For some time I have wanted to write about a couple of extremely interesting blog posts by Justin Irving on his excellent blog Economic Sophism.
In three blog posts (here, here and here) Justin has explained how he has estimated a real-time model for next year’s US NGDP growth.
I don’t want to get into details of Justin method, but overall Justin is using so-called Principal Component Analysis to identify a common “trend” in different financial asset prices to estimate a real-time forecast for expectations for next year’s US NGDP.
The graph below shows a week of data for US NGDP expectations according to Justin’s model.
As far as I know Justin’s model produces new forecasts every three-minutes. That of course provides analysts, commentators, reporters and policy makers with an extremely interesting tool.
Just imagine how the tool can be used during a FOMC meeting. First how will the indicator react when a policy decision is announced? And then later during Ben Bernanke’s Q&E session. We will actually be able to in real-time to evaluate whether Bernanke’s comments and guidance is tightening or easing monetary conditions.
So I certainly hope that Justin will make the real-time available during the next FOMC meeting. Maybe Justin should be tweeting real-time during the next FOMC meeting and Bernanke’s Q&E? Otherwise I will do it…(btw you find me on Twiiter here)
PS I have earlier suggested using market data to estimate NGDP expectations in the US. See my blog post “Markets are telling us where NGDP growth is heading”
Justin Irving
/ July 25, 2013I’m glad that you find it worthwhile Lars. The system is *extremely* buggy, having broken down from a downloading problem yesterday. Still, I am publishing the 3-minute updates here: https://docs.google.com/file/d/0B6v3_6dcHgorTVRqWTZVLUZnWWc/edit
(note that July-24 is missing)
Looks to be working today, though with only bonds and currencies trading, it’s not so interesting. I can’t promise it won’t go haywire…yet!