One of the most pressing matters facing the world of finance today is finding a way to measure and hedge risk. Previously, numbers and statistics was used to attempt to quantify risk into a single number or variable. This method has been exposed as inadequate. One of the culprits behind this inadequacy is Var.
The risk measure that most banks used, is referred to as Value at Risk (Var). Var has long been an accepted measure of risk in the marketplace. Basically, Var measures the value that is at risk should certain events transpire. The process involves running statistical tests to determine what value could be lost in the course of business. Unfortunately, any prediction is based on historical values which have no explicit influence on future events. As a result, any predictive qualities of Var may be insignificant.
Perhaps more importantly, it also fails to recognize extraordinary events and can therefore drastically underestimate high risk low probability losses. As an example, consider a car with airbags. If the airbags work 99% of the time, except during the event of a wreck, then they should be considered useless. No one would consider giving that car a five star safety rating; such is the state of Var as an accurate measure of real risk.
Low Var numbers bred complacency and contributed to the excessive risk taking by the market as a whole. Risk is now more convoluted than ever with the introduction of credit default swaps, derivatives and other advanced financial securities. Pablo Triana, a derivatives consultant says that, “Var provided a convenient Nobel-backed theory-sanctioned alibi for the types of super complex super-leveraged punts that ultimately took down Wall Street.” It has become apparent, with the recent events in financial markets the Var is no longer a viable measure of risk.
So, it is easy to assume that a measure is useless after the catastrophe we’ve experienced over the last eighteen to twenty four months. Everyone is suddenly an expert. Adding true value will involve putting the pieces back together. In terms of risk, what will replace the measure? The problem with Var will plague any other attempt to measure comprehensive risk with one number. Quants can no longer run Wall Street like a high school science experiment. Unlike fields such as physics, you cannot isolate variables in the real world. That is the main reason predicting values in economics is such a struggle.
The real problem lies in trying to quantify a science that relies on human behavior. Any attempt to do so will be met with failure because humans change, adapt and act irrationally often in no discernible manner. It is difficult to really assess risk based on one all encompassing number. That is why I believe the true revolution in finance will deal more with changes in incentive systems rather than a new way to measure risk.
Quote from: Pablo Triana, “A real culprit behind this almighty crash,” Financial Times, 24 Nov. 2008: 13.
