tag:blogger.com,1999:blog-19649274.post907905847885227346..comments2023-02-02T12:24:16.707+05:30Comments on The Leap Blog: Crisis watch, 29 December 2008Ajay Shahhttp://www.blogger.com/profile/03835842741008200034noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-19649274.post-68758870776099941872009-01-01T22:28:00.000+05:302009-01-01T22:28:00.000+05:30The first graph superposes the policy rate with th...The first graph superposes the policy rate with the rate that would be computed off a `proper' Taylor rule. There are several hurdles which have to be crossed to get there.<BR/><BR/>1. Operationalising a Taylor rule requires having estimates of the stuff on the r.h.s.: we'd need to have a forecast of inflation and a forecast of the output gap. The models which make these forecasts haven't yet been figured out in India.<BR/><BR/>2. Making the graph requires agreeing on what Taylor rule to use. One strategy is to estimate what's RBI's Taylor rule. In my knowledge, this has been done twice so far: by Vineet Virmani (who was then a Ph.D. student at IIM Ahmedabad) and in a BIS working paper by Mohanty & Klau which dates back to 2004. These parameter values could be utilised.<BR/><BR/>3. But the trouble is, there are obvious problems with the Taylor rule that's suggested by these parameters. E.g. we know that monetary policy is procyclical when the inflation coefficient is below 1. But the above estimates show that RBI's Taylor rule has an inflation coefficient of well below 1. So if you want to make Graph 1 of Taylor's paper, which is supposed to compare the actual course of policy against some normative yardstick, do you use RBI's Taylor rule? It would make more sense to use standard parameter values from the literature such as 1.5 for the inflation coefficient and 0.5 for the output gap.<BR/><BR/>That'd give us 3 lines on the curve: Actual behaviour; behaviour under RBI's Taylor rule; behaviour under a normative Taylor rule with sensible properties.Ajay Shahhttps://www.blogger.com/profile/03835842741008200034noreply@blogger.comtag:blogger.com,1999:blog-19649274.post-30352895419191254692008-12-29T13:54:00.000+05:302008-12-29T13:54:00.000+05:30Sincerely grateful for referring to the paper by ...Sincerely grateful for referring to the paper by John Taylor. Just too good. <BR/><BR/>Anything similar to that first graph (loose fitting) available for India somewhere ?<BR/><BR/>Best regards,<BR/>Anonymous1Anonymousnoreply@blogger.com