Calibrating Real-World Drift for Potential Future Exposure

Summary

The document asks how to estimate real-world drift parameters when generating scenarios for potential future exposure (PFE). It distinguishes scenario generation under the real-world probability measure from revaluation under the risk-neutral measure, then raises the practical question of when historical averages are a reasonable drift estimate for a risk factor.

It also asks how to handle factors whose drift is difficult to estimate from observed history and requests references. No answer, calibration procedure, empirical evidence, or comparison of alternatives is included, so the document is best read as an open modeling question rather than guidance. Its main lesson is the need to consider drift estimation separately by risk-factor class and to recognize that historical data may be insufficient for some factors.

Key ideas

• PFE scenario generation uses real-world dynamics, while valuation of the resulting exposures uses risk-neutral pricing.
• The document questions when a historical average is an adequate estimate of drift.
• It highlights that drift may be difficult to estimate reliably for some risk factors.
• No recommended method or supporting evidence is provided.

Tags

# Calibration of real-world drift for PFE


# Calibration of real-world drift for PFE












It is well-known that for the calculation of potential future exposure (PFE), real-word probability measure should be used to scenario-generation and risk-neutral measure is used for revaluation; see e.g. J.Hull "Financial Risk Management and Financial Institutions"-chapter 7. Now, the point is how we should go about calibrating real-world parameters, e.g. drift. I know that this task is not obvious. Can anyone explain to me the following points: a) For which classes of risk factor, it is OK to just calibrate drift to historical data, e.g. using simple average? b) For risk factors that are hard to obtain estimates of drift from historical data, what is the best practice to tackle this? Any good reference would be highly appreciated.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.