Belief Updating Under Ambiguity : A Numerical Simulation Analysis

We compare Bayesian with the two non-Bayesian updating under ambiguity in its model specification and numerical implications. For an ambiguity averse (seeking) decision maker, we find that Gilboa and Schmeidler (GS) updating leads to a lower (higher) certainty equivalent (CE) than Dempster-Shafer (DS) updating, and the change of historical information has a bigger (smaller) impact on GS updating than on DS updating. In addition, the results shows that GS updating is more sensitive to the ambiguity attitude than DS updating. The application of numerical analysis in finance especially implies that the sensitivity of price to information depends on the ambiguity attitude of the market