There have been a lot of attempts by both practitioners and academics to develop a repeatable and consistent investment strategy that can outperform a relevant benchmark in absolute terms, whilst minimising the associated risk measures, be it volatility or maximum drawdown ( the peak to trough percentage) , to deliver superior risk-adjusted returns. Our interest in this topic focusses specifically on the listed real estate (REIT) sector, which is one of 11 separate equity sectors accounting for around 3.5% of the global equity market. The generally accepted benchmark for REIT performance is the FTSE/EPRA/NAREIT Index which comprises the largest REITs, weighted according to their free float market capitalisation (i.e. investible size). Its composition is therefore agnostic to the type of asset (retail, office, industrial etc.) owned by the REITs. The asset composition of the companies in the index is extremely important to fund managers who specialise in the real estate sector, so it is important to isolate the impact on performance of selecting companies by asset type not size (as measured by free float market capitalisation). Companies who have a dominant asset type in their portfolio are designated Specialist (as opposed to Diversified) REITs. In the benchmark the changes to weightings are determined by the change in the free float market capitalisation. A key element of the paper is determining whether an automated trading strategy can be developed to change the portfolio weightings, to improve performance, independent of changes in market capitalisation.
In this paper two specific elements have been examined relating to the performance of portfolios of specialist REITs compared to that of a diversified benchmark:
Firstly, whether automated portfolio weightings of sector specific REITs can be created that can outperform a diversified free-float market capitalisation based FTSE/EPRA/NAREIT EPRA benchmark. In contrast to the free-float market capitalisation weighting of the benchmark index, four alternative portfolio weighting strategies are considered, namely; Equal Weight, Minimum Variance, Maximum Sharpe and Risk Parity. These result in very different portfolio weightings to the index. The size of the US REIT market allows a high degree of specialisation amongst the individual companies, which enables investors to assemble portfolios with weightings based on individual property types rather than merely market capitalisation
Secondly, given that Maximum Drawdown for single sectors (be it asset specific such as Offices, or indeed equity specific such as REITs) is a major concern for practitioners we investigate whether the application of automated Trend Following strategies as an overlay to the initial portfolios can improve risk-adjusted results. The data used for this study is from NAREIT the US for the period 1995-2015 and from EPRA for the extended dataset which includes Europe and Asia for the period 2007-2015.