;

Forecasting Lodging Demand

This article evaluates recent changes in the often-cited correlation between lodging demand and GDP.

Corresponding with the most recent national economic recession, lodging demand in the U.S. declined precipitously beginning in August of 2008. A rebound in demand became apparent in many markets in 2010. As lenders, investors, owners, and developers formulate near-term plans, their assumptions about growth in lodging demand will be critical in determining whether their strategies succeed or fail. Conventional wisdom points to U.S. gross domestic product (GDP) as a leading indicator for lodging demand trends. This article suggests that the correlation between GDP and lodging demand in the U.S. has weakened substantially during the past decade. However, a significant correlation remains between Gross Private Domestic Investment (GPDI), a component of GDP, and lodging demand in the U.S.

Historical Trends in Lodging Demand
The authors analyzed lodging demand trends between January 1987 and January 2010 on a quarterly basis. The trend indicates an average annual growth rate of 1.4% over this 23-year period. During this same period, demand has increased in 19 of 23 years, and declined in only four of these years.

The following chart illustrates annual changes in U.S. lodging demand since 1987.



Demand declines are clearly apparent during the past three recessionary periods. Extended periods of demand growth appear during periods of economic expansion. This seems to support some relationship between lodging demand and U.S. economic growth generally.

Taking a closer look at shorter time intervals may provide greater insights pertaining to such relationships. Because demand is seasonal in the U.S., we analyzed quarterly periods of demand data, which better reflect trends in leisure and business travel as well as the seasonality of meeting and convention demand.

The following chart illustrates quarterly changes in lodging demand since January, 1987.

Demand measured on a quarterly basis also appears to be related to economic conditions generally, but clearer records of deceleration and acceleration of demand declines and rebounds also are evident.

An extensive series of research has identified numerous factors that influence lodging demand in the United States. Variables considered to affect lodging demand locally and nationally include the following:

  • Population growth
  • Employment trends
  • Cost of air travel
  • Gross Domestic Product (GDP)
  • Gross Private Domestic Investment (GPDI)

This analysis will focus specifically on GDP and GPDI as potential indicators for lodging demand. The authors tested the relevance of GDP and GPDI as potential indicators of lodging demand nationally. Although GDP was once thought to be among the best indicators of lodging demand, the correlation between GDP and lodging demand has diminished substantially during the past two business cycles. In contrast, a relatively more significant correlation between lodging demand and GPDI has remained in effect historically, including the period covering the most recent business cycles.

Historical Trends in GDP and GPDI
The Bureau of Economic Analysis tracks GDP and GPDI in the U.S. on an annual and quarterly basis. The authors focused on quarterly data to test the correlations between lodging demand and GDP and GPDI.

Between year-end 1987 and year-end 2009, GDP increased from $4.7 trillion to $14.1 trillion. This trend indicates an average annual compounded growth rate of 5.1% over this 22-year period.

The following chart illustrates annual changes in GDP since 1986.



GDP registered declines on an annual basis in only two of the three most recent recessions. However, GDP, like lodging demand, is seasonal in the U.S. and reflects trends in business and agricultural cycles, holiday shopping seasons, and the school year, among other factors. Therefore, a closer look at quarterly changes in GDP may reveal movements within annual periods.

The following chart illustrates quarterly changes in GDP since January, 1987.


Quarterly GDP data indicates declines during each of the past three recessions. This reveals shorter periods of economic decline that were not apparent in the annual data.

A brief literature review of academic and government research indicates numerous factors that influence GDP in the U.S. Variables known to affect GDP include the following:

  • Consumer sentiment
  • Business sentiment
  • Interest rates
  • Money supply
  • Demographic trends

As such, GDP is often used as a broad proxy for overall economic growth in the country. Government analysts and private-sector economists frequently publish forecasts of GDP in the U.S. which are readily accessible. To the extent that these forecasts are reliable, they may serve as indicators of future trends in lodging demand. Therefore, the authors of this article attempt to determine to what extent lodging demand is correlated with GDP.

Correlation between Lodging Demand and GDP
Despite covering three distinct national economic recessions, GDP continued to increase, or remain flat, in 20 of the past 22 years. This is in contrast to national lodging demand, which declined in four of the past 22 years. Further inconsistencies are apparent when analyzing quarterly data. This suggests that GDP may not be highly correlated with lodging demand, at least not in recent decades.

To measure the correlation between quarterly GDP and lodging demand, the authors employ a regression analysis, using the following simple, linear regression formula:

Yi = β0 + β1X1 + ε i


In this formula, the preceding symbols are defined as follows:

  • Yi = room nights sold (dependent variable)
  • β0 = constant term
  • β1 = coefficient (or slope of regression line)
  • X1 = GDP (independent variable)
  • ε = error term
  • i = number of observations

Observing the R2 coefficient of determination from this regression analysis can provide an indication of how well the regression model predicts the real data points. For the period between 1987 and 2009, using GDP as the independent variable produces an R2 value of 0.86, suggesting a highly significant relationship between GDP and lodging demand over this period. However, for the period between 2000 and 2009, the R2 value declines to just 0.34, suggesting a much less significant relationship between GDP and lodging demand during the past two business cycles.

The component parts of GDP may provide important insights into why the correlation between lodging demand and GDP has broken down in recent years. During the 22-year observation period, we established that GDP has increased in 20 of these 22 years, and declined only twice, in 1991 and 2009. Similarly, personal consumption declined in only one of these years, in 2009, and government consumption increased in all 22 years. In contrast, Gross Private Domestic Investment (GPDI) has declined in seven of the past 22 years, while increasing in 15 of these years; moreover, the declines corresponded specifically to periods of economic recession (1990-1991, 2001-2002, and 2007-2009).

Using a similar regression formula, the authors tested the correlation between lodging demand and GPDI. For the period between 1987 and 2009, using GPDI as the independent variable produces an R2 value of 0.91, suggesting a highly significant relationship between GPDI and lodging demand over this period. Moreover, for the period between 2000 and 2009, the R2 value was 0.76, suggesting a continued significant relationship between GPDI and lodging demand during the most recent business cycles.

Conclusion
One important strategic implication from this analysis is that forecasts of GPDI may be an increasingly relevant tool for forecasting lodging demand growth. One possible explanation for the declining correlation between GDP and lodging demand could be the increased use of fiscal stimulus to off-set the two most recent national economic recessions in the U.S. Trends in consumer spending and exports and imports could also be contributing to the results we observed. Therefore, the practice of using GDP forecasts as an indicator of future lodging demand growth could become less relevant, especially during periods of significant fiscal stimulus or rapid changes in consumer sentiment.

Lodging demand data presented in this article was provided by Smith Travel Research. Gross domestic product and gross private domestic investment data were provided by the U.S. Bureau of Economic Analysis.

0 Comments

Submit a Question or Comment