Belmont Club

The devil in the dark

L. Gordon Crovitz focuses on one of the fundamental challenges to restoring ‘confidence’ in the economy in a WSJ op-ed. “We need to know what’s mucking up the financial system.” In an earlier post I compared the financial meltdown to the situation facing a database administrator facing serious database corruption. It’s no good simply rolling back the system to its last stable state unless he can identify and remove the processes which caused the corruption in the first place. Otherwise things will go right back to being bad. Crovitz asks the financial equivalent of the question: ‘Just how did assets wind up misvalued and how can we prevent it from happening again?’.  The obvious place to start is to find out what the true value of the records in the system are, i.e. what the real value of the “toxic assets” is.

Addressing this basic issue was the original purpose last fall of the $700 billion government bailout program, but the Troubled Asset Relief Program didn’t live up to its name, leaving the size of toxic debts unquantified.

Plan B is to go back to Plan A. Regulators urge using new bailout funds to return to the original goal of discovering the true value of these securities. “A continuing barrier to private investment in financial institutions is the large quantity of troubled, hard-to-value assets that remain on institutions’ balance sheets,” Federal Reserve Chairman Ben Bernanke said in a London speech earlier this month. “The presence of these assets significantly increases uncertainty about the underlying value of these institutions and may inhibit both new private investment and new lending.”

Banks can’t resume lending because they don’t know how unsound they are. Private investors can’t know how bad bank debt is, so they hesitate to invest in banks.

It was never clear to me, as a layman, whether the purpose of the Troubled Asset Relief Program (TARP) was to quantify the size of the toxic assets or to simply finesse the problem by assigning them an arbitrary value by acquiring them at a fixed price.  In other words, whether the purpose of TARP was to discover how far down the hole went or simply to insert a floor under under it. It would be like running an update query on records whose values you don’t actually know by setting them to a value arrived at by consensus. True, this would be paid for by decrementing values from other records whose values were actually known; by getting Peter the Taxpayer to pay Paul, but once the update query was run the doubtful records would no longer have null values. They would have definite values. Then we could put the database online again and start the transactions rolling.

But the problem with this approach — what Crovitz called Plan A — was twofold.  There was concern that in updating the records a lot of bandits would make out like bandits; that TARP would paper over the places from which people stole, or lost money. It would reward financial mismanagement with taxpayer money. But worst of all it would leave these very same financiers who crashed the system, solvent, and in a position to play the game again. This is the obvious human meaning to not fixing what’s broke with the financial system. Leaving the same people in the same places. But Crovitz’s point goes beyond that and into the question of how we derive our information.

He correctly argues that we haven’t solved the problem of why we can’t read the value of the records; why certain assets are not transparently priced; and hence why we have to treat them as “null” (because we don’t know what they’re worth) and have to update them to an arbitrary value instead. Crovitz says, “before banks can get back on their feet and start lending again, financial professionals need more confidence that they know what went wrong and how to avoid more mispricing of risk.” In other words, the financial system must find a way of exposing the “true information” of assets so that people will trust each other again. After all, unless a pilot has confidence that his instruments are fixed after an emergency landing following his compass spinning like a top, he may be reluctant to fly on instruments again. So how do we go about fixing the instruments? Crovitz suggests a few ideas, but none so fascinating as his desire to quantify the distortion imposed on information systems by politically motivated agendas.

Bankers now recall the fine print of VaR analysis, which is that it always includes low but real risk that some new element could make the historical data a poor measure of the future. The late Dennis Weatherstone, the J.P. Morgan chairman who led this initiative, used to remind his team that the math of VaR alone could not measure risk in the outlying parts of the bell curve of probabilities. “The reason we pay as much as we do to traders is to manage the risk in the tails of the distribution,” derivatives expert Mark Brickell, a former J.P. Morgan managing director, quotes Weatherstone as saying. “That’s the hard part. For events inside the tails, it is not so difficult nor so remunerative.” It’s now clear that the data that banks used were distorted by years of government initiatives to promote homeownership. Government-mandated loans led house prices ever higher and house-price volatility ever lower. When the VaR models looked back, they wrongly modeled a low risk of default. Wall Street shouldn’t make the mistake again of ignoring the impact of politics on economics — and politicians should find ways to achieve social goals without undermining the integrity of markets.

It’s really a shame that Crovitz doesn’t go any further. But he makes a point I’ve often made in previous posts. That confidence in a failed system is seldom restored without changing the people who were responsible for the failure. It’s hard to conclude that the “impact on politics on economics” is over if the same politicians are still in charge or have even been promoted. And that brings us back to Plan A.  Perhaps the most politically convenient thing about TARP was precisely to prevent the markets from discovering the true size of the toxic assets. That answer would have been supplied by the markets. The trouble is, nobody would have liked the answer. But if a veil could be dropped across that scene of disaster — if we could run an update query to replace the ghastly values with something a little more bearable — then temporary confidence would have been restored. Unfortunately it would have been restored at the cost of undermining the very thing Crovitz feels we must achieve: “we need to know what’s mucking up the financial system.” I wonder to what extent the “stimulus” package is designed to avoid answering this basic question. Because from a certain point of view, we don’t want to know.

Additional readings for discussion:

Peter Schiff Was Wrong

I have seen your future, America, and it doesn’t work