As the sub-sectors of the US REIT market move in a more homogeneous manner so the potential for diversification within the REIT sector has diminished as Simon Stevenson reports

The growth of the global real estate investment trust (REITs) market in the past decade has led to considerable discussion about the investment dynamics of the sector. The US REITs market, in particular, is a totally different beast now to what it was 20 years ago. In the late 1980s it was a niche sector characterised by a small number of firms and thinly traded; the interim years have not only transformed the sector but have also changed its dynamics.

In the early 1990s the sector developed in part because of the introduction of the UPREIT structure. This led to a doubling in size of the equity REITs sector, with 95 REIT IPOs in 1993 and 1994. In the late 1990s, REITs attracted further institutional investment. In the early 1990s the average institutional shareholding in US REITs was approximately 15%. By the end of the decade it was touching 40%. This was added to by S&P's 2001 decision to allow REITs to be constituents of its indices, leading to investment by index funds and increased investor awareness.

The strong performance of the sector in the first half of this decade led to a major increase in the inflow of funds. From 2000 through to mid 2007 REITs consistently outperformed the overall US equity market, while in the five years to the end of 2006 REITs delivered a cumulative return of 184.9% compared with 39.6% from equities. This increased institutional investment in one of the few equity sectors to produce strong returns during this period contributed to changes in the sector's investment behaviour.

Even ignoring the large-scale trading of the past three years associated with the current financial crisis, trading volume in US REITs had risen considerably, particularly since 2000. SNL financial data reveal that in 1993 the average daily aggregate trading volume was 2.5m shares across the sector. By 2000 it had increased to over 16m shares a day. However, by 2006 it was in excess of 50m shares.

Many of these factors have had major positive effects on the REITs; they have increased liquidity and broadened the investor base. But there is a price to be paid with such change. It is well documented in financial markets that trading volume is a key determinant of volatility, and recent research has shown that the increase in trading volume has contributed significantly to the higher volatility of REITs.*

In addition to the impact on volatility, the changing nature of REIT investors has had a major effect on the sector. One of the key issues in the analysis of any investment asset is to consider who the investors are and what is the nature of their comparison. During the last 15 years, REITs have become increasingly dominated by mainstream equity traders, to the extent that their natural comparison is not with private real estate but with other equity sectors.

An example of this was from 1998-2003. In 1998 and 1999 US REITs produced an average monthly return of -1.12% in comparison to an average figure of 2.06% for the overall market. In comparison, from 2000-03 the corresponding average returns were +1.11% for REITs and -1.21% for the overall equity market.

Many people have viewed this period as one where REITs became detached from the broader stock market. However, one can take the perspective that the sector was actually being influenced more by equities than ever before, and that during the 1998-2003 period REITs were effectively acting as a counter-cyclical defensive sector. Simply put, why would an equity manager invest in a relatively low-growth income sector like REITs during the dot.com boom? Likewise, the sector provided a safe haven during the uncertainty of the three years after the 2000 technology crash, which also included 9/11 and the subsequent weakening of the US economy.

Possibly the most interesting shifting dynamic has been the relationship between the sub-sectors in the US market. It is well known that US REITs tend to be focused in terms of property type. In October 2009, of the 113 listed equity REITs only nine were classified as diversified. Early studies showed a strong relationship between the performance of REITs and the underlying private real estate sectors. This is one of the key issues in the argument in favour of focus, namely that it allows investors to make the diversification decision for themselves. However, over the last few years, as the market has matured and become more mainstream distinct trends have been observed.

Figure 1 shows the rolling correlations between office REITs and the other sectors. A clear upward pattern in the correlations can be seen, together with a tightening of the coefficients. This effect is seen across all sectors and is not just unique to office REITs. Effectively, REITs sectors have increasingly moved in a more homogeneous manner. This has obvious implications for the diversification potential within the REITs sector. In addition, it would indicate that the markets are increasingly viewing and pricing REITs more as a sector in its own right.

It is important to realise that there are differences in performance of the sub-sectors. The trends have not resulted in the total absence of real estate factors, whether at a property type or individual firm level. Whereas the difference in performance in the early 1990s was more absolute in nature, today it is more relative. It is perhaps a more subtle relationship than it once was, but while the movement of the overall sector may be increasingly determined at an equity market level, the real estate issues are still present.

It is therefore vital for any investor, whether coming from a real estate or equity perspective, to appreciate these changing dynamics. The nature of REITs, as stocks whose business is real estate investment, means that they will always share some characteristics with both the broader equity market and the underlying property market. The key for an investor is to appreciate how each influences the REITs market.

*Cotter & Stevenson (2008), ‘Modelling Long Memory in REITs', Real Estate Economics, 36:3, 533-554