Common sense brews a safer culture
Thursday, 6 December 2007
David Wighton JOHN Thain says one of the first things he has to do as the new chairman and chief executive of Merrill Lynch is revamp the Wall Street bank's risk management system. Over at Citigroup, they have launched a review of risk management processes and UBS has committed itself to improving its risk control procedures. Quite right too, you might say, given that these banks have warned they face combined writedowns of more than $20bn on their US subprime mortgage-related holdings.
One of the many qualifications Mr Thain has for the Merrill job is that he was once chief financial officer of Goldman Sachs and so in charge of a risk management operation that is the envy of Wall Street.
Goldman's reputation has been further enhanced by its extraordinary performance - so far, at least - during the credit market turmoil. Mr Thain would seem to be just the sort of person you need to oversee the revamping of a risk management system.
Yet, while it is obvious there has been a massive failure of risk management across most of Wall Street, it is less clear that there was always a failure of such processes. In many cases, senior executives claim that the procedures actually worked very well. What went wrong were the assumptions behind the procedures. Most people simply did not conceive of the possibility of a significant nationwide fall in US home prices. And then they did not conceive that this could cause the evaporation of liquidity in mortgage-backed securities. The risk machine worked fine - it was the raw material that was wrong. "There was a poverty of imagination," says Jim Wiener, a risk specialist at consultants Oliver Wyman.
Consider the case of Merrill Lynch. It had a large and very profitable business packaging mortgage-backed securities into collateralised debt obligations. During the summer, demand for the super senior tranches of CDOs dried up. So Merrill just held on to them. It may be that risk management best practices were not observed. But it may just have been that those super senior tranches were viewed as being as safe as houses, as it were.
At Morgan Stanley, the bank's proprietary traders rightly decided a year ago that subprime was going south and took out a huge bet that subprime-related securities would fall. They paid for the cost of that bet by acquiring super senior tranches of CDOs. These offered a nice yield, comfortably above the traders' cost of funds.
Unfortunately they did not conceive that the subprime meltdown could get so bad it would eat into the value of the super seniors. They lost a cool $3.7bn in two months.
Part of the problem may have been that banks had too much confidence in their fancy risk management procedures and failed to apply common sense.
David Allen, a risk management consultant at Accenture, says there was too much reliance on risk models based on historical information. When a market event happened that was not in the historical data, the models fell apart. You also have to use "human judgement", he says, and apply old rules such as "when things are looking too good to be true, perhaps they are".
But beyond the reminder to use common sense, there are a few obvious lessons to be learned. The first is to limit exposure to one underlying risk factor and ensure that you have identified everything that could be affected by that factor. Banks would probably not have allowed themselves to be exposed to one foreign currency to the extent they were exposed to US subprime mortgages.
Executives admit that they are still not good enough at looking across "silos" in their businesses and Gary Crittenden, chief financial officer of Citigroup, has said that it needs to get credit and market risk teams working more closely together.
The ultimate value of super senior CDO tranches may turn out to be higher than values the banks have put in their books after a brutal haircut. But that will be too late to save the jobs of Chuck Prince at Citigroup or Stan O'Neal at Merrill Lynch. Indeed, it is hard to see how any normal risk management procedures can be applied with securities that are so hard to value at times of no liquidity. When Merrill revised up its estimate of third-quarter mortgage-related writedowns to $7.9bn, Mr O'Neal said the figure was within the range of assumptions it had used less than three weeks earlier when its estimate was $4.5bn.
That suggests that, despite all its sophisticated risk systems, Merrill simply did not have a clue what its real risk position was. Another factor that is very difficult to measure is reputational risk. Many of the banks had set up special purpose vehicles that they had no obligation to support and of which they provided little or no disclosure.
But faced with the possibility that these vehicles would go under, some banks have decided the damage to their reputation might be worse than the damage to their balance sheets.
A final lesson is perhaps the importance of a risk-focused corporate culture, led from the top. It is surely no coincidence that Lloyd Blankfein, chairman and chief executive of Goldman Sachs, bangs on about risk the whole time.
..........................................
— FT Syndication Service
One of the many qualifications Mr Thain has for the Merrill job is that he was once chief financial officer of Goldman Sachs and so in charge of a risk management operation that is the envy of Wall Street.
Goldman's reputation has been further enhanced by its extraordinary performance - so far, at least - during the credit market turmoil. Mr Thain would seem to be just the sort of person you need to oversee the revamping of a risk management system.
Yet, while it is obvious there has been a massive failure of risk management across most of Wall Street, it is less clear that there was always a failure of such processes. In many cases, senior executives claim that the procedures actually worked very well. What went wrong were the assumptions behind the procedures. Most people simply did not conceive of the possibility of a significant nationwide fall in US home prices. And then they did not conceive that this could cause the evaporation of liquidity in mortgage-backed securities. The risk machine worked fine - it was the raw material that was wrong. "There was a poverty of imagination," says Jim Wiener, a risk specialist at consultants Oliver Wyman.
Consider the case of Merrill Lynch. It had a large and very profitable business packaging mortgage-backed securities into collateralised debt obligations. During the summer, demand for the super senior tranches of CDOs dried up. So Merrill just held on to them. It may be that risk management best practices were not observed. But it may just have been that those super senior tranches were viewed as being as safe as houses, as it were.
At Morgan Stanley, the bank's proprietary traders rightly decided a year ago that subprime was going south and took out a huge bet that subprime-related securities would fall. They paid for the cost of that bet by acquiring super senior tranches of CDOs. These offered a nice yield, comfortably above the traders' cost of funds.
Unfortunately they did not conceive that the subprime meltdown could get so bad it would eat into the value of the super seniors. They lost a cool $3.7bn in two months.
Part of the problem may have been that banks had too much confidence in their fancy risk management procedures and failed to apply common sense.
David Allen, a risk management consultant at Accenture, says there was too much reliance on risk models based on historical information. When a market event happened that was not in the historical data, the models fell apart. You also have to use "human judgement", he says, and apply old rules such as "when things are looking too good to be true, perhaps they are".
But beyond the reminder to use common sense, there are a few obvious lessons to be learned. The first is to limit exposure to one underlying risk factor and ensure that you have identified everything that could be affected by that factor. Banks would probably not have allowed themselves to be exposed to one foreign currency to the extent they were exposed to US subprime mortgages.
Executives admit that they are still not good enough at looking across "silos" in their businesses and Gary Crittenden, chief financial officer of Citigroup, has said that it needs to get credit and market risk teams working more closely together.
The ultimate value of super senior CDO tranches may turn out to be higher than values the banks have put in their books after a brutal haircut. But that will be too late to save the jobs of Chuck Prince at Citigroup or Stan O'Neal at Merrill Lynch. Indeed, it is hard to see how any normal risk management procedures can be applied with securities that are so hard to value at times of no liquidity. When Merrill revised up its estimate of third-quarter mortgage-related writedowns to $7.9bn, Mr O'Neal said the figure was within the range of assumptions it had used less than three weeks earlier when its estimate was $4.5bn.
That suggests that, despite all its sophisticated risk systems, Merrill simply did not have a clue what its real risk position was. Another factor that is very difficult to measure is reputational risk. Many of the banks had set up special purpose vehicles that they had no obligation to support and of which they provided little or no disclosure.
But faced with the possibility that these vehicles would go under, some banks have decided the damage to their reputation might be worse than the damage to their balance sheets.
A final lesson is perhaps the importance of a risk-focused corporate culture, led from the top. It is surely no coincidence that Lloyd Blankfein, chairman and chief executive of Goldman Sachs, bangs on about risk the whole time.
..........................................
— FT Syndication Service