Thread by @prometheusmacro

Dissecting CPI 🧵👇

1. While over 300 line items drive CPI that we have visibility into, we can condense these measures into four broad categories that account for the bulk of the variations: food, energy, transportation, and shelter. We show this composition below:

2. Both economically and statistically, these categories explain about 85% of the monthly variation in CPI. Therefore, we think it makes sense to approach our dissection of CPI by evaluating these areas.

3. We believe that food & energy prices can continue to contribute to a softening in CPI, but the swing factor will likely be transportation inflation. So far this cycle, transportation has been a net support to the disinflation we expected over the year.

4. As we can see in the previous visual, transportation has accounted for approximately 1.5% of headline disinflation in CPI from the peak. Drilling down further, most of these declines have come from transportation goods, i.e., primarily car & truck prices.

5. As shown above, transportation goods have been a deflationary force in CPI over the last year. This
disinflation was driven primarily by a decline in used car prices, while new carprices continue to be modestly additive.

6. This decline in used car prices is likely behind us, as used car demand was primarily a function of auto shortages and stimulus demand. Before we look ahead, we drill down into the other side of transportation inflation, i.e., services.

7. As we can see above, transportation services are seeing broad based inflationary pressures. These inflationary pressures are driven primarily by motor vehicle leases and insurance, which are linked to higher new car prices and interest rates.

8. Due to their contractual nature, transportation services tend to show more trend persistence in their monthly changes than transportation goods.

9. Furthermore, deflationary prints are far more uncommon in transportation services, with 75% of months over the last ten years showing positive prints.

10. Weighing these factors, if we are to see further disinflationary
pressure on CPI from transportation, it will likely need to come from further declines in new and used vehicle prices, which will eventually bring down lease, rental, and insurance prices.

11. Whether this can be achieved remains in question, as COVID-19 distortion to automobile supply chains still remain in place.

12. These dynamics have created a dearth of automobile inventories for businesses, which has resulted in businesses, particularly retailers, significantly bidding up automobiles to increase their inventories. We show this below:

13. This persistent inventory demand comes alongside moderating consumer demand for automobiles, albeit from contractionary levels:

14. As we can see above, periods of contracting automobile sales have coincided with and, to some degree, facilitated recessionary disinflation.

15. Looking ahead, it will likely be the balance between a weakening of automobile credit to consumers as rates increase flow through to automobile loans and how much retail dealers continue to demand automobiles, even at increased supply.

16. Below, we show how automobile borrowing rates have accelerated, which will be positive for lease payments but negative for auto prices.

17. Considering these unique dynamics, the direction of the automotive sector remains unclear relative to typical cycles. While cyclical dynamics favor a further deterioration in automobile prices depleted business inventories can keep automobile prices elevated.

18. On balance, there will likely be some degree of stabilization in the used vehicle deflation, potentially moderating new car inflation and persistent transportation service inflation.

19. Cumulatively, these forces will likely be modestly supportive of inflation well over the Federal Reserve's objectives.

20. On the other side of these inflationary pressures from transportation will be food and energy, which will likely follow the tone set by raw commodity prices, which we show below:

21. As we can see above, food and energy prices tend to move in lockstep with the trend in raw commodity prices, with a modest lag. Thus, declines in energy and agricultural prices will likely flow through to their respective components in CPI over the next one to three months.

22. Finally, we turn to shelter, which remains the largest contributor to CPI inflation. As we have detailed many times over the last year, CPI
lags behind real-time measures of home prices due to a computational smoothing process applied by the BLS.

23. This feature creates a somewhat predictable path for the shelter component of CPI. Below, we show how our forecasts imply a softening in the shelter component of inflation:

24. While our forecasts imply housing disinflation is likely ahead of us, the magnitude of the disinflation will likely pick up only as we enter 2024, which will likely keep shelter's contribution to inflation alone at more than 2%.

25. While the Fed does not target CPI, this is still a significant inflation boost across measures.

26. Considering persistent automobile #inflation, food and energy deflation, and slow disinflation in the shelter component, inflation will likely remain well above the Fed's 2% objective, barring a #recession.