Terrific l o n g article in the Sunday Times Magazine by Joe Nocera, titled Risk Mismanagement. Its all about how Wall Street developed and still uses VaR — Value at Risk.

The application of VaR remains hotly debated today. Did it contribute to the credit crisis — or was it ignored/misapplied/distorted, and THATS what was a key factor.


Risk managers use VaR to quantify their firm’s risk positions to their board. In the late 1990s, as the use of derivatives was exploding, the Securities and Exchange Commission ruled that firms had to include a quantitative disclosure of market risks in their financial statements for the convenience of investors, and VaR became the main tool for doing so. Around the same time, an important international rule-making body, the Basel Committee on Banking Supervision, went even further to validate VaR by saying that firms and banks could rely on their own internal VaR calculations to set their capital requirements. So long as their VaR was reasonably low, the amount of money they had to set aside to cover risks that might go bad could also be low.

Given the calamity that has since occurred, there has been a great deal of talk, even in quant circles, that this widespread institutional reliance on VaR was a terrible mistake. At the very least, the risks that VaR measured did not include the biggest risk of all: the possibility of a financial meltdown. “Risk modeling didn’t help as much as it should have,” says Aaron Brown, a former risk manager at Morgan Stanley who now works at AQR, a big quant-oriented hedge fund. A risk consultant named Marc Groz says, “VaR is a very limited tool.” David Einhorn, who founded Greenlight Capital, a prominent hedge fund, wrote not long ago that VaR was “relatively useless as a risk-management tool and potentially catastrophic when its use creates a false sense of security among senior managers and watchdogs. This is like an air bag that works all the time, except when you have a car accident.” Nassim Nicholas Taleb, the best-selling author of “The Black Swan,” has crusaded against VaR for more than a decade. He calls it, flatly, “a fraud.” . . .

What will cause you to lose billions instead of millions? Something rare, something you’ve never considered a possibility. Taleb calls these events “fat tails” or “black swans,” and he is convinced that they take place far more frequently than most human beings are willing to contemplate. Groz has his own way of illustrating the problem: he showed me a slide he made of a curve with the letters “T.B.D.” at the extreme ends of the curve. I thought the letters stood for “To Be Determined,” but that wasn’t what Groz meant. “T.B.D. stands for ‘There Be Dragons,’ ” he told me.

Best line in the article: “When Wall Street stopped looking for dragons, nothing was going to save it.”

I particularly loved the graphics and illustrations that were part of it:

Cover Image


Risk Mismanagement
Joe Nocera
NYT Magazine, January 4, 2009


Category: Credit, Data Analysis, Derivatives, Markets, Mathematics, Quantitative, Really, really bad calls

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

28 Responses to “Risk Mismanagement & VaR”

  1. donna says:

    “All the triple-A-rated mortgage-backed securities churned out by Wall Street firms and that turned out to be little more than junk? VaR didn’t see the risk because it generally relied on a two-year data history.”

    How can you rely on a two year history of mortgage data? This is the kind of stupidity that just kills me…. anyone familiar with the housing market knows it is a cyclical business.

  2. sixx says:

    Most of the problems with VaR as risk measure come from the fact that it assumes normal distribution of returns which turns out not to be the case since extreme events appear more often in reality than the Gaussian distribution would predict.

    Yet current advances in finance start looking more often at fat-tailed VaR which accounts for non-normality, extreme events and asymmetry of returns. For more info on the subject you can check here:


  3. gorobei says:


    the problem is not so much the use of a short history, but more that the mathematical framework is just plain not valid. Most, if not all, of these timeseries have the paradoxical property of making you less, rather than more, confident as you examine more datapoints. Examine enough datapoints, and you become convinced (in a mathematical sense,) that the models have zero predictive power: the data doesn’t refine the model, it undermines it.

  4. gorobei says:


    I guess that stuff is better than nothing if you are trying to sell product, but it’s just another attempt to stuff the data back into a normal distribution framework. If you have no means in your distributions, it doesn’t do a lick of god in the long run.

  5. Andy Tabbo says:

    At commodity trading desks we always used to laugh at these stock and bond traders….we would see Bob Pisani on the floor talking about “how much volatility we have here today….” Up until 2008, it was always amusing to us. You want to see volatility? Go sit on a Natgas trading desk for a few months…or go trade Product cracks….the volatility in some of those markets is insane. And I’m not even talking about the “flat price.” The inter-month time spreads are sometimes way more volatile than “flat price” alone.

    We used to talk regularly about fat tail phenomenon in commodities….basically every commodity “spread” was a 3++ std deviation event waiting to happen. For a recent example of the pitfalls of VaR, see the Jan/Feb WTI spread that expired on 12/19. Throughout most of the history of that particular spread Feb was trading in a steady range $.50-$1.00 premium…On 12/10/08 that particular spread settled at $.99 with Feb over Jan. On 12/19, that spread settled at a mind blowing $8.50….so that was a nice little 850% return in SEVEN trading days (35,000% return annualized). There is NO Var Model that would have been close to evaluating the risk of that trade….not even remotely close. Those sorts of events transpire regularly in many markets.

    I’m actually growing tired of having debates with these Academic types about “efficient market theory”. The evidence to the contrary is so overwhelming….

    - AT

  6. gorobei says:


    Yep, we used to laugh that any commod desk worth its salt would blow itself up every few years. We can all be happy it took the magic of modern derivatives to let all the other desks lever up to the same risk profile (well, except the muni guys – that little tax thing kinda puts a brake on serious synthetics.)

  7. Moss says:

    Having the model is one thing but:

    ‘saying that firms and banks could rely on their own internal VaR calculations to set their capital requirements. So long as their VaR was reasonably low, the amount of money they had to set aside to cover risks that might go bad could also be low.


  8. Jnavin says:

    “VaR didn’t see the risk because it generally relied on a two-year data history”

    This sounds similar to the Long Term Capital Management mistake: they only took the last few years or so of bond history, as I recall.

    I’m guessing that the model works perfectly if you use the 2-year data, but the model doesn’t work at all if you use the 30-year…therefore, use the 2-year.

  9. danm says:

    Yup. VaR would have told you to buy Nortel at 120$.

  10. danm says:

    These models use historical data when theory clearly stipulates expected returns, expected risk…

  11. AGG says:

    The irony of our present situation as Americans is that , with all the wailing and nashing of teeth about risk, the biggest risk to our future as a healthy country and world leader in human decency is being studiously ignored by the media:

    We must not confuse Israel’s Zionist government with world Jewry, just as we must not confuse the American people with the war criminals in the Bush Regime.

    Consider, who do you trust with your civil liberties, the US Department of Justice or the ACLU’s phalanx of Jewish attorneys?

    We must avoid the mistake that was made by blaming the German people for Hitler. It was the aristocratic German military that tried to remove Hitler. In contrast, Democratic Speaker of the House Nancy Pelosi blocked the attempt to impeach George W. Bush and Dick Cheney. Pelosi is a discredit to California, but shall we blame all of America for Pelosi’s defense of war criminals? How can we do so when US Rep. Dennis Kucinich courageously read out the articles of impeachment on the House floor?

    Are all Americans guilty because Kucinich did not prevail?

    Paul Craig Roberts was Assistant Secretary of the Treasury in the Reagan administration. He is coauthor of The Tyranny of Good Intentions.He can be reached at: PaulCraigRoberts@yahoo.com

  12. danm says:

    The irony of our present situation as Americans is that , with all the wailing and nashing of teeth about risk, the biggest risk to our future as a healthy country and world leader in human decency is being studiously ignored by the media:

    Honestly, you can say the same thing about the people in every country. Even Iraq! Why does it always have to be all about the US of A?

    If Americans were so decent it wouldn’t always be all about them. The downfall of the US will be that the world is just not big enough for them.

  13. AGG says:

    In a culture that celebrates mendacity and roguish behavior, the biggest liers avoid punishment while the ones who “crack” (tell the truth when exposed) are persona non grata.
    Donna, they are lying about which “data set” they consulted. It’s all a piece like saying that our favorite scape goat, MR. NO ONE, is the “one” who foresaw the problems.
    If the assets of all the liars and crooks at the top of this pyramid aren’t seized by the government, sold at PUBLIC auction and deposited in Social Security or Medicare, people will simply stop cooperating with the system. The first ones to do so will be the police. Rich crooks had better run a cost benefit analysis on how much peace of mind and body will cost them if they don’t stop bullshitting the public or else no hideout in Montana or Idaho will save them.

  14. ButtoMcFarty says:


    SEC Said to Examine More Ponzi Schemes After Madoff (Update1)
    Email | Print | A A A

    By David Scheer

    Jan. 2 (Bloomberg) — U.S. regulators working to untangle Bernard Madoff’s alleged $50 billion Ponzi scheme are probing other money managers suspected of using similar tactics, two people with knowledge of the inquiries said.

    The U.S. Securities and Exchange Commission is pursuing at least one case in which investors may have been cheated out of as much as $1 billion, according to a person, who declined to name the manager and asked not to be identified because the probe isn’t public.

    Regulators may discover additional Ponzi arrangements as declining stock markets prompt investors to withdraw their cash and they question how their money is being managed. This week, the SEC said it halted what the agency described as a $23 million scam targeting Haitian-Americans, and said the Florida- based operators as recently as last month sought more investors.

    The new cases ,“signal it’s become an enforcement priority,” said University of Rochester President Joel Seligman, who wrote a history book on the SEC. After Madoff’s alleged fraud “you’ve got to check hard and see if there is more like it in the marketplace.”

  15. AGG says:

    You are right. However, I’m an American and have lived all my life in the USA and one must clean up one’s own house before judging others.
    All the money we spend on “defense” could have been spent on space travel, mining ore on the asteroids and colonizing other planets with just as many jobs created but it was more fun to kill and lock people into a culture of killing (hey, a man needs a job) along with women proud to produce lots of cannon fodder. I spent time at the USMA so I inderstand the “feel good” part of this culture that just goes on and on. It will end badly. It always does.

  16. ButtoMcFarty says:

    Now it’s a priority…

  17. Darkness says:

    “U.S. regulators working to untangle Bernard Madoff’s alleged $50 billion Ponzi scheme are probing other money managers suspected of using similar tactics, two people with knowledge of the inquiries said.”

    Whoa, we’re going to get a few weeks of work out of the sit-on-their-hands ideologues that fill the Bush executive branch? On the upside, forcing them to actually work may make them far less likely to try to hold onto their “career” job when someone qualified is found to replace them after the 20th.

  18. Bill Werner says:

    Any time you are dealing with probability distributions you are gambling. But in finance it is even worse because not only the operating environment for the product is changing but also the rules (government intervention) and products themselves are always changing. Therefore there will never be enough data on the actual circumstances. But even if there was enough data a manager somewhere will get greedy and order the “outliers” (unexpected data) be thrown out of the model. In finance it is as if sometimes the jokers are left in the deck and sometimes not. And when the jokers are in the deck sometimes they are wild cards and some times they are dead cards and many times the player does not know until he plays the hand. Even for financial markets with more or less stable rules and products the fact that they do not have a normal distribution but have fatter tails (higher kurtosis) than a normal distribution increases the risks (and rewards) by definition.

    For the sake of simplicity let us assume a best case scenario where we have a market with stable rules and products and a nice neat normal distribution. Even under those ideal (and unrealistic) circumstances not having a risk of ruin component separate from the normal operating model is pure negligence. The reason finance does not take risk seriously is because they always assume a “regression to the mean” will takes place (things will return to normal) and save them, not to mention a deep faith in government bailouts, golden parachutes, master of the universe bonuses etc. In industry the risk of ruin component in a plant is called an emergency shutdown system (ESD, aka Safety Instrumented System (SIS)) and is always separate from the process control system controlling production. The ESD is typically double or triple redundant and much thought goes into making these systems idiot (and manager) proof. If industry behaved like finance Three Mile Island nuclear accidents (1979), Bhopal toxic releases (1984) and Texas City refinery explosions (2005) would be common place.

    And if all gamblers behaved like Wall Street firms there would be no professional gamblers. The same could also be said for independent traders. Or as Market Wizard Ed Seykota once put it “There are bold traders and old traders, but no bold and old traders.” So honor your stops and Laissez les Bons Temps Rouler (Let the Good Times Roll)!

  19. KJ Foehr says:

    @Bill Werner

    Wiser words I have rarely read. Thank you for that.

  20. gorobei says:


    Bill Werner is an optimist.

  21. Bruce in Tn says:

    Risk mismanagement on a governmental level:


    When all you have is an elephant gun, everything looks like a charging elephant…

    good night.

  22. Mike in Nola says:

    Thanks. Very cogent article. When I saw how many pages it was, I almost gave up, thinking here’s someone failing at trying to explain something very complicated. But, he did a good job. Taleb sounds just like I pictured him after reading his books.

    The thing about the quants is that people often confuse hard work and knowledge of a narrow discipline with intelligence. They are not the same. I remember some of the Ph.D’s in the math dept. where I studied were really not very bright, but they did have a Ph.D. Quants may be able to do all those calculations, but they may not be able to tell you when they are worth anything.

    My two cents:

    VaR is like many other things where people try to apply complicated mathematics to situations for which they aren’t meant. It looks impressive to those who don’t know any better and gives them a feeling of certainty which is something many like.

    I remember after getting my M.S., I interviewed with a small company that did predictions of future energy needs for whatever Entergy used to be called 30 years ago. They did it by fitting curves to past numbers and thinking that the curve was predictive. Of course, the curve only told you what had happened and not what would happen. Real BS

    I tend to agree with those who said that VaR was good for showing you when things were getting abnormal. For that, it’s probably a useful tool.

  23. David Merkel says:

    We need to move from finance based models of risk to actuarial models of risk. The pity is that the actuarial societies have been silent because of a misbegotten inferiority complex because of the seeming success of Wall Street models.

    I’ll write about this at my blog soon. The main idea is to have risk models that work even when markets freeze up. If you can’t live with the markets being closed, or even wide bid-ask spreads, your risk model does not work.

  24. danm says:

    The thing about the quants is that people often confuse hard work and knowledge of a narrow discipline with intelligence
    So true. When I was doing my math degree, there was a high functioning autistic math genius in the program who only got 99s and 100s. I guess the best comparison I could make is that he was like Raymond in Rainman, somebody had to direct him and tell him where to use his intelligence.

    Our society admires genius and barely notices a well-rounded person with above average intelligence in all areas… Diversification is not only key in portfolio management but in most areas of life, yet is rarely valued.

    It’s no wonder risk is mismanaged, we don’t value the right types of people and intelligence for our own good.

  25. danm says:

    We need to move from finance based models of risk to actuarial models of risk
    Actuarial models are entirely based on past experience and based on the same math principles. As a portfolio manager, we used those tools/models because the consultants (actuaries) NEEDED to see numbers!

    No model will ever catch the black swan event.

    There is one basic principle to keep a system stable…

    1. Consistency.

    Your breakfast is not 4 toasts when you are thin and 2 toasts when you are trying to lose weight. The key to maintaining your weight is to get the number of toasts right sticking to it.

    Other examples:
    You don’t lower rates by 6% in in year
    You don’t go from 20% down to 0% down
    You don’t use historical data when the types of mortgages have changed

    The list goes on.

    Obviously consistency is not attractive nor exciting.

  26. Moss says:

    Also prudence

    What exactly was the black swan event???
    No single event represents this crisis.
    A number of factors led up to it.

    In the end it is human behavior.
    What motivated the behavior? Ego, hubris, greed, Ideology, group think.

  27. JG says:

    The limitations of VaR were widely known and considered. The main issue on Wall Street was the lack of investment in alternative analytical tools and systems to measure risk, and the discounting of opinions of those who predicted an end to the party.

  28. hjortsbergster says:

    If you liked this article you must read:


    Which discusses the article above.