Price-volume Correlation in the Housing Market: Causality and Co-movements (P.2)

Introduction
Housing markets play an important role in the economy. For example, Bertaut and
Starr-McCluer (2002) show that residential properties accounted for about one
quarter of aggregate household wealth in the United States in the late 1990s, and
Tracy and Schneider (2001) show that housing wealth accounts for about two-thirds
of the wealth of the median U.S. household. Changes in home prices and trading
volume seem to have significant economic impacts on builders, brokers, lenders,
appraisers, furniture consumption as well as local property tax collections and
related local government budgets, in addition to local affordability and wealth. A
rapid surge in home prices and trading volume after 2000 has been seen across many
areas in the United States, and it is followed by a recent decline in house prices and
trading volume. These phenomena generate a lot of discussion regarding whether the
US has been in a “housing bubble” (see, e.g., Case and Shiller 2003). Despite the
importance of the housing markets and the economic and policy implications of
changes in home prices and trading volume, some important aspects of housing
markets are not well understood.
A well known pattern in the housing market is that prices and trading volume
seem to correlate with each other: trading activity tends to be more intense (i.e.,
more transactions and less time on the market before sale) when prices are rising
compared to falling markets. The positive correlation between prices and trading
volume appears to be inconsistent with standard rational expectation asset market
models, in which housing prices are present discounted values of the future service
streams (see e.g. Poterba 1984). A conventional interpretation of the correlation is
that price changes cause changes in trading volume. The causal relation is built on
one of three factors: equity constraints,1 nominal loss aversion (homeowners are less
willing to sell their homes in a falling market to avoid realized losses), or the option
value of homeowners (homeowners wait to sell when the upside benefits exceed net
carrying costs, see Cauley and Pavlov 2002). Stein (1995), Genesove and Mayer
(1997), Lamont and Stein (1999), and Chan (2001) provide theoretical and empirical
evidence for equity constraints of home sellers. Genesove and Mayer (2001), Cauley
and Pavlov (2002), and Engelhardt (2003) provide evidence for nominal loss

aversion.
Although research regarding the causal relation between prices and trading
volume greatly improves our understanding of the dynamics of the housing market,
a few important questions have not been satisfactorily answered. First, is a positive
price–volume correlation widely observed across markets? It is striking that there is
mixed evidence regarding the relation between prices and trading volume, and the
evidence is from either aggregate national level data, or from small panel data (up to
22 metropolitan areas). While a positive price–volume correlation is found by Stein
(1995), Berkovec and Goodman (1996), Andrew and Meen (2003), and Ortalo-
Magné and Rady (2004), a negative relation is found by Follain and Velz (1995) and

1 Falling prices reduce homeowners’ home equity values. Therefore, when homeowners want to sell their
houses, to ensure that the proceeds from selling would be sufficient to repay their mortgages and provide
down payments on new homes, they need to ask for higher prices, which increases the time on the market
and reduces the trading volume.

Next Page 3



Leave a Reply

You must be logged in to post a comment.