Comments on: A neat application of Robustness http://www.ambrosini.us/wordpress/2009/11/a-neat-application-of-robustness/ Sharpening my knife Tue, 18 Oct 2011 07:32:30 -0700 http://wordpress.org/?v=2.8.6 hourly 1 By: pushmedia1 http://www.ambrosini.us/wordpress/2009/11/a-neat-application-of-robustness/comment-page-1/#comment-8305 pushmedia1 Thu, 26 Nov 2009 05:57:20 +0000 http://www.ambrosini.us/wordpress/?p=1304#comment-8305 AC, sounds like your ready to pick up the Hanson and Sargent text. AC, sounds like your ready to pick up the Hanson and Sargent text.

]]> By: Agent Continuum http://www.ambrosini.us/wordpress/2009/11/a-neat-application-of-robustness/comment-page-1/#comment-8304 Agent Continuum Thu, 26 Nov 2009 00:27:17 +0000 http://www.ambrosini.us/wordpress/?p=1304#comment-8304 Minimax is about their loss function. I get best forecast of the worst-case scenario vs. best forecast of most likely scenario and all, but then I suppose you just call it something else. It can't be pure minimax either, since we can certainly think up worse scenarios than their forecast. Minimax is about their loss function. I get best forecast of the worst-case scenario vs. best forecast of most likely scenario and all, but then I suppose you just call it something else.

It can’t be pure minimax either, since we can certainly think up worse scenarios than their forecast.

]]> By: pushmedia1 http://www.ambrosini.us/wordpress/2009/11/a-neat-application-of-robustness/comment-page-1/#comment-8301 pushmedia1 Wed, 25 Nov 2009 22:22:39 +0000 http://www.ambrosini.us/wordpress/?p=1304#comment-8301 They're minmaxing so you can read the FOMC's forecast as a worst-case scenario. Presumably their forecast has a low forecast error in the universe where those worst-case scenarios played out. There's no reason to think the most important moment forecast is the mean. They’re minmaxing so you can read the FOMC’s forecast as a worst-case scenario. Presumably their forecast has a low forecast error in the universe where those worst-case scenarios played out. There’s no reason to think the most important moment forecast is the mean.

]]> By: Agent Continuum http://www.ambrosini.us/wordpress/2009/11/a-neat-application-of-robustness/comment-page-1/#comment-8294 Agent Continuum Wed, 25 Nov 2009 01:41:12 +0000 http://www.ambrosini.us/wordpress/?p=1304#comment-8294 I'm not familiar with the Robustness literature but, presumably, it should be about the decision maker's loss function and her optimal actions rather than about her forecasts. Why would anyone want to have anything but a zero mean and minimum variance forecast error, model uncertainty and objectives aside? I’m not familiar with the Robustness literature but, presumably, it should be about the decision maker’s loss function and her optimal actions rather than about her forecasts.

Why would anyone want to have anything but a zero mean and minimum variance forecast error, model uncertainty and objectives aside?

]]>