New Perspectives on Inflation Shocks and Public Real Estate Equity Returns
Introduction
Inflation has again reared its ugly head in the United States. The December 2021 CPI inflation rate, 6.8%, was the highest recorded in 39 years, and producer prices were 9.8% higher than a year earlier. Relative to the past several decades, inflation is rising and survey evidence shows that actual inflation exceeds the expected inflation rate for most people. For investors, including real estate investors, this raises significant practical questions. Among those questions are:
- Will inflation destroy the real value of my property portfolio?
- Isn’t real estate a hedge against inflation?
- What changes might I make now, if I am concerned about inflation?
- Should my actions change, if higher inflation is universally expected?
This article highlights new research bearing on some of the central issues facing investors in publicly-traded real estate equity. The analysis focuses on how equity markets respond to inflation innovations (‘shocks’), defined as quarterly changes in one-year-ahead expected inflation. We measure these innovations using data from the Survey of Professional Forecasters. The data from 4Q-2021 are particularly sobering: the expected inflation over the next year nearly doubled from 2.6% in the previous quarter to 4.6% this quarter.
The major findings of this research project are:
- The equity return response to expected inflation shocks changed in 1997, so what we learned in the 1970’s and 1980’s isn’t predictive now. The good news is that after 1997, all major property types appear to hedge against expected inflation shocks.
- Virtually all the comovements between inflation shocks and real estate equity returns happens in weak economic periods, i.e., in recessions. There is no systematic relation between them in strong economic periods, i.e., in expansions.
- REIT equity is about as sensitive to expected inflation shocks as the aggregate equity market in the great majority of comparisons for the post-1997 period. In a few settings, REIT equity is marginally better than public equity generally in hedging inflation shocks.
- Not all public real estate equity functions equally well as an inflation hedge. There is one notable difference in property type-based equity returns relative to expected inflation innovations, and it appears to be related to lease contract length.
Data
The work here relies on several data sources. REIT index returns are drawn from the CRSP/Ziman Real Estate Data Series, compiled by the Ziman Center for Real Estate at the University of California – Los Angeles and distributed by the Center for Research in Security Prices (CRSP) at the University of Chicago. The underlying data are monthly returns which we convert to a quarterly basis to match the timing of the inflation-related data. We use quarterly returns in part to avoid the noise in monthly stock returns.
Our primary measure of expected inflation shocks is the change in one-year-ahead forecasted inflation from the Survey of Professional Forecasters (SPF, hereafter) that the Federal Reserve Bank of Philadelphia has run for nearly four decades. Our SPF-based inflation shocks are based on subjective expectations of survey participants, whereas much of the research on inflation relies on shocks defined by the difference between predictions from statistical models of inflation and realized values. Evidence in the literature shows that surveys like the SPF provide the best out-of-sample inflation forecasts, and this weighs heavily in our choice. (Importantly, we draw the same conclusions with other inflation shock measures.) Available every quarter, the SPF-based inflation shocks have a predictive focus: we rely on the idea that changes in expected inflation should have longer-term economic importance to the degree they reflect views about future inflation, and thus affect quarterly returns on equity.
The reasoning behind the focus on inflation surprises, not on the level of inflation is simple: inflation shocks change behavior, but stable inflation is not likely to have that effect. Everyone can adapt to predictable inflation by adjusting all nominal prices over time by the stable inflation rate. Inflation shocks – really, surprises – are more problematic because plans were made, prices were set, and investments were undertaken based on a particular inflation rate. When expected inflation changes, economic agents recognize that adjustments must be made. The key to the ‘adjustment’ step lies in understanding why inflation expectations are changing and what the best choices are going forward. It is the consequences of these adjustments for asset prices that we are picking up in quarterly equity returns.
Figure 1 — Source: Center for Research in Security Prices at the University of Chicago and Survey of Professional Forecasters from the Federal Reserve Bank of Philadelphia
Major Findings
Figure 1 illustrates the first two major findings. It shows the relation between expected inflation shocks pre-1998 vs. post-1997 and contrasts the relation between recessions and expansions. The two charts on the top row show the relation between 1981 and 1997, while the bottom row shows the 1998-2019 relation. The left side shows the recession relation, and the right side shows the relation during expansions.
Before 1998, there is a negative relation between inflation shocks and the REIT market index in recessions (the left chart) and no statistically or economically reliable relation during expansions (the right chart). If investors expected higher inflation, REIT returns would most likely be negative, conditional on being in a recession. Put simply, higher expected inflation in a recession was bad news for REIT returns. After 1997, there is a still no relation between expected inflation shocks and REIT returns in expansions (the right chart), but now there is a positive relation in recessions (the left chart). This means that higher expected inflation would most likely be accompanied by higher REIT returns, conditional on a period of economic weakness. We show that this contrast in performance also holds for the broader equity market.
Figure 2 develops a deeper understanding of how inflation shocks are related to real estate equity returns sorted by property type. Each chart shows the estimated comovements of expected inflation shocks with REIT returns by property type in strong and weak economic times. The left chart shows the pre-1998 period while the right chart shows the post-1997 period. We report the exact values in Table 1.
The entries are interpreted as follows. A one-standard deviation change in expected inflation over the next year (0.304 for the pre-1998 period and 0.183 for the post-1997 period) is associated with a 5.82% higher quarterly return for the All-REIT index in the post-1997 period, but a 3.97% lower quarterly return for the pre-1998 period. From the table, it is plain to see that for the post-1997 period, hotel/lodging REITs show double the sensitivity to expected inflation shocks vs. equity REITs generally and a broad stock market index. By comparison, these same hotel/lodging REITs have very modest quarterly returns, 1.02%, if there is an inflation shock but no recession.
In summary, for the 1981-1997 (1998-2019) period, positive expected inflation shocks are associated with negative (positive) REIT return performance, regardless of the property type. More recently, hotel/lodging REITs post the strongest performance in response to expected inflation shocks, but it is particularly strong in recession periods.
Comovement Factors by Property Type
| Period | All | Equity | Mortgage | Hotel | Residential | Retail | Non-REIT Equity |
|---|---|---|---|---|---|---|---|
| Pre-1998 | -3.97 | -3.93 | -3.52 | -1.63 | -4.41 | -4.03 | -1.63 |
| Post-1997 | 5.82 | 6.10 | 4.58 | 12.24 | 5.45 | 5.79 | 5.79 |
| Baseline | -0.33 | -0.32 | -1.77 | 1.02 | 0.07 | -0.51 | -0.61 |
Table 1 — Source: Center for Research in Security Prices at the University of Chicago and Survey of Professional Forecasters from the Federal Reserve Bank of Philadelphia
Figure 2 — Source: Center for Research in Security Prices at the University of Chicago and Survey of Professional Forecasters from the Federal Reserve Bank of Philadelphia
Explaining the Findings
As we said, earlier, inflation surprises may signal changed views about the future and new choices, and what we find is that this effect is found only in recessions. One natural question is why this is happening, and whether there are implications for investors now.
Based on Irving Fisher’s analysis of inflation and asset returns nearly 100 years ago, the prevailing idea as of 1970 was that absent any impediments, agents would adjust quickly so that over longer investment horizons, equity investment returns would generate a real component and an inflation compensation component. In this simplified world, real returns compensated appropriately for risk and what was left constituted an inflation-based adjustment that preserved real returns. Accordingly, assets priced like this functioned as an inflation hedge (early research studied exactly this property of asset prices). This means that if inflation was unexpectedly high and was expected to persist, asset prices would fall in value so that new owners would earn a higher inflation premium. If you simply held the asset, you took a capital loss. No wonder inflation surprises were unwelcome!
Economists recognized that this was too simple, and developed several additional perspectives. One important idea first discussed by Nobel Laureate Eugene Fama held that inflation surprises may convey information about the state of the economy. In the data since the 1960’s, there are (at least) two major ways investors might look at inflation and the state of the economy. Before the stagflation of the 1970’s, investors might have expected higher inflation to accompany higher economic growth. The implication is that higher inflation forecasted higher economic growth, and so asset prices would likely rise because of cash flow growth (i.e., there is a positive correlation of asset returns and inflation). The experience of the 1970’s exemplified the alternative: the US had high inflation and lower economic growth (so-called stagflation), and there was a negative correlation between asset returns and inflation. If Fama’s way of thinking was to be useful (and operational), investors needed an understanding of how inflation expectations were related to economic growth expectations.
The other idea to consider addresses inflation and financial risk. Simply put, if higher inflation makes cash flows riskier, the risk premium paid to investors must rise. This implies that asset values must fall. There is recent research evidence establishing that inflation shocks are a source of risk affecting equity pricing in US markets. This implies a negative relation between inflation shocks and asset returns, but this effect operates through a risk (i.e., discount rate) channel, not through a cash flow channel where an inflation shock implies lower future economic growth.
The upshot of this is that whether an asset is an inflation hedge depends on how inflation shocks are related to expected economic growth and the risk premium earned by investors. Since recessions are followed by stronger economic growth, inflation in a recession may be viewed as a signal that the recession will end in the near term and the economy will improve. This forecasts higher cash flows (at least) and so stock prices rise. It also implies that interpreting the data requires careful controls for the state of the economy and variation in risk premiums paid to investors.
When we study the real world data, we find evidence that both cash flow and discount rate effects on average have a hand in accounting for value changes. While what we find holds in the aggregate and for REITs with different property types, it is clearly true that the details of specific property holdings may lead to some differences. We turn to this question now.
Looking Ahead
The last quarter of 2021 looks to be at odds with the findings of our research: expected inflation rises by the largest amount seen in the Survey since 1981, economic growth is positive (so, no recession), and REIT index returns are positive. How can this be reconciled with the evidence discussed earlier? In contrast to the data analysis presented earlier, what we have to say here is more speculative. That said, we think the answer lies in recognizing two things: public real estate equity is priced based on cash flows and discount rates, and there are some very unusual dimensions to current capital market conditions not seen before in US capital market data.
REIT pricing in 2020 arguably reflected expectations of lower cash flows. Actual cash flow performance has been more positive, and as investors adjust their initial negative expectations to reflect this, there is upward pressure on REIT shares. To the degree that REITs are successful in passing through inflation to lessee’s and/or can control cost increases, we expect the cash flow part of REIT pricing to be positive. This implies positive stock performance.
The discount rate picture is where we think the action really is. As of the 4Q-2021, US capital markets are significantly affected by financial repression. Despite the US Treasury borrowing unprecedented amounts, Federal Reserve policy has depressed Treasury yields. This has produced real returns that are actually negative: the Treasury TIPS rate is negative and the real return on 10-year Treasuries – the 10-year yield minus the inflation rate – is also strongly negative. The implication is that the discount rate used to price REIT cash flows is very low. To the degree that Treasury yields remain near to current levels, we would expect strong performance of REIT shares: good cash flow news will have a strong impact on share prices. If the Fed moves to raise interest rates and/or abandons bond-buying entirely, the upward pressure on Treasury yields will be a drag on REIT share pricing.
Nothing like the present circumstances appears in the data underlying our results. Accordingly, applying the earlier results to the present case requires careful consideration of the fundamentals of cash flows and discount rates. That brings us to property basics. Properties with rent rates fixed for the longer-term will not generate higher cash flows that offset inflation. Only resort/lodging properties seem especially well-poised to offset higher inflation with higher rental rates. Multi-family properties will follow but with a lag as the rent roll changes.
Beyond this, we expect that properties in markets with strong demand will fare better in getting higher rents to offset inflation elsewhere. To the degree that the property risks in those markets are smaller, we expect the risk premium built into the discount rate will remain relatively smaller. Marginal properties will likely have trouble getting rent increases, and we expect discount rates to rise because property risks are higher as nominal results suffer.