The NY Fed has published an article titled “The Global Supply Side of Inflationary Pressures” by O. Akinci and coauthors.
Please find below our review, our comments and the possible implications for the Fed staff.
To keep in mind
Domestic inflation appears to be influenced by global factors, such as the global supply-chain and fluctuations in commodities prices. However, most of the passthrough is from global factors to domestic energy (PPI and CPI) items. Net of the energy sectors, the passthrough is limited (if any) and the empirical evidence remains controversial. For this reason, in our view the Fed staff continues to be blinded when forecasting the impact of supply-chain disruptions on US consumers prices.
What the paper does
The authors postulate that global supply-side forces have an influence on domestic inflation. The authors consider two global supply-side measures: (i) the Global Supply Chain Pressure Index (GSCPI) which measures the disruption in global supply chains recently developed by NY Fed economists, and (ii) a measure that explains energy fluctuations (essentially, proxies for oil demand and supply). Based on this identification strategy, the authors estimate the correlations between PPI/CPI inflation for the United States (and the Euro area) and the global factors. The regressions are run using monthly data stopping the estimation before the pandemic. Then, based on the estimated coefficients, the authors estimate the predicted evolution of US PPI (and CPI), conditional on the evolution of the global supply-side proxies.
Results
The authors claim that global supply factors are very strongly associated with producer price index (PPI) inflation across countries, as well as with consumer price index (CPI) goods inflation. Specifically, the authors stress three results. First, changes in the global supply chain pressure index (GSCPI) had a meaningful impact on PPI and goods CPI inflation both for the U.S. and the euro area. Second, oil supply factors play a larger role in driving the overall fit of the regressions compared to the other two explanatory variables. Third, the set of global variables appear to be unrelated to movements in services CPI inflation.
As an example of the authors’ results, we report their main findings for US CPI goods inflation. The chart below shows the evolution of US CPI goods inflation (the light blue line), together with the predicted evolution from three versions of the authors’ model: a version that includes only the GSCPI (the red line), and two versions including GSCPI and oil supply/demand proxies. According to the authors, the chart illustrates that global supply factors (the dark blue line) have played a substantial role in driving observed US CPI goods inflation (the light blue line), even though they do not fully capture all observed CPI goods inflation moves.
Comment
While it is reasonable to think that -at least to some extent- the evolution of domestic prices is influenced by global conditions, we caution against the authors’ results for three reasons. First, as the authors explicitly admit, there is no pass-through from global supply-side conditions to services prices. Put it differently, two thirds of the CPI do not seem to respond to global factors (indeed, they are mainly non-tradable). Second, the authors’ analysis considers total goods CPI, meaning that it includes energy products. Therefore, the authors results appear to be driven by the energy component (which is typically irrelevant for monetary policy). Third, even abstracting from oil prices, the correlation/causality between the authors’ GSCPI (the red line in the figure above) and the US goods prices excluding energy items (not shown) is unclear (if any).
Implications for the Fed staff
In our view, the Fed staff medium-term forecast continues to be based on the calibrated fundamentals. Therefore, any effect from global supply-side issues is not folded in the forecast (in case, it would be parsed as “other factors” –that is, factors unrelated to the economic fundamentals that should not influence monetary policy). Not only, but in our view the staff continues to think that supply-side issues will solve quickly in the second half of the year. However, reality is different: nobody seems to have a good model/forecast of the global supply chain and most forecasters underestimated the supply chain issues last year. The empirical evidence (as shown above) unfortunately does not help solving the puzzle. For this reason, we continue to believe that the Fed staff forecast is exposed to upward surprises going forward.