Trading and Volatility in Dual Market: Theory and Evidence from Real Estate

Abstract

This article investigates the informational role of trading volume and the impact of transaction costs on trading volume and the return dynamics in public and private markets, using real estate as an application. The model extends Kyle (1985) in a framework with two markets associated with asymmetric transaction costs across markets and over time. It derives equilibrium and predicts that at the market downtime, investors’ incentive to trade in the public market increases while that in the private market declines based on a return-cost trade-off, which is when transaction cost in the private market accelerates. Trading volume induces return volatility. The marginal effect of trading volume on price increases monotonically with uncertainty in asset values and transaction costs. An assessment using a simulation method and data on public and private real estate markets in the United States from 1996 to 2014 validates the model’s propositions. We offer two explanations for the increased real estate investment trust (REIT) return volatility during the financial crisis: increased trading volume in the REIT market and a larger marginal impact of volume on price during market downtime.

Keywords:

• Transaction cost
• return volatility
• trading volume
• REITs

Notes

1 Subject to various exceptions, in each taxable year, a REIT must derive at least 75% of its gross income from the following sources: 1) rents from real property; 2) interest on obligations secured by mortgages on, or interest in, real property; 3) dividends and gains from sale of shares of other REITs; 4) abatements and refunds of taxes on real property; 5) income and gains derived from foreclosure property; 6) amounts received for entering into agreements a) to make a loan secured by a mortgage on, or interest in, real property, b) to purchase or lease real property; 7) gains from the sale or disposition of real assets, not “prohibited transaction”; 8) income from investment of the proceeds of an offering by the REIT of certain types of securities, including common stock, during the one-year period beginning on the date the proceeds are received.

2 See Williams (1995) for a discussion of decentralized trade in real asset markets.

3 Besides real estate, commodities such as gold, oil, and agricultural products are traded in both physical markets and in the corresponding derivatives markets. The private versus public market differences generally apply to these cases as well.

4 Illiquidity costs in different asset classes have been documented in the literature: the bond market (Amihud & Mendelson, 1991; Chen et al., 2007), equities (Cox & Peterson, 1994), initial public offering (IPO) (Bajaj et al., 2001), options and futures (Brenner et al., 2001), and real estate (Krainer et al., 2004).

5 The liquidity of a market is typically aided whenever a financial intermediary has the strength to take on unwanted real or financial assets, redistribute them to interested buyers, and inject added liquidity into the market. If a market can be seen as in dynamic equilibrium, with financial excess in one area creating financial need in another, financial intermediaries can exist as brokers of sorts in diverting capital from one party to another.

6 We acknowledge that price discovery in the dual market is unlikely to be completely independent given that it is the same CRE assets being priced. However, the difference still exists due to the uniqueness of each CRE asset and the much lower transaction frequency in the private market. Besides, even if the transaction price is covered by the media, much detailed information is still not publicly available. For example, in private CRE market, the negotiation process on private property is relatively confidential, where a relatively small group of participants got involved. The price discovery highly depends on the negotiation power between market participants and the detailed cash flow information on a property. So, the trading behavior and price discovery process is different from that of the public market.

7 This “noise trader” assumption is standard in microstructure models. It reflects the difficulty noted by Milgrom and Stokey (1982) that, in the presence of traders with better information, uninformed traders acting solely on speculation would be better off not trading.

8 It is the additional cost paid on top of the fundamental price. This assumption leads to an explicit linear equilibrium without loss of generality.

9 The market maker collects the transaction cost and spends it to facilitate the transactions in market i. Thus, transaction costs do not affect a market maker’s profit and the optimal price rule.

10 It holds for an assumption of a gross transaction cost range of (1,2). See proof in Appendix B.

11 The return measures the percentage change in stock values over a period of time and takes into account the effects of splits and dividends. Splits and dividends are relatively infrequent events, so return for most days is simply the relative or percentage change in daily price.