Josh Bivens at PPE recently presented a decomposition of price changes into those attributable to price-cost (i.e., roughly, profits) margins, labor and non-labor input prices, namely:
Source: ENPApril 2023.
Former Fed Governor Brainard as well as Paul Krugman have commented on this price-price spiral (although I think the latter is a little less definitive as to whether he thinks this constitutes the majority of the inflationary impulse).
While the decomposition is interesting (it is a decomposition after all), I’m not sure Bivens’ argument that it’s not about demand pressures (i.e. overheated). While in very early NK models the elasticity of demand for differentiated goods is constant over the business cycle, so the price markup is constant, more recent work (e.g., Nakardo and Ramey, JMCB 2020) notes that depending on the type of shock, the profit margin can be procyclical.
It might be useful to think about price changes in the presence of viscosity. When inflation is rapid, deviations from the optimal price at any given time between price revisions will be greater, inducing a greater loss of profit. The assumption of faster expected inflation in the current episode than in previous periods then implies that firms are redefining prices faster and in larger increments. From this perspective, I would expect a larger mechanically defined contribution from earnings, particularly if companies overestimate inflation.
The previous argument is based on a Calvo pricing view. If price changes are staggered, another interpretation is that the strategic complement that slows price adjustment during periods of low inflation would be dampened as firms reach consensus on a faster rate of inflation.
These are not rigorous (ie general equilibrium) arguments for a high profit margin; these are just ways of saying that we are not sure that the high profit margins are not due to higher aggregate demand.
How fast did corporate CEOs expect inflation to be? Coibion and Gorodnichenko provide the answer in their Survey of Firm Inflation Expectations.
Figure 2: Actual CPI inflation, y/y (black), New York Fed consumer inflation expectations for the year ahead (red), Atlanta Fed unit cost inflation for the year (green) and firm inflation expectations for the year ahead (sky blue squares), all in %. The NBER has defined peak-to-trough recession dates as shaded. Source: BLS, New York Fed, Atlanta Fed, Coibion-GorodnichenkoNBER and author’s calculations.
The last observation we have for firm expectations is the January 2023 forecast for January 2024 year-on-year inflation. Interestingly, in the November and January surveys, expected corporate inflation exceeded expectations consumers (which, in turn, exceeded those of economists). In other words, companies were expecting inflation to be fast (but not as fast as the actual results, until the April figure). Note that the Atlanta Fed’s measure of expected unit cost inflation is not from the same sample as that of business inflation, so a direct comparison is not possible.
None of the above means that the profits currently enjoyed by companies are “good”. Instead, I would simply say that it is not clear that one can infer from profit margins whether demand shocks drove the outcome or not.