The present value model with time-varying discount rates: Implications for commercial property valuation and investment decisions

Abstract

A vector autoregressive model is developed for predicting cash flow and returns in the private (unsecuritized) commercial property markets. The model predicts both of these variables quite well during the sample period. The forecasting model is then used to develop a simple “buy/sell” rule for identifying property market value peaks aud troughs. An improved present value model, taking account of the predictability of property returns, is described and found to track historical market values much more closely than does either the appraisal-based index or the traditional present value model with constant expected returns. Analysis in this paper suggests that most of the change in commercial property market values has been due to changes in expected returns, rather than to changes in expected future operating cash flows. Key Wordsr valuation, investment, present value model, timing, cycles, discount rates The present value model underlies all of modern financial economics, and lies at the heart of commercial property valuation and real estate investment decision making. Traditionally this model is applied by forecasting property net cash flows and discounting those cash flows at a constant discount rate. In this model the discount rate is meant to represent the expected return (that is, the internal rate of return or total return) to an investment in the property, thereby reflecting the opportunity cost of capital. In practice, the discount rate used in the present value model as applied to property valuation has not changed much over time, largely because analysts have not known how to quantify changes in the market’s expected return on pr0perty.i Recently, evidence has mounted that asset value changes in the securities markets are not consistent with the constant discount rate model or the constant expected return assumption? Campbell and Mei (1993), among others, have found that changes in stock prices over time are due more to changes in the market’s required total returns (including price changes), than to changes in the market’s cash flow expectations.3 Furthermore, changes in the market’s expected return can be forecasted to some extent, as has been demonstrated by Liu and Mei (1992, 1994) in the case of REITs, small stocks, and large stocks. This implies that asset price changes are more predictable than was previously thought, and that turning points in asset price cycles may be somewhat identifiable in advance!