The Returns to Single Family Rental Strategies

In the wake of the 2008 housing bust, Single Family Rental (SFR) strategies have become popular investment vehicles for private equity funds and hedge funds. Industry estimates place the number of homes purchased by SFR investors from 2011-2013 at over 350,000, which is about 10% of all transactions. However, these purchases are not evenly distributed geographically. Homes purchased for SFR strategies are in fact concentrated in states that experienced larger house price declines from 2007-2010 because investors have focused purchases on distressed and foreclosed properties. In these areas, then, well over 10% of recent transactions are SFR related. The emergence and evolution of SFR strategies poses a number of interesting questions for housing capital. First, are investor purchases quantitatively important for explaining the observed 11% nationwide house price appreciation (HPA) in 2013? Distressed markets such as Las Vegas, Phoenix, Miami, Detroit and Riverside all exhibited HPA over 10% from 2011-2013. Second, are single family rental strategies a sound investment, and if so, what are the main drivers of SFR portfolio returns? In this paper, we propose to study these questions using detailed, monthly zip code level data on single family home prices as well as monthly zip code level rents by home type and number of bedrooms along with a simple discounted cash flow model of SFR returns. In our initial work, we document the following stylized facts: (1) In the cross section, as well as in the time series, the returns to SFR strategies are driven mainly by house price appreciation. (2) Long run HPA is determined by a cointegration relationship with income. We build on work in house price forecasting (as performed by practitioners such as Case-Shiller, Moody's, Core Logic, and Goldman Sachs) to show that SFR strategies thus far have succeeded mainly due to corrections to short run deviations from trend house prices. (3) There is substantial heterogeneity at the MSA level in house price appreciation. We find that this appreciation is related not only to overall income growth and land supply inelasticities, but also to which percentiles of the income distribution the MSA serves. We argue that this is because, in recent decades, only the top percentiles of the income distribution have experienced income growth. As a result, other moments beyond the mean or median of the income distribution matter for the impact of income growth on house prices, consistent with within city assignment models of house prices. (4) The heterogeneity in HPA within an MSA across zipcodes is at least as large as the heterogeneity in HPA across MSA's. We explore the extent to which traditional risk measures, gentrification effects, and/or innovations to housing finance, can explain cross-zipcode patterns in HPA, and relate our findings to the prior literature.