Does trading increase real estate portfolio performance? Piet Eichholtz finds the case in favour to be far from overwhelming

A potential source of manager value added in real estate is the ability to pick property type or geographic winners and losers and to time the market. Clearly, if a manager can accurately predict, for example, the optimal time to sell office buildings in London and buy retail properties in Amsterdam, risk-adjusted portfolio returns can be significantly enhanced. Until recently, the academic literature did not give much guidance regarding the extent to which active property portfolio managers can add value by pursuing such tactical, that is opportunistic, asset selection and investment timing strategies. A paper I wrote recently with Dirk Brounen of Erasmus University Rotterdam and David Ling of the University of Florida sheds some light on this issue.

We define active real estate managers as those who pursue tactical market timing and asset selection strategies for their private property portfolios. More specifically, we examine the effects of portfolio management intensity on the risk and return characteristics of publicly traded real estate companies that purchase their assets in private markets. Although a company's stock performance is likely to be more volatile than the performance of its underlying real estate portfolio, this approach provides direct evidence of the shareholder wealth created by active real estate portfolio strategies. Such direct and complete measurement of performance is not possible in the unlisted property market.

The research employs a sample of publicly traded firms that spans eight years (1996-2003) and includes the three largest public real estate markets in the world: the US, the UK and Australia. Historically, listed property companies in these three countries were generally not managed in a very active fashion. Rather, real estate companies were typically investment vehicles that provided investors access to a liquid and diversified portfolio of properties acquired in private direct real estate markets. In the last decade, however, public real estate firms in all three countries have become more actively managed and focused - either geographically and/or by property type.

We measure firm-level performance in several ways. First, we calculate excess (ie, risk-adjusted) stock returns and market betas using both single factor and multifactor asset pricing models. The excess returns (or alphas) and market betas from these models are then regressed on several indicators of portfolio management intensity, including the frequency of property acquisitions and dispositions, the firm's portfolio turnover rate, and the extent to which the firm has expanded the size of its real estate portfolio, as well as a number of control variables. We also estimate the relation between our management intensity indicators and two measures of firm-level operating performance: return on assets and return on equity.

We analyse the relation between our measures of management intensity and performance in two ways. First, we place property companies into quartiles on the basis of portfolio trading activity, and look at a set of performance indicators for the bottom and top quartile. None of these univariate statistics shows a consistent relation between trading activity and risk-adjusted stock return performance.

We next develop multivariate regression models to explain variations in excess stock return performance and systematic risk. In the first stage of the regression analysis, both single-factor and multifactor (Fama-French) asset pricing models are estimated using monthly firm-level excess returns as the dependent variable. The alphas from these time series models are then regressed in a second stage on our measures of portfolio trading intensity. In a second set of cross-sectional models, we regress our estimated firm level market betas on our proxies for management intensity.

These second stage regressions of alphas and betas also control for variables found in previous studies to be helpful in explaining the cross-section of excess returns and systematic risk. When the property type focus of the firm is included as a control variable, none of our indicators of portfolio management intensity are statistically significant in explaining return performance. Our attempts to explain cross-sectional variation in firm-level systematic risk and operating performance using our measures of management intensity are equally unsuccessful.

Our empirical results suggest that it is difficult for managers of public real estate companies buying properties in private markets to generate outperformance on the basis of the active trading strategies quantified in this paper. This suggests that even if private commercial real estate markets do exhibit the inefficiencies often attributed to them by practitioners, these ¬inefficiencies might not be large enough to justify active portfolio management strategies.

Several limitations of our research design are worth noting. First, we cannot observe or proxy for variation across managers in deal execution and ongoing property and asset management skills. However, our inability to carefully control for other sources of manager value-added may impede our ability to isolate the effects of trading intensity. Second, trading intensity may in fact play a role in the determination of a firm's excess returns and systematic risk. However, the effect may be too small economically when compared with the importance of diversification and the efficient provision of ongoing property and asset management.

Besides that, it is important to note that the literature regarding performance of non-property mutual funds generally shows that portfolio trading activity is associated with significant and consistent underperformance. In comparison, our results for property companies are rather positive. We conclude that property companies' trading activity is just successful enough, on average, to cover the additional transaction costs incurred by this trading.

Piet Eichholtz is partner with Utrecht-based Finance Ideas and professor of real estate finance at Masstricht University