Wednesday, December 13, 2017

How is the CPI forecast holding up?

Let's update the dynamic information equilibrium CPI (all items) forecast graph with the latest data (previous update here):



Here is the year-over-year inflation version as well:



2 comments:

  1. Just followed your link to your September forecasts. They all look good except Bitcoin. What do you have to say about that?

    ReplyDelete
    Replies
    1. Hi Russ,

      I wrote up a post on that failure:

      https://informationtransfereconomics.blogspot.com/2017/10/bitcoin-model-fails-usefulness-criterion.html

      The data is still consistent with the dynamic equilibrium model, it's just too buffeted by shocks (which are unpredictable because they're based on human behavior like bubbles or groupthink) for the model to be useful for forecasting. Measures like the unemployment rate, GDP and CPI have fewer shocks (unrate every ~ 8 years with a recession, the latter two have only a couple significant shocks over the post-war time series) which makes the dynamic equilibrium forecast useful. However bitcoin has shocks every few *months*, and every shock adds a couple parameters (location, size) meaning that the model would never be validated on bitcoin data alone (we'd need a confluence of data from various sources and then only looking at bitcoin after the fact).

      At the link, I've been tracking the estimate of the size of the latest shock. What's interesting is that just like the unemployment rate model, the estimate of a shock in real time first underestimates its size and then over estimates its size. I want to see if that plays out with the recent bitcoin shock, so I'm continuing to follow it.

      Delete

Comments are welcome. Please see the Moderation and comment policy.

Also, try to avoid the use of dollar signs as they interfere with my setup of mathjax. I left it set up that way because I think this is funny for an economics blog. You can use € or £ instead.