Risk and Investment Banking Compensation

The erstwhile bosses at financial majors Citigroup, Merrill Lynch and Countrywide, were in Washington at Congressional hearings, defending their erstwhile pay packages. The New York times reports it here.

I have written before about the challenges in setting CEO compensation before [link]. In the capital markets it gets even more complex.

Raghuram Rajan and Martin Wolf have a couple of great pieces in the FT. These are behind pay walls so I’ll see if I can bring out the key points here with a thought experiment.

Let’s say we create a highly simplified situation where investing in a financial asset was like rolling dice. A firm can invest $1B and there are six possible outcomes that are equally likely. Five of them have the firm making a 5% return on its invested capital. In the sixth one, it loses 30% of its investment.

Let’s also assume that the manager at the firm is entitled to 10% of the profit the firm makes on these investments. If the sixth event occurs, he is fired.

The interests of the firm and manager are divergent. [The spreadsheet is here Banker Compensation]. The expected value of the gain on each investment is:

Loss for the firm – $ 16.67 M
Bonus for the manager – $3.33 M

Any leverage will amplify the difference.

Actually, the expected value of each investment is perhaps not the right metric to look at. What we should be looking at is the expected cumulative gain of the firm and the manager assuming that the investments continue until the long-tail event (30% loss of investment) occurs. By my calculations the expected cumulative gain is:

Loss for the firm – ~$57 M
Bonus for the manager – ~$25 M

At the heart of the compensation issue for bankers is the possibility that managers can take advantage of this divergence to personally gain, while the firm loses. There are at least a couple of ways that I can see this happening.

One way would be to belittle the risk – either the probability of its occurrence or the magnitude of the loss. Given that managers and not shareholders, assess the risk, this is always going to be a possibility.

Another way would be to influence the pricing of non-exchange traded assets. If the asset is still on the books at the end of the bonus assessment period, it is in the manager’s interests to inflate the notional gain. Again, the pricing of these assets is done by managers not shareholders. If this game is taken further, a firm’s book could disproportionately comprise of long maturity, illiquid assets because managers find it easier to manipulate notional gains.

At the root of it, this problem is about managers getting rewarded on short-term performance while shareholders are left picking up the pieces in the long-term. This is an old problem that is not the exclusive domain of the financial industry. There are no easy solutions (see my earlier post).

Martin Wolf makes the case that bankers’ bonuses should be regulated. Investment banks are in the unique situation where most governments will bail out a distressed financial institution instead taking the pain of letting it go under. Since the taxpayers’ money is going to go towards rescuing a bank it is only fair, says Wolf, that in exchange, the managers of the bank have certain regulations on how their pay is structured.

I think it is highly unlikely that any regulation on pay will be put in place. There are so many problems with any regulated form of pay that having no rules around compensation is probably the least worst solution.