April 30, 2008
Risk taking, remuneration and leverage
Mervyn King, Governor of the Bank of England, is correct in linking the reckless lending by banks and other financial institutions that, together with the matching reckless borrowing, lay at the roots of the current financial crisis, to remuneration structures that rewarded extreme risk taking on poorly designed financial products. The diagnosis is fine. What to do about it is less obvious. These remuneration packages did not fall to earth from the moon. They are the result of a distorted economic environment. The key distortions, unfortunately, cannot be remedied, because they have highly desirable consequences as well as the dysfunctional ones highlighted by the crisis. Let’s consider some of them:
(1) Limited liability. A great social invention for getting people to put up risk capital for ventures with uncertain returns. However, an enterprise with limited liability by definition introduces a serious asymmetry in the payoff function. Losses are limited to the capital provided to the company. Gains are unbounded from above.
(2) No more debtors’ prison. We have moved on quite far sinnce Mr. Micawber’s statement in Charles Dickens’ David Copperfield :” Annual income twenty pounds, annual expenditure nineteen nineteen six, result happiness. Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.” A company that borrows and goes broke will not have its management team slung into the slammer until the debts have been paid off. With the growth of non-recourse loans for households and legal personalities, only the collateral is at stake in case of default. Non-recourse lending is, from the borrower’s point of view, a form of limited liability borrowing. The combination of limited liability and cheap default for corporates further accentuates the asymmetry of the pay-off function.
(3) Free labour. In most countries chattel slavery and indentured or bonded labour (including debt bondage) have been abolished de jure if not necessarily de facto. One consequence of this has been that human wealth (the present discounted value of the future expected stream of after-tax labour income) is lousy collateral for loans. Not only is it illegal to sell oneself or someone else into slavery, the free labour doctrine means in practice that labour contracts are binding (have legal consequences if breached) only on the part of the employer.
An employee therefore cannot credibly commit himself not to leave if a competitor makes him a better offer. The employee’s inability to sign a contract that is legally binding on himself means that opportunistic behaviour is constrained only by reputational considerations. In such a world there are likely to be at least two kinds of equilibria. One is where opportunistic behaviour is the exception, your reputation for ‘loyalty’ follows you around and people that renege on (non-legally binding) commitments are pariahs who don’t get hired by the best employers. The other equilibrium is where opportunistic behaviour is the rule, a reputation for loyalty means you are viewed as a ’sucker’, employers ruthlessly raid their competitors to poach their best staff and employees are forever dropping handkerchiefs and have both eyes open for the main chance. The first equilibrium is based on Hirschman’s ‘voice’ and ‘loyalty’, the second on ‘exit’ or the outside option. There is no doubt that the world has moved comprehensively to the ‘exit’ equilibrium, where long-term commitment is the exception rather than the rule.
(4) No negative bonuses
The absence of negative bonuses is a form of limited liability for employees. It’s custom-based rather than legally based, as far as I know.
(5) Bonuses based on short-term performance
The combination of asymmetric pay-off functions for corporations, favouring risk taking, and the inability of employees to make a long-term commitment to stay with their employer means that contracts based on a long history of earnings, profits or some other measure of individual or company-wide performance are going to be few and far between. Should such a contract be signed, a competitor is likely to buy it out on generous terms, if the employee stuck with the contract has just the characteristics the competitor urgently thinks it needs.
(6) Excessive leverage
Finally, leverage permits the scaling of everything to incredible levels, including bonuses and other rewards. When irrational exuberance rules the roost, all promises become credible. Credit is based on the sale of promises. When promises are cheap, leverage explodes and bonuses with it.
What is to be done?
I don’t see an early return of chattel slavery and indentured labour in the financial sector. Nor are we likely any time soon to witness the abolition of limited liability and the re-introduction of debtors’ prisons. The fear of God, or of the devil, instilled by the current crisis will restrain risk taking and excessive and ill-considered remuneration in the financial sector for at most a couple of years (the half-life of memory in the City and on Wall Street). Another attempt at reforming corporate governance - across the board, not just in the financial sector - may help a bit, but I am not holding my breath.
I therefore only expect any real progress from limiting or penalising (through progressive capital and liquidity requirements), all highly leveraged institutions above a certain minimum size, regardless of what they call themselves, regardless of the sector they are formally located in and regardless of their ownership structure. Leverage amplifies the good, the bad and the ugly. Given the many asymmetries favouring excessive risk taking, restricting leverage and regulating it across the board, seems a promising way to tackle some of the worst excesses.
The above contains about the silliest argument I’ve ever seen. Throw in chattel slavery with bonuses paid on short term gains - why, what’s the difference? that’s not even a funny smoke screen for the upper classes - that’s more like a drunk’s maudlin spiel.
Actually, it would be quite easy to devise a tax code that would swoop down and expropriate those excessive bonuses, and that would simply make it illegal for any company to pay the tax of any of its employees. Hey presto, you got yourself an incentive structure going. You can build it yourself, or you can listen to the sophisms of the defenders of the upper crust as they make ever more implausible arguments for maintaining an ever more inequitable and, apparently, inefficient system.
Posted by: roger | April 30th, 2008 at 10:08 pm | Report this commentJust to add:
1) Pay for a typical City worker is convex: he or she receives a flat wage, but after meeting target (revenues, profit), pay increases. Pay therefore has the characteristic of an option: just replace “target” by “strike price”, and “revenues or profits” by “underlying share price”.
Those who work in the city know all too well that options are most valuable when the underlying share price is volatile. For pay this means that City workers have a perverse incentive to create swings in revenues and profits. These swings hurt all of us! Moreover, we should not be suprised to see banks reporting the biggest losses ever, as these losses will not affect base salaries. Currently, banks’ managers have strong incentives to exaggerate losses so that next year’s earnings and bonuses will be super stellar again.
The key point here is that these ubiquitous convex pay contracts create massive externalities: a few city workers benefit at the expense of the world economy. Real people suffer, just because some City workers have a good time creating serious disruptions.
We cannot do away with convex pay deals, but remuneration boards should be aware of the devastating consequences of the deals they approve.
2) Many people don’t learn. I am amazed to see executives receiving big salaries only to be fired after, say, three years. We should not accept that these executives are the victims of some sort of bad luck. They are smart and they know very well how they will perform. Boards should price these contracts with a long term focus in mind.
Posted by: iconoclast | April 30th, 2008 at 11:16 pm | Report this commentWillem Buiter has world-class skills in starting from intelligent and sensible observations, ruthlessly stretching them in the process, and then arriving at ridiculous conclusions. To use a metaphor, he leveragess the sensible observations excessively. He should follow his own advice at the end of this article.
Posted by: FX | May 1st, 2008 at 11:15 am | Report this commentPerhaps King is comparing the merits of the City bonus system with the renumeration system that he oversees in the Bank of England.
The Bank’s approach to renumeration appears back-end loaded, and thus incentivises employees to promote non-default of the sponsor (admittedly here the sponsor is HMT, i.e. the rest of us). This aligns interests over the long term but at the cost of increasing fixed costs and the cost of people moves.
A cursory glance at the BoE’s accounts to the year ending March 31, 2007 shows DB pension contributions on the part of the Bank at 44% of salaries. This compares very well to the average private sector DB contributions of 16% of salary, and extremely well compared to average private sector DC contributions of 6% of salary. King’s own personal pension pot is valued in the accounts at around £4m (where did the £1.5m cap go that applies to the rest of us before the super-tax kicks in?).
Posted by: fi-geekier | May 1st, 2008 at 1:09 pm | Report this commentThe remuneration packages may be part of the problem, but are certainly not the key factor behind this current crisis. Let’s not forget that crisis of similar nature and magnitude has been around for decades and centuries, whilst the said “excessive pay packages” have not.
I have 2 answers to the current problems:
1) we simply accept with Minsky (hence, Keynes) that in our world of financial capitalism there is no such thing as an “equilibrium”. A cyclical upswing in credit and real output ends abruptly with an asset meltdown, driven by the growing unability of debtors to service their increasingly riskier financing arrangements (hedge->speculative->Ponzi, as described by Minsky). Government and central bank content themselves to clean up the mess and make sure that things deflate not too far down and spiral out of control, but other than that, there is not much more that they could do.
OR 2) we establish a more stringent credit framework, e.g. require the banks to put up far more capital for loans that fall into speculative or Ponzi catergories. The ABS-financed conduits and SIVs clearly were Ponzi in nature. The very fact that the FED and the SEC were happy that these conduits would be kept off banks balance sheets and there backed by only a tiny amount in capital, to me clearly is a behaviour best described as “want to see no evil, speak no evil, hear no evil”.
In stark contrast to that, it would have been easy for central banks and other regulators to sa y “NO” to certain practices, like for instance the Bank of Spain did when it came to conduits for spanish mortgage debt.
Posted by: weissgarnix | May 2nd, 2008 at 12:42 pm | Report this commentIf regulators are concerned about the asymmetric relationship between compenation and risk, there is a relatively straight forward way in which they can tackle this without directly imposing on compensation policies.
The amount of capital a bank is required to hold is a function of the BIS ratios. However a regulator can require a bank to hold excess capital if it feels that a bank’s risk management processes are not sufficient for the operations that it is running. Why not consider compensation policy as a risk factor to be considered and give the regulator the ability to require excess capital if it feels that the bank’s compensation structure puts the organisation at risk.
The bank will then cost this extra capital Vs the returns they are making and change its behaviour accordingly
Posted by: david, london | May 2nd, 2008 at 6:09 pm | Report this commentPreliminaries: commenting upon comments above:
1) The problem with Roger’s tax ‘em and flog ‘em approach is that it tries to optimize for a resentment function, won’t work since jurisdictions compete for the proceeds (tax, other) of mobile economic activity, avoidance is easy, as it is in general, and governments are not smart enough to keep up with profit-motivated bankers. This good old Labour stuff, was debunked a long time ago.
2) Iconoclast, a typical city worker settles trades, accounts for P&L or fixes system problems. Of the traders/originators/sales people, most earn very good money but not the seriously excessive stuff that some rant over. A small number of elite performers make the big money. Your traded option analysis is flawed (not floored!): the share price cannot be less than zero, profit can. Volatility of the share price means increasing chances of hitting the strike, volatility in the employee example means, ceterus paribus, less. Senior execs are often hired in to achieve something (e.g fix a problem) that the board thinks will be very difficult. When results are slow, there is a pause before both sides move to negotiating an exit so that the next candidate can try to address the self-inflicted problem.
Main point: competition to deliver profit can result in super-optimal risks being taken, super-optimal competition for employees and remuneration. The latter two are symptoms not causal factors. Conceptually, solutions require companies to publish something like a Sharpe Ratio that shows return over risk (and they publish VAR today) so that we can judge whether there is excess risk. Practically, it won’t work. Institutions will play with numbers, eg. capital arbitrage, inventing new concepts like SIVs. The idea behind Goodhart’s Law applies too: an observable stable relationship between two things breaks when used for policy targeting purposes.
Conclusion: regulation is a necessary evil required to damp down the excessive fringes of innovation. Innovation is net-net good for us, always has been, always will be. Let’s not worry too much about a 2 year retrenchment in capital markets as we learn the consequences of bespoke hi-leverage over genetically engineered assets. The process we are going through is the medicine, the learning experience, the place to lock in that which we have understood. The regulators’ job is to follow along behind, cleaning up and being wise after the event.
Posted by: Andy, London | May 5th, 2008 at 8:58 am | Report this commentMy reading of Prof. Buiters’ post is that the primary causes of this round of financial market turmoil are assymetries of incentive structures in the legal framework of business law, financial market regulation and labor markets. In short, existing structures (law) and processes (information dissemination and gathering) favor risk taking and penalize risk avoidance. Research in psychology, economics and game theory has established that people value what they have more than what they can get (”bird in hand”), but, skepticism (risk aversion) shrinks as potential reward grows. In reviewing the options for remedies, some knowledge of history will help us rediscover some effective ones.
There is historical precedent for this situation in the U. S. — (what Paul Krugman calls) the Extended Gilded Age from (roughly) 1880 through 1930. An example of regulatory lassitude is the margin requirement for stock purchases: only 10% equity was required. Another example is that liquidity and capital required for deposit-taking lending institutions was virtually nil. We are familiar with the 4-year downward, self-reinforcing cycle that followed the stock market crash of 1929.
I agree with Prof. Buiters. The rewards of greed and narrow self interest are too great and the risks of failure are too little. It is now possible, even plausible (if unreasonable), to hope to earn enough money in a few years (less than 5, even less than 3) to secure one’s quite comfortable financial future. And, the consequences of failure are so small that they do not discourage continued efforts even after successive failures in risky ventures. We are Wile E. Coyote chasing the Roadrunner in this cartoon of law and regulations, forever bouncing back from disaster only to engender another for ourselves.
What to do? First, indeed, limit leverage for enterprises and individuals of all types to limit the financial risk (vs. operating risk) that they take on(a complex task, indeed). After the crash, during the Roosevelt administration, the American financial authorities (SEC, when created and empowered) expanded initial margin requirements to 50% of inital purchases and 35% equity to total portfolio value (stocks). The SEC also set minimum capital requirements for non-deposit taking financial institutions. The Federal Reserve Bank increased the required capital and deposit/liability ratios for deposit taking institutions. It worked and would continue to work, if extended to non-deposit-taking financial institutions (including industrial captives). While mortgage lenders insisted on 10% down with PMI or 20% down without, the default rate on mortgage loans remained small, even during economic recessions. When non-deposit-taking lenders entered the market and assumed the role of Fannie, Ginny and Freddie, without similar regulation, real property leverage skyrocketed, prices skyrocketed, investment became indistinguishable from specualtion and, well, we know what happened (duh!).
Second, increase marginal tax rates at the top and decrease them or eliminate them at the bottom. Tax inheritances as if they were ordinary income. Eliminate all regressive tax schemes (including U. S. Social Security and Medicaid taxes, sales taxes, VATs, etc.) and simplify the tax codes to eliminate the distinction between ordinary income and capital gains as well as all deductions and exemptions of any kind. Such a scheme would reduce the rewards to risky behavior (not least among many other consequences) much as it did during the years following the second world war, an era of increasing prosperity and growing standards of living world wide. We must be careful not to remove all the rewards, but we do need to find a better place for the fulcrum.
Third, seek international cooperation in devising and enforcing all such schemes. Without such cooperation, financial capital will flow to those states with favorable tax regimes and little or no disclosure requirements. Businesses would arbitrage financial and labor markets in pursuit of superior economic outcomes and, thereby, capture most, if not all, future gains from greater labor and capital productivity. Individuals who are not owners would be left “holding the bag” while shareholders and corporate executives would be racing off with the money.
We are not completely rational beings. No markets are even weak-form efficient or anything resembling “free” or “unregulated”. To regulate or not to regulate is not the question; markets are already regulated. To tax or not to tax is not the question; taxation is fact and fabric of life. The questions are how to regulate who and how much and whom to tax and how much?
Posted by: Charles | May 5th, 2008 at 8:08 pm | Report this comment