Timothy Taylor excerpts from a book on macroprudential regulation.
Paul Tucker: “Legislators have typically favoured rules-based regulation. That is for good reason: it
helps to guard against the exercise of arbitrary power by unelected officials. But a static rulebook is the meat and drink of regulatory arbitrage, which is endemic in finance. Finance is a ‘shape-shifter’.
Rules do not work, because banks figure out a way to manipulate the rules. Tucker gets this. So does Wolf Wagner, also quoted by Taylor.
Also, discretion does not work, in my opinion, because discretion tends to be procyclical, doing exactly the wrong thing at the wrong time. In good times, regulators ease up, and in bad times, they tighten up. Just look at how regulators behaved before and after the housing crash. Or compare Ben Bernanke’s discussion of bank supervision before and after he knew about the crisis.
I think that principles-based regulation might work better. That is, pass a law saying that managers and directors of financial institutions are responsible for prudent management. Require auditors to flag questionable practices.
Also, I think that incentives are important. Casual observation suggests that investment banking was more cautious when investment banks were partnerships rather than limited-liability corporations. We should look for ways to give bank executives more skin in the game in their institutions. Suppose you have a bank that goes bust in 2025. All of its top executives over the preceding 10 years would be held personally liable those losses, in proportion to the compensation that they received over that period.
(For each year, take the five most heavily compensated executives, and put their total compensation into a hypothetical pool. Add these to get a company total for fifteen years. Then divide each executive’s total compensation over the 10 years by the company total to get the fraction of losses for which that executive is liable.)
Actually, I don’t think that the formula needs to be perfectly “just.” The point of any such system is to make executives manage banks as if they were risking their own money, because they would be.
What we really need is principles-based Academics, with that kind of liability incentive scheme in place. This is mostly for Academics who say they have results that bear on public policy, and some kind of empirical relationship and implicit forecasting model. If you publish a paper like that, you should have to invest in your claims, say, in “tenure points”. The betting odds are exactly the same as your published claims.
The forecasting model should be expressed explicitly in terms that can be (and must be by the publication) easily verified about every year. Every time any one your results falls outside the, say, 1 standard deviation, you lose 5% of your tenure points, divided by the total number of your published results.
There’s a 31% of that happening for each result for year, but no worries, so long as you don’t make it to 100% lost tenure, you are still eligible, and the vast majority of people with correct claims will make it. Of course, you are still
And each year, if an old result is starting to bite you in the butt, you always have the option of publically recalling it, calling the result unfounded, and admitting you were wrong. Maybe it’s not even your fault, because some major circumstance or assumption changed beyond your control. Fine, no problem. Still, recall the publication with that explanation, “Results no longer valid going forward due to a change in circumstances.” You earn 2% positive tenure points for each result you recall, to help people get over their pride and do the right thing.
But if you were, for example, to be one of the publishers of those global climate model forecasts that has been over 1-SD off for 10 years now, and you’ve done ten similar papers since you were in graduate school by the time you’re up for tenure, well, you’re in big trouble. One could always adjust the tenure points to be properly tailored to the typical publication frequency of the field.
A 50% equity rule for banks would be hard to get around. Maybe the rulebook should focus on just the things that really matter; that would keep it short and more easily monitored.
Their first task would be to produce a hedge allowing them to offload this risk to third parties and increase their compensation to pay for it. They aren’t dumb.