In his new biography of Alan Greenspan, Sebastian Mallaby says some things I agree with, but he also rides a number of hobby horses that I take issue with.
Where I agree:
1. I agree that it is hard to achieve financial soundness through regulation. Financial markets are too flexible and adaptive to prevent institutions from gaming the system. If you want to see that point made at greater length, read my essay The Chess Game of Financial Regulation.
2. From 1970 to 1990, we got rid of interest rate ceilings on deposits, restrictions on bank branches, and futile attempts to distinguish commercial banking from investment banking. The process was long and grueling, with lobbyists engaged in furious rent-seeking battles all along the way. What Mallaby points out, and that I hadn’t considered, is that when the dust settled, we had a more rational, integrated competitive financial sector, but we had the same archaic, fragmented regulatory structure. So we had a separate regulator for thrifts, even though institutions with thrift charters were doing things that the thrift regulator had never seen before. The same with commercial banks, insurance companies, and investment banks. It was a regulatory structure that was set up to fail.
Where I disagree:
1. Mallaby buys into the theory that Brooksley Born should have gotten her way and had all derivatives trading moved to exchanges. I disagree. It is possible to trade derivative contracts in Treasury bonds and bills on exchanges, because the underlying securities are generic and liquid. Traders can benchmark prices and cheaply engage in arbitrage transactions. What AIG and others were doing involved creating a separate credit default swap for each security. In effect, Born would have been asking the exchanges to set up hundreds of different markets, most of which would have been illiquid in terms of the underlying securities.
If Greenspan was reluctant to wade in with financial regulatory proposals, that may have been because he thought that the issues were over his head. In fact, that may be what I most respect about Greenspan. Regulators generally do not see their own limitations. Brooksley Born would be a prime example of a regulator willing to take on a task while lacking sufficient knowledge.
2. Although I agree with Mallaby on the challenges of reining in financial excesses using regulation, he takes the view that monetary policy can and should be used to prick bubbles. He writes as if a major lesson, perhaps even the main lesson, of the financial crisis is that central banks should raise interest rates to pop bubbles. He writes as if this is obvious, when in fact very few economists see it that way, even now. In fact, Timothy Taylor recently pointed to an IMF study saying that global debt is at an all-time high, and only on the extreme right are there economists suggesting that monetary policy needs to be tightened. The other day, I got to attend a talk by Mallaby and I posed this issue. He agreed that his views were not widely shared by the mainstream (the people who complain about low interest rates as a threat to financial stability tend to be on the far right), but he said that one of the perks of writing the book was putting his opinions out there. Fair enough.
3. Mallaby blames the crisis in part on inflation targeting. He sees this policy as the mindless result of Fed officials’ not-entirely-rational preference for low, stable inflation. He could have pointed out that it was the overwhelming consensus of academic economists of the 1980s and 1990s that low, stable inflation was exactly the right objective for monetary policy. They believed that demand-driven recessions were the result of the public’s errors in expectations about inflation. Get rid of those errors by stabilizing inflation, so the thinking went, and you would eliminate recessions. This was known as the so-called Divine Coincidence, because it meant that the Fed could just focus on keeping the rate of inflation steady and let full employment take care of itself.
4. Mallaby takes a cheap shot at the Basel II approach to risk-based capital requirements, in which regulators were to use a bank’s model of its risks to gauge the amount of capital it should have. He compares this to giving a teenager the keys to the Mercedes. (a) I think that Greenspan had retired before Basel II was widely implemented. Most banks, perhaps even all banks, were still on Basel I, which used risk buckets. (b) Rather than being silly, using models was a good idea. The Basel I approach treated a bank that hedged its risks and a bank that went unhedged as identical. Basel I had no coherent way of dealing with derivatives or securities with embedded options, such as mortgage-backed securities. You need to use a model to solve both of those problems. And because every bank codes its portfolio differently, it is impractical to try to input the data into any model other than the one that the bank itself uses. Since you cannot try other models on the data, the best you can do is audit the way the bank goes about its modeling process.
It’s not a perfect way to regulate, but there is no obviously better way. At his talk, Mallaby emphasized that he did not think that any regulatory policy could truly rein in risk-taking. This gets back to point 1 under “Where I agree.”
The logic is convincing on model-based regulation, but it just seems like a bad idea. There’s way too much incentive to tweak the models. There’s gotta be a way to regulate that doesn’t require the regulator to have that kind of information.
“Macroprudential monetary policy,” insofar as that means using monetary policy to prick bubbles, seems to have credibility outside the far right.
Under CCAR, the regulators both review the models that the banks use and apply their own models based on standardized models. The offsite models suffer from the inherent difficulty of each portfolio being idiosyncratic, as you point out, but at least it provides a comparable floor so that the bankers have limited room to maneuver. See also the current debate about putting floors on Basel models.
The models of “new behavior” will ALWAYS be wrong, because they will be based on history, and the new behavior won’t include enough history. In particular, it won’t include a bubble and a popped bubble in the new financial instrument.
Instead, the new behavior will start out small, will be successful, will be duplicated and emphasized and copied by others, probably with some variations — and this increased new behavior will create a financial situation where the “new behavior”, which started out being successful, instead has created an environment of over-investment / mal-investment, and the model’s predictions will fail to capture the collapse.
Therefore, higher capital requirements are called for, especially for larger firms, whose failures could be the cause of some systemic failure. Smaller companies & funds would be allowed more speculative low-capital requirements BUT a stronger “no-bailout” policy. The moral hazard of the Long Term Capital Management bailout was being addressed by allowing Lehman to fail — but the whole “financial system” was full of Too Big To Fail companies, all with too little capital. So the bailout to the incompetent super-rich to avoid them losing the money they were so recklessly speculating with, even if they didn’t know themselves how much they were speculating.
No “model” of the MBS was going to show the systemic failure. Dr. Michael Burry “knew” it would fail, but not when.
Specifically higher capital requirements on the larger financial institutions, and less on partnerships as compared to public companies, is the best path to follow now. Financial partnerships should especially be freed to compete on their own speculative internal models, with less capital requirements, but where the investors are uninsured and unprotected by the gov’t.
The slow growth afterward is due to high taxes and especially high regulations, which increasingly act as drags on the whole economy, but most particularly on new business formation.
Agree. Remember the Value-at-Risk model that led to fragile leveraged positions?
The reductionism of such models can provide insights but cannot be a “precise and accurate” policy lever. Harkening back to Adam Smith’s Theory of Moral Sentiments, the proper place for the regulators is “commutative justice.” Instead, our regulators exhibit arbitrary power with loose reference to seemingly scientific models.