In page navigation

Allais paradox

Updated: Aug 20, 2021

An example of choice under uncertainty where the outcome for most experimental subjects violates the axioms of expected utility theory. In the first experiment subjects are requested to choose between winning £1 million for sure and a gamble in which the prizes are £1 million with probability 0.89, £5 million with probability 0.10, and nothing with probability 0.01. Most experimental subjects would choose the option of £1 million for sure. In the second experiment the choice is between two different gambles. For the first gamble the prizes are nothing with probability 0.89 and £1 million with probability 0.11. The second gamble has prizes of nothing with probability 0.90 and £5 million with probability 0.10. Most subjects choose the second gamble. These typical choices violate the independence axiom of expected utility theory: if the first option is chosen in the first experiment then the first option should also be chosen in the second experiment. The Allais paradox and other similar examples have motivated the search for alternatives to expected utility theory.

Reference: Oxford Press Dictonary of Economics, 5th edt.

Sources & references
Risk disclaimer

Invezz is a place where people can find reliable, unbiased information about finance, trading, and investing – but we do not offer financial advice and users should always carry out their own research. The assets covered on this website, including stocks, cryptocurrencies, and commodities can be highly volatile and new investors often lose money. Success in the financial markets is not guaranteed, and users should never invest more than they can afford to lose. You should consider your own personal circumstances and take the time to explore all your options before making any investment. Read our risk disclaimer >

James Knight
Editor of Education
James is a lead content editor for Invezz. He's an avid trader and golfer, who spends an inordinate amount of time watching Leicester City and the… read more.