The recent price surge for many goods and services in the U.S. raises the prospect for an extended period of persistent high inflation. Conventional measures of inflation track either specific sets of prices that consumers pay, notably the Consumer Price Index (CPI), or for economywide spending - Personal Consumption Expenditures (PCE). How these measures align with the movements in revenues, expenses and profits of different types of businesses has, for the first time in four decades, become an important planning consideration. Stated another way, the revenues and expenses of every business react differently to the general price level movements measured by inflation indexes. These specific business financial-to-inflation index relationships are rarely one-to-one.
Hotels occupy a unique position within the commercial real estate investment universe for two reasons. First, the rents charged for the use of property spaces, commonly known as room rates or average daily rates (ADRs), are subject to high frequency adjustment. Aside from corporate and institutional (e.g., hospitals, universities) contracts, hotel guests do not enter into legally binding contracts for periods of demand. In high and rapidly changing inflationary environments, the possibility of keeping rents aligned with inflation is seen as a distinct advantage of owning hotels rather than other property types.
Second, hotel properties house operating businesses often controlled by property owners, unlike other property types that house businesses controlled by others. The income generated from hotel investments therefore depend on the following:
Changes in period-to-period hotel property NOIs with low inflation occur because of changes in ADRs and expenses that are almost entirely determined by the economy which affects demand and supply (i.e., non-inflationary real factors) and managements’ abilities. During periods of high inflation, changes in hotel ADRs and expenses become more a function of increases in the general price level.
In Part 1 of this series, I report on the long-run statistical associations between general inflation and hotel revenues, expenses and profits. The statistical results support intuition that ADRs and inflation have a significant positive relationship but shed doubt on the notion that net cash flows adjust to inflation. Part 2 of this series will provide additional insights on the relationship between inflation and hotel operating performance that may surprise some readers.
The data for this analysis come from two sources. First, hotel operating performance histories are taken from a file developed and maintained by CBRE Hotels Research titled The Financial History of the U.S. Lodging Industry. This file contains annual averages on a per available room (PAR) basis for ADR, occupancy, total operating revenues, total operating expenses, and operating profits from 1930 to present. Note that these averages extend across all hotels operating during each year and are not ‘same store’. Second, the macroeconomic data come from FRED, a large repository of time-series measures of economic activity around the world assembled by the St. Louis Federal Reserve Bank.
During most years since the early 1980s inflation consistently hovered around 2%, as shown in Figure 1. The all-urban consumers CPI achieved a 7.0%year-over-year (Y-o-Y) increase in December 2021, the largest year-end change since 1981. During Q1 2022 CPI growth rose to 8.0%Y-o-Y. Continuation of this trend suggests that inflation will make more sizeable contributions to future nominal hotel financial performance than during recent decades.
Figure 1: Consumer Price Index Annual History, 1946-2022 Q1
The underlying assumption is that changes in the operating performance of hotels derive from changes in lodging demand and supply conditions (i.e., real effects), changes in the general price level, and management efforts. Thus, from an economic perspective hotel profit can be expressed as,
where HP is hotel profit, DS denotes relative demand and supply relationships caused by general economic conditions, I is inflation, and ME represents management efforts to maximize profits. The symbol, Δ, indicates changes in each of the variables. Equation 2 presents the standard financial expression for movements in profits – changes in profits result from changes in revenues minus changes in expenses.
where ΔHR indicate changes in hotel total revenues and ΔHE is the change in hotel total expenses. Revenues and expenses also can be cast in the same way as ΔHP in Equation 1. Specifically,
Different demand and supply conditions exist for ΔHR versus ΔHE. For changes in HR, the determinants of ADR and occupancy are numerous and independent from labor market demand and supply drivers for hotel employee wages and salaries that contribute so much to changes in HE. Because the period-to-period ΔME is not available in a time series dating back this far, this effect is treated as a residual contribution of ΔHR and ΔHE and thus ΔHP as well.
The following testable hypotheses emerge from this set up over the long term:
Hypothesis 1: The effect of inflation on total hotel revenues and ADR is positive and statistically significant.
Hypothesis 2: The effect of inflation on total hotel expenses is positive and statistically significant.
Hypothesis 3: The effect of inflation on hotel profits is statistically insignificant because inflationary effects on revenues and expenses offset one another.
To inform about how much inflation affected historical hotel performance, the following set of regression equations are estimated:
ΔADRt = α + β1ΔCPIt-1 + β2ΔEMPt + et (5)
ΔHRt = α + β1ΔCPIt-1 + β2ΔEMPt + et (6)
ΔHEt = α + β1ΔCPIt-1 + β2ΔEMPt + et (7)
ΔHPt = α + β1ΔCPIt-1 + β2ΔEMPt + et (8)
The purpose for including total nonfarm employment, EMP, is to represent general economic conditions thereby capturing real effects on hotel performance without that variable being highly correlated with CPI. The correlation coefficient of ΔEMP and ΔCPI during the study period is only .29.
Given data availability, the study period extends from 1946 through 2020 (annual). The Δs are Y-o-Y percentages, the ts represent time, and e is the error term of the regression equations incorporating random effects. Management contributions are embedded in the intercept, α. The all-urban consumers, all-cities CPI is introduced in all estimating equations. It is highly correlated with other inflation measures including the PCE. The one-year lagged value of CPI provides the best fit of the data, and based on intuition, it is assumed that the full effect of inflation in the immediately prior year is unlikely to be entirely present in hotel operating results until the current year. This statistical relationship implies that the full effect of inflation on hotel performance experienced in 2022 will not appear until 2023.
Figure 2 presents the descriptive statistics for the data involved in estimating the regression equations. All the average Y-o-Y percentage growth rates fall within a narrow range of about 2.5% to 4.5%. The standard deviations are more dissimilar across variables than the means. Extremely negative minimums come from 2020 operating results. Removing 2020 data has almost no impact on the statistical results.
The regression results appear in Figure 3. Across the top row of the table are the four left-side variables presented above in equations 5-8. Down the left margin are the explanatory variables from the same equations. Focusing on the three hypotheses, the β1s on the inflation variable, ΔCPIt-1, are positive and statistically significant at the .05 level or greater in the total revenue, ADR and total expense equations but not significantly different from zero (even at the .10 level) in the profit equation. These results exactly align with the stated hypotheses (#3) that the rise in total revenues related to inflation is largely offset by the increase in total expenses attributable to inflation resulting in no statistically significant effect of inflation on hotel profits.
The intercepts are significantly different from zero and negative except in the ADR equation. Negative ADRs would not occur even when inflation and employment are zero. The employment variable β2 s are highly significant and positive, thus ΔEMP is serving the intended purposes. The AdjR2s are not outstanding, but the F-statistics confirm that the equations are statistically sound. Adding more controls may boost the AdjR2s, but also likely introduce more collinearity within the explanatory variable set.
The sizes of β1s on the inflation variable, ΔCPIt-1, have some interesting implications. These coefficients represent how each hotel performance measure has reacted to inflation (on average) over many hotel cycles. A 100-basis point (bp) movement in CPI, for example, is associated with a specific number of bp movements in each hotel performance measure. The β1s in Figure 3 are these numbers: .6824 (68.24 bps) for total hotel revenues, .8480 (84.80 bps) for ADR, and .7126 (71.26 bps) for total hotel expenses. ADR is the most sensitive performance measure.
A major reason profit is insensitive to inflation is the simple arithmetic of both total revenue and total expenses being positively related to inflation - they offset at the bottom line. Operating leverage might have resulted in a statistically significant inflationary impact on profit if the β1s of total revenue and total expenses were equal. However, the results indicate that the sensitivity of total expenses to inflation (71.26 bps.) exceeds that of total revenues (68.24 bps.).
Commercial real estate investors in their pursuit of hotel acquisitions would be wise not to overweight the benefits of high frequency ADR adjustment for hedging inflation. The ADR is an important, albeit not the only, component of total revenue, but hotel expenses increase as well with general price movements. The evidence presented herein indicates that net cash flows and inflation are not statistically related in the long run.
I offer some caveats and extensions. The analysis reported here is performed with data at a high level of aggregation. These data are annual, they represent the entire population of hotels which changes each year, and they reflect the idiosyncratic characteristics of hotel markets at different times. These characteristics include excessive supply, changes in consumer preferences causing shifts in hotel offerings (e.g., more select service property development relative to full-service), and technological advancements such as the emergence of online travel agencies. CBRE Hotels Research offers customized econometric analyses of inflation effects within MSA markets, submarkets, and for individual hotels. See contact information below.
Returning to Figure 1, U.S. inflation experiences since 1946 can be separated into two sub-periods. The first, before 1983, inflation was high and persistent. The second, after 1983 inflation was low and reasonably under control.
In Part 2 of this series, I rerun the analysis examining each sub-period separately to determine if the results reported above are robust under different inflation regimes. The most interesting result will come for the ‘hot’ inflation sub-period given some similarities to today’s macro-economic environment.
Click here to view the original version of this release.
For more information about CBRE Hotels, visit pip.cbrehotels.com
3550 Lenox Road, Suite 2300
Atlanta, GA 30326
Phone: (404) 812-5024
Jack B. Corgel
Jack Corgel is Managing Director, CBRE Hotels’ Americas Research. Jack holds the Robert C. Baker Chair (Professorship) in Real Estate at the Cornell University School of Hotel Administration and holds the position of the director of graduate studies for the Baker Program in Real Estate. He was the first director of the Center for Hospitality Research at SHA. After receiving undergraduate and PhD degrees from the University of Georgia in real estate and corporate finance, he held faculty positions in several business schools at major universities before joining Cornell. Corgel serves as senior advisor to PKF Hospitality Research (PKF-HR), where he helps the firm develop products for the hotel industry based on property-level financial and real estate performance information, including Hotel Horizons econometric forecast of U.S. hotel market performance. Corgel published 80 articles in academic and professional journals, mainly on the subjects of real estate finance, investment, valuation, and hospitality real estate. His research appeared in the most prestigious journals in real estate (Real Estate Economics), urban economics (Journal of Urban Economics), insurance (Journal of Risk and Insurance), business law (Journal of the American Business Law Association), and hospitality management (Cornell Quarterly and International Journal of Hospitality Management). In addition, he has written for nearly every national journal read by real estate professionals. His textbook, Real Estate Perspectives (with Smith and Ling), was used throughout the nation for introductory real estate courses. He co-edited and wrote chapters for The Cornell School of Hotel Administration on Hospitality: Cutting Edge Thinking and Practice. Corgel’s current research interests include the relationship between the macro-economy and hotel markets and real estate price and capitalization rate forecasting.
Senior Advisor - PKF Hospitality Research
Phone: +1 678 910 4816