Home > Main > Did inflation shoot up in December, or was it January?

Did inflation shoot up in December, or was it January?

February 4, 2011

People all over the place are questioning the validity of the GDP deflator from the last GDP report.  The GDP deflator doesn’t seem to match the anecdotal evidence, and with commodities going hog wild, people are right to be skeptical.

But I was looking at the numbers from the Billion Prices Project, and it appears that the spike in inflation really didn’t start until January. This would help explain the outcry over the perceived inflation and the printed inflation GDP deflator.  GDP deflator is the prices paid over the entire quarter, not just the last day.

I cannot get more granular than this with the charts the BPP provide, but it looks like the spike of inflation started late in Q4 and perhaps in Q1 of 2011.

In this chart, there is a huge downward spike for a few days in the BPP data in late December.  What could have caused that?  Intense last minute holiday discounts is a reasonable answer, but I don’t have access to the core data. Speculation on a cause for the fall is all I can provide.

The methodology is that this index provides Year over Year inflation levels. For December, inflation was only running at 1% YoY, and through most of the quarter it was running weaker than 1%.  This means that right now we’re running at an inflation level of 2% for the last few days of January.

These are online prices, for products that are sold online.  These products must be skewed away from goods that are part of the CPI and to electronics very heavily.  Maybe not.   I just found a very, very recent report that goes into more detail on how they collect the prices.  I’ll have time to read it next week.

If they are then adjusting the numbers to accounting for this in some way, I get very suspicious.  I’ve used everything from simple moving averages to kernel regression and kalman filters, and adjusting the numbers with a mathematical transformation always skews the underlying data in ways that can be difficult to explain.   Every filter has a corresponding data structure that exploits its weak points.


Categories: Main
%d bloggers like this: