Kamal Armel & Frédéric Planchet (2020)
Working Paper. ISFA – SAF Laboratory. University of Lyon
Abstract. The « economic » valuation of liabilities under Solvency 2 corresponds to the best-estimate (discounted future cash flows) plus an adjustment to compensate for the immobilization of the solvency capital required for non-hedgeable risks. The valuation framework under IFRS 17 is similar, in that it proposes the valuation of liabilities as the sum of a best-estimate (discounted future cash flows) and an adjustment for non-financial risks. The economic valuation approach for liabilities imposed by Solvency 2 and IFRS 17 can therefore only be fully applied when one has the ability to define: (1) the future cash flows of the liability, (2) an appropriate probability measure and (3) the discount rates. The application of this approach to savings contracts in € (and more generally in the presence of a profit-sharing scheme) leads practitioners to the now classic modelling structure in which a « risk neutral » economic scenario generator feeds a flow projection model to allow an approximation by simulation of the value of the best estimate. In this article, we present a critical analysis of the economic valuation process of life insurance liabilities. We discuss the modelling and probability measurement choices made by the market and their implications. We also provide recommendations for relevant modelling.
Extract. “Indeed, in the standard general equilibrium paradigm, it is assumed that the coordination of agents is ensured by prices. Given the assumption of rationality, contracts drawn up by agents are complete and optimal. They perfectly ensure their coordination and are therefore the only necessary means of coordination.
However, according to the economics of conventions (see BATIFOULIER & al.  and EYMARD & al. ), prices are indeed the efficient means of coordination when all assumptions of general equilibrium theory are verified. They no longer ensure perfect inter-individual coordination when, in particular, there is uncertainty and significant information asymmetries (cf. section 220.127.116.11).
Absent from the Walrasian theory, the convention would be an element of coordination complementary to the price mechanism. For conventionalist authors, price is therefore one means of coordination among others, effective in a particular space, i.e. the market. This market is structured by rules. SEARLE  distinguishes two categories: regulatory rules and constituent rules (cf. BATIFOULIER & al. ).” (p. 19)