Note on two applications of the CEVR utility function

✍ Scribed by Georges Bernard

• Publisher: Springer US
• Year: 1978
• Tongue: English
• Weight: 181 KB
• Volume: 9
• Category: Article
• ISSN: 0040-5833
• DOI: 10.1007/bf00131775

✦ Synopsis

NOTE ON TWO APPLICATIONS OF THE CEVR UTILITY FUNCTION

In [1] we proposed a constant elasticity of value and risk utility function for a better explanation of actual behaviours in the presence of uncertainty. This note analyses two 'paradoxes' by means of this CEVR function. They are the examples proposed by Allais in [2], in 1953, and by Tversky and Kahneman in [3] and [4], in 1974 and 1975. I. Allais considered the following stochastic variables or bets: L1. A 1 9 a sure gain of $ one million A 2: a gain of $ five million with probability 0.1, of nothing with probability 0.01 and of $ one million with probability 0,.89 B 1: a gain of $ one million with probability 0.11 and of nothing with probability 0.89 B 2:

a gain of $ five million with probability 0.1 and of nothing with probability 0.9 Several persons to whom he proposed these bets, if they preferred A 1 to A 2, preferred B 2 to B 1. These behaviours are inconsistent with the expected utility criterion of yon Neumann and Morgenstern and others. Hence the 'Allais Paradox'.