Derivados

The risk of value at risk
Nicholas Pratt. Financial Times. London (UK): Oct 18, 2010. pg. 7
Resumen
"Var is still a credible measure but it must be properly used," says Laurent Pasquier, head of investment risk at Axa Investment Managers. "It cannot be used in every circumstance or if it is not well understood. In particular Var is of little use in cases of extreme market behaviour because ofthe lack of historical data and the limitations in backtesting. Therefore the higher the level of volatility, the lower the level of confidence in the Var figure."
La variaciòn es aùn un medida de credibilidad, pero debe ser usada propiamente, dice Laurent Pasquier, cabeza de riesgo de inversión de Axa Investment Managers “

The models are subjected to both basic stresses, such as singlefactor interest rate shifts, and more complex ones - such as bear steepening stresses for fixed income portfolios. "The models are periodically reviewed and whereas the basic models will stay the same, the more complex models will be changed according to market conditions," says Mr [Simon Bray-Stacey]. Aviva also uses multiple risk models. "This has always been the case with certain portfolios but isnow in place across the board."
It is a similar story in the hedge fund world. "If anything the focus has been to place less reliance on the models," says Matthew Weir, chief risk officer at BlueCrest Capital Management. "All models include flawed assumptions so we run fairly basic models and ensure that we understand the limitations. We then use stress tests to assess where the tail risk liesand try to make sure these are extreme yet plausible."
» Saltar al índice (detalles del documento)
Texto completo
(831 palabras)
(Copyright Financial Times Ltd. 2010. All rights reserved.)
Assessment models ; The widely used snapshot picture of overall exposure must be applied correctly, writes Nicholas Pratt
The risk management role was one of the first to be placed in the spotlightfollowing the financial crisis.
It was clear that the various risk models had failed to foresee the market's crash but less clear how this lack of foresight could be corrected. One theory focused on value at risk, the widely used measure for calculating the risk threshold of a portfolio.
Value at risk is designed to work within a stable environment and is best suited to instruments such asequities that display daily changes in risk. But in the pre-crisis period of prolonged stability, Var was often applied to other less suitable or liquid instruments, such as credit, and began to be treated as definitive, forming the basis of portfolio risk management for many firms. Since the crisis the reliance on Var, rather than Var itself, has been called into question.
"Var is still a crediblemeasure but it must be properly used," says Laurent Pasquier, head of investment risk at Axa Investment Managers. "It cannot be used in every circumstance or if it is not well understood. In particular Var is of little use in cases of extreme market behaviour because of the lack of historical data and the limitations in backtesting. Therefore the higher the level of volatility, the lower the levelof confidence in the Var figure."
Consequently Var models are being subjected to far greater scrutiny through stress-testing and shock scenarios rather than simple historical analysis.
Simon Bray-Stacey, portfolio risk manager at Aviva Investors says: "The crisis has shown us that history tends not to repeat itself so having a series of well-defined single and multi-factor portfolio stressesgives us a lot of insight into the sensitivities of the portfolio at an individual stress level and gives us some insight into the further tails of the distribution."
The models are subjected to both basic stresses, such as single factor interest rate shifts, and more complex ones - such as bear steepening stresses for fixed income portfolios. "The models are periodically reviewed and whereas...