk, you are compensated more. For example: if IBM bonds have a small chance of defaulting, but Wal-Mart bonds have a smaller chance, then IBM's bonds (of a given maturity) will be rated say AA and pay 6 percent and Wal-Mart will be rated say AAA and pay 5.4 percent.

Now bond buyers aren't stupid. Aunt Flo isn't a big player in this market; these are institutions and hedge funds and a lot of sophisticated money managers ... take a look at some bond-pricing formulas, study a little duration, and get back to me in the morning if you doubt what I'm saying.

The point I'm making is this: if these yields flipped, there's a good chance that the "fire sale" explanation is probably the weakest one to make. Because look: even if you're a fire sale buyer, scared as hell of what's going on in the financial markets, your IQ doesn't instantly fly out the window. Relatively speaking, you still prefer AAA over AA. If my fire sale price for AAA is 90 cents on the dollar, my fire sale price for AA isn't going to be 93 cents on the dollar. Because that's leaving free money on the table. That's your first lesson in Bonds 101. No one does that.

So how to explain what happened? Here are a couple of hypotheses that I think are probably more believable: (1) There was a lot of mis-rated AAA crap that really deserved to be graded A or below, and for some reason, the AA securities tended to be rated more accurately, or maybe they were mispriced somewhat too, but it took longer for the problems to be evident. (2) Something else -- maybe even fraudulent -- is going on here. Look at his chart and see how nonsensical it is -- in early March of last year, a good four months after the peak of the crisis, AA's were paying about 2.25 percentage points more than AAA's! By way of comparison, that's more than double the amount that AAA's ever exceeded AA's between January of 2007 and November of 2009.

So this is what Gorton asks us to believe: four months after the "panic button" phase of the financial crisis, there was a sudden move to massively offload AAA's at any price -- sellers were willing to get scorched on the asking price and buyers didn't recognize the value that they were getting compared to AA's and never pushed the spread back together. I mean, 2.25 percentage points?!?! That makes no sense to me.

I'm willing to bet there's something else going on there and I'm surprised that Gorton didn't sniff harder to try to find it.

So while I'm glad that Gorton's writing about shadow banking -- we all need to understand its role better -- he comes up a bit short by my yardstick.