Compensation in financial services (II)
By Denis on Wednesday 28 January 2009, 21:49 - Permalink
Bonuses have been an integral part of the compensation in financial services and clearly a bonus which isn't just a bonus translates abuse...
For example, abuse by the employers who, by defining normal salary as bonus, thus reduce the average severance package they have to pay to "let people go" (few actually have golden parachutes, and normal severance packages are based on basic salary only). If the system abuses workers, in avoiding to provide legitimate protection the workers are entitled to, is it surprising workers abuse the system? I'm afraid not. This does not legitimate abuse, but that may explain why it appears.
It is often said that the incentives are wrong in investment banking because employees have a call option: they get a share of the profits they may generate but they do not share the load of potential losses. This is a legitimate concern. However, funnily enough, financiers know how to value a call option. They know the value increases with volatility (higher profits and higher losses translate on average into higher bonuses since only profits count) hence the wrong incentive of always looking for more risks and leverage. It is easy to counteract this (while also ending the abuse mentioned in the previous paragraph): swap a substantial or integral part of the bonus with a fixed premium (the option price). If traders were to earn fixed and high salaries (instead of a low base + large bonus), the incentive would naturally move to keeping the job. To keep the job would mean finding the balance between two antagonist consequences of risk: one's job exists if it is profitable (which in finance requires taking some risk), one's job exists if it is profitable (which means risks are controlled enough that they cannot blow). If traders lose their job when they lose more than some pre-allocated capital (not the current state of the world), risks in the industry will be controlled tightly.
I addressed earlier why bonuses were useful to motivate talented financiers to work for others rather than only for themselves. A solution different from bonuses is simply to pay these financiers decently for the hours they put in to keep risks controlled (no joke: try discussing with an adoption agency about your working hours when you're a trader. The agency will immediately consider you're working too much, to an unacceptable degree, and that you're probably not suitable parenting material. You didn't even start the assessment process yet!). Again, the fair price is the value of the call option (on the notional the talented individuals can trade if they're independent, so we're unlikely to talk of a call worth millions).
Comments
It is often claimed that regulating compensation is hard, and that market forces determine the right level of pay.
I think we know by now that market forces are not enough: boards do not control enough CEO's pay (or strategy for that matter), market forces do not keep the "market price of risk" in line with the risks, etc. Can government or legislative powers determine what is the (fair) value of work in a particular job? Few people would agree (ask workers in public health services, teachers, firemen, policemen... or bankers). So let's assume the pay cannot be set by law. What can be done?
At the moment, the debate only comes into play because banking bonuses became politically unacceptable. Funnily enough, there are solutions regarding such bonuses, powerful yet not preventing compensations for the best talents:
The key legislative point would be: compensation should be set by the owners of the company (shareholders), not by a board or by managers who don't lose anything (directly) when compensating others. Realistically, it cannot be asked to shareholders to vote on the compensation of each individual employee but setting a envelope or voting on the top management should be an acceptable compromise. This would require only a bit more involvement of shareholders, delegation would still be the norm, but this would allow more subtle ways to handle managers than simply firing them and much better alignment of incentives. Long hours on a regular basis would be recognised as such - in fixed salary hence in severance packages too - at the risk of loosing talents willing to work less (see their family more) for the same price, performance at the top would be controlled rather than assumed, bonuses would become back a real motivational tool and bankers would drop the sense of entitlement that shocked the general public...
"On February 4th 2009, the U.S. president announced a set of executive compensation limits aimed mainly at firms that are the recipients of federal aid under the Troubled Asset Relief Program (TARP). The rules place a $500,000 cap on salaries. Any additional compensation will have to be in the form of restricted stock grants that will not vest until after taxpayers are repaid. In addition, all banks must accept new limits on golden parachutes, requirements that shareholders be able to review and vote on compensation packages (the vote would be non-binding), and tougher disclosure rules for spending on travel, office renovations and entertainment." [Knowledge@Wharton doc]
I fail to see why the shareholders vote should only be non-binding. Shareholders should take responsibility, and they own the company. Why should the vote be only informative to the board? This maintains the current black-or-white situation where shareholders either keep the top management or "let it go". It should be possible to say: we're happy with your job, but (given the conditions?) we do not agree on a pay rise of e.g. 20%. Another vote would follow, allowing maybe 10%, or 5%. Why should this be an issue?
Moreover, first reports indicate shareholders can opt out when things are going well... Well, before this crisis, things seem to go well... shareholders should not opt out and should keep their responsibilities theirs. They should not even have the option to opt out, overlooking pay should be a new duty for shareholders.
Most commentators admit such a $500k limit (+ deferred bonuses) makes hard to recruit CEOs to save an institution in trouble. The job is hard and, unless the CEO is already rich, it takes serious compensation to motivate somebody to take on a (career) suicide mission. Most commentators also agree on the fact that the limit will have to be temporary, for the reason just given and also because it is hard to let government set salaries for private companies. Either companies are private or public, but it is hard to hold a position for long in the middle. For the moment, I prefer my maximum percentage of total compensation in bonuses (and the limit being set by shareholders) to Obama's announce. It gives "private"-typed flexibility on rewards, while ensuring the system does not go dramatically against the general public's interests by creating incommensurate incentives for short-term views.
update: Are Bankers Over-Paid? by T. Philippon on the Stern on Finance blog from the Department of Finance of the Leornard R. Stern school of business (New York University) is worth reading. The full paper is also available.
In the series "xxx reads my blog", I had the nice surprise today to see that the Democratic Senator from Connecticut, Chris Dodd, Chairman of the Senate Committee on Banking, Housing and Urban Affairs, got amendments in the stimulus bill resulting in a
US bonus cap set at third of annual
salary.
Obviously, we're again hearing of brain drain and foreseeable loss of talent in the industry.
We're also hearing of penalising traders and investment bankers who may have nothing to do with companies' losses or strategic mistakes. That's ridiculous. If isolating their business was possible, it would have been done. Each investment bank have a myriad of subsidiaries already. The "profitable businesses" were not isolated because they needed the "critical mass" of a big group of businesses, the synergies of a big group of businesses, the financial strength of a big company. That is fine, but then the "separable profits" argument does not stand. The profit would not have existed in the first place without the businesses being part of a big company... so if the company is not profitable, there is no money to reward local performance. It is not to deny any reality to the mentioned performance... it is just that there is no money.
It is all the more funny that these profitable businesses could have had a bad year, and (I've seen it happen multiple times) they would have asked solidarity from the profitable units, to retain talents! So the flawed argument remains: profits are mine, losses are others'. Some people never learn...
I keep my view that bonuses should be bonuses, and if it is legitimate for a banker to earn more than its basic, then the basic should be increased. Market forces will still play, only with less abusive flexibility in compensations.
In the series "xxx reads my blog", UBS acknowledged that pay rise had been implemented as bonuses and reviewed the compensation structure. Reportedly, this practise was implicitly recognised as being wrong since
UBS Raises Top Bankers’ Salaries
as Bonuses Fall.
The FSA joined to the on-going debate, stating that
bonuses must reflect risk. This is
very nice in principles and very hard to implement (if not downright
counter-productive). However, the FSA also contributed that fixed salaries
should be kept large enough that groups can withhold bonuses entirely in years
when they make losses... i.e, implicitly, that fixed salaries should
be higher and bonuses lower.
Let me clarify why it is hard and possibly wrong to assess risks when setting bonus levels: up to the discovery of the fraud, traders having invested in the Madoff fund (or other low-vol high-return quantitative strategies that blew up) would have been awarded large bonuses... Low volatility is generally perceived as an indicator for high predictability hence low risk, high returns for (well) high returns, so risk-adjusted profits would still have been perceived favourably for decades! The deferral of two-thirds of each bonus would have had negligible impact over these decades. As the FSA stresses that bonuses should not be based solely on revenue or turnover or on a single year’s results, surely the depth of due diligence could have been taken into account but let's not hold our breath here. Measuring risk is notoriously hard. Nassim Taleb (writing in the FT about bonuses
here) has made his name on this:
should something seemingly predictable with a good level of certainty be seen
as barely risky or highly risky? It is just impossible to say if one doesn't
understand how the observed entity works. But funds never really give access to
their strategies, the bare-bones of their competitive advantage... so
risk-adjusted performance? Will the FSA tell the street how to measure
risks?
Some commentators (e.g. Hugo Dixon from breakingviews) go against "higher basis, lower bonuses" by linking the previous compensation culture with flexibility, itself perceived as an economic blessing favouring adaptation. The argument "higher base = less flexibility = harder to stand the next crisis" does not work though if flexibility is only for appearance's sake (as seen recently): most bonuses paid for 2008 have been so on the justification of retaining talent. And with the comment that the said talents would not work the long hours they do for their base salary. In such a case, the flexibility these commentators talk about simply does not really exist and only comes down to abusing the rights of employees when disputes arise or contracts are terminated.
If the raise in fixed salary is the un-compressible amount of "bonus" that had to be paid to retain people, and if bonuses subsequenty are just bonuses (with a real possibility of being null in bad years at the corporate level), the higher base is obviously not a higher cost and it is no less flexible (except marginally, at termination)... Smaller bonuses (no matter what, including major success) also means smaller gambling incentives; the game becomes about staying in the long term and moving up the social ladder, like in any other business line.