Evaluating managers diversification of real estate portfolios: evidence from Nigeria.

ABSTRACT. While there is evidence on the effectiveness of diversifying real estate portfolio geographically, or by property type, there is lack of empirical evidence to justify whether diversification of managers is worth pursuing. This study produces evidence on the effectiveness of diversifying by managers and property types. The study collected data on capital and annual rental values from three (3) main Property Investment and Development Companies in Lagos Metropolis, Nigeria. From the data so collected, annual total returns (IRR), on residential properties, for a period of between 1997 and 2001 on the managers' portfolio were calculated. Under the assumptions that investments are held long and that constant correlation model or excess return to standard deviation represents the covariance structure of assets' returns, the study's analyses suggest that diversification of managers and property types produce improved performance. It also opens the possibility that an efficient portfolio developed by using constant correlation analysis may not be more efficient than a naively diversified portfolio as some of the naive diversification strategies are found to be effectively efficient.

KEYWORDS: Evaluation; Managers diversification; Real estate portfolio; Constant correlation model; Naive diversification

SANTRAUKA

NEKILNOJAMOJO TURTO PORTFELIO VALDYTOJO DIVERSIFIKACIJOS VERTINIMAS: ARODYMAI IS NIGERIJOS

Nors yra irodymu, kad geografine nekilnojamojo turto portfelio diversifikacija arba diversifikacija pagal nuosavybes rusis yra efektyvi, truksta empiriniu duomenu, patvirtinaneiu, kad verta siekti valdytoju diversifikacijos. Sis tyrimas pateikia irodymu, kad valdytoju diversifikacija ir diversifikacija pagal nuosavybes rusis yra efektyvi. Tyrimo metu surinkti duomenys apie kapitala ir metines nuomos vertes is triju pagrindiniu Lagoso miesto (Nigerija) nuosavybes investiciju ir pletros bendroviu. Pagal surinktus duomenis apskaieiuota 1997-2001 metu metine bendroji graza (vidine grazos norma) is valdytoju portfelyje esaneios gyvenamosios nuosavybes. Tariant, kad investicijos ilgalaikes, o pastovios koreliacijos modelis arba perteklines grazos ir standartines deviacijos santykis sudaro turto grazos kovariancinfae struktura, tyrimo metu atlikta analize rodo, kad valdytoju ir nuosavybes rusiu diversifikacija leidzia padidinti rezultatyvuma. Be to, analizes metu pastebeta ir tai, kad naudojant pastovios koreliacijos analizae suformuotas efektyvus portfelis gali buti ne kiek ne efektyvesnis uz paprastai diversifikuota portfeli, nes paaiskejo, kad kai kuri paprastosios diversifikacijos strategija yra labai efektyvi.

1. INTRODUCTION

Throughout the ages, investors have approached the problems of investment decision in a number of ways. Some prefer the strategy, which according to Hargitay and Yu (1993) was formulated by Andrew Carnegie, whose maxim says "put all your eggs in one basket and then watch the basket". The dictum "do not put all your eggs in one basket" appears to be the belief of many. Hence, the idea that the risk of loss can be minimised by not putting all of one's assets in "one basket" has been around for a very long time. Since Markowitz (1952, 1959) foundation works on Modern Portfolio Theory (MPT), authors and professionals have examined almost every possible ways of diversifying within the stock market although two decades elapsed before MPT was first applied to real estate. Attempts to transplant this methodology have long been frustrated due to the peculiar nature of property market, especially the lack of adequate time-series data. Meanwhile, the first research began in early 1980s (quoting from Mueller, 1993) with the presumption that a properly diversified real estate portfolio should help to partially overcome the illiquidity and mobility problems inherent in real estate.

Studies such as, Hadaway, (1978); Miles and McCue, (1982); Hartzell, Hekman and Miles, (1986); Grissom, Kuhle and Walther, (1987); Giliberto and Hopkins, (1990); Mueller, (1993); Pagliari, Webb and Del Casino, (1995); Brown, (1997); Brown, Li and Lusht (2000); Conover, Friday and Sirmans (2002); Lee (2005); Lee and Stevenson (2005) and Adair, McGreal and Webb (2006) have generally demonstrated that diversification benefits may be captured by combining different classes of real estate assets in different locations or by acquiring different property types or using both strategies. In other words, real estate diversification has traditionally been studied along two dimensions, namely, geographic/economic grouping and property types. However, the comment of Cheng and Liang (2000) on diversification of managers and property types as well as the results of Ajala (2001), in Nigeria, opened the question of whether diversification by managers and property types would produce comparable (or improved) performance. The answer to this question is necessary because authors such as Del Casino, (1995), Olaleye (2000) and Ajala (2001) have noted the importance of differing managerial skills on the overall performance of property portfolios.

The concept of managers and property type diversification holds that investors may combine investments and achieve good portfolio performance by investing in various property types focusing on the different management firms. The idea derives from the fact that the varying skills and experiences possessed by various property managers can have influence on the performance of a particular property or portfolio in terms of risk-return trade-off. In addition, it is believed that the differences among property types relate to the time and expertise necessary to manage them as investments. Thus, Cheng and Liang (2000) reported that pension funds in U.S.A. have been advised to seek diversification of managers and property types.

Since early 1980s, studies were conducted with various techniques and databases used to examine the benefits of diversifying real estate geographically and by property types. They were also based on a variety of levels including national, regional, metropolitan areas, and even smaller spatial definitions. Miles and McCue (1982) tested diversification strategies in United States of America by dividing the country into four geographic regions and comparing this with a strategy that diversified the portfolio by property types. They found out that diversification by property type showed better risk return characteristics than did a four region geographic strategy. In the same vein, Hartzell, Hekman and Miles (1986), Hartzell, Shulman and Wurtzebach (1987), Grissom, Kuhle and Walther (1987), Giliberto and Hophins (1990) and Mueller (1993) studies examined efficient frontiers for various diversification schemes and compared them against naively diversified portfolios. The studies found evidence (although mix evidence) to show that geographic and or property type diversification brings marginal improvement in portfolio performance. Besides, Pagliari, Webb and Del Casino (1995) findings suggest that while MPT yields optimal ex post portfolios, its use as an ex ante portfolio allocation strategy can lead to mixed results. Cheng and Liang (2000) improved on pervious studies on optimal diversification by answering few questions that centred on whether improvement is significant in a statistical sense or not. The study found evidence to support the fact that an efficient portfolio is statistically more efficient than a corresponding naively diversified portfolio when the portfolio period is the same as the period used for testing the difference in efficiency. Brown, Li and Lusht (2000) study on intracity geographic diversification divide Hong Kong into sub-markets and concluded that efficient portfolios outperform many naive strategies based on equal allocations across districts, and outperform most, but not all, of those based on "all your eggs in one district" strategies. Viezer (2000) found evidence to suggest, among others, that more dimensions of diversification are better than fewer dimensions and that the best strategy used sixteen dimensions (four property types in four geographic regions). Recent studies such as Steinert and Crowe (2001), Conover et al. (2002), Lee (2005), Liow, Ooi and Gong (2005) and Lee and Stevenson (2005) have also evaluated and determined the benefits of diversification from both local and foreign/global real estate investments. Adair, McGreal and Webb (2006) also established the diversification effects of direct versus indirect real estate investment in U.K. These studies have shown that different diversification strategies come with different portfolio benefits.