Unitless units. 

Europe Doesn't Want Bankers to Worry About Pay

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

There's a popular theory that banking bonuses are bad because they make bankers do risky or evil things. The theory starts with these stylized facts:

  • If you make a lot of money for your firm, you will get a big bonus.
  • If you make a little money for your firm, you will get a little bonus.
  • If you make no money for your firm, you will get no bonus.
  • If you lose a lot of money for your firm, you will also get no bonus.

The payoff is, as they say, nonlinear. I guess they also say "asymmetric," as in "asymmetric incentives." I drew you a graph:

This is the payoff graph of a call option, and everyone knows that the value of a call option increases with volatility. So if you own a call option, and you can increase volatility, you should do that. Your expected value from doing nice safe deals, making a little money and getting a little bonus, is lower than your expected value from doing risky evil deals with the prospect of either making a lot of money and getting a huge bonus, or else blowing up and getting zero bonus. Heads you win, tails you don't really lose that much, is perhaps a helpful mnemonic here, sort of. So you flip a lot of coins.

This theory has some flaws -- for one thing, if you lose a sufficient amount of money, you'll be fired, so the bonus doesn't work in isolation ; for another, externally imposed incentives have second-order inference effects -- but it's a popular theory. It's made its most significant progress in the the European Union, which has a rule limiting banking bonuses to 100 percent of base salary (or 200 percent with shareholder approval), "in order to avoid excessive risk taking."

Prior to that rule, which was issued in 2013, lots of bankers got bonuses that were much bigger than their base salaries. So the rule required changes. But the two obvious possible changes were both bad. The banks could cut bonuses while leaving base salaries the same, reducing total compensation. But this would be bad because banks are run by bankers and bankers like money, so reducing their money would be bad.

Or the banks could cut bonuses and raise base salaries, leaving total compensation the same but making more of it fixed. But this would be bad because banks tend to have pretty variable revenues, and compensation tends to be a big component of their costs. That's why they like compensation to be a variable rather than a fixed cost. If you have high fixed costs and highly variable revenue, when the revenue is low, you tend to go bankrupt. Going bankrupt is also bad.

So the banks came up with what they thought was a clever solution, which was to try to increase base salaries without increasing fixed salaries. The idea is, you get paid 100 of "salary," and 100 of "special allowance" and zero to 200 of "bonus." The bonus is a bonus; your boss decides it at the end of the year, based on your performance, and it could be zero. The salary is a salary, and it's fixed; basically everyone in your associate class or whatever gets the same salary.

The allowance is paid sort of like base salary -- twice a month, the same amount each time, etc. -- but is otherwise variable. The specifics differ a bit from bank to bank, and here is a European Banking Authority reporton how allowances work at 39 institutions. But, for instance, different people will get different allowances, depending on how good they are. Your allowance is only fixed for a year, and can go up or down the next year. It can even go down mid-year, if you mess up, or if the bank runs into trouble and there's not enough money to go around. It might be paid in stock rather than cash, or get escrowed for a period so that the bank can take back some allowance that it already paid you if you subsequently mess up.

This seems good. On the one hand, the banks have some flexibility in their cost structure: If things get tough, they can cut payments to bankers rather than to, you know, creditors. On the other hand, bankers still get paid a lot, which they like. And there's one more good thing: If your allowance can go down, that reduces the asymmetry a bit. Just a bit! I mean, you still can't go below zero, so you still have incentives to take risks. But this is always true if you're making decisions with someone else's money. The point is, though, that if you are currently getting paid 200, will get up to an extra 200 if you make a lot of money, but will have 100 taken away from you if you lose a lot of money, then you no longer have quite the same option payoff. You have some downside. Your risk is slightly more symmetric.

All of this is very obvious but, I mean, read that report, and the accompanying opinion of the EBA. The EBA has concluded that these allowances are Not OK, because they are variable, and variable compensation causes risk-taking. Here's paragraph 36 of the report:

Staff cannot assume that the amount of the role-based allowance is fixed remuneration in a similar way as basic salary over the long term as it can be expected that in years where the institution does not perform well, allowances may be reduced, cancelled or not renewed. As no criteria for the review or adjustments are set or made transparent, the remuneration policy is neither predetermined nor transparent and does not guide staff behaviour. To avoid remuneration reductions in a downturn, staff might take additional or even excessive risks to try to ensure that an allowance is maintained.

That's not crazy exactly. I mean, they might, right? There's always an incentive to make money, and one way to make money is by taking risks. On the other hand, to avoid remuneration reductions, staff might avoid risks that would create the possibility of reducing an allowance. That's sort of the more natural expectation, right? Giving people something makes them averse to losing it. If you believe in incentives, the way to make bankers take less risk is to take something away when their risks go badly. And the allowances sort of do that!

But the EBA doesn't like them. It's moving to ban the allowances, which would throw bankers back to the bad choice of paying themselves less (ha!) or risking their banks' solvency with high fixed compensation costs. Weirdly, it doesn't even take a stance on which of those it prefers. Most opponents of the banking bonus culture also tend to think that bankers should just be paid less, bonus or otherwise. The EBA doesn't say that. Instead, its objection to variable compensation sounds ... it sounds a little like the EBA worries that allowances aren't fair to bankers?

The EBA is of the view that taking into account the above analysis, role-based allowances which are discretionary, not predetermined, not transparent to staff or not permanent should not be considered as fixed but should be classified as variable remuneration, in line with the letter and purpose of the CRD. The amounts paid are not fixed, are not permanent (i.e. not maintained over a period tied to the specific role and organisational responsibilities for which the allowance is granted), are not set out in a predefined objective manner, nor are they transparent to staff and therefore they do not promote sound and effective risk management but are based on other contractual conditions which do not form part of routine employment packages.

Your compensation shouldn't depend on your boss's whim, says the EBA. It should be predefined and objective. You should have a lot of certainty about how much you'll be paid. Nobody should be able to take your money away without due process. That's only fair.

Part of this is that the EBA is a European Banking Authority. So it internalizes European labor protections:

It is common under national labour and contract laws in the EU that fixed remuneration can only be changed with the consent of the employee or in collective bargaining and that it is paid over the full length of the contract. The amount may change with promotions or be renegotiated. National labour law may under some exceptional conditions allow unilateral changes by the employer.

That is, obviously, not generally the law in the U.S.

But it's tempting to also note that the EBA is also the European Banking Authority. And banks are Marxist paradises run for the benefit of their workers. And those workers, despite the stereotype of risk-living bankers, tend to prefer certain predictable (high!) compensation to risky uncertain compensation, just like most other workers. And the EBA has concluded that banks shouldn't be able to take money away from their workers just because those workers mess up and lose money. In that, the EBA and the bankers are in accord.

  1. Not true! But, you know, we're simplifying.

  2. I guess some people don't know that. One rough hypothesis I have is that regulators should want those people to become bankers. Though there are downsides there too. The ideal banker, from a regulator's perspective, would understand incentives, except for her own.

  3. Even if you're fired, you still have an asymmetric-compensation problem, as getting fired is still just an extreme form of getting paid zero. Even if they claw back your prior pay -- and they won't! -- you still end up with zero, which is still an asymmetric incentive. The only way to make things linear is if you could end up with less money than you had when you started in banking. (Or equivalent badness, e.g. criminal penalties.)

    This is important for the world, but not really for the bonus debate. Like, if you said, "One needs to think holistically about compensation to really understand asymmetric incentives," I would agree with you. But if you said that you wouldn't be so mad about bonuses.

  4. For them!

  5. For everyone! That's, like, a theme.

  6. A lot of these things are sort of true of salary, too, in the abstract; it's just that they tend not to be in practice. If your employer cuts your salary, that's a big deal. (Especially in Europe, where there are, you know, labor laws.) Just by calling it an "allowance" and creating different expectations around it, you make it more flexible.

  7. I mean, you have an option payoff, but it's well in the money, so you have downside risk. Also you don't have a call option any more, you have a call spread, since your bonus is capped. If you lose $20 million you will have your allowance taken back. If you make $20 million you will get a bonus of 200. If you make $2 billion you will also get a bonus of 200. So there's much less incentive to take outsize risks for outsize rewards: Your bonus can only be so big, no matter how big your profits are. This is not a feature of the allowance, of course; it's a feature of the bonus cap.

  8. The EBA acknowledges the high-fixed-cost problem in this paragraph, which I don't really understand:

    Some role-based allowances might only have been introduced to comply with the bonus cap introduced by the CRD IV while retaining some cost flexibility. Cost flexibility is of importance where the performance of the institution or a business unit is no longer considered adequate. When there is the need for cost flexibility, the review of allowances would usually take account of the costs and benefits of an allowance and the performance of relevant business units.

    But then it trails off. The next paragraph says "These allowances are also at least indirectly linked to performance as they are intended to retain cost flexibility."

  9. On the other hand, the European approach does involve clawbacks of bonuses. But note the contrast between the European approach (bonus caps, no allowances, lots of predictability) and the U.K. approach (bigger clawbacks, a distaste for bonus caps). The European approach is to eliminate asymmetry by flattening the upside. The U.K. approach is to steepen the downside.

  10. I don't know who originated that joke, or half-joke, but here's a 2013 Matt Klein usage anyway.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Matthew S Levine at mlevine51@bloomberg.net

To contact the editor on this story:
Zara Kessler at zkessler@bloomberg.net