Talk:Two-moment decision model

the utility function
I know that there are two types of utility that we can have mean variance analysis. What are they? they are not mentioned in the article. Jackzhp (talk) 19:51, 19 May 2011 (UTC)


 * Under the condition discussed in the article (location-scale-related random variables), any utility function can use mean-variance analysis. I'll clarify that in the article. In addition, without the condition on the random variables, a quadratic utility function could use mean-variance analysis; but no one would ever do that, since the quadratic utility function is not monotonically increasing and also because it exhibits the undesirable property of increasing absolute risk aversion. So quadratic utility is irrelevant and I think it does not belong in the article. Duoduoduo (talk) 20:08, 19 May 2011 (UTC)

Double meaning of E?
I have no objection to today's edit changing EW to $$\mu_W$$, but I'm curious about the motivation: I don't see the double meaning of E -- it always meant "the expected value of what follows". Either way is standard notation in this literature. Could you clarify the motivation? Thanks. Duoduoduo (talk) 23:15, 19 April 2013 (UTC)


 * It looked like a mistype for Ew, particularly next to &sigma;w, so that it looked like E was both a quantity and (later) an operator. Note that Wikipedia standards for maths/stats/probability is to use non-italic for operators. I didn't change lower case random variables to upper case, but that might also make things clearer. Melcombe (talk) 11:18, 20 April 2013 (UTC)