Empirical tests of real estate market efficiency

Abstract

Recent empirical research using real estate data has supported the weak and semi-strong forms of the efficient markets hypothesis. Previous studies have not included an estimate of expected appreciation into the tests of market efficiency, thus raising a question about the reliability of the results. We first use a market model to test for market efficiency with results similar to those reported by others. We next use a dynamic multiple indicator, multiple cause (DYMIMIC) model, which extracts a vector of expected appreciation from the price data, to test market efficiency. This approach produces superior results and a stronger conclusion about the efficiency of housing markets. The results indicate limited adjustment delays which can be explained by the existence of high transactions and search costs.Keywords: Efficient markets hypothesis, real estate market efficiency, housing market efficiency, DYMIMIC model and real estate, DYMIMIC model and housing market efficiencyThe common perception of real estate markets is that they are less efficient than financial markets. This would seem to be especially true for the housing market where many properties trade only infrequently, making it difficult for participants to determine "correct" prices. The unique characteristics of real property and the local orientation of the market, which require specialized knowledge of the factors that affect risk and return, also may contribute to inefficiency. Furthermore, transaction and financing costs as well as tax considerations make it difficult to exploit potentially profitable opportunities even if they can be identified.There is a growing literature testing for market efficiency using real estate databases. Previous researchers have taken one of two approaches when testing for market efficiency. Gau (1984Gau ( , 1985, Rayburn, Devaney, and Evans (1987), Mclntosh andHenderson (1989), andCase andShiller (1989) primarily use a forecasting approach, where an inability to predict future prices is interpreted as evidence supporting market efficiency. Linneman (1986), and Guntermann and Smith (1987) use a market model approach analogous to earlier tests of efficiency in securities markets. These studies generally find some evidence of market inefficiency but typically conclude that trading rules or strategies cannot be developed to earn abnormal rates or return either because of high transaction and information costs or because of the difficulty of accurately forecasting real estate prices.

Housing Price Dynamics in Time and Space: Predictability, Liquidity and Investor Returns

“…Inertia in the price adjustment process in the housing market has been welldocumented; many studies have concluded that returns in housing markets are predictable by demonstrating that time series estimates of aggregate prices exhibit inertia (Hosios and Pesando 1991;Guntermann and Norrbin 1991;Gatzlaff 1994;and Malpezzi 1999).…”

The Subsequent Market Value of Former REO Properties

“…Guntermann et al . () examine the activity of “flippers” or short term residential real estate investors in the Phoenix and Scottsdale, Arizona area. They utilize a restrictive methodology to distinguish between “flippers” who repeatedly invest short term in residential real estate, and other individuals who are not experienced or professional flippers.…”

“…Guntermann et al . () control for property improvements using permit data that includes information on the increase in assessed value resulting from the permitted improvements. Their data only includes property improvements that require permits, and thus their permitted sample is a relatively small (66 permits or 0.22%) component of the 30,328 identified flip sales.…”

“…Our article complements the analysis in Guntermann et al . (), in that we provide an alternative approach to measure improvements made by investors who buy houses out of foreclosure. We suggest that building permit data may significantly undercount properties that have undergone renovation.…”

“…Guntermann et al . () in their study of Phoenix metropolitan area find that foreclosures are purchased at a 20% discount within a 1988–2010 sample period. They suggest that flippers might have the ability to identify mispriced houses and thus purchase properties at a discount based on a possible information advantage; however discounts on foreclosures and cash purchases are available to all individuals, and are not due to flippers’ information advantage.…”