Saturday, 27 March 2010

Were There Widespread Fire Sales of Assets During the Financial Crisis?

Last week I argued that Gary Gorton failed to prove there were fire sales of assets during the financial crisis. Of course he chose a bad indicator, and so set himself up for a formidable challenge, sort of like trying to hit a home run with a plastic whiffle bat. He looked at spreads of investment-grade corporate bonds -- specifically, at periods during the crisis when, inexplicably, investors demanded higher yields for AAA corporates than for AA. But, alas, the inexplicable proved all too explicable on closer examination.

Still, that's a narrow example. What about those big U.S. banks that refused to sell securities, citing "fire sale" prices they were being offered. Was that accurate at least?

This is worthwhile to ponder because the Treasury and the Fed completely bought into the "fire sale" story. They fashioned a response to the crisis appropriate to a situation where the biggest problem was not bad assets, but investors with bad (irrational) thoughts about assets that were good ... or at least not all that bad. Remember these words from the Treasury rollout of PPIP?
The ... need by investors and banks to reduce risk triggered a wide-scale deleveraging in these markets and led to fire sales. As prices declined, many traditional investors exited these markets, causing declines in market liquidity.
For starters, it's worth disentangling some self-interest here. The banks had a huge vested interest in having us believe the "fire sale" thesis. Because, if we did, that meant: (1) They could justifiably refuse to sell the assets at the "fire sale" price and not confront the fact that they might be insolvent. (2) They not only could keep the asset on their books, but also they could justify fudging the price -- after all, if a market isn't rational, shouldn't you inject your own rationality? (3) If the problem lay not in the asset, but in the broader market, they could slough off blame for having made a bad investment. (5) Not only could they shed blame, but they also could make a play for sympathy: "vultures" who prey on the distressed by seeking "fire sale" prices aren't very sympathetic figures. (6) They could wait for a bailout that they could already see coming on the horizon.

So the "fire sales" thesis was a very, very powerful one, in many ways, for the big banks.

But again: to what degree was it true?

I thought about this for a while and came up with a back-to-basics approach to understanding where the truth lies. It starts with a typical "fire sale" example from real life.

Joe, in Detroit let's say, just lost his job Friday. It's Saturday. He needs to make rent Monday. He's on the front lawn next to a large hand-lettered sign that says, "Yard Sale." All around him: the armoire, his old Spider-Man comics, a few boxes of hand tools, and other odds and ends marked really cheap. Consider the armoire alone. Say its "true" price, secondhand, should be $100, but Joe's selling it for $20.

80% off! A real "fire sale" price.

But what is meant by its "true" secondhand price? "True" in this context is a slippery word. So let's define further -- okay, a bit arbitrarily, but some benchmark of value must be established. Let's stipulate the "true" price represents what a used-furniture dealer in the middle of Detroit would typically get for the item within a one-month time frame, were it offered for sale in his showroom.

This example allows three important factors to be isolated, in determining whether something is being subject to a "fire sale" price:

1. Time urgency -- The quicker something needs to be sold, the more "fire sale" pressure on the price, all else being equal. If Joe had more than two days to sell the armoire (the furniture dealer typically counts on a month), chances are good he could get a better price.

2. Breadth of universe of buyers -- Joe is counting on finding a buyer among the people who happen to drive by his house, and who at the same time happen to be looking for an armoire. The furniture store, on the other hand, has more relevant buyers by virtue of the fact that there's a regular flow of clients that cross the threshold expressly looking for pieces such as what Joe is selling.

3. The "money like" nature of the asset -- The less "money like" the asset, the more likely it will be subject to "fire sale" pricing pressure. Joe's armoire is very "un-money like." But if Joe was selling a $100 savings bond coming due in six months (and, to keep the example simple, we assume a zero interest and inflation environment over that time), he should receive close to $100 for it.

Now consider a residential mortgage-backed security in the fall of 2008. A big U.S. bank holding the asset says it's worth 90 cents on the dollar. A buyer says it's worth 40 cents. So is that a "fire sale" price? How does the above criteria apply here?

1. Time urgency: Does the asset need to be sold immediately? Small hedge funds and thinly capitalized speculators did have to dump assets at stress points during the crisis, trying to meet margin requirements or cover large withdrawals. But the big banks didn't appear to be in dire straits. They freely turned away buyers. So there really wasn't time urgency that the buyer could leverage for advantage.

2. Breadth of universe of buyers: Markets these days are increasingly global. U.S. subprime dreck, after all, found a home in insurance company portfolios in Taiwan. Likewise, if there really was a liquidity crunch in the U.S., the big bank's assets could have been offered up around the world. The Chinese have huge dollar reserves. Well before 2008, they complained publicly about having to hold so many Treasuries. If the "fire sale" assets were really woefully underpriced by investors in the U.S. and a surefire bet to offer reasonable returns at 90 cents on the dollar, China's state investment vehicle could have snatched them up.

3. The "money like" nature of the asset: The RMBS is a bond backed by streams of payments from mortgage holders. So it's fairly "money like." Granted, you do need to crank through calculations of expected defaults etc. But that all translates into a degree or risk, for which you demand a premium. At the end of the day, you still get compensated with money.

So what happened? How can you get so fast from 90 cents on the dollar to 40 cents without a fire sale? There must be some irrationality wrapped up in that low price, right?

Maybe not much. Consider that a buyer of the asset will demand some discounts, for sensible (not "fire sale") reasons:

1. The underlying assets, as home prices plunged, were starting to rot out, even if homeowners were at that moment current in their payments. Negative equity loomed.

2. The assets were further suspect because it was becoming obvious that the ratings agencies had improperly bestowed AAA ratings on many of them when they shouldn't even have been rated investment grade.

3. The buyer would have to do a certain amount of due diligence on a complex asset to become comfortable with the risk contained in the thousands of underlying home mortgages, and would naturally need to be compensated for this information gathering.

4. The broader RMBS asset class was tainted and so the asset was no longer as valuable for use in the huge repo market (just as, in the repo market, a bond that becomes "special" becomes more valuable in a quantifiable way, so the reverse is true -- when it becomes "stinky special" it's worth less for repo transactions and so a bond that's heavily repo'ed -- as AAA securities were -- should drop in price).

5. The very complexity of the asset, and its reputational damage, would impair the ability of the buyer to resell it, so the perceived illiquidity would impair value.

6. The complicated nature of the asset created possible minefields for the buyer, minefields that would most likely start blowing up if the asset were to deteriorate further, so taking on the risk related to complexity demanded some discounting.

7. There was significant systemic risk in the world at large, and this led to risk profiles being adjusted upwards for even apparently safe tranches of dicey securities.

So of the 50 percentage point gap, what can be attributed to "fire sale" pressure? Of course it's impossible to say without a concrete example. But I think if you went through these assets for the big banks, one by one, the "fire sale" factor would generally account for a small to very small part of the discount.

If you want another take on this issue -- basically agreeing with me, but using a capital asset pricing model where equity and credit markets are compared (and using a lot of formulas) -- check out The Pricing of Investment Grade Credit Risk during the Financial Crisis. Its conclusion:
Many analysts appear to be looking at large recent price changes and concluding that we must be witnessing distressed pricing and widespread market failure. This conclusion is based on intuition that fails to appreciate the extreme nonlinearity in the risks of credit securities, especially those manufactured by securitization (i.e. CDO tranches). Our analysis suggests that the dramatic recent widening of credit spreads is highly consistent with the decline in the equity market, the increase in its volatility, and an improved investor appreciation of the risks embedded in these securities. From this perspective, policies that attempt to prevent a widespread mark-down in the value of credit-sensitive assets are likely to only delay -- and perhaps even worsen -- the day of reckoning.
Now, after having expressed all this skepticism, I'm going to do a bit of a pivot here and move in the other direction: I do think that "fire sale" risk is a growing danger going forward.

If the first of the three stress points for "fire sales" is time urgency, that means when everyone beelines for the exits simultaneously, you're in deep trouble. Now the way the modern financial system has been evolving -- quants who seem to be copying each other's homework and modeling the same assumptions, super-computers that trade at blistering speeds, a global system where money flows easily across borders, interconnected networks of growing complexity -- it's becoming easier to imagine a situation where everyone does try to cash out all at once, and that causes the system to lock up. And a systemic regulator should be looking hard at this issue.

But in late 2008 and early 2009, I don't think the "fire sales" thesis explains the huge pricing gaps for securitized assets such as RMBS the big banks were trying to sell, especially considering the Fed's activist role during this period. Rather, this was simply a clever decoy that the banks used to redirect attention away from the truth. It's what they want you to believe happened because it lets them off the hook ... and helps create the rationale for the Great Hidden Bailout of 2009 that we'll all be paying for, in ways large and small, for years to come.