Is Bigger Better? A Re-Examination of the Scale Economies of REITs

Article excerpt

Executive Summary. This study re-examines economies of scale for the equity real estate investment trust (REIT) industry, particularly the use of the translog model in empirical tests. Three cost functional forms are tested using data from 120 equity REITs in 1997: translog, simple quadratic and quadratic semi-log. The commonly utilized translog model is found to have little significance in explaining the scale effect, which suggests that the economy-of-scale parameter results derived from a translog model are irrelevant. However, the simple quadratic and quadratic semi-log functions are found to be significantly concave, supporting the existence of economies of scale in equity REITs.

Introduction

"Is bigger better?" That is one of the questions puzzling the academic researchers and practicing professionals in the real estate investment trust (REIT) industry. Recent mergers and acquisitions in the industry have certainly fueled this debate. Compelling cases have been made by both academic researchers and practicing professionals that bigger is better, primarily because of the increased economies of scale achieved by relatively large firms in other industries. Many large REITs are projected to become even larger in the future.1 More real estate will be concentrated in fewer public real estate firms. It is also argued that size efficiency may come from management economies of scale, access to lower-cost capital, brand imaging and increased bargaining power with customers and suppliers potentially associated with large companies.

However, some researchers question this notion.2 They argue that the characteristics of REITs may inhibit scale economies. For example, diversification over geographic areas and property types may limit management efficiency. Economies of scale may decrease when the size of a firm exceeds a certain level, as found in the banking industry.3 Lastly, the mega-mergers in the REIT industry may not be in the best interest of shareholders if economies of scale do not result.

Do large REITs enjoy economies of scale? Do they benefit from their branding and reputation, which stem potentially from their relatively large scales? If so, what factors contribute to the scale of economy? Is there an optimal economy of scale? This study carefully examines these issues, including the methodology employed in the economy of scale studies for REITs.

Recent research has begun to empirically address these issues. Ambrose, Ehrlich, Hughes and Wachter (2000) study the economies of scale for the REIT industry in the early 1990s. With a sample of forty-one multifamily equity REITs, they construct shadow portfolios that simulate each REIT's exposure to changes in local market conditions. Using the shadow portfolio, they utilize an ordinary least squares model that includes the shadow growth rates of net operating income and dummy variables for branding and geographic concentration. The authors find no size-economies or significant benefits for REITs in image branding and geographic specialization.

On the contrary, Bers and Springer (1997, 1998) directly test the scale economies of REITs by using translog models and find supporting evidence.4 In the 1997 article, they study the economies of scale in the REIT industry and particularly the sources of scale economies. They investigate five categories of cost sources: general and administrative expenses, management fees, operating expenses, interest expense and interest expense as a percentage of total liabilities. Using a translog cost model and REIT data from 1992 through 1996, they find that overall economies of scale exist for all aforementioned cost figures except interest expense. General and administrative expenses and management fees demonstrate the largest economies of scale; operating expenses show only a modest effect and interest expense shows diseconomies of scale. They argue that these findings are significant evidence supporting the existence of scale economies. …