Boxwood Capital Research LLC is a REIT investment research affiliate of Boxwood Means, Inc. Our general research strategy and methodology are outlined below.
The research strategy that underpins the model design is to uncover the factors or issues to which the market is currently responding - and not to reveal the issues to which we think the market should respond.
The assumption of factor-based quantitative models like ours is that investment styles come and go, but do so in a slow evolutionary way, not with quick starts and stops. That is, the dominant styles (there is usually more than one) in the market tend to persist from month to month. The persistent style(s), once identified, yields information that may be an effective predictor of forward retums.
Our research strategy pursues the following:
Our quantitative models incorporate massive quantities of real estate market data and securities-related information on roughly 120 REIT firms. We initially estimate asset-level REIT retums based on property income, rental growth, occupancy rates, employment and other variables based on time-tested models of historical and projected real estate market and macroeconomic conditions that are observable and quantifiable.
Such real estate metrics are then combined with roughly 40 company financial factors that are correlated with REIT price movements. Our factor models evaluate all REIT stocks and quantify a score or estimated forward one-month return for each one of the companies. The resulting relative REIT rankings inform decisions on long positions in the highest-scoring REITs and short positions in the lowest.
In addition, the Boxwood scores are the foundation for constructing optimized portfolios customized to the risk tolerances of institutional investors.
Our REIT investment research is provided to clients on a monthly subscription basis.