By Jack Corgel, Ph.D.
Hotel real estate investors need to measure both the absolute and relative financial performance of their properties. A 10% annual return may appear solid unless a collection of comparable investments earned 15%. Risk-adjusted returns measured against the appropriate index provide a way to assess managers’ abilities to generate excess returns. Measures of investment return come in many flavors.
The multi-period IRR can be computed as a realized or unrealized percent. Equity dividend rates are an example of a single-period, income return counterpart to internal rate of return (IRR). Calculating equity dividend rates and IRRs for commercial real estate is straightforward for one property but generating IRRs for a multi-property index presents problems.
The compromise reached more than four decades ago for reporting commercial real estate returns to pension funds by fund managers captures both the income and capital realized returns over a single holding period. These holding-period total returns for individual properties then can be compared to a return calculated in the same way for a collection of similar properties.
Under ERISA in the United States, pension fund managers are required to diversify plan assets in their control to minimize risks. Because plan assets historically consisted of equity and debt securities, real estate professionals following the enactment of ERISA in 1974 developed an institutional infrastructure to support the use of commercial real estate investments as diversifiers.
This infrastructure included the creation of the National Council of Real Estate Investment Fiduciaries (NCREIF) organization, the membership of which are asset managers of ERISA-qualifying plans, and the development of total return indexes for commercial real estate.
The historical series of indexed returns has been useful for demonstrating the diversification benefits of commercial real estate and allows pension plans to evaluate managers based on how they perform relative to an index (i.e., benchmarking). The array of indexes reported by NCREIF allows for comparisons of property risk-adjusted returns to the appropriate indexes.
NCREIF's HOTEL INDEX
In Q4 2020, NCREIF’s National Property Index (NPI) included 9,289 U.S. properties worth over US $700 billion. The NPI has many variants of the national aggregate index, including commercial property type indexes for apartment (1,943 properties), hotel (69 properties), industrial (4,387 properties), office (1,610 properties), and retail (1,280 properties). Asset managers provide quarterly income and appraised value information to NCREIF for properties under their management. The organization only includes information voluntarily provided for member-managed properties in these indexes.
The large number of properties included in each of the indexes, except hotel, allows NCREIF to report returns by combinations of property type and geographic region (e.g., office, Midwest) and market segment (e.g., retail, neighborhood). The number of hotels in the NPI is quite small compared to other property types. Why? The answer lies in the fact that NCREIF members do not control nearly as many hotels as the major property types (the ‘four food groups’ as they are affectionally known). The number of hotels in the index ranged from six in 1982 to 243 in 2012.
The unfortunate consequence for hotel investors of so few indexed hotels is that the NCREIF hotel index is suspect for Markowitz-style diversification analysis, performance reporting to passive investors, performance benchmarking, loan underwriting, etc. With under 100 reporting hotel properties the aggregate returns may be skewed to certain geographic areas, market segments, and price points. Importantly, sub-index reporting by hotel categories (e.g., MSAs and chain scales) is impossible with any degree of reliance. The NCREIF organization recognizes the small sample size issue for this index, so it does not report hotel returns by subdivisions of the national hotel index.
Despite few underlying properties, the NCREIF Hotel Index tracks closely with the all-property NPI and shows hotel returns to be more volatile than commercial property returns in general, as expected. For the period 1983 through 2020, the means are 8.27% and 7% for the NPI and NCREIF Hotel Index and the standard deviations are 7.25% and 10.86%, respectively. The relationships among these statistics raise basic risk/return questions.
The unfortunate reality about the NCREIF Hotel Total Return Index is that not enough properties are included each quarter to enable full usage by hotel investors. The aggregate (i.e., national) hotel index return patterns are as expected over time; however, hotel users likely would want disaggregated index numbers for benchmarking chain scales, locations, and MSAs.
A robust alternative to the NCREIF Hotel Index certainly is possible. The underlying property sample for an alternative would be large, consistent over time, and representative of many different hotel types. The alternative index would need to be computationally comparable to NCREIF and would afford hotel investors many options for return performance benchmarking.