Core PCE to End 2023 Above 4.0%

Summary

• Due to lower imports, goods inflation has only limited downside over the remainder of 2023. Longer term, due to protectionism, reshoring, and industrial policy, goods price inflation is likely to remain positive.
• Historically, very low vacancies suggest limited downside to housing inflation.
• Solid growth momentum suggests continued tight labour market and limited downside to wage growth and services inflation. 
• Solid growth momentum in turn reflects loose fiscal and monetary policies.
• The recent decline in energy prices masks strong underlying inflation momentum.

Market Implications

• The market continues to underprice the Federal Reserve (Fed).

Inflation Is Not Trending Down

Central bankers generally view core inflation as a better indicator of trend than headline inflation. That is because headline is influenced by food and energy prices, which are believed to be more volatile than other categories. 

In reality, as noted by the Saint Louis Fed, the volatility of inflation components, and their predictive power, change over time. For instance, the Personal Consumption Expenditure (PCE) component with the highest SNR (signal to noise ratio, the correlation between YoY headline inflation and the 12-month lagged component divided by the component standard deviation) is healthcare. The Consumer Price Index (CPI) component with the highest SNR is food, followed by healthcare.

However, SNRs are based on past data and are likely to prove unreliable when the long-term drivers of inflation are changing, which I believe is happening. I therefore prefer measures such as trimmed mean or median price inflation. These filter out the largest contemporaneous price changes (see here for an academic discussion of median price vs core inflation).

Based on median price inflation, there has been no obvious downtrend (Chart 1). In this note I use bottom-up and top-down approaches to explain why that is and why I expect no significant downtrend over the remainder of 2023. 

Protectionism to Keep Goods Prices Inflation Positive

I start my bottom-up analysis by discussing core goods prices. Domestic core goods inflation reflects both the costs of producing goods domestically (i.e., the PPI) as well as the cost of imported consumer goods. 

Historically, the consumer goods PPI has grown faster than the price of imports that, before the pandemic, was flat or falling (Chart 2). Partly as a result, the CPI has been growing more slowly than the PPI.  

During the pandemic, imported goods prices rose, and the share of imports in consumption fell. This contributed to the acceleration of consumer goods price inflation.

Going forward, protectionism and reshoring suggest the share of imports in consumption is unlikely to recover to pre-pandemic levels. As a result, the CPI could become more influenced by the PPI and less influenced by import prices. In addition, protectionism and industrial policy suggest slower productivity gains and therefore faster PPI inflation.

A return to pre-pandemic goods price deflation therefore seems unlikely (Chart 3). Overall, and accounting for the recovery in used car prices, I expect core goods PCE to remain above 2% for the remainder of 2023.

Low Vacancies Imply Limited Downside to Rent Inflation

The second component of my bottom-up discussion is housing costs. Since the pandemic, a plethora of alternative measures of rents have received strong investor attention. These have several limitations, for instance sample representativity, construction, qualitative adjustments, and short time series. For a detailed discussion, see the BLS’s paper, ‘Disentangling Rent Index differences: Data, Methods, and Scope’. 

In my view, vacancies are a simpler and more reliable indicator of rent dynamics. Vacancies do a reasonable job of explaining OER (Chart 4):

• Vacancies of owned homes did not rise during the pandemic, are currently the lowest since the 1970s, and are still falling.
• Vacancies of rented homes seem to have risen in Q1 but remain low by historical standards. In addition, only a third of Americans rent their home.
• Current OER inflation (about 8% YoY) was last seen in 1982, when vacancies were comparable to current levels. Therefore, OER inflation is not out of step with market ‘fundamentals.’

A marked economic slowdown would be required for vacancies to rise. Vacancies are driven by a combination of supply and demand factors. On Chart 5, I have used as a demand proxy real consumption of housing in the personal spending data. And as a supply proxy, I have used the BEA estimate of the net real housing stock divided by the number of households. 

Housing demand contracted in the aftermath of the GFC due to a decline in disposable income as well as a flattening out of household formation. It only recovered from 2014 onwards and did not slow much during the pandemic.

The net real housing stock rose relative to the number of households up to the GFC and subsequently fell because of a collapse in residential investment. Underinvestment in real estate could explain the post-GFC combination of falling vacancies and falling demand. 

Over the remainder of the year, the housing stock is pretty much fixed and higher vacancies would have to come from lower demand for housing. In turn, this would require a marked slowdown in real household income, which seems unlikely based on current growth trends (see below). Overall, I see limited downside to the current 8% YoY OER inflation.

Wage Growth to Remain High

The third component of my bottom-up discussion is core services excluding housing. As Fed Chair Jerome Powell stressed, this inflation component is driven by wages, and a slowdown would therefore require lower wage inflation. However, the slowdown in wage inflation has been modest so far (Chart 6). 

A further decline in this component would require an easing of the labour market and therefore GDP growth significantly below trend. Bottom-up indicators, however, do not signal slower growth:

• The Atlanta Fed GDP nowcast is hovering around 2% (i.e., trend growth).
• Labour market indicators generally remain stronger than before the pandemic (May NFP Review).
• Forward-looking spending (e.g., consumer durables and capex) is rising (Chart 7).
• Growth in real household income excluding transfers is accelerating, which historically is a good indicator of expansion, possibly because US GDP is 70% consumption (Chart 8).

Overall, bottom-up indicators are not signalling an imminent and meaningful slowdown in inflation. 

Neither do top-down macro indicators.

Policy Mix Remains Loose

The resilient growth momentum discussed above largely reflects loose fiscal and monetary policies. On a 12-month basis, the May 2023 deficit was 8% of GDP, despite very low unemployment (Chart 9). Fiscal consolidation implemented in FY2022 did not last.

Similarly, monetary policy remains loose, with the Federal Funds Rate (FFR) still well below the Taylor rule (Chart 10). This is shown in, for instance, the recovery in residential real estate prices and spending, when residential real estate is the most interest rate sensitive demand component!

Furthermore, there is limited evidence of a tightening of credit conditions beyond what is usually associated with a Fed tightening cycle. Bank credit has been flat since early April. And the decline in deposits, which reflects largely QT, is offset by a decline in banks’ cash and bond holdings rather than by a decline in credit (Chart 11). In any event, at this stage in the recovery, growth is not dependent on credit.

Altogether, this analysis suggests core PCE could remain above 4% by end-2023. Yet, if anything, the loose policy settings appear inconsistent with inflation stability since 2022. This apparent inconsistency could well reflect the ongoing negative energy price shock.

Negative Energy Price Shock Hiding Strong Inflation Momentum

In recent research the BIS has shown that inflation dynamics are very different depending on whether inflation is low or high. In a low inflation regime, economic agents tend to practice ‘rational inattention,’ (i.e., ignore relative price shocks because overall inflation is low). In this regime, the components of CPI tend to be uncorrelated, wages and prices tend to be uncorrelated, and inflation shocks tend to be self-correcting.

By contrast, in a high inflation regime, economic agents’ behaviour is influenced by their expectations of inflation. In this regime, agents’ expectations of inflation tend to converge, the components of the price index become strongly correlated, and a feedback loop between wages and prices develops. Relative price shocks, such as energy price shocks, tend to spill over across the whole price structure (i.e., energy and core inflation become correlated).

This model of inflation fits the data well. Core and energy PCE were strongly correlated in the 1970s and ‘80s when inflation was high (Chart 12). The correlation disappeared in the 1990s and up to the pandemic, when inflation was low.

The correlation reappeared with the pandemic, when the energy shock of 2020-22 spilled over into core inflation. But core inflation lowered when the energy price shock reversed from mid-2022.

The decline in energy prices has allowed nominal wage growth to slow while real wage growth has accelerated. It has also helped stabilize inflation expectations.

The risk is that because of the strong growth momentum and loose policy settings, a renewed increase in energy prices could lift core inflation onto a higher path.

Market Consequences

Since the Silicon Valley Bank bankruptcy, 10-year BEs have remained in a 2.2% to 2.3% range, and 2-year BEs have fallen to about 2.1% from previously 2.5%. The above analysis implies that markets are too optimistic on medium-term inflation prospects, especially as the Fed appears much more focused on employment than on the inflation mandate .

Eventually the Fed is likely to hike much closer to the Taylor rule than current plans, but this is unlikely to happen until core inflation has risen significantly.

At the same time, in the short term, the strong underlying inflation dynamics suggest the Fed is highly unlikely to cut rates in 2023 and instead is likely to implement the two additional 2023 hikes of the June SEP. By contrast, the market is pricing only one hike up to November and a cut afterwards.

Dominique Dwor-Frecaut is a macro strategist based in Southern California. She has worked on EM and DMs at hedge funds, on the sell side, the NY Fed, the IMF, and the World Bank. She joined Macro Hive soon after its inception.

Photo Credit: depositphotos.com

(The commentary contained in the above article does not constitute an offer or a solicitation, or a recommendation to implement or liquidate an investment or to carry out any other transaction. It should not be used as a basis for any investment decision or other decision. Any investment decision should be based on appropriate professional advice specific to your needs.)