The ALM approach has lost its relevance
VERY POPULAR WITH PENSION FUNDS, THIS MODEL IS INCREASINGLY CONTROVERSIAL
Very popular with pension funds, this model is increasingly controversial
“The decision of the SNB has invalidated many models.” Cronos asked on Wednesday about the increasingly controversial relevance of the ALM (Asset Liability Management) model for pension funds. of Actuarial Sciences at HEC Lausanne, François Dufresne, insisted on the need to regularly recalibrate the established models. “When they are built, they are supposed to resist weak shocks, like the” Arab Spring “for example.” On the other hand, and because they have a lasting impact on the macro-financial environment, major shocks like the 2008 crisis require revisions. and new data collections, according to him.
For Aminata Wane, an analyst at Synopsis, this tool, based on the balance between the two parts of the balance sheet (assets and liabilities), lacks flexibility, but does not deserve to be abandoned. Just reviewed in its design. In her case analysis of a popular approach since its introduction in the 1990s, Aminata Wane was struck by the fact that “many ALM reports are based on overly optimistic assumptions,” causing yellow smiles in an audience especially composed of professionals in the field visibly confronted with the differences between envisaged and actual performances.
The analyst went on to describe, step by step, the shortcomings resulting from the typical construction of an ALM management model. “The economic risk is identified only in relation to assets,” she noted, for example. Or: “The extreme scenarios are not included in the methodology.” After launching some lines of thought, such as the possibility of a strategic allocation to two or even three levels, Aminata Wane also conceded that “the practice is otherwise more difficult to grasp than the theory presented here “.
“The actuarial is fortunately – or unfortunately – not an exact science,” added Jacques-André Monnier. The co-founder and boss of Cronos wished to remind us that academic models can hardly anticipate the psychological effects on the market of certain events, planned or not.