Testing for asymmetric information in the viager market

A viager real estate transaction consists in selling a property in return for a down payment and a life annuity that the buyer has to pay until the seller dies. This paper tests for the presence of asymmetric information in this market using notarial data on transactions in Paris between 1993 and 2001. We first derive and test empirically a no arbitrage condition stating that the price of a viager has to be equal to its expected returns. Using this condition, we identify the type of the seller as a sum of weighted death probabilities. By comparing these sums with analogously defined national-level sums we find that sellers have on average shorter survival times than persons in the general population. We then develop a model for a viager sale and derive testable predictions under symmetric and asymmetric information. Our test for asymmetric information consists in regressing the contract parameters (down payment and annuity) on the inferred type of the seller, and comparing the estimates with the predicted outcomes. The hypothesis that information is symmetrically distributed between buyers and sellers is accepted.

► Exploration of an original market (the viager market) reminiscent of the annuity market.
► A no arbitrage condition allows to recover some information about the death probability of the sellers.
► At first sight, a surprising selection effect: viager sellers die sooner than the average population.
► Sellers have information about their death probabilities and share it with the buyers (symmetric information).
► A theoretical model is developed and tested econometrically.