Academic journal article Real Estate Economics

U.S. Office Market Values during the Past Decade: How Distorted Have Appraisals Been?

Academic journal article Real Estate Economics

U.S. Office Market Values during the Past Decade: How Distorted Have Appraisals Been?

Article excerpt

Edward J. Kane [**]

Patric H. Hendershott [*]

The leading time series of real estate returns is the Russell-NCREIF (RN) Property Index. The RN series tracks returns, cash flow plus appraised capital gains, for multiple property types. To evaluate the accuracy of the capital-gains component of the office-market return series, this paper constructs two benchmark measures for the present value of projectable office-market cash flows from 1982 to 1991 and compares these with a real value series based on the RN capital-gain component. The RN-based series runs 30% above the highest of the benchmarks throughout the 1986-1989 period. While this overstatement is consistent with the development of a price bubble, failure of the bubble to burst until 1990-1991 is implausible. Real estate experts recognized overvaluation in assessments as early as the spring of 1986.

The RN Office-Market Index was slow to register price declines when the markets first weakened and then overstated the rate of decline once the market began to bottom out. This pattern likely reflects incentives for appraisers to smooth potentially temporary price volatility and for investment managers to maintain appraised values in declining markets. It traces as well to systematic differences in the character and condition of the properties that tend to trade at different stages of the real estate cycle. These incentives and differences provide reason to believe that other RN Indexes were similarly distorted.

Throughout most of the last decade, commercial real estate markets have been distressed. The one market with solid data on vacancy rates and real rents is that for office buildings. For this property type, vacancy rates have stood at unprecedented levels since 1984 and real rents have halved relative to 1982. Motel/hotel and shopping center markets have shown parallel signs of distress Shulman (1990).

Accounting evidence of falling commercial real estate values appeared first on thrift balance sheets and in Federal Savings and Loan Insurance Corporation (FSLIC) bailout appropriations. Writedowns then spread to commercial banks and insurance companies. These value declines are a response to the substantial overbuilding that was fed by what can be described as a lending frenzy Hendershott and Kane (1992). Perhaps surprisingly, value declines were slow to register in appraised values (Salomon Brothers 1992). In part because of the delayed downturn in appraisal-based returns, the deep writedowns of real estate values that began in the middle 1980s proved something of a surprise to deposit-institution regulators and others.

Calculations developed in this paper from the Russell/NCREIF (RN) Property Index show two curious features. First, the office market component of the RN Index was higher on a real constant-quality basis at the beginning of 1987 than it had been at the beginning of 1982, in spite of an increase in the vacancy rate from 5% to 18%. Second, the retail component failed to decline until 1990.

Our analysis uses data from the national office market to assess the accuracy of the RN office market index during the 1980s. This analysis employs a numerical valuation model to estimate, on an order-of-magnitude basis, the ratio of value to replacement cost in the office sector at the beginning of 1992: The model is indicative and forward-looking and depends primarily on the projected future path of vacancy rates and the path of real rents the vacancy-rate time path predicts. While this paper applies the model to the national office market, it would be relatively straightforward to extend the analysis to local markets (see Hendershott, 1995).

To generate a time series of the value/replacement-cost ratio, we shift the valuation date backward a year at a time and repeat the calculation. Nine shifts generate an annual time series covering the beginning of 1982 to the beginning of 1992. Two benchmark series are computed, one based on perfect foresight and another based on the more optimistic assumption that lending would be curtailed sufficiently in the near future to cause vacancy rates to decline. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.