Covid distorts the stats

 After rebuilding my programs and updating my data banks, I have started to once again be able to calculate world industrial production and world GDP.  The way I do that is to calculate IP and GDP for each continent.  

I was the first person to calculate monthly world industrial production back in 1992.  I did it to explain movements in commodity prices, because OECD industrial production didn't cover China, Brazil, Russia, etc.  By factoring in these fast-growing economies, the surge in commodity prices made sense.  Now, of course, including large developing economies in our calculations is standard.

To create my early version of world industrial production, I did a lot of trawling through old physical copies of the UN Monthly Statistical Bulletin and other paper data sources to obtain industrial production data to feed my indices, and I have kept those time series more or less up to date, though recently, various factors led to my letting them get a couple of years out of date.  

I have now updated these records, and can once again calculate world industrial production and GDP, as well as regional indices.  

The chart below shows GDP-weighted IP and GDP for South America,  extreme-adjusted, which helps reduce the Covid-related  downwards spike in 2020.  I have also expressed the indices as a percentage of their long-term moving trend.  As economies develop, the trend growth rate gradually slows.  Expressing the underlying series as a percentage of its own moving trend allows us to understand the business cycle better.

One interesting result is that industrial production has rebounded much more strongly relative to trend and on a year-on-year basis than GDP, despite both falling year-on-year by similar percentages.   Otherwise, as you can see, industrial production usually tends to have very similar cycles to GDP.

I should be able to calculate world GDP and industrial production tomorrow, and I'll post the results then.