“What goes up must (eventually) come down.”
That, in general terms, is the Fed’s argument why the current inflation surge should be transitory as prices normalize (somehow that logic does not apply to markets, but that’s where the Fed’s balance sheet comes in).
But is that indeed the case? Do higher prices go down on their own?
As we discussed last week in “How Average Inflation Hit A Red Hot 2.4%“, a large portion of the recent rise in inflation is being aided by re-opening dynamics and supply chain bottlenecks — drivers which Goldman expects are likely to be transitory.
However, in a follow up note published overnight by Goldman, the bank’s economists highlight how the expected normalization of prices later this year and into next is likely to be uneven across categories. The bank clarifies that in its year end-2022 core PCE inflation forecast of 2.0% it assumes that a third of the bottleneck-induced inflection in core goods prices reverses at that horizon. Specifically, the bank expects price reversion to be larger in categories of relatively larger goods like used cars and car rentals, where pandemic and policy effects have driven what Goldman believes is a “temporary” wedge between supply and demand. On the other hand, whereas we have already seen a sharp slowdown in some price categories such as lumber, a quick look at container shipping rates shows that this clearest indicator of broken supply chains has yet to peak, let along normalize and drift lower.
In any case, going back to Goldman’s note, which after doing a deep dig into supply chain disruptions and reopening effects, focuses on exploring historical divergences in category-level inflation from a statistical perspective. As shown in the chart below, used car prices rose 38% year-on-year in May, and they are now well above statistically derived estimates of the underlying trend.
According to Goldman, full normalization to trend in this category alone would lower year-on-year core PCE inflation by 0.35pp—or even more if recent price outliers have introduced an upward bias to the filter.
To analyze the tendency for category-level prices to converge back to their medium-term trends, Goldman incorporated a “deviation from trend” variable into its bottom-up core PCE models. As shown in the grey bars below, there is evidence of normalization in the majority of categories, including many of today’s inflation outliers (the left 7 categories). However, with the exception of car rentals, and as should be blatantly obvious to anyone, Goldman “finds” that prices on average do not fully converge back to trend.
Even with the benefit of hindsight—and therefore ignoring price inflections that in fact reflected a permanent change in trend— we find that just over half of the price deviations close over the next year (58% on average across all core PCE categories and 55% across the seven currently elevated categories.
The second set of columns reflects the same test using real-time price trends. They show that the predictive power of price deviations is much weaker when they are measured in real time. In other words, while categories prone to temporary supply-demand imbalances such as car rentals and used cars continue to show significant normalization tendencies, the mean-reversion coefficient for the majority of the other categories loses statistical and economic significance.
The last chart looks at the divergence episodes since 1997 for the seven categories that are particularly elevated today. Here, a more optimistic Goldman argues that the historic imbalances caused by the 2020 lockdowns and demand swings suggest that today’s divergences could resolve more quickly or fully, and as a result, the bank expects “price reversion to be relatively larger in categories like used cars, where pandemic and policy effects clearly drove a temporary wedge between supply and demand.”
Maybe: it is true that according to the chart, prices in that category fully normalized two years after the Cash for Clunkers program generated a similar price bubble during the financial crisis. But in contrast, the sustained increase in appliance prices during the last housing boom suggests scope for prices in that category to remain elevated long after after the chip shortage ends.
Taken together, Goldman concedes that its results suggest that we should not get carried away with forecasts of rapid or complete price normalization, because it is often difficult to identify bubbles without the benefit of hindsight, and because imbalances often require more than a few quarters to resolve. Which, of course, is precisely why the Fed will continue to argue that soaring prices are “transitory”… while the artificial, Fed-induced all time highs in the stock market will keep rising into perpetuity.