Can we take the risk of capitalization?

Summary The purpose of this thesis is to evaluate the trajectory, liquidation and longevity risks associated with funding. After a first chapter reviewing the literature, Chapter 2, which is essentially historical, is devoted to calculating the replacement rates that funding could have provided from 1900 to 2003. We show that the risk reduction provided by diversification is small and very costly (in terms of replacement rates). Chapter 3 is devoted to the evaluation of the path risk - measured by the probability that the replacement rate falls below a certain threshold - by the parametric bootstrap method. We estimate the probability of reaching these threshold replacement rates, as well as the savings rate required to reach them with a given probability. The minimum replacement rate imposed by the Fillon law is out of reach with the current funding and contribution rates. The advantages of the multi-pillar system are quite limited. The main one is the reduction of extreme risks. We also show that the probability of failing to reach the pay-as-you-go replacement rate with funding increases with the duration of contributions. In Chapter 4, we compare different strategies for programmed withdrawals to a life annuity. A risk/return approach is adopted where the return corresponds to the expected annuities and possible bequests and where the risk is measured by the probability of not obtaining a certain pension level. We show that this risk can be very high, regardless of the drawdown strategy. No alternative distribution strategy can compensate for the risks associated with the capital accumulation phase.