|
MODEL VALIDATION AND BACKTESTING
Experts in finance and mathematics often develop sophisticated mathematical
models to value and hedge transactions involving many millions of
dollars. Because of the inherent complexity of these processes,
managers at many firms simply assume that these "black box"
models are realistic and accurate. Unfortunately, volatile and illiquid
market conditions often reveal that many untested models have fundamental
flaws that can produce large losses.
The risk specialists at Black Swan have been on the forefront of
research on developing both quantitative and qualitative techniques
to assess the predictive accuracy of pricing and risk measurement
models. These techniques can help ensure that the models used to
make decisions affecting shareholder value are in fact producing
accurate results.
Black Swan Risk Advisors risk professionals have extensive expertise
and advanced tools to identify and measure model risk. Based on
our analysis, we provide guidelines for managers and risk practitioners
in order understand model risk and develop the know-how and processes
to minimize it and prevent it.
Click here to view a sample of articles on Model Validation and Backtesting authored by BSRA principals
Sample projects. Backtesting of market risk models using advanced methods. Communication with regulators regarding model risk. Mitigation of model risk. Policies and procedures around use of derivatives pricing and risk measurement models. Technical Risk Audits.
| Case Study - Model
Error
In 1997 many derivatives traders noted that NatWest Markets,
part of one of the largest banks in the UK, was aggressively
pricing interest rate options and swaptions. It is believed
that NatWest's pricing models incorrectly assumed that a key
valuation parameter - implied volatilities - were the same
for all options on a particular asset.
In reality, market professionals have long known that implied
volatilities are almost always higher for out-of-the money
options. The failure of the bank's pricing and risk management
models to incorporate the "volatility smile" effect
led to a significant over-valuation of the portfolio.
Losses from these trades eventually totaled £90.5 million.
The bank was taken over by the Royal Bank of Scotland in 2000.
|
For more information, please contact us at services@blackswanrisk.com.
|
