Search interesting materials

Monday, October 12, 2009

Did we get a good IIP number today? No.

It seems that we got a good IIP number today? Here's what the year-on-year growth of IIP Manufacturing says:

Mar 2009
Apr 2009
May 2009
Jun 2009
Jul 2009
Aug 2009

So it looks like a recovery is gaining traction, with yoy growth going up in each month from March 2009 till August 2009, other than a slight decline in July 2009?


The year-on-year growth is, unfortunately, a highly misleading measure. Here's what we get when we look at the seasonally adjusted levels and the annualised growth of these:

Mar 2009
Apr 2009
May 2009
Jun 2009
Jul 2009
Aug 2009

This shows a very different picture. It shows one big month -- June 2009 -- where seasonally adjusted IIP jumped from 299.09 to 319.78. This was an annualised growth rate of 80.32%. This was the month in which the recovery kicked in, where we jumped back from the low state to a better state.

After that, growth has dropped back to anaemic levels, with annualised growth of 3.84% in July and 4.98% in August. So today's data release was actually not so hot - it was just 4.98% growth (annualised) of seasonally adjusted IIP manufacturing.

Each value of the year-on-year growth rate is the moving average of the latest 12 values of the month-on-month growth. So this one big value of 80.32% is going to elevate each reading of yoy growth up, all the way until we calculate the yoy growth from May 2009 till May 2010. After that, we'll get to the yoy growth from June 2009 till June 2010 and this jump will go away.

You only get the true picture by looking at point-on-point changes of seasonally adjusted data.


  1. Any theories as to why is there an 80% growth rate in June?

  2. 80% annualised is still huge growth. Are there any reasons for such growth rates? Or are we to question the IIP itself.

  3. Is it not the responsibility of the govt. to give accurate data rather than feel good data?

  4. It is the responsibility of the government to produce the public good of the statistical system. And there are big concerns about the mechanisms used for making IIP.

    In this case, the choice between year on year (YOY) vs. point on point seasonally adjusted (POPSA) is not one of accurate vs. feel good. It's one of faster versus slower. You always want a faster picture - of what happened last month. YOY always delivers the average of the changes of the last 12 months. Sometimes the YOY can be more pleasant to contemplate, and sometimes the POPSA can be more pleasant to contemplate.

    I think the biggest problem lies in the user community. Once the government has released information, it is possible to do seasonal adjustment externally. The key is that users should be careful in not accepting YOY growth as a reasonable way to think about this.

  5. I think that the seasonally adjusted numbers are a great thing and it's very good that you are doing it.

    As a suggestion, could you put tables indicating the actual numbers and seasonal growth rates alongside the graphs. Maybe 3 columns, month, y-o-y growth, seasonal growth.

  6. Rishi,

    I think the table you are looking for is in the PDF file which is on the website. Can you look at that?


Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.