Moving Past Peak: Time to Reallocate?

 

The results shown in Exhibit 2 highlight one of the main challenges with inflation analysis. When applying our criteria we have identified seven (7) distinct episodes, of which only six (6) have fully completed. We must therefore be mindful of our “confidence intervals” given that we are analyzing so few observations. Another challenge we face is the difficulty in isolating the impact of inflation from the broader economy, and by extension, the market response. Not only are growth and inflation closely linked, but their relationship is also circular – inflation affects growth, and growth affects inflation. Therefore, the extent to which the market is reacting to inflation at any point in time is difficult to ascertain.  While remaining cognizant of these challenges we proceed to the next step in our analysis.

Establishing the baseline

In order to evaluate the significance of reaching peak inflation we first need a baseline for comparison. Exhibit 3 shows the annualized, median^[6] active^[7] factor returns for the Russell 1000 Index for the six full inflationary episodes over our time horizon (refer to Exhibit A of the appendix for full details and MSCI World Index data).

The results reported in Exhibit 3 are consistent with recent factor performance, as factors have outperformed the market on average during inflationary regimes.  Perhaps not surprisingly value has performed the best historically, consistent with the notion that value stocks do well when future discount rates rise. Rising inflation often leads to rising interest rates as central banks attempt to reign in prices. PE compression tends to be greater in growth stocks than value stocks in a rising interest rate environment.^[8] The min and max values imply significant dispersion in factor returns, as the range of active returns for size, value, and momentum, is greater than the range of total returns for the market itself (the Russell 1000 Index). This indicates heightened cross-sectional variation during inflationary episodes, as certain companies (industries) are better positioned than others to adjust to inflationary pressures.

"Inflationary episodes are accompanied by high cross-sectional volatility."

The impact of inflation normalization

Having established a baseline, we can now measure the marginal impact of moving past peak inflation.  Exhibit 4 repeats the analysis shown in Exhibit 3, including only the period from peak inflation to the end of the inflationary episode (when the inflation trend falls below 3%).  The median active factor returns of the full inflationary episode are included for comparison (refer to Exhibit B of the appendix for full details and MSCI World Index data).

When comparing the median active returns after peak inflation to the full episode, we note that size now joins value among the top-performing factors.  This result supports the view that smaller companies have less pricing power than larger firms, and therefore benefit more from falling inflation. The fact that the median value premium remains high is somewhat counterintuitive given the strength of the current relationship between value and interest rates^[9]. As inflation begins to fall one may reasonably assume that investor expectations of future interest rates are also falling. Declining interest rates ought to disproportionately favor growth stocks whose present value of future earnings is quite sensitive to future discount rates. One potential explanation for value’s sustained performance is the simple recognition that investors revise their expectations of growth as well as interest rates. Conventional wisdom alleges that high inflation begets slowing growth, which can lead to reductions in forward earnings.  These revisions may more than offset the impact of falling discount rates, leading to continued multiples compression for the growth cohort. Conversely, in some cases, the adverse impact of rising inflation may have been overestimated. Only after inflation begins to normalize are the prospects for these value firms adjusted accordingly^[10]. Of course, these explanations conform to the classic behavioral interpretation of the value premium which asserts that investors tend to extrapolate information too far, becoming too optimistic (or pessimistic) in their projections. This rationale is consistent with the data presented, as the biases responsible for the value premium are agnostic to the level or direction of interest rates. Relative to our baseline, the range of returns reported in Exhibit 4 has increased considerably, suggesting that the transition to inflation normalization is both difficult to characterize and warrants caution from tactical investors. The minimum active return for momentum has decreased appreciably, consistent with the tendency for momentum to underperform during changing market regimes.

We next evaluate the inflation transition period to contrast the impact of a “hard landing” to that of a “soft landing.”^[11] We begin our analysis by focusing first on aggregate equity market performance in the context of growth.  For the purpose of our analysis we define the “transition period” to be the 1-year period spanning the last 6 months of the inflationary episode and the first 6 months after the inflationary episode (i.e., the first 6 months of the inflation trend remaining below 3%).  Exhibit 5 reports the following data for each inflationary episode over the analysis horizon: GDP growth rates leading into and during the transition period, the return of the Russell 1000 Index, and recession start and end dates (if applicable).

The data reported in Exhibit 5 shows that only 2 of the 6 episodes avoided recession. The inflationary transitions of 1984 and 2005 both experienced slowing growth, but the economic expansion was sustained. The equity markets responded favorably to these “soft landings”, returning 22% and 9%, respectively.  The other four regimes were each accompanied by a recession, but the timing varied. In 1979 and 1990, the economy was coming out of recession as inflation normalized, whereas in 2000 and 2007 the economy was falling into recession during the transition. This highlights the circular relationship between growth and inflation that was referenced earlier. In the “tech wreck” and global financial crisis, inflation was falling in response to collapsing growth. In 1979 and 1990, growth was rising in response to falling inflation. For tactical investors, these are important distinctions, as evidenced by the extreme range of equity market returns associated with these “hard landings.” Most investors equate the current inflation regime to the 1980s, and the consensus appears to be forming that a recession is looming. Should these assumptions hold, Exhibit 5 suggests that an economic recovery may precede inflation normalization. Exhibit 6 reports factor performance for the inflation transition period.

In general, the results reported in Exhibit 6 (inflation transitions) are similar to those reported in Exhibit 4 (after peak inflation), with all median active factor returns remaining positive.  The range of active factor returns remains remarkably wide affirming that inflation transitions provide differentiated responses among factors^[12], while further underscoring the risk of tactical positioning. As an alternative to factor timing, Exhibit 7 illustrates the benefits of a multi-factor approach using a naïve, equally weighted blend of the five factors (refer to Exhibit C of the appendix for full details and MSCI World Index data).

[1] The US Consumer Price Index was updated through June 2022 on July 13, 2022.

[2] The US Consumer Price Index breached 3% on a year-over-year basis on March 31, 2021.  From March 2021 through June 2022 the active factor returns of the MSCI World Index were as follows: Value 2.8%, Dividend Yield 11.0%, Low Volatility 10.1%, and Quality 3.8%.  For full methodology notes refer to the Appendix.

[3] See Hodrick and Prescott (1997).

[4] The HP filter also removes the cyclical (business cycle) component of a macro time series which was not used in this analysis.

[5] 3% represents the median inflation trend over the reported time period.

[6] Median returns are reported in lieu of average returns in order to mitigate outliers which skew the results.

[7] Active returns denote returns in excess of the market cap-weighted index.

[8] See “The Great Inflation, Factors and Stock Returns” by OSAM (2021).

[9] Refer to our paper What’s Driving Value? for more information.

[10] Both potential explanations may also apply to the increase in the size premium after peak inflation.

[11] “Hard” and “soft” landings both refer to economic outcomes that are perceived to be a result of central bank tightening.  Per a congressional research report (2022), “soft landing” refers to economic conditions whereby growth is moderate but positive, while a “hard landing” is associated with an economic recession.

[12] See Gupta, Kassam, et al. (2014).

References

Hodrick, Robert, & Prescott, Edward C. (1997). Postwar U.S. Business Cycles: An Empirical Investigation. Journal of Money, Credit, and Banking, 29(1), 1-16.

O'Shaughnessy Asset Management. (2021). The Great Inflation, Factors and Stock Returns. White Paper.

Wang, Di, & Lehnherr, Rob. (2022). What’s Driving Value? Northern Trust Asset Management. White Paper.

Gupta, Abhishek, Kassam, Altaf, Suryanarayanan, Raghu, & Varg, Katalin. (2014). Index Performance in Changing Economic Environments. MSCI Research Insights.

Congressional Research Service Insight. (2022). Where is the U.S. Economy Headed: Soft Landing, Hard Landing, or Stagflation?