Asset and Liability Management: The Banker’s Guide to Value Creation and Risk Control, Second Edition

Earnings-at-risk in ‘normal’ times

In this measure of risk, you report the impact of a change of interest rate likely to cover most of the potential cases (for instance, 95% of the cases^[4]). Historical data are usually used to analyse the ‘normal’ volatility in a specific country. A standard measure of volatility in statistics is the standard deviation.

^[4] The confidence level of 95% is arbitrary. Some banks work with 97.5% while others use 99%.

EAR_95% = |gap| × (Δ_95% rate)

If this ‘95% confidence’ information is useful, management should also be informed of the potential losses in case of rare, but still possible, big shocks. With reference to those who have experienced, and survived, periods of very high volatility, they have called this ...