Increased volatility in capital markets, the significant growth of derivative instruments and other developments have motivated market professionals to decide on a composite global indicator of financial risk.
Thus, the VaR (Value-at-Risk), initially only applicable to banks, has become the essential standard for financial risk evaluation in the asset management sector.
Development of the regulatory framework
The old method of assessing risk, based on a linear approach to commitments by converting simple derivative financial instruments into their equivalent
underlying instruments is now confined to funds holding less than 10% in complex derivatives. Other funds, so-called sophisticated funds, are subject to a new probability-type approach based on calculation of VaR.
SGSS offers its clients an independent risk measurement and attribution service. Carried out by our financial engineering team, this includes regulatory VaR calculations. The parameters (confidence levels and calculation horizon) meet the regulatory authorities' requirements but can also be adapted to clients' needs.
Regulatory risk measurement
This first service was designed to handle regulatory VaR calculations, obligatory in France from 1 January 2007 for the sophisticated funds.
This second approach completes the measurement of VaR/CVaR and enables potential losses in a crisis situation to be estimated: volatility exit, yield curve distortion, macro-economic events, stock market crises, LTCM-type bankruptcies, etc.
Multi-criteria risk attribution (VaR, Tracking Error, Volatility)
The third element involves analysis of the contribution of management decisions to global portfolio risk. In this sense, it is an ideal complement to performance attribution and enables managers to overlap the results obtained.
Combining expertise with flexibility
The technical principles on which SGSS services are founded give them the transparency required by the regulatory authorities: definition of the mathematical formulae used, setting out the advantages and limitations of the models and clear explanation of the sources employed, as well as the alternative rules and strict procedures for checking the results.
In this context, VaR may be determined in accordance with the following approaches:
- the analytic method;
- the historical simulation method.