A quick look at the abstract reminds us of the dangers in tightly packed formal language (e.g. slightly dominate)* in addition to getting the gist of the paper without the mind-numbing effect of the more technical sections. It looks for short term/horizon exchange rate predictability using the Taylor Rule and other complicated forecasting methods I’m not familiar with. Another important distinction is that of real-time data and vintage data, the later having been revised in some way and the former being what you might consider raw.
Historically research hasn’t been able to predict change other than a no change hypothesis, i.e. they can’t reject the null hypothesis of 0.** In an attempt not to dig myself an even larger hole of blatant ignorance I’ll quote the conclusions of Molodtsova’s paper.
“In contrast with previous results in literature, the evidence of short-term predictability with (an) asymmetric Taylor rule model in a panel framework is striking. The evidence of out-of-sample predictability is found for 5 out of 10 countries.”So, good work team!
The most useful item I picked up, from my perspective of complete ignorance, was a cursory understanding of the Taylor rule. It is an item used by central bankers to help target the nominal interest rate (read federal funds rate) and includes things such as inflation and target inflation, output and potential out, and other things that seem to require ghastly assumptions such as the equilibrium real interest rate. Also included for some countries (not the US) is a targeted exchange rate.
*I included sentence interruption specifically because McCloskey thinks its crap and I happen to be quite fond of interrupted sentences.
**At this point let me elaborate on my almost complete ignorance on this topic. I’m an undergrad who has had two legitimately macro oriented classes. This is the first time I’ve been exposed to any type of macro forecasting.
No comments:
Post a Comment