I’m sure you know the old saw about the drunk looking for his keys. A cop comes across a drunk crawling on all fours under a street lamp. The cop asks what he is doing. The drunk says he is looking for his keys. The cop looks about and can’t find any keys, and asks the drunk if he is sure this is where he lost them. The drunk says, “No, but this is where the light is good.”
Financial Times columnist John Dizard depicts a Fed that bears a strong resemblance to the storied drunk. Dizard starts with a discussion of how the Fed is recognizing that some of its model aren’t what they are cracked up to be, but this discussion, while interesting, is a circumloquacious way of getting to a more serious observation. While the Fed may understand that it has comparatively little influence (it’s hard to miss that 16 interest rate increases had just about zero impact on the long bond), the Fed doesn’t have a good enough grasp of what is going on in the credit markets to know how it would influence them.
We’ve spoken before about the Fed’s limited power and the fact that it offers benign reassurances about the advantages of financial innovation, when between the lines, it’s evident that the Fed is a peripheral observer of phenomena it does not fully understand. But Dizard points to the Fed’s failure to take even some basic, simple steps to be more savvy about the credit markets.
One bit of cheery news: Dizard thinks the credit selloff is overdone and represents a short-term buying opportunity.
The subprime mortgage crisis has finally shocked the central bankers out of a complacent trust in their economic models. That, not last week’s downgrade announcements by the ratings agencies, or a few dozen points of widening in the credit curves, is the real change in the situation on the ground. This message was communicated in typically dense and indirect language in Federal Reserve chairman Ben Bernanke’s July 10 speech to a group of academic economists in Cambridge, Massachusetts.
That’s good news. Unfortunately, when Mr Bernanke and his counterparts move to pull the levers to flood the markets with liquidity, they could find that the levers aren’t connected to anything any more.
The Bernanke speech was a classic of the central banker/academic economist genre. The key message was to be found under the heading of “Inflation Forecasting at the Federal Reserve”. While most of the section was taken up with a description of just how the Fed has been improving on its Dynamic Stochastic General Equilibrium models, the key point was in two sentences. Those models, which had been used to justify and tune the tightening monetary policy, are, Mr Bernanke said, “unlikely to displace expert judgment”. The DSGE has to be combined with “anecdotal” and “extra-model information”, along with that expert judgment.
In other words, the models aren’t our models any more. I think that’s a welcome recognition of the reality that the DSGE isn’t computing.
This doesn’t mean that Mr Bernanke intends to fly by guesswork and dead reckoning. He has another model ready, one he outlined in a June 1983 journal article. There’s nothing more satisfying to an academic than winning a 24-year-old argument, even though it might take a world financial crash to provide the final proof.
Mr Bernanke’s point, back in the early 1980s, was that the standard monetarist account of the origins of the Great Depression of the 1930s was inadequate. Simply correlating the downturn to a contraction in monetary aggregates did not fully account for all the data. Instead, Mr Bernanke hypothesised, it was necessary to also include the effect on what he termed the “CCI”, or the Cost of Credit Intermediation. This is the price and efficiency of the intermediation provided (back then) by the banking system. Because, he added, with a swipe at Eugene Fama of the University of Chicago, markets are “incomplete” or imperfect, and therefore need the “real service provided by the banking system” of the “differentiation between good and bad borrowers”. In the paper, Mr Bernanke goes on to select and test various measurements of the CCI, and concludes that yes, they “. . . improved the purely monetary indicators . . . “
The problem that the Fed, and the other central banks face is that the lending officers and credit committees at the banks aren’t, in many cases, there any more. Which means giving them free money, so to speak, in a crisis won’t work so well. The classic banking mechanism has been replaced with the production of pools of loans or notes, which are then sliced up by securitisers and rated by organisations such as Moody’s and S&P. The risk of this stuff is measured day-to-day by the volatility of the prices of the pieces, not by hands-on monitoring of the underlying assets.
Of course, the Fed staff and management know this. But they don’t know enough to oversee the new processes. Last week, I spoke to Numerix, a New York analytics firm that has about half the market for the risk management, pricing and product development models used by credit traders and investors on Wall Street. I asked how many government agencies, including the Fed, had bought their products. Answer: none. Oh.
However tragic the subprime mortgage crisis is for the defaulting homeowners or some Bear Stearns risk managers, the numbers are small in relation to the economy.
But that’s not the point. “This is a bigger issue than it seems, because of the way margin or leverage has been deployed through the prime brokers and the repo desks,” says a credit strategist at a major dealer. “You have to worry about the knock-on effects of the greater discrimination and higher collateral requirements. It’s the contraction on the heels of the subprime mess that has the multiplier effect.”
How big is that effect? I don’t know, which isn’t that important. What is important is that the Fed doesn’t know either. Which means that when they pull a lever for a 50, 75 or 100 basis point Fed Funds rate cut, it might not have the desired or necessary result. That’s one reason why I am rethinking my negative call on gold for the balance of this year. The Fed, and its counterparts elsewhere, may wind up throwing an awful lot of liquidity on the fire.
For the moment, I think the credit correction is overdone, as a trading situation. You could probably profit over the next month or two by going long credit indices such as the LCDX, which could well be oversold as a consequence of frantic dealer hedging and re-hedging. But levered credit as a long-term investment strategy, or as a prop to the US, and, by extension, global economy? Probably not.
Perhaps Mr Bernanke should blow the dust off that CCI model and get some staffers to update it with new information.