The Origin of Risk Bombs Published October 2008 by Ralph McKay, author of Risk Mechanics, financial derivatives, finding and fixing risk holes. The collapse of the US banking system, engulfing the world, is the explosion of a massive risk bomb. A risk bomb of such epic proportions is not created overnight, it forms over a period of many years. How did this massive risk exposure sneak up and surprise governments, central bankers, regulators and policymakers? How did so many borrowers, lenders, rating agencies, academics and almost every analyst in the system get it so wrong? The only possible explanation is that the risk stakeholders and those interested in the risks believed the risks were known and not extreme. Nothing else can explain their silence before the explosion. The risks were not seen because they were hidden by a flawed industry risk standard called VaR (Value at Risk). VaR simplifies risks to a naive near meaningless expression. A misleading risk measure is worse than no risk measure. The false sense of security created by a flawed risk measurement process allowed the risk bomb to balloon out of sight. This is the real root cause of the crisis. VaR failed absolutely yet it's still the industry standard for risk measurement. A sensible, robust risk measurement technology ensures risks are transparent. Risk transparency ensures that high risk derivative instruments cannot be packaged as safe investments. The key point here is transparency in the risks as measured by price sensitivity to economic parameters. This is not the same as a simple disclosure of the face value of the instruments. The whistle blowers did their work many years before the event, but were ignored. A report published in Euromoney's Corporate Finance in 1996, exposed devastating flaws in VaR. It was a response to Wall Street's aggressive marketing of VaR in the 1990s establishing VaR as the industry standard of market risk measurement. The report, entitled, VaR is a dangerous technique, warned that VaR "hides large risk holes ... managers consequently oblivious to the actual risk can increase systemic risk ... VaR assumes well-behaved markets but ignores the impact of real worry, the financial markets' Tunguska events". I was the report's principal author. The term "Risk Tunguska" was coined after a mysterious and massive aerial explosion 8km above Eastern Siberia on June 30, 1908. It devastated 500,000 acres. My intention was to draw attention to the magnitude of the long term threat from VaR, now exploded closer to home in 2008. I published many other short papers in the 90s on the same theme including the book Risk Mechanics, financial derivatives, finding and fixing risk holes. Whatever flawed risk measurement process was behind the sub-prime debacle, it was in nature, a derivative of VaR. Senior managers, investors and regulators alike felt relaxed and comfortable with a risk measurement process with a name like "Value at Risk" unaware that its name betrayed its function. VaR hides the many complexities of market risk. It attempts to represent risk as a single statistic. It relies heavily on past statistics as a guide to the future. It uses negative correlations to minimize the apparent risk when in fact correlations frequently increase during market shocks. VaR throws away essential data required for reshaping of risk. Its most serious flaw is that it ignores elements of risk associated with catastrophic loss. VaR measures the expected, not the unexpected. It failed because risk management is about managing the unexpected. Imagine a main street fruit and veg trader attempting to value a concealed bag of fruit and veg by its weight only. He has no idea if the bag contains potatoes, a mixed bag or bad apples. The single statistic of weight alone does not enable the fruit and veg trader to price the bag or understand the risk of holding it. A mortgage derivatives risk manager relying on VaR is no more informed than the fruit and veg trader attempting to value his investment on weight alone. Less is not always more. How should the risks have been measured? First the various economic variables that impact the derivatives value are isolated. In the case of mortgage backed securities this includes uniform shifts, tilts and flexes in interest rate term structure, housing prices related to inflation and various other economic data which may affect the ability of borrowers to service mortgages. The derivatives portfolio is then modeled and stress tested to various extreme moves in the variables. The stress testing continues virtually in real time throughout the life of the instruments. It's a straightforward engineering approach. Guided by accurate risk measurement the risks are then either avoided or reshaped and distributed with the use of other financial instruments. A long term solution to this crisis must also answer why so many investment banker "rocket scientists" and academics embraced a risk measurement method as naive as VaR. Why were voices of reason ignored? A failure in the market of opinions preceded the failure in financial markets. Whatever the motive or reason for so many clever people getting it so wrong, the rise of VaR made it easy to sell financial instruments with catastrophic elements in their payoffs, risk holes. Promoters of these instruments are rewarded on a time scale much shorter than the expected time scale of these catastrophes. "Rocket scientists" with the ability to solve equations of risk and create risk algorithms often lack a good feel for the nature of risk. Risk management demands a strong physical intuition and vision for the unexpected. The global warming crisis is extremely difficult to manage and in dispute because the precise mathematical relationship of warming to the relevant variables is unknown. However, the price sensitivity of financial instruments to the relevant economic variables is well known and not in dispute. No parties involved in the creation, distribution or regulation of these instruments have any excuse for misrepresenting the risks. The crisis was totally avoidable. The principle of securitization is sound. Derivatives are wonderful tools for spreading risk and allowing the economy to grow. The proper use of these instruments is just a matter of sound engineering. There is no place for economic forecasting in risk management. Back