Watch the full version of this recorded webinar by clicking on the play button above.
To view the slides, register here to become a Quants Hub member. This will also give you access to a host of previously recorded webinars in our exclusive Members Area.
Presenter: Alexander Sokol: Head of Quant Research, CompatibL
Estimation of Real World Drift in Interest Rates and Credit Spread using Local Price of Risk
*Includes material from a recent paper by Hull, Sokol, and White
http://ssrn.com/abstract=2403067 and Risk Magazine, October, 2014
- The estimation of real world drift in interest rates and credit spread is usually based on ad-hoc techniques or expert opinion
- The popular ad-hoc techniques include assuming zero real world drift or substituting the risk neutral drift for the real world drift
- Both ad-hoc techniques can cause large backtest errors which we will demonstrate on several realistic portfolios
- We will then review a simple, practical, and non-subjective method of estimating the real world drift from the historical data
- The method is based on a new concept of “local price of risk”
- The relationship between local price of risk and market price of risk is similar to the relationship between local volatility and market implied volatility
- The new method is shown to significantly improve backtest accuracy in models for limits, liquidity, and regulatory capital