January 2013 Industrial Production

Yesterday’s industrial production numbers had some good, some bad (log annual and monthly changes; seasonally adjusted; 2000=100):

The good news is that in annual terms, industrial production is looking up a bit; the bad news is that the monthly numbers are flat to negative except in electricity production.

I can’t discount the impact of CNY here, as last year the holiday fell right in the middle of January, whereas this year it was a bit later. That suggests, even allowing for seasonal adjustment, these figures might overstate economic activity a little.

Nevertheless, the implied growth rate is about 5.5% for 1Q2013 – bearing in mind we’re extrapolating from just one month’s data:

If last quarter’s forecast level appears to have hit the mark exactly, it’s because I’ve re-estimated the IPI-based model, which would incorporate the latest data. In fact I did that for the whole slew of time series models I use – the weighted average forecast is between 5.4%-5.8% real GDP growth for 1Q2013, with a 95% confidence interval of plus-minus 1%.

FWIW, I’ve been studying forecast evaluation methodologies, and the forecast error for 4Q2012 was the highest since 1Q2009, right in the middle of the recession i.e. statistically, the 4Q2012 result is likely to be an outlier. I hope to eat those words this year.

Technical Notes:

January 2013 Industrial Production Index report from the Department of Statistics (warning: pdf link)