The European Commissionâs UCITS rules recommend that VaR calculations be conducted using ârecent volatilities (i.e. NO MORE THAN one year)â. However, when this was transposed into national regulations in Luxembourg, Ireland and Germany it became a stipulation that the historical period used should âNOT BE LESS THAN 1 yearâ.*
This, says EM Applications, is as though an EU speed limit of â70â had been set, which in some countries was then legislated in kilometres per hour and in others in miles per hour! While both kilometres per hour and miles per hour are perfectly sensible ways to measure speed, they will produce very different results and lead to confusion amongst drivers about what the exact meaning of the speed limit is.
But this contradiction at the heart of the UCITS rules is not due to an error on the part of either the Commission or the Luxembourg, Irish or German regulators. It is a recognition that the historical period used is fundamental to the purpose of the VaR calculation.
What has been overlooked in the ongoing debate about the usefulness of the âValue at Riskâ (VaR) measure is that there are many ways to measure VaR, which will generate very different answers. Typically, VaR is determined by an investmentâs historical behaviour. Using a short historical âlookbackâ period to estimate VaR is appropriate for those seeking to estimate potential losses in the immediate future, as it will be more reflective of current market conditions. However, for long-term management of an investment portfolio with a certain type of risk profile, a longer term measure, using historical data from several years in the past, is more appropriate as an indication of the portfolioâs intrinsic risk.
EM Applications believes that neither approach, short term or long term VaR, is âsuperiorâ to the other, as both measures provide useful information. EMA recommends that the UCITS rules be changed to require both measures â a long term one as a way of characterising the risk level of a fund, and the short term one to ensure that losses in excess of the fundâs assets cannot happen in reasonably foreseeable market conditions.
This issue is also addressed in the Turner review in relation to bank capital. In it, Lord Turner wrote: âMeasures of VAR were often estimated using relatively short periods of observation e.g. 12 months. .... At very least much longer time periods of observations need to be used.â (P44)
Peter Ainsworth, Managing Director of EM Applications, said: âWhether by design or accident, the UCITS Regulators have come up with the solution to the main issues raised about VaR â collectively they have required investment managers to calculate both a short term and a long term VaR. Had the banks been doing this they would not have operated with so little capital and would have suffered much less harm as a consequence of the sub-prime crisis. All that is needed now is for the EU to amend the UCITS rules to require all jurisdictions to compute both long and short term VaR and we will have a practical way forward that addresses a major shortcoming of the previous regime.â