In the last installment, we looked at replacing REITs with other common asset classes, but what if we stretch a little farther? Utilities ETFs offer promise at doing all the things people say they want REIT ETFs to do…but better. Again, check out the backtests!
The theory behind why REITs are included in some Risk Parity portfolios goes a little something like this:
REITs are a sub-asset class of equities that offer pretty good returns but with low correlation to the core of your equity portfolio, thereby diversifying the portfolio and providing smoother, but still strong, returns. They offer a way to track the real estate market, with all its unique twists and turns, and so make up a separate return stream. Plus, they have the added bonus of offering some dividends. What’s not to like?
First off, nothing and REITs are fine. I have absolutely nothing against REITs like I do against corporate bonds (don’t get me started) or like Frank Vasquez has against TIPS (don’t get him started).
But I am trying to uncover whether the conventional wisdom on REITs is actually all that wise. REITs are and have long been a part of my actual portfolio and my thinking on Risk Parity portfolios, so I wanted to examine whether REITs actually do what we think they do, and if there might be other ways to do them better in a RP portfolio.
Given what we look to REIT ETFs for, my mind turns to a sector ETF that might fit the bill better than REITs: Utilities. These are funds of publicly-regulated but privately-held companies in electricity, natural gas, waste management, and other municipal infrastructure services. They are typically large companies with restrictions on how much they can charge for whatever service, but also enjoy monopolies in their areas of service. This leaves them without a lot of upside, but also less downside: people are always going to need to keep the lights on and the water running. Here is a good explainer on utilities and their pros and cons.
As an investor, there is also a plausible reason for why they would behave differently than the market as a whole. In the post I did on REITs and Dalio’s three criteria, I cited one study of the 11 different sectors of the stock market and noted that REITs were the third lowest. Utilities were the lowest, and by a fairly big margin: .4, compared to REITs’ at .59. At the same time, the price appreciation of Utilities ETFs has outpaced that of REIT ETFs.
To demonstrate, here is the data from Portfolio Visualizer. I used State Street’s Utilities Sector SPDR (XLU) for this, since it is six years older than Vanguard’s VPU, though for the backtests I’ll go over later, I used VPU. Fidelity’s FUTY is another option. Optimized Portfolio has put together a nice comparison, if you’re interested. The data begins in October 2004.
Looking at the important metrics for judging asset classes: positive return, volatility, and correlation with other asset classes, we can see that compared to VNQ, XLU is notably less correlated with the stock market, more profitable (without accounting for dividends - see below), and less volatile.
But What About Dividends?
Leaving off the discussion about whether dividends are relevant anyway, they are cited as a positive reason to invest in REITs. When comparing VNQ and XLU for dividend yield, the two funds are quite close, though for most years, VNQ does have a slight edge.
So an edge for VNQ here, but one demerit (for American investors) that has to be accounted for: REIT ETFs pay out regular dividends, which are then taxed as ordinary income. Utilities, though, pay out qualified dividends as long as you keep them the requisite time. If you are paying taxes on dividends, then you’d much rather have XLU. Whether it’s enough to overcome the slight edge that VNQ has in payouts is impossible to say, since it depends on circumstance, but clearly, it matters.
How Would These Work in a Risk Parity Portfolio?
Using Portfolio Visualizer, I then ran two backtested portfolios below to explore utilities as an alternative to REITs. Usually I like to use Portfolio Charts for backtests, since the time frame is so much longer, but in this case I’m looking at specific assets, so the “Backtest Portfolio” tool on PV is the best method. One important note: the data is constrained by the relative youth of UPRO, the 3x leveraged S&P 500 fund which started in January 2010. That is not nearly enough data for me to feel confident with, but it will have to do.
Once again I am using the RPC Income as my baseline portfolio. Since I just did the breakdown of what’s in the RPC Income in the previous post, I’ll refer you to that, but briefly:
- 25% Equities, but functions like 45% since 5% each are dedicated to UPRO and TNA which are 3X leveraged.
- 30% Bonds, but functions like 60% since 15% is in TMF, which is 3X leveraged.
- 15% Commodities
- 15% Gold
- 15% REITs in the form of VNQ.
For the backtest, I switched out VNQ for VPU. Importantly, Portfolio Visualizer allows us to reinvest dividends in their portfolios, so we get a fuller appreciation of that aspect for both VNQ and VPU versions. There was also annual rebalancing for the portfolios to bring them back to their target allocations.
The results:
It’s not a big difference in results, for the simple fact that you are only changing 15% of it. Still, there is a clear advantage for the portfolio with VPU - better in all five metrics. The lower correlation of VPU seems to be doing wonders.
For a second opinion, and to get a richer sense of the impact of VNQ and VPU, I also ran a quick check for those assets exclusively. The data for that goes back to January 2005, and in this case, the superiority of VPU is much more obvious:
Whoa. That’s pretty much a clean sweep of everything we supposedly look to VNQ for in a portfolio. Keep in mind that VPU has a lower correlation to equities as well, so these results have a meaningful impact when considered in the context of an entire portfolio.
When you take into consideration the correlation numbers, the return profile, the dividend history, the type of dividends paid out, and then importantly, the data on what REITs or utilities have done inside portfolios, then simply put, VPU seems just better at doing the things we want VNQ to do.
In the next post in this series, I’ll present another alternative to REITs that involves really thinking outside the box - away from ETFs entirely.