Eggheads and long tails
May 17th 2007 From The Economist print editionInvestment banks are a high-wire act. How good are the safety nets?
BANKERS can be surprisingly colourful when describing the risks that keep them, their shareholders and their supervisors awake at night. One Spaniard likens the financial system to a “spider's web” over which central banks have less control than they did in the past; another talks about a “constellation” of risks and approaching asteroids. A French banker refers to a “soufflé” of off-balance-sheet exposures, whipped up by leverage. A British counterpart, speaking over breakfast, compares the process of transferring risk around the financial system to spreading butter on toast. With luck, he says, the butter will not be rancid.
Mr Isaacs at Lehman Brothers in London notes that over the past century of world wars and cold wars the financial system has proved more resilient than might have been expected. Like many bankers, he thinks that the biggest danger to the financial system at present is avian flu. But he agrees that there are new difficulties in managing a system where risk is transferred from banks to investors. Other bankers point out that although investment banks may have fewer loans on their balance sheets, they now carry more complex exposures, such as collateralised lending to hedge funds.
Risk, though spread far more widely, has not disappeared. In a crisis such exposures may surface in ways that are hard to predict, especially because so many banks are exposed to the same counterparties. “The risk is disseminated, but trying to re-aggregate it becomes an almost impossible task,” says one senior investment banker.
Banks are not helpless in the face of such uncertainty. One of the landmark changes of the past 20 years has been the development of risk-management systems that estimate potential market, credit, liquidity and operational losses and seek to impose appropriate limits. Investment banks, because of the complexity of their exposures, have been at the forefront of efforts to measure and manage risks, encouraged further by the approach of Basel 2. They have employed mathematicians and spent billions of dollars on computer systems. Barclays Capital has turned whole walls into whiteboards to accommodate the hugely complex formulas involved. But the results of all this effort are not quite as brilliant as might have been hoped.
The bond-market turmoil leading to the liquidation of Long-Term Capital Management (LTCM) in the summer of 1998 not only exposed the risk-management negligence of that institution, staffed by Nobel laureates and storied traders; many investment banks were caught out too. The lesson, learned at an exorbitant cost, was that models are only as good as what is fed into them. So risk management is as much about human judgment as about mathematical genius.
And there is always the million-to-one chance, far along one tail of the bell curve, that something will go disastrously wrong. There is a parallel here with climate change, where many previous sceptics are now in favour of taking preventive action—not so much because they think the world could not cope with the most widely expected rise in temperature, but because they fear that things might just possibly turn out very much worse.
Thinking the unthinkable
An assessment of the investment banks' risk-management systems must start from the premise that as they have grown, and diversified their sources of revenue, they should be less exposed to losses capable of wiping out the firm. That, after all, is the point of diversification.
Also, trading floors now pay much more attention to risk management than they used to, and although there are inevitable tensions between risk-takers and risk-controllers, almost everyone appears to take the issue seriously.
But because risk management has become so complex, understanding the systems in place even in one bank could take days. Those who have tried, such as Moody's Investors Service, are impressed with the progress made by Wall Street banks, though they think that levels of disclosure are higher among some of the big European banks. They have also found a fascinating mixture of risk-management methodologies which provide a rare glimpse into the internal culture of these secretive organisations.
In a report last year Moody's looked at Goldman Sachs, the most profitable investment bank and, in many respects, the envy of its peers. Describing its risk philosophy, Moody's said Goldman saw risk not only as the counterpart to revenues or profit, but as “the source of profit”. According to Moody's, “most firms might ask themselves if they are taking too much risk; Goldman Sachs's most senior management—more aggressively than others—constantly asks the question, 'are we taking too little?' and 'are we taking the right kind of risks?'”
In its annual report Goldman says that all risk-control functions ultimately report to the management committee (though, in common with other Wall Street firms, not to the board of directors). The checks and balances then cascade down to the level of each trading-desk head, the front-line where decisions to buy and sell are made almost instantly. The main tool for assessing the short-term losses traders are exposed to is value at risk (VAR). Goldman, for example, would expect to lose more than its VAR in the course of a single day about once a month. Last year its daily VAR averaged $101m, against $70m in 2005. But VAR is clumsy, relies on historical data and works best when markets are stable. It does not capture the effects of highly stressed markets, nor what would happen if assets could not be quickly hedged or liquidated.
What bankers use for those things are stress tests. According to Moody's, the overall assumption behind Goldman's scenario planning is that the firm is unable to hedge or sell positions and must hold them from peak to trough. One of its fixed-income tests is a replay of the 1998 LTCM and bond-market meltdown. In its equity division it uses a “supercrash test” that assumes an instantaneous fall in the price of equities of 50%. The Armageddon experiment combines the worst that has happened in both fixed-income and equity markets over the past 30 years. Besides market risks, Goldman also assesses credit risks, based on whether a counterparty might default on a loan or fail to honour a derivative contract, and liquidity risk. But the firm does not disclose much detail on any of its simulations, nor do most other banks. France's Société Générale is an honourable exception.
Moody's is impressed by many of Goldman's techniques. Although the risks have increased, the ability to generate earnings has gone up even more. But the rating agency also has concerns, both about Goldman and about its Wall Street peers. It worries about the opacity, the exposure to illiquid investments and the failure to spell out worst-case scenarios. Goldman has large, illiquid off-balance-sheet exposures, too, known as Variable Interest Entities. These include pools of debt securities such as mortgages. Last year these assets surged to $72.5 billion from $54.1 billion, and Goldman's maximum exposure to loss from them doubled to $20 billion.
Mr Viniar, the firm's chief financial officer, acknowledges that the models cannot be perfect: “We know with 100% certainty that we can't know everything.” But the main backstop, what he calls “the lifeblood of Goldman Sachs”, is liquidity. He thinks that liquidity problems are the main source of risk in investment banking. They brought down Drexel Burnham Lambert in 1990 and Barings five years later. Goldman keeps $50 billion-60 billion of government securities ready to be deployed if a liquidity crisis strikes. The amount increases in proportion to the firm's assets and capital. And Goldman finances all its long-term illiquid assets with long-term debt and equity.
Simmering pots
The universal banks conduct their risk management along similar lines to Wall Street. But there are differences in methodology, and in each firm's risk appetite. Credit Suisse and UBS, two Swiss banks, are among the most forthcoming with their risk-management assessments.
Wilson Ervin, Credit Suisse's chief risk officer, appropriately works in a building that is a living testament to risk. Credit Suisse's New York headquarters, at 11 Madison Avenue, was intended to be the world's tallest building when Metropolitan Life started work on it in the 1920s. But the 1929 stockmarket crash put an end to that ambition and construction stopped at 28 floors. That is how the tower has remained to this day, stubby but stunning.
Mr Ervin is ready to clamp down on exposures that might dent the bank's balance sheet or earnings, but most of his job involves “watching the pots simmering”. He says it is important to concentrate on good information and clear rules, not to micromanage the decisions of traders and credit officers. However, he notes that risk appetites are changing across the industry, especially in the use of investment banks' own capital to support private-equity transactions. “Things we used to say were verboten are now considered an acceptable risk-return trade-off as liquidity and hedging markets have developed.”
Maria Jeeves
Credit Suisse's rival, UBS, can hardly forget the risk-management failure of its predecessor, Union Bank of Switzerland. It made disastrous investments in the late 1990s, including in LTCM, and was taken over, in tatters, by Swiss Bank Corp. Now, as at Credit Suisse, risk is managed at the top of the bank. Responsibility is delegated by the board of directors to Walter Stuerzinger, group chief risk officer.
Mr Stuerzinger describes UBS's approach to risk in stark terms: zero tolerance for fiefs; beware of tail risks, risk concentrations, illiquid risks and legacies; avoid risks that cannot be properly assessed or limited; and never be hostage to a single transaction or client. UBS sets its overall risk capacity on the basis that it wants to be able to pay dividends out of current earnings, not retained ones. But thanks to earnings growth and the divestment of its private-equity portfolio, its ability to take risks has grown much faster than its actual risk exposure. Indeed, it is sometimes criticised for not taking large enough punts. Earlier this year Ken Moelis, an investment banker prominent on Wall Street, left the firm, reportedly because he felt it was too conservative in using its own capital in private-equity deals.
A question of culture
In the end, for all the fancy technologies, it is culture that counts. This boils down to the people a firm employs, the responsibilities it gives them and the way it rewards them for putting the company's interests above their own. One well-established technique is to give them shares in the company, which can make up a large part of their pay. Some Wall Street firms maintain a partnership culture even after they have gone public. Goldman, for example, has a “partnership pool” into which a large share of its profits go.
Given the rapid staff turnover at investment banks, keeping the firms' “culture carriers” on board is a perennial problem. The best way is to pay them generously. But success is also a powerful incentive, so employees need to be allowed to take risks and not be stifled with controls. That may mean giving ambitious twenty-somethings their heads. Mr Blankfein at Goldman acknowledges that some of the young traders at the firm have no memory of a business cycle. Clearly he does, but he says he respects their views and has not been disappointed so far.
Above those traders are wiser heads who should be prepared to extend a restraining hand. Between them, the 21 members of Goldman's management committee have 300 years of experience within the firm. Yet even that massed experience cannot guarantee that they will always be able to keep trouble at bay.
May 17th 2007 From The Economist print editionInvestment banks are a high-wire act. How good are the safety nets?
BANKERS can be surprisingly colourful when describing the risks that keep them, their shareholders and their supervisors awake at night. One Spaniard likens the financial system to a “spider's web” over which central banks have less control than they did in the past; another talks about a “constellation” of risks and approaching asteroids. A French banker refers to a “soufflé” of off-balance-sheet exposures, whipped up by leverage. A British counterpart, speaking over breakfast, compares the process of transferring risk around the financial system to spreading butter on toast. With luck, he says, the butter will not be rancid.
Mr Isaacs at Lehman Brothers in London notes that over the past century of world wars and cold wars the financial system has proved more resilient than might have been expected. Like many bankers, he thinks that the biggest danger to the financial system at present is avian flu. But he agrees that there are new difficulties in managing a system where risk is transferred from banks to investors. Other bankers point out that although investment banks may have fewer loans on their balance sheets, they now carry more complex exposures, such as collateralised lending to hedge funds.
Risk, though spread far more widely, has not disappeared. In a crisis such exposures may surface in ways that are hard to predict, especially because so many banks are exposed to the same counterparties. “The risk is disseminated, but trying to re-aggregate it becomes an almost impossible task,” says one senior investment banker.
Banks are not helpless in the face of such uncertainty. One of the landmark changes of the past 20 years has been the development of risk-management systems that estimate potential market, credit, liquidity and operational losses and seek to impose appropriate limits. Investment banks, because of the complexity of their exposures, have been at the forefront of efforts to measure and manage risks, encouraged further by the approach of Basel 2. They have employed mathematicians and spent billions of dollars on computer systems. Barclays Capital has turned whole walls into whiteboards to accommodate the hugely complex formulas involved. But the results of all this effort are not quite as brilliant as might have been hoped.
The bond-market turmoil leading to the liquidation of Long-Term Capital Management (LTCM) in the summer of 1998 not only exposed the risk-management negligence of that institution, staffed by Nobel laureates and storied traders; many investment banks were caught out too. The lesson, learned at an exorbitant cost, was that models are only as good as what is fed into them. So risk management is as much about human judgment as about mathematical genius.
And there is always the million-to-one chance, far along one tail of the bell curve, that something will go disastrously wrong. There is a parallel here with climate change, where many previous sceptics are now in favour of taking preventive action—not so much because they think the world could not cope with the most widely expected rise in temperature, but because they fear that things might just possibly turn out very much worse.
Thinking the unthinkable
An assessment of the investment banks' risk-management systems must start from the premise that as they have grown, and diversified their sources of revenue, they should be less exposed to losses capable of wiping out the firm. That, after all, is the point of diversification.
Also, trading floors now pay much more attention to risk management than they used to, and although there are inevitable tensions between risk-takers and risk-controllers, almost everyone appears to take the issue seriously.
But because risk management has become so complex, understanding the systems in place even in one bank could take days. Those who have tried, such as Moody's Investors Service, are impressed with the progress made by Wall Street banks, though they think that levels of disclosure are higher among some of the big European banks. They have also found a fascinating mixture of risk-management methodologies which provide a rare glimpse into the internal culture of these secretive organisations.
In a report last year Moody's looked at Goldman Sachs, the most profitable investment bank and, in many respects, the envy of its peers. Describing its risk philosophy, Moody's said Goldman saw risk not only as the counterpart to revenues or profit, but as “the source of profit”. According to Moody's, “most firms might ask themselves if they are taking too much risk; Goldman Sachs's most senior management—more aggressively than others—constantly asks the question, 'are we taking too little?' and 'are we taking the right kind of risks?'”
In its annual report Goldman says that all risk-control functions ultimately report to the management committee (though, in common with other Wall Street firms, not to the board of directors). The checks and balances then cascade down to the level of each trading-desk head, the front-line where decisions to buy and sell are made almost instantly. The main tool for assessing the short-term losses traders are exposed to is value at risk (VAR). Goldman, for example, would expect to lose more than its VAR in the course of a single day about once a month. Last year its daily VAR averaged $101m, against $70m in 2005. But VAR is clumsy, relies on historical data and works best when markets are stable. It does not capture the effects of highly stressed markets, nor what would happen if assets could not be quickly hedged or liquidated.
What bankers use for those things are stress tests. According to Moody's, the overall assumption behind Goldman's scenario planning is that the firm is unable to hedge or sell positions and must hold them from peak to trough. One of its fixed-income tests is a replay of the 1998 LTCM and bond-market meltdown. In its equity division it uses a “supercrash test” that assumes an instantaneous fall in the price of equities of 50%. The Armageddon experiment combines the worst that has happened in both fixed-income and equity markets over the past 30 years. Besides market risks, Goldman also assesses credit risks, based on whether a counterparty might default on a loan or fail to honour a derivative contract, and liquidity risk. But the firm does not disclose much detail on any of its simulations, nor do most other banks. France's Société Générale is an honourable exception.
Moody's is impressed by many of Goldman's techniques. Although the risks have increased, the ability to generate earnings has gone up even more. But the rating agency also has concerns, both about Goldman and about its Wall Street peers. It worries about the opacity, the exposure to illiquid investments and the failure to spell out worst-case scenarios. Goldman has large, illiquid off-balance-sheet exposures, too, known as Variable Interest Entities. These include pools of debt securities such as mortgages. Last year these assets surged to $72.5 billion from $54.1 billion, and Goldman's maximum exposure to loss from them doubled to $20 billion.
Mr Viniar, the firm's chief financial officer, acknowledges that the models cannot be perfect: “We know with 100% certainty that we can't know everything.” But the main backstop, what he calls “the lifeblood of Goldman Sachs”, is liquidity. He thinks that liquidity problems are the main source of risk in investment banking. They brought down Drexel Burnham Lambert in 1990 and Barings five years later. Goldman keeps $50 billion-60 billion of government securities ready to be deployed if a liquidity crisis strikes. The amount increases in proportion to the firm's assets and capital. And Goldman finances all its long-term illiquid assets with long-term debt and equity.
Simmering pots
The universal banks conduct their risk management along similar lines to Wall Street. But there are differences in methodology, and in each firm's risk appetite. Credit Suisse and UBS, two Swiss banks, are among the most forthcoming with their risk-management assessments.
Wilson Ervin, Credit Suisse's chief risk officer, appropriately works in a building that is a living testament to risk. Credit Suisse's New York headquarters, at 11 Madison Avenue, was intended to be the world's tallest building when Metropolitan Life started work on it in the 1920s. But the 1929 stockmarket crash put an end to that ambition and construction stopped at 28 floors. That is how the tower has remained to this day, stubby but stunning.
Mr Ervin is ready to clamp down on exposures that might dent the bank's balance sheet or earnings, but most of his job involves “watching the pots simmering”. He says it is important to concentrate on good information and clear rules, not to micromanage the decisions of traders and credit officers. However, he notes that risk appetites are changing across the industry, especially in the use of investment banks' own capital to support private-equity transactions. “Things we used to say were verboten are now considered an acceptable risk-return trade-off as liquidity and hedging markets have developed.”
Maria Jeeves
Credit Suisse's rival, UBS, can hardly forget the risk-management failure of its predecessor, Union Bank of Switzerland. It made disastrous investments in the late 1990s, including in LTCM, and was taken over, in tatters, by Swiss Bank Corp. Now, as at Credit Suisse, risk is managed at the top of the bank. Responsibility is delegated by the board of directors to Walter Stuerzinger, group chief risk officer.
Mr Stuerzinger describes UBS's approach to risk in stark terms: zero tolerance for fiefs; beware of tail risks, risk concentrations, illiquid risks and legacies; avoid risks that cannot be properly assessed or limited; and never be hostage to a single transaction or client. UBS sets its overall risk capacity on the basis that it wants to be able to pay dividends out of current earnings, not retained ones. But thanks to earnings growth and the divestment of its private-equity portfolio, its ability to take risks has grown much faster than its actual risk exposure. Indeed, it is sometimes criticised for not taking large enough punts. Earlier this year Ken Moelis, an investment banker prominent on Wall Street, left the firm, reportedly because he felt it was too conservative in using its own capital in private-equity deals.
A question of culture
In the end, for all the fancy technologies, it is culture that counts. This boils down to the people a firm employs, the responsibilities it gives them and the way it rewards them for putting the company's interests above their own. One well-established technique is to give them shares in the company, which can make up a large part of their pay. Some Wall Street firms maintain a partnership culture even after they have gone public. Goldman, for example, has a “partnership pool” into which a large share of its profits go.
Given the rapid staff turnover at investment banks, keeping the firms' “culture carriers” on board is a perennial problem. The best way is to pay them generously. But success is also a powerful incentive, so employees need to be allowed to take risks and not be stifled with controls. That may mean giving ambitious twenty-somethings their heads. Mr Blankfein at Goldman acknowledges that some of the young traders at the firm have no memory of a business cycle. Clearly he does, but he says he respects their views and has not been disappointed so far.
Above those traders are wiser heads who should be prepared to extend a restraining hand. Between them, the 21 members of Goldman's management committee have 300 years of experience within the firm. Yet even that massed experience cannot guarantee that they will always be able to keep trouble at bay.