Friday, October 29th, 2021

Adding REITs to the Mix

We are getting lots of questions from clients about how they can protect their portfolios from inflation and still generate some income – hardly surprising given where bond yields are at the moment. The obvious asset of choice is real estate. Because we only analyse exchange-traded securities, we will focus on REITs, but individual clients may want to consider funds or physical real estate as well, depending on their mandates. Coming straight to the point, it is clear from our work that diversification into REITs can offer a superior long-run returns on both an absolute and a risk-adjusted basis. The question is how to do it.

The simplest way is to take a constant 50/50 equity bond portfolio in the US and add a fixed exposure to real estate. For reasons which will become apparent later, we have settled on 12%, giving us a new 44/44/12 portfolio. Since inception in 1995, this has produced an annualised return which is 50bps better than the 50/50 portfolio, but 60bps more volatile. The return on risk is slightly worse (0.92 vs 0.93) but most investors would probably regard this as a reasonable trade-off, given the extra optionality/flexibility of the enlarged portfolio.

However, this approach denies us the main advantages of owning real estate via REITs (and better still an ETF based on a REIT index), which are better liquidity and much lower friction costs. REITs were designed to be actively traded, so we shall. If we use our normal risk-adjusted momentum process, the excess annual return goes up from 50bps to 80bps and the excess annual volatility drops from 60bps to 30bps. This gives us a return on risk of 1.00 vs the original 0.93 – i.e., a diversification strategy which ticks the two most important boxes: higher absolute returns and higher risk-adjusted returns.

So far, so good. The next question, is what REITs should we own. Is it better to focus on the top-level REIT index or can we generate extra returns by looking at the sub-indices? For the purposes of this exercise, we focus on the three largest in the US, which are Industrial, Retail and Residential. Clients may prefer to include other categories such as Offices or Hospitality, but the conclusions will be broadly similar.

We find that the annual total return generated by rotating between the three sub-indices is 170bps better than the broad REIT index, but the annualised volatility is 190bps higher. The return on risk is better, but not by much. More importantly, we find that the rotating REIT portfolio is more correlated with the 50/50 portfolio, which means that combined portfolio loses some if its risk-efficiency. In English, the risk-adjusted returns of the rotating REIT portfolio are marginally worse than those of the broad REIT index, when combined with a 50/50 portfolio.

However, this is not the final test. We want understand how an actively managed portfolio of REITs can be integrated into an actively managed portfolio of fixed income and equities – our basic portfolio. In its own, this has produced better absolute and risk-adjusted returns than the standard 50/50 model. If we add our rotating REITs portfolio, we take the annualised return up 10.7% for an annualised volatility of 10.4% and a return on risk of 1.03.

However, we find that it is essential to control the maximum potential exposure to REITs because the drawdowns, when they occur, are very much greater than in equities or fixed income (75% at worst and frequently 30-40%).  We could use a hard limit of say 15-20%, but we find that using a scalar, (i.e., multiplying the unconstrained weight by a constant factor) produces better risk-adjusted returns. Using a scalar of 25%, we find that our average recommended weighting to real estate is just over 13%, compared to a benchmark of 12.5% (i.e., 50% x 25%).

If we set the scalar at a higher level, we find that the annualised volatility increases by more than the annualised return and that risk-efficiency decreases. More importantly, the maximum drawdown increases disproportionately, which is not what clients pay us to design. All these factors persuade us to err on the side of caution and go for a 44/44/12 benchmark.

So far, this discussion has concentrated on the US, partly because the data quality is better and partly because REITs have been more widely used as a diversification tool for longer. US investors can get all the diversification they need without going outside their own currency. This is not true for euro-based investors. It is possible to construct a rotating REIT portfolio in Europe, but the publicly available indices only go back to the middle of the last decade.

The bigger issue is that US and Asian REITs have performed significantly better than those in Europe, so euro-based investors must decide if they want to take advantage of this. They must consider how much currency risk they want in their overall portfolio, and whether investing in international REITs will restrict their ability to invest in international equities and/or international bonds. As our last note made clear, we are in favour of prioritising the asset class decision over the currency decision, but there are times when this can lead to an uncomfortable concentration of currency risk.

Nonetheless, we are confident that it is possible to create a euro-denominated portfolio including REITs which would outperform our benchmark and our basic portfolio. In the US, the result of all the refinements outlined above is to take the overall return from 8.1% on the 50/50 up to 10.7%. In addition, there is a style-timing model which takes it up to 11.1%, for a return on risk of 1.12. Over 25 years, this is the difference between a cumulative performance of 1530% vs 750% (base = 100% on 31/12/1995).

So, diversification into REITs really can work, provided clients (1) take advantage of the liquidity they provide, (2) follow a systematic process, and (3) limit their maximum exposure.

And finally, a note of caution. Many investors are considering this strategy and judging by recent returns in the US, many have done the trade already. The broad REITs index has generated a total return of 32% YTD, while our rotating portfolio is up by 43%. Investing in REITs is a successful diversification strategy, but that doesn’t necessarily mean that you want to buy a full weighting immediately.

Leave a Reply

Full Blog Archive:

Other Recent Posts