Arnold Kling  

Against Risk-Based Capital

PRINT
Selection Bias in Blogging... Against the Big Stimulus...

Jon Danielsson writes,


Risk-sensitive capital can be dangerous because it gives a false sense of security. In the same way it is so hard to measure risk, it is also easy to manipulate risk measurements. It is a straightforward exercise to manipulate risk measurements to give vastly different outcomes in an entirely plausible and justifiable manner, without affecting the real underlying risk. A financial institution can easily report low risk levels whilst deliberately or otherwise assuming much higher risk. This of course means that risk calculations used for the calculation of capital are inevitably suspect.

This is true enough. But it reminds me of the argument against mark-to-market accounting. Both risk-based capital and mark-to-market accounting have their flaws. But there is no flawless system, and anyone who lived through the savings-and-loan crisis knows that failure to mark to market and failure to differentiate for risk is a fatally flawed way to regulate banks. Just because risk-based capital faces implementation problems does not mean that reverting to simple leverage ratios would be a solution.

I keep coming back to the following points:

1. All centrally-designed incentives systems degrade over time. Within a firm, the goal of the manager is to maximize the change in employees' behavior and to minimize the expense of getting that behavioral change. The goal of the employees is the opposite. As employees learn how to game the system, any given set of compensation rules becomes dysfunctional and has to be changed. The same goes for regulation. If you keep a system in place long enough, banks will naturally learn to game it and it will become dysfunctional.

2. If we are going to have financial institutions that are given government backing, then they cannot be given free-market privileges. You cannot have Freddie Mac, Fannie Mae, or deposit-insured banks given complete freedom. We also cannot count on letter-of-the-law regulation to keep them from abusing their privileges, for the reasons given in the preceding paragraph. Therefore, I believe that government-backed financial institutions also need spirit-of-the-law rules. Executives of those institutions should feel the threat of prison if their firms fail, even if they obey the letter of regulations. You want government-backed institutions to attract executives who value safety and are naturally risk-averse. Let the risk-takers gravitate toward institutions that do not expect a government bailout if their gambles go awry.


Comments and Sharing


CATEGORIES:




COMMENTS (10 to date)
Matt C writes:

"Therefore, I believe that government-backed financial institutions also need spirit-of-the-law rules. Executives of those institutions should feel the threat of prison if their firms fail, even if they obey the letter of regulations."

Can you explain how a person defends himself if he is charged with violating a "spirit-of-the-law rule"?

This sounds to me like making it a punishable offense to incur the ill will of a government official.

If you get it in finance, it won't stop there.

Arnold Kling writes:

Matt,
If you want to restrict government to letter-of-law regulation, then expect government-backed institutions to have catastrophic failures fairly regularly.

We don't live in a world where letter-of-law regulation is compatible with safety and soundness of government-backed institutions. We live in a world where there is a trade-off.

dWj writes:

In re point 1, suppose bankers were angels, then a regulatory regime is still going to favor banks that have an entirely incidental tendency to optimize what is being measured. If there are multiple ways a reasonable person might try to run a financial institution, those ways that give the best metrics will perform the best in that regulatory environment. The degradation of regulatory institutions doesn't even require that bankers be trying to degrade it.

Matt C writes:

"If you want to restrict government to letter-of-law regulation, then expect government-backed institutions to have catastrophic failures fairly regularly."

That is what I expect in any case.

I do not think that giving this kind of discretion to regulators will prevent the next catastrophic failure. It will only change who gets punished after the fact. Those with good connections will be safe. Those who are political outsiders, guilty or innocent, will be hanged.

"We don't live in a world where letter-of-law regulation is compatible with safety and soundness of government-backed institutions. We live in a world where there is a trade-off."

Empowering regulators to discard the letter of the law and punish whatever they please will not increase safety and soundness. It will only increase corruption.

Jim Gatti writes:

The S&L debacle was more a result of the failure of FSLIC and the FHLMBB to close quickly or more tightly supervise/control S&Ls when it became apparent that they were in trouble. Even without mark-to-market accounting it was clear which ones were market-value insolvent or close to it. The problem was compounded by the fact that the legal authority of regulators to act based on estimates of market value net worth was questionable. Even had MTMA been in place, regulatory forbearance would have prevented action.

On the other hand MTMA breaks down when you have a systemic collapse of markets even if there is no regulatory forbearance. When markets break down, regulators need the authority to suspend MTMA.

Friedman and Schwartz have a great quote from W.R. Burgess (pp. 355 - 356)and follow-up discussion relating to this point.

Marcus writes:

It seems to me "spirit of the law" misses an important fact: regulators didn't see this bubble coming either. In fact, they cheered it on.

So, under "spirit of the law" can regulators go to jail too for being wrong?

Philo writes:

I second Matt C, except where he says (about the CEOs of government-backed institutions): "Those with good connections will be safe." When the going gets *really* tough, even good connections may not save you. That is why it is odd for Arnold to say: "Let the risk-takers gravitate toward institutions that do not expect a government bailout if their gambles go awry." Only a daring risk-taker would take the top job at a government-guaranteed institution under Arnold's scheme: he'd be setting himself up for a possible prison sentence.

I also second Marcus. Why doesn't Arnold want the regulators to go to jail, including those who strictly and conscientiously apply the letter of the regulations but nevertheless allow disastrous failures? (Failure, I should think, is contrary to the "spirit" that Arnold wants enforced.)

RobbL writes:

Arnold, you write;

"Within a firm, the goal of the manager is to maximize the change in employees' behavior and to minimize the expense of getting that behavioral change. The goal of the employees is the opposite."

Can you name a firm that operates like that? Can you name a manager that has such a goal? I have worked for many companies, big and small. I have been a worker bee and high in management. Nothing like that ever went on as far could see. Every time I have suggested that we examine incentives and try to adjust them to bring about a desired behavior, I am met with bewildered silence.

Clearly the regulators don't try to produce incentives for "good" behavior and that is precisely the problem.

stanford writes:

Arnold,

The problem in the past is that we were under BASEL I. Risk weights were calculated based on ratings-agency ratings. Moving to BASEL II, where banks mostly individually determine the ratings, will align incentives better.

The problem with BASEL I as it related to the housing market, though, was that all mortgages were given a 250% risk weighting. This incentivized riskier lending. You could charge a higher interest rate, but still hold the same amount of capital. In theory, this means more profits.

We need to move to BASEL II.

eccdogg writes:

I have been thinking recently about financial regulation and have hit upon an idea that I have not had sufficient time to flesh out, but thought I would post anyway.

Most schemes seem to focus on measuring and limiting risk with return as the free variable. So if you score low on a risk measure but get a high return then you are doing a good job. So institutions strive to maximize their return on a given risk measure. Of course this leads to gaming as institutions find ways to take risk that are compensated for in the return but not measured by the risk measure.

What if instead the regulators limited the return you could make. In that case to maximize value to share holders you would have to minimize risk (it its true not measured sense) so that your return on risk capital would be higher. The assumption would be that all excess returns are because of risk premium so the only way a regulated institution made an excess return was by taking risk as judged by the market. If after the fact it turned out that some institutions were better at assessing risk than others then they would make a higher after the fact return on risk capital.

Comments for this entry have been closed
Return to top