US inflation is no 1970s retro act (for now)

Key points 

• Is the surge in inflation and central bank policy responses to it seen in the 1970s, further to an oil price shock, still relevant today? 
• Our research shows that there are some parallels with events in the 1970s, but also crucial differences in how inflation is behaving this time. Not least, the impact of soaring inflation rates has had a limited impact on the labour market so far this time 
• The composition of inflation in the 2020s is starkly different from five decades ago. Goods constituted over 60% of the consumer basket in 1972, while they account for below 40% of today’s basket
• While we expect inflation to ease to below 3% by the end of the year, there remains a risk that energy prices surge again. For those looking to hedge against this risk, strategies with returns that are uncorrelated to macro variables, such as managed futures, can help to provide a useful buffer against the effects of price shocks on portfolio values

While Europeans’ living standards are being squeezed by the most in decades, which exerts considerable risk to inflation in its own right, US price hikes are slowing. With a fuel price shock and surge in inflation reminiscent of 1970s in many ways, we look at what parts of policymakers’ 1970s  playbook are useful for investors.

The path for US inflation this year is pretty clear, according to Bloomberg consensus surveys: a combination of base-effects (as the relevant earlier data falls off comparisons) and a slump-induced cooling of inflationary pressure is expected to see inflation rates drop to 3.0% at the end of 2023 from 7.1% at the end of last year. As such, peak inflation should be behind us, with the surge in energy costs and core inflation having reached their highest reading in June and September, respectively.

As inflation worries become yesterday’s news and investors contemplate what the effects of last year’s interest rate hikes will be for the US economy, we examine the latest inflation dynamic, 2020s style, and glance back to the 1970s for any hints as to what might be in store this time. 

The 1970s playbook

Last year we analysed the inflationary surges of the 1970s for comparisons with the latest inflationary shock and lessons for that may now play out (see Inflation: Learning from the 1970s). At first, we studied readily available consumer prices index (CPI) components to understand the dynamics of typical oil-price induced inflationary shocks. We found that: 

• First, booms in the energy price preceded those for goods and services inflation
• Second, interest rate hikes could accompany accelerating energy price inflation, as long as the economy did not crash into a recession
• Third, significant negative output gaps and strong rises in unemployment were needed to curb non-energy inflation

Reducing inflation complexity

We then carried out a principal components analysis, to dig deeper. Instead of using readily-available aggregates, like “energy” or “core inflation”, standardised quarterly changes were run for all available subcomponents of the consumer price index (CPI) baskets in the US – items such as “housekeeping supplies” or “motor fuels” –  through our model to extract the underlying factors behind the inflation dynamics¹.  

The above analysis allowed us to capture over 70% of the variation in all subcomponents of the CPI in the first three principal components PC1, PC2 and PC3. According to the dominating loadings for each of these components, we named them “broad cycle” (PC1), “fuel” (PC2) and “necessities” (PC3).

For this article, we add the fourth “services” principal component (PC4) to the group. While the latter did not add much explanatory power when looking at inflation in the 1970s, it is important to understand the differences between the 1970s and today.  

Our analysis of 1970s inflation suggests that the fuel component leads the cyclical one by several months in a typical oil-price induced price shock. It also confirms that the broad cyclical inflation component (think of items such as apparel, furniture or new vehicles) only eased when unemployment rates start to rise (see chart).

Spotting parallels and differences with the 1970s 

When the model that we trained on the 1970s is applied to the same CPI subcomponents from 2020 (see chart), we find parallels but also differences. Until September 2022, the parallels prevailed. The most notable was the sequence of peaks which remained the same: the cycle peak followed several months after the peak in energy inflation. 

The differences started to emerge in last year’s fourth quarter: first, neither broad unemployment rates or those of the most vulnerable parts of the labour market showed significant increases as the cyclical component peaked. Second, while price pressures in the broad cyclical factor seem to have relaxed, inflation is still simmering on price increases from the services factor (PC4).

Understanding what the differences with the 1970s mean

Understanding the reasons for the differences in factor behaviour can help to get to the bottom of the inflation dynamics. First, the change in the consumer baskets comes to mind. While all CPI subcomponents are as consistent across time as the data provider, the Bureau of Labor Statistics (BLS), can provide, their relative economic importance can change considerably. 

Goods constituted over 60% of the CPI basket in 1972. By 2021 their share was below 40%, according to BLS data². The budget freed up from goods went largely into shelter, which used to be one component of services, among others, and now makes up a third of the whole basket. 

Second, not only has the consumer basket changed, but so has the reliance of the US on fuel. This reduced pressure could lag the effect of monetary policy intervention on the economy compared to the 1970s setting. This could lead to a softening of the finding from the 1970s that links peaks in the cyclical factor to a start in the rise of the unemployment rate.  

Third, in addition to this lowered reliance, the American economy has one additional tool at its disposal that was borne out of the 1973 oil price shock: the Strategic Petroleum Reserve (SPR). The release of over 20% of the SPR between November 2021 and July 2022 is assumed to have contributed significantly to reducing the strain of rising fuel prices on producers³. At the end of January 2023, inventory levels of the SPR were over 60% off their peak and back to levels that were last observed in 1983.  

Keep the script in reach

We believe that the remaining inflationary pressure that is seen in the services component will fade as the US labour market tightens this year (see chart). However, while it may have exhausted its usefulness for now, investors might not want to put the 1970s script too far out of reach.

The 1973 oil shock, and its economic effects, etched itself into the memories of consumers, producers and market participants. By contrast, in 1979 it did not take a large price increase for oil to send another shockwave around the globe. While we expect US inflation rates to fall below 3.0% by the end of 2023, another surge in energy prices, whether caused by geopolitical actions or otherwise, remains a side-scenario to consider. 

When inflation abounds, there are few places to hide

Investors may get a chance to fortify their portfolio against further inflation surprises this year. Outside of commodities, there are usually very few placed to hide from inflationary shocks in the absence of strong accompanying growth. 

In such an environment, strategies with returns that are uncorrelated to macro variables can provide a buffer. Managed futures, for example, thrive on strong trends, no matter where macro data or markets are headed. This could provide a useful hedge against a repeated shock to energy prices.