Oil, demand, and inflation

It is reasonable to ask if the recent high correlation of inflation among nations is due to oil. I don’t think the evidence supports this claim – though for sure, oil is highly correlated with inflation at any point in time, and being a globally traded good, that correlation has a contemporanous impact on inflation in all markets.

The first, and best, reason to doubt this is that it doesn’t look right on a chart. I see oil leading inflation. However, inflation is only ~5% of most CPI baskets, so to assert that the correlated increase in CPI is due to the pass through of higher oil prices to the other 95% of prices is to assert that there is a highly similar inflation process in all markets.

I think that’s too much to assert given how much difference there is between the underlying units in our sample. It would be a tremendous coincidence if that were the case.

This assertion is also at odds with other research. IMF has done the work more carefully than I have time (just now), and they find that a one percentage point increase in global food prices causes a ~25bps increase in Australian food prices; similarly a one percentage point increase in energy prices causes domestic energy prices to increase by ~10bps.

The reason for this is that, in the IMF’s calculations, the output gap does more of the heavy lifting.

Additionally, the IMF’s work suggests that even if you targetted an exclusion measure, you would not ignore food and fuel price inflation. The above chart shows IMF estimates of the bleed into core inflation from a 100bps increase in food and fuel prices. Note that the estimated passthrough is different between markets.

So what’s going on? Why do we see such co-ordinated swings in global inflation, which appear to be preceded by oil, given that oil is only 5% of the CPI basket and that the transmission from oil to core CPI varies so much between nations?

Have a look at the recent lead-lag relationship, which i’ve zoomed in a bit in the below chart.

Note in particular the earlier decline of oil prices in the financial crash, and the earlier increase in oil prices during the recovery — this is the first clue to what is going on here. Note also that the drop and increase in CPI was globally coordinated, and that it followed oil with a lag. Finally, correlation remains high but it is lower than the correlation between national CPIs (note i switched to 2yr rolling averages, as it better shows up the difference in correlations when there is a phase shift).

I think the correlation is due to a third – common – force … the global business cycle (or equivalently, the Global Business Cycle). Because oil supply is rigid in the short run, small movements in demand cause large movements in its price.

Oil leads CPI, in part, because variations in demand lead variations in CPI – as is typically the case.

Policy is probably playing a big role lately. What i think is occurring is that appropriately easy US monetary policy is being imported by developing markets, due to their inappropriate FX policies, and that this has pushed their economies above their long run supply potential.

After a decade of more of exporting deflation, these nations are exporting inflation. They are the lowest cost suppliers, so it is hard to substitute away from them.

Those in the West, who are not linked into the parts of the global economy that benefit from Asian growth, are paying an inflation tax – their personal terms of trade are declining. This is a break on income formation and activity, and yet inflation is rising despite weak growth.

You can see this in the US GDP figures released last night. Those nations suffering an adverse terms of trade movement can look forward to stagflation. indeed, the US and UK already have it.

Those lucky enough to sell something Asia wants, such as Australia, get a two speed economy.

What we all are going to get is a period of higher rates, given any set of real economic indicators, and probably also higher economic volatility