U.S. CPI reached a peak in June 2022, exceeding 9% YoY, but then started normalizing downward. It increased by 6.5% YoY in December and is likely to fall further in the coming months. In absence of external shocks, proxies suggest U.S. CPI growth could converge towards 3% YoY (or less) by June 2023.
Market rents will keep retracing in the coming months. Household formation and rental demand are slowing in response to macro conditions. As a matter of fact, Apartment market conditions weakened in the National Multifamily Housing Council’s (NMHC) Quarterly Survey of Apartment Market Conditions for January 2023. The Market Tightness Index came in at 14 this quarter—well below the breakeven level (50)—indicating looser market conditions for the second consecutive quarter.
In the meantime, housing supply will gain traction, with a peak expected around 4Q23. Latest data showed housing under construction reached a new record high in December 2022.
The collision of these factors will probably result in a contraction of market rents on a YoY basis at some point in 2H23. In the past, the CPI’s measure of rents for homeowners has typically lagged other measures (9 to 12 months looks appropriate for this cycle) because of data construction. Recently, Brian Adams, Lara Loewenstein, Hugh Montag, and Randal J. Verbrugge published an article “Disentangling Rent Index Differences: Data, Methods, and Scope” where they created the New Tenant Repeat Rent (NTRR) Index and All Tenant Repeat Rent (ATRR) Index. According to Joey Politano, “using the same underlying BLS microdata that composes the housing component of the CPI, the NTRR uses information on lease turnover to track rent growth in units that change tenants. The ATRR covers all housing units but attributes rent changes to when they happened, as opposed to the official CPI data which tracks price changes when units are surveyed“. In this context, the index already points to a significant slowdown in rents’ growth (see chart below).
Proxies also point to a downward normalization of food prices’ growth by the end of 2023. Agricultural commodity prices — which are usually leading CPI food prices by 12 months — retraced over the past few months.
In addition, fertilizer prices — another leading indicator — are also down ~50% from the peak reached in March 2022.
Since a few months, supply chain disruptions have eased significantly. Several indexes point to a partial normalization at the global level and even a return close to normal in the U.S. (at least in the manufacturing sector). As an example, the ISM Manufacturing survey’s measure of supplier deliveries fell to 45.1 in December, from 47.2 prior. It declined below the 50 threshold in October for the first time since February 2016. A reading below 50 indicates faster deliveries to factories.
In the meantime, wholesalers of finished goods have faced an extremely large rebound in inventories due to weak real consumption and will probably implement discounts. As a result, core goods’ prices are expected to contract on a YoY basis at some point in 2023.
Separately, there are already signs used cars and trucks’ prices could fall further on a YoY basis amid significant supply chain improvement for new vehicles. In January, the mid-month Manheim Used — a leading indicator — was down 13.7% from the full month of January 2022.
On a YoY basis, gasoline prices are expected to become a drag on CPI headline from March 2023 if prices remain unchanged (I know it’s a very strong assumption subjected to high uncertainty). Base effect is expected to reverse significantly (from positive to negative) and hit a maximum in June 2023.
There are also signs CPI Services (less rent of shelter) growth could ease slowly on a YoY basis. This component is more sensitive to the labor market. Since a few months, the latter has weakened while wages’ growth peaked.
Each year, seasonal adjustment factors are recalculated to reflect price movements from the just-completed calendar year. The next update will mark the beginning of the transition to yearly weights. “Starting with January 2023 data, the BLS plans to update weights annually for the Consumer Price Index based on a single calendar year of data, using consumer expenditure data from 2021. This reflects a change from prior practice of updating weights biennially using two years of expenditure data.“
Several analysts already flagged downside risks linked to these changes stating that it will give more weights to the sectors that experienced large increases in 2021 and that are supposed to weaken in 2023.
In this context, market participants already expect CPI YoY to slow markedly by June 2023 with swaps pointing to a level below 3% (and even closer to 2%).
Even if U.S. CPI YoY should slow in the coming months, its speed will also depend on external factors that could turn into upside risks. First of all, it’s very unlikely that gasoline prices remain unchanged. They could move in either directions but China reopening could be supportive especially in a context where officials look determined to boost growth using fiscal and monetary policies. In addition, we can’t exclude that OPEC+ will remain conservative regarding its supply (as it did last year). Weather conditions could also affect electricity and gas prices in a context where geopolitical tensions remain elevated. Lastly, although the labour market has weakened, a full normalization in 1H23 is excluded, especially given the current level of initial jobless claims (below 200K).
Whatever, I think the Bloomberg consensus of 3.9% for 2023 CPI is a bit too high and is likely to be revised downward soon. Yet, my guess is Fed policymakers will maintain a “confirmation bias” in the short term and will refrain from sending any sign of an upcoming accommodative policy starting in July 2023 (as expected by market participants). As a reminder, policymakers are still focusing on CPI services (less rent of shelter), a component which is expected to ease slightly compared to others.