Hedge Effectiveness Testing Revisited
Publication: The Journal of Derivatives
This article provides two recommendations for hedging practitioners attempting to qualify for special hedge accounting treatment. First, we propose an alternative measure to the traditional dollar offset ratio, in which we prescribe division by the starting value of the hedged item rather than division by the change in the value of the hedged item. Our proposed measure provides similar information to the dollar offset ratio, but it is less prone to the potential problem of division by zero in times of low volatility. As a result, it is less likely to exceed acceptable boundary conditions during periods of calm markets (i.e., when small price changes occur). Second, we propose an alternative to the R-squared of a regression analysis as a measure of anticipated hedge effectiveness, which documents the proportion of the total risk that is actually mitigated given the hedge ratio chosen by the hedger, as opposed to the proportion of total risk that would be mitigated if the hedger used the regression slope coefficient as the hedge ratio.
This article addresses two problems often encountered by practitioners who endeavor to qualify for special hedge accounting treatment.
The first problem has to do with the traditional dollar offset ratio (DOR), which is constructed as the ratio of gains or losses on the hedging derivative relative to gains or losses associated with the risk being hedged. During periods of low price volatility, the denominator of this ratio (i.e., the change in the value of the hedged item) is likely to be close to zero, making the DOR prone to exceed the upper bound that serves as a limiting condition for applying hedge accounting (i.e., above 1.25).
As a result of this problem, hedgers who rely on DORs for effectiveness testing may find that hedge accounting is unfairly disallowed during periods when price changes have limited economic effect. We propose an alternative measure to the DOR: the percentage offset ratio (POR), which prescribes division by the starting value of the hedged item, rather than division by the change in this value. Our proposed measure provides similar information to the DOR regarding the anticipated effectiveness of the hedge, but it is less prone to being inflated in periods of low volatility due to division by a price change near zero.
The second problem has to do with the R2 measure of a linear regression analysis, which is often used by hedgers to document anticipated hedge effectiveness. The R2 statistic measures the proportion of the total risk that would be eliminated by implementing a hedge for which the size of the hedge position (i.e., the hedge ratio) is dictated by the slope of the associated regression. In general, however, the actual hedge ratio implemented will differ from this slope value. We propose an alternative measure we call the “R2 analogue,” which correctly reflects the proportion of the total risk being hedged (i.e., the total variation in the price change or price level of the hedged item) that is actually mitigated by the hedge ratio put in place.
Ira Kawaller, Managing Director
Heidi Lindahl, Marketing Manager