Liquidity Dynamics in Commercial Real Estate

Article excerpt

Executive Summary.

This paper examines private real estate market liquidity dynamics using the recently introduced family of Transactions-Based Indices (TBIs) calculated from the NCREIF database. The paper derives a Liquidity Index from a combination of the TBI Price Index and the TBI Supply Index. Liquidity in the private real estate market is shown to comprise more than the traditional bid-ask spread; it includes a term for the probability that eligible buyers and sellers meet. An econometric model calibrates the joint behavior of price and liquidity, and the price adjustment process is shown to be less smooth than the liquidity adjustment process.

Although liquidity is crudely defined as "market activity" or the inverse of expected time-on-themarket, no uniformly accepted metric of national or sector liquidity exists. Indices that measure real estate prices have recently become available in variable liquidity and constant liquidity varieties, but heretofore little effort has been made to develop an index measuring the characteristic at the very core of the real estate asset.

Given the current condition of the real estate market in the United States, a standardized measure of liquidity would provide researchers, practitioners, and investors with a better understanding of today's environment and the possibility of predicting it for the near future. Until now, market liquidity levels have only been inferred from other indicators. A liquidity metric would reduce, to some extent, such uncertainty.

A measure of liquidity would also propel research into, among other areas, the interaction between price and liquidity. Unlike the markets for stocks and bonds, real estate markets exhibit highly variable liquidity. As a result, in real estate "observed transaction prices are conditional on overall market liquidity at the time of sale (i.e., price and liquidity are jointly determined)" (Clayton, MacKinnon, and Peng, 2008). To date, little research has been done on this simultaneous relationship in the commercial real estate context due to limited data availability.

To arrive at a standardized Liquidity Index for all commercial real estate in the U.S., this paper builds on the Transactions-Based Index (TBI) family introduced in the pioneering works of Fisher, Geltner, and Pollakowski (2007), hereafter FGP (2007), and Fisher, Gatzlaff, Geltner, and Haurin (2003), hereafter FGGH (2003). Using the econometric tool known as cointegration, an equilibrium relationship is unveiled between the TBI Price Index and the TBI Supply Index. This paper shows that a linear combination of the parent indices can be interpreted as a measure of private market liquidity. As a result of the search model framework underlying the TBI family, liquidity is defined as the product of the buyer/seller spread and the probability that eligible buyers and sellers meet.

The paper is organized as follows. First, the paper explains the data sources and the TBI family. The paper then conjectures a Liquidity Index representing ln(Liquidity) by simply subtracting the constant liquidity TBI Supply Index from the variable liquidity TBI Price Index. The coefficient of cointegration is formally estimated and found to be statistically different from one for these few observations. After explaining how the uniqueness of the Liquidity Index is rooted in the properties of the coefficient of cointegration, various unit root tests and a simulation study establish cointegration of the parent indices. Next, an AR(2) process is fit to the stationary Univariate Liquidity Index, and a forecast is attempted.

The Liquidity Index is then re-derived in real terms and applied to a vector autoregression (VAR) with exogenous variables to model the endogenous relationship between price and liquidity. This framework explores the simultaneity between price and liquidity that defines market conditions. The statistical analysis concludes that the estimated effect of liquidity on the change in price depends on how much time has passed, and that the estimated effect of the change in price on liquidity is positive. …