The Association Française de la Gestion financière (AFG) – the association representing French management companies – published on 7 May 2015 guidelines regarding the use of stress tests by management companies (Guidelines).The Guidelines are aimed at helping the AFG’s members set out the regulatory stress tests applicable to collective investment schemes (CIS), which have been implemented by EU regulators under their national laws. These regulations were derived from the stress tests implemented for the banking industry.2

Indeed, stress tests have become mandatory for UCITS using the Value at Risk (VaR) method, since the publication of CESR level 3 measures contained in CESR’s Guidelines on Risk Measurement and the Calculation of Global Exposure and Counterparty Risk for UCITS (July 2010). Stress test requirements were extended in January 2014 by the Autorité des Marchés Financiers (AMF, the French Financial Markets Authority) to all French registered alternative investment funds within the meaning of the AIFM directive3.

The Guidelines provide management companies with useful recommendations regarding the implementation of internal procedures appropriate to ensure efficient methods of risk valuation for each CIS under management. The Guidelines are divided into three main parts – objectives, applicable regulation, and AFG’s principles concerning stress testing practices. The AFG’s principles focus on five main concerns that are directly related to the efficient implementation of regulatory stress tests: proportionality; correlation; tolerance level; thresholds for reverse stress tests; and disclosure of results.

Proportionality of Stress Tests

The Guidelines indicate that the regulatory stress tests must be implemented in a way that complies with the principle of proportionality – that is, the stress tests must be calibrated to reflect the complexity of risk of the CIS. Furthermore, the “performance drivers” (the techniques and/or methods of investment used by a CIS that mainly impact its performance) set forth in the CIS’ prospectus and marketing documents also represent sources of risk that should be the subject of stress tests. In particular, the AFG notes that the use of arbitrage strategies represents a performance driver as well as a source of risk, and therefore a CIS utilizing arbitrage will have a complex risk profile. As a result, the stress tests should include all arbitrage positions and imperfectly hedged risks. On the other hand, the AFG considers that the calibration of the stress test for a non-complex (i.e., plain vanilla) or narrow-focused (i.e., equity benchmark) CIS may be simpler.

Stress Tests Without Correlation Between Risk Factors

The Guidelines recommend that management companies perform at least one stress test without correlation between risks factors (a “worst case” scenario) for those CIS using the VaR method. Indeed, AFG notes that extreme risks are not taken into account in the VaR method and that, consequently, the data derived from historical stress tests are not sufficient to measure the real risk exposure of such CIS. Therefore, the AFG specifies that at least two different scenarios should be considered by management companies in connection with the severity of the event as measured by the stress test. One scenario should be calibrated for situations that are unfavorable to the CIS, but probable. The other scenario should be designed to test the consequences to the CIS in the event of a severe market crisis.

Tolerance Level for Stress Tests

The AFG stresses that, according to the applicable regulations, stress tests are to be integrated into the CIS’ risk monitoring process. In particular, the stress tests results should be assessed and analyzed by the company’s risk management personnel and reviewed by the executive board of the management company. Therefore, such results must be taken into account in the CIS’ investment process, and management companies should be encouraged to take appropriate measures in case the results reveal a particular vulnerability to a set of circumstances (i.e., by hedging or reducing the exposure of the CIS). In addition, the AFG recommends that management companies establish a loss tolerance level for each CIS subject to stress tests, which can be based either on the SRRI (a specific risk-reward indicator) or the VaR method, and must be communicated to the executive board of the management company.

Thresholds for the Applicability of Reverse Stress Tests for CIS using the VaR Method

The AFG highlights that the main purpose of reverse stress tests is to measure any potential depreciation of the CIS’ value arising from unexpected market movements. The AFG considers that such reverse stress tests should be provided only for those CIS whose tolerance level for stress tests is higher than 50% and where the stress test results are close to such level.

Disclosure of Stress Test Results

The Guidelines describe conditions for the disclosure of stress tests results. The AFG recommends that, in order to minimize the possibility that clients may be confused when reviewing these results (i.e., pertaining to risk factors or market shock levels), management companies should also make available (i.e., on their website) the hypotheses upon which the testing was based.