• Core CPI landed at 0.2 again with mixed details
  • Breadth of price pressures is nevertheless still rising
  • UofM sentiment plunged, inflation expectations moved up…
  • ...but unemployment expectations soared…
  • …and they have an uncanny knack at forecasting this
 
  • US CPI / core CPI, m/m %, SA, March:
  • Actual: 0.9 / 0.2
  • Scotia: 1.0 / 0.3
  • Consensus: 0.9 / 0.3
  • Prior: 0.3 / 0.2

US core cpi was up by 0.2% m/m SA without rounding and hence a tick beneath most expectations while matching the prior month’s reading. US headline CPI was up by 0.9% and on the screws. It’s too early to get any transmission into core so personally I would fade that reading from a policy standpoint.

Annualized m/m core CPI inflation is tracking close to 2% (chart 1). 

Chart 1: US Core CPI Inflation

Rising breadth will concern the hawks on the FOMC (chart 2). 

Chart 2: US Inflation Breadth

Core goods inflation picked up a bit over the past two months while core services inflation has been ebbing (charts 3, 4).

Chart 3: US Goods Inflation; Chart 4: US CPI Core Services Ex-Housing

The SA factor for core CPI was unusually high—in fact, a historic high comparing like months of March over time (chart 5). Because it’s driven by a recency bias skewed to the pandemic-onward period, this means it probably overstated core inflation.

Chart 5: Comparing US Core CPI SA Factors for All Months of March

All that said, the BLS is continuing to use a very high proportion of proxy methods for gathering prices by drawing upon substitute products and substitute markets where their resources and sampling methodologies are unable to get data (chart 6). At 39%, a large portion of the CPI basket is suffering from low data quality.

Chart 6: BLS Use of Alternate Estimation Methodology in US CPI

The next few pages offer plenty of charts to consider but we’ll keep this one brief. March CPI won’t influence a thing at the Fed and barely drew any market reaction. It’s a first stab at the war’s effects, it’s too early for core transmission to occur, and there remains uncertainty around duration and magnitudes of the shock and how it impacts inflation and jobs.

Chart 9: US Food Prices; Chart 10: US CPI: Gasoline; Chart 11: New vs Used Vehicle Inflation; Chart 12: US Motor Vehicle Insurance
Chart 13: Housing Inflation; Chart 14: US Rent Inflation; Chart 15:US Airfare; Chart 16: US CPI: Household Furnishings
Chart 17: US Apparel
Chart 18: March Changes in US Headline CPI Categories; Chart 19: March Weighted Contributions to Monthly Change in US Headline CPI ex Gasoline
Chart 20: March Weighted Contributions to the 12-Month Change in US Headline CPI; Chart 21: March 12-Month Changes in US Headline CPI Categories

Consumer Sentiment Plunges

University of Michigan consumer sentiment fell sharply to 47.6 in March from 53.5 on a combination of current and expected conditions. That’s worse than expected by consensus.

Consumers’ inflation expectations moved higher (chart 7).

Chart 7: U. Mich. Measures of Inflation Expectations

What consumers expect to happen to unemployment in future remains concerning (chart 8). It’s soft data, but has never sent a false signal. Consumers are closer to their employers’ plans, the water cooler and Teams talk etc relative to markets and forecasters. I hope they’re wrong this time. Otherwise, what they are signalling is one of the few recession signals that points to massive job losses in which case nuts to inflation, the Fed may pivot more aggressively than our –50bps forecast spread over Q4/Q1. Consumers may be one part antsy about the cycle, the composition of policy and market risks, and AI’s coming effects.

Chart 8: U.S. Consumer Anxiety Consistently Predicts Unemployment Trends
Table: US Inflation Component Breakdown
Table: US Inflation Component Breakdown
Table: US Inflation Component Breakdown