Wednesday, July 23, 2014

Japan Update: We've Waited, Let's See

A few months ago, Stephen Williamson posted this unfortunate picture.

It turned out that the large drop in inflation that he thought was evidence in favor of his views on monetary policy was actually the result of an accidental misuse of Japan data (as I pointed out in this post and in the comments on his blog).  He corrected the mistake and adopted a wait and see attitude, "Thus, the Bank of Japan seems to have had some difficulty in producing inflation - but this time may be different. Given that the Bank seems intent on holding the short-term nominal interest rate at zero (essentially), I don't see it, but I'm curious to see how the data unfolds."

Well, we've waited a bit and can see a little more.  Let's update the graph:

Wow!  It's almost directly opposite of the original.  Japan yoy inflation is up close to 4%.  What liquidity trap?

It's interesting to note that way back in April of 2013, we noticed that markets were predicting a rapid rise in the second year of Abeonomics.  The market predicted .35% for year one (until April 2014) and then 2.3% the following year.  Well, the market estimate was obviously low.  Inflation actually averaged about 0.89%, but we may be seeing that the dynamics of back loaded inflation is substantially correct.  Now will the BOJ be able to hit their target?  What is their target?

In response to a comment, here are the index values: (Data)


  1. Charlie, interesting. How does this compare with the plot here (made earlier this month):

    It's hard for me to tell if that slightly earlier plot is missing any data points.

  2. Tom,

    It looks like he has the two month uptick. I'm not sure exactly what he is using though. In a previous post he used core, which is 100.6 in April and 100.7 in May. In my post, I used year over year inflation rate on CPI-all items, as this is what Williamson used. The April CPI is 103.1 and May is 103.5.

    He may have slightly different numbers, because the series he is using has a different base year. I will add the index numbers for CPI and Core.

    1. Yep, I was using the core with 100.6 and 100.7 in April and May, respectively.

      This jump appears to be due to the VAT increase predicted by Sumner almost a year ago:

    2. Jason,

      Yes, I discussed the VAT on the previous posts. Unlike that Sumner post though, the market doesn't seem to be predicting a backslide back to below 2%.

      Is that what your model predicts?

    3. Yes, according to the model, inflation should head back below 1% (unless the log-linear extrapolation of NGDP growth or currency base growth is off significantly). If NGDP growth is large, then inflation will be slightly above 1%. I'm excited to see if the model works :)

  3. I put the market break even rates from today on a new post. Would be curious how they compare to Jason's model.