There seems to be at least a plausible chance that Deutsche Bank avoided requiring a bailout by refusing to mark-to-market a very large derivative position during the height of the financial crisis. Why is this a problem?
Tom Braithwaite, Kara Scannell and Michael Mackenzie of the Financial Times have provided details of a story first reported by Matthew Goldstein in June 2011. The gist of the story is this: Some disgruntled former Deutsche employees have accused the German bank of engaging in accounting chicanery to hide losses.
As Team FT tells the story, two risk managers and one trader have told U.S. officials that Deutsche had in effect hidden billions of dollars of paper losses, allowing the bank to appear stronger than it otherwise would have. If Deutsche Bank had accounted for the losses as the former employees think they should have, it might have been in such dire straights that it would have required a government bailout.
This is a good place to pause in the narrative. One indication that something very odd is going on here is the fact that Deutsche Bank was able to stay solvent by not recognizing its losses. Why does Deutsche Bank's solvency depend on whether or not it believed in and declared losses?
The obvious answer is that Deutsche Bank is subject to regulatory capital requirements. Its banking license depends on it having a level of capital compared to overall assets set by regulators. So a bank can become insolvent not because it is unable to meet its commercial obligations but because it is unable to meet its regulatory obligations.
And, indeed, if we take a closer look at the charges against Deutsche Bank, it becomes apparent that this whole thing revolves around capital regulation.
Let's start by looking at the trade.The biggest unrecognized loss is said to come from a portfolio of leveraged super senior trades. The bank arranged to buy credit insurance on a portfolio of credits of very safe companies from Canadian pension funds. The bank paid an insurance premium to the Canadian funds, and the funds put up a small amount of collateral against the notional amount insured — as little as $100 million on $1 billion of insurance, according to the FT.
The bankers didn't think more collateral was necessary because, as the FT explains, the "chance of several safe companies, such as Dow Chemical or Wal-mart, all going bankrupt at the same time was infinitesimally small. It might require a nuclear war. The chance of the investors having to pay out on the insurance appeared impossibly remote. The chance of their collateral being used up was inconsequential."
This is a good moment to pause in our storyline. If there were no chances of the investors having to pay out on the insurance, why on earth was Deutsche Bank buying it? We don't normally think of Deutsche Bank being in the business of handing out free money to Canadians.
What was happening is that this deal with the Canadians was one half of what folks in finance call a "negative basis trade." The bankers were also taking another position in which they were sellers of a slightly different (but only slightly) insurance contract for which they charged a bit more than they paid the Canadians. This creates a revenue stream for them without really creating any new risk, since if you have to pay out on the one contract, you are getting paid on the other.
Now you're probably asking: "Wait a minute! I thought there was only 'nuclear war' risk to begin with. Risk so infinitesimal that the bankers didn't care about the collateral Why would the bankers be willing to buy the protection from the Canadians — which cost them almost the entire income stream earned by selling protection — if the risk was so low?"
This is a very important question. The answer is that this is a form of regulatory arbitrage. Selling the insurance would create a risk asset on the books of the bank, against which it would have to have regulatory capital. By buying the insurance from the Canadians, the bankers creating a balanced book that required little to no regulatory capital.
Notice that there is still something not quite right here. Why is it that Deutsche Bank can sell its insurance cheaper than the Canadians? If they really are perfect substitutes, shouldn't efficient market processes result in them having exactly the same pricing?
Well, yes. If the Canadians are receiving less for the insurance they are selling it must be because that insurance is worth less — probably because the Canadian pension funds were deemed a greater credit risk. So it appears that what Deutsche Bank was really doing was profiting from the difference in the way the market saw risk and the way regulators so risk.
What the Deutsche Bank whistle-blowers are arguing is that when the market signaled that the insurance from the Canadians was becoming riskier, Deutsche Bank should have recorded losses on the asset. This is what Goldman Sachs — where the primary whistle-blower earned his stripes before coming to Deutsche Bank — would have done, marking the assets to market each day.
If you make the (perhaps heroic) assumption that the bankers behind the trade weren't idiots, however, you come to a different conclusion. The trade itself was designed to take advantage of a gap between market prices and the regulatory view of risk. That is, the bankers setting it up knew in advance that price was telling them one side of the trade was riskier than the other. They also knew that regulators didn't care about this gap. So they knew that the trade was regulation resilient. It wouldn't matter if prices moved against them because the thing wasn't ever going to be marked to market.
In other words, the bankers designed the trade on the premise that it didn't have to be marked-to-market. If you were going to mark this trade to market, you'd never do it in the first place.
The odd thing about this, however, is that the regulatory view of risk turns out to have been better than the market view. The losses that market prices appeared to predict turned out to be very much exaggerated —thanks in no small part to the intervention by governments around the globe. Perhaps the market would have been right if not for the intervention of governments, but that's a bit like saying that Mr. Lincoln would have enjoyed the evening at the theater very nicely except for the intervention of Mr. Booth.
Regulators looking into this matter should probably at least consider the macroprudential view: Would we have been better off if Deutsche Bank had recognized losses deep enough to force it into the arms of the government? What purpose would this have served?
The standard answer to this is gauzy talk about moral hazard. If Deutsche Bank had been forced to fail because of accounting, then other bankers in the future wouldn't be so cavalier.
The problem with this is that the Deutsche Bankers weren't cavalier. They were correct in their view that the market wasn't appropriately pricing their assets. This wasn't a risky trade, the losses did not materialize, and the bonuses that the bankers who built the trade were paid were actually earned.