Featured Research

The 43-day U.S. government shutdown has materially degraded the quality of October’s inflation data. With BLS field operations suspended for the duration of the collection window, almost no primary price data were gathered. Any October CPI release will therefore rely heavily on imputation rather than observed prices, and the White House has already suggested that other October reports, such as employment, may never be released in full. The CPI dataset faces the same limitations.

October CPI will be a number, but not a reliable signal. For markets, the question is not whether the print is high or low, but how much weight it deserves. Turnleaf maintains a medium-term view of inflation hovering around 3% YoY through the next twelve months, but in the near term the data degradation shifts the analytical burden to evaluating the quality of the October print and the implications for November, which will be the first clean datapoint after the shutdown (Figure 1-PAID). Hereafter we present some scenarios:

October in line with expectations  (2.9-3.1%YoY on both headline and core)  The market should treat the result as low-information. The Federal Reserve is likely to do the same, effectively discounting the release and placing greater emphasis on the November and December prints before reassessing the policy path. Breakevens may firm slightly, less because of any perceived change in inflation dynamics and more because uncertainty around the data tends to widen risk premia. Real yields and nominal Treasury yields may drift modestly higher, reflecting the absence of a downside surprise rather than the presence of new information. In this case the narrative remains stable, and the shape of November becomes the dominant print.

October above expectations  (>3.1%YoY) — Part of the strength could reflect genuine price pressure, particularly from tariff pass-through (e.g. removal of $800 de minimus, reignited trade tensions with China) or stickier components of core services, but part of it could also arise mechanically from imputation and a limited survey base. Nevertheless, markets cannot ignore a firmer print. The risk of renewed inflation momentum would increase and the Federal Reserve would likely adopt a more cautious tone, delaying any discussions of policy easing. Breakevens would likely widen, TIPS would find support from hedging demand, and nominal yields would move higher on both the inflation and real-rate components. Forecasting November in this environment requires a dual view that includes one path where October strength is genuine and inflation remains firm, and another where the number reflects measurement noise and reverses as data quality normalises.

Softer-than-expected October print (< 2.9%YoY)— should be approached with caution. Missing data tends to bias inflation downward, particularly in rent and services components, where non-response forces heavier reliance on carry-forward imputations that understate month-to-month price changes. The Federal Reserve would acknowledge the softer reading but is unlikely to adjust its stance without confirmation from subsequent releases. Breakevens could narrow modestly, and nominal yields could drift lower if markets lean toward a disinflation narrative, though any such move is vulnerable to reversal once more reliable data become available. Analysts would need to incorporate the possibility of a rebound in November if October softness proves to be a data-collection artifact.

Underlying all three scenarios is the issue of imputation. When the BLS cannot collect prices, it fills the gaps by substituting price changes from similar items within the same area, widening the geographic scope if necessary, or simply carrying forward the previous month’s price change when no comparable data exist. Housing components are imputed differently, using rent-class and unit-type adjustments. These methods preserve continuity in the index but degrade the signal when imputation rates rise sharply, as they did during the COVID-19 pandemic and are likely to again following this shutdown. Backfilling may correct some of these values once collection resumes, but the process is incomplete and slow, and many imputed values are never replaced. As a result, a high-imputation CPI print, even one close to expectations, cannot be treated as a clean read on the underlying inflation trend.

The overarching conclusion is that October CPI is an impaired release with limited policy relevance. Markets should avoid over-interpreting the headline number and instead position around the likelihood that November provides the first credible signal on inflation momentum. Measurement uncertainty is the dominant feature of the near-term landscape, and pricing should reflect that reality more than the level of the October print itself.