As already noted in the recent – and closely related – InFocus piece on US inflation: Past the peak but then what? (May 2022), headline inflation as measured by the CPI fell to 8.3% y/y in April. At the same time, core inflation fell to 6.2% y/y. While these rates represented a decline on the rates observed in March, they are far above the Federal Reserve’s 2% objective.1
There is understandably much focus on headline and core inflation (and sometimes also on the food and energy prices that represent the difference between these measures). However, it is also informative to disaggregate the CPI one step further. By doing so it is possible to distinguish between food at home and food away from home, between energy commodities and energy services and, furthermore, to split core inflation into commodities less food and energy and services less energy.
The table and graph below show that the prices of energy commodities (comprising fuel and motor oil) have risen most sharply, almost 40%, over the last year. Energy services and commodities less food and energy commodities have also risen markedly, 12%, in the same period.
Investment Insights
5 min read
US CPI Inflation: volatility and weights
At long last, the US inflation rate appears to have peaked, but at a much higher rate and much later than anticipated. Why have these large forecasting errors been made and what now is the outlook for inflation? In this Macro Flash Note, EFG chief economist Stefan Gerlach looks at these issues.
Figure 1. CPI inflation, April 2022
Source: BLS.
Of course, the importance of any subcomponent of the CPI in accounting for variations in inflation depends on two factors.2
- Its volatility, that is, how much the subcomponent varies over time. Plainly, subcomponents that are prone to large swings are more likely to impact the overall inflation rate.
- The weight attached to the subcomponent. Subcomponents with a large weight will be more important for the overall inflation rate.
It is easy to see that it is the combination of these two factors that matters. In the graph below, the volatility of the subcomponents is measured on the vertical axis and the weight of the subcomponent on the horizontal axis.3 The size of the spheres is given by the product of the weight and the volatility of the subcomponent.
Figure 2. US CPI: Weights and volatility
Source: EFG calculations using BLS data.
The graph tells an interesting story. The largest sphere is that of energy commodities. While these have a minimal weight in the CPI, 4.9%, they are highly volatile.4 They therefore contribute disproportionally to movements in the CPI.
This is demonstrated in the graph below, which shows headline CPI inflation together with the annual percentage change in the index for energy commodities. The relationship is stunningly close (the correlation coefficient is 0.9). Indeed, as a first approximation almost all variation in CPI inflation is due to changes in prices of energy commodities.
Figure 3. Headline inflation and energy commodities
Source: BLS.
The second largest sphere is that for commodities less food and energy. That component is very stable, which reduces its importance, but it has very large weight (21.5%) in the CPI. The third largest sphere, that for services less energy services, is almost as important. It is even less volatile but has an even larger weight, 56.8%.
The graph below shows the rate of change over 12 months of these components. Both are exceptionally stable in the decade before the Covid pandemic, suggesting that they have been of little importance in explaining fluctuations in inflation, but they then surge from spring 2021 onward. The rate of change of prices for commodities less food and energy reached a peak in July 2021, before falling back and then rising again to a new peak in February 2022.
These increases were driven by a combination of shifts towards goods and away from contact-intensive services as Covid started, additional spending increases as the economy reopened, coupled with price increases arising from supply-side tensions.
The behaviour of the prices of services less energy, which is essentially shelter, is similar. They rose at a very steady rate before Covid, before moderating as the pandemic took hold. They rose very sharply during the recovery. While the pickup is small – from 2.7% in August 2021 to 4.9% in April 2022 – their weight in the CPI is so large that the overall inflation rate was pushed up by about 1.3%. This effect has more than offset the fall in the rate of increase of commodity prices less food and energy from 12.4% to 9.7% between February and April 2022, which has lowered overall CPI inflation by merely 0.6%.
Figure 4. Commodities and Services
Source: BLS.
The main conclusions to draw from this analysis are straightforward. Fluctuations in inflation have historically been due largely to changes in energy prices that move erratically. However, during the last year the prices of commodities less food and energy and services less energy, which in the past have risen at slow and very stable rates but which have large weights in the CPI, have increased in highly unusual ways, putting upward pressure on inflation. While the rate of increase of commodity prices has started to fall, the rate of growth of the prices of services less energy continues to rise. This suggests that inflation will take longer to decline than previously expected although exactly how long is difficult to quantify.
1 Importantly, the Fed’s objective is for PCE inflation, which generally runs about 0.5% below CPI inflation.
2 This is a slight simplification: the correlations between the different subcomponents also matter.
3 The volatility is computed by calculating the standard deviation of the change from month to month of the annual rate of change of the subcomponent, using data starting in January 2005.
4 This is the weight applicable to April 2022.