This one is really an experiment - comparing three equity funds, VPU, PFF and USMV, with somewhat low correlation with the S&P 500. Putting them through their paces to get a sense of whether low(er) beta equity funds may be another arrow in the RP quiver.
Of all the asset reviews and comparisons I have done, this one represents the most wide-open inquiry. Following Dalio’s maxim that the holy grail of investing is to put together multiple, uncorrelated asset streams, the pursuit is on for funds that deliver strong returns while behaving differently than the typical equities you probably already have.
When correlation with the S&P 500 gets below about .5, you can start to see some real diversification benefits in a portfolio: smoother performance, higher lows, and better risk-adjusted return. That range is typically reserved for “alternatives” - gold, commodities, managed futures, and the like – but are there equity funds which can get in that range?
Consider this the first step in figuring out if pursuing “low beta” equity funds can improve your portfolios. There is a big question in there of whether they will or not which I’d like to come back to in a future post. In this one, I’m going to tackle the concrete question of which equity fund(s) are the best representatives of this strategy.
This will be round one of the search, as I’m sticking with some of the funds I already know about or even use in my test portfolios. Later, I’ll broaden the search and approach it from another angle. But for now…
The Candidates
VPU
Vanguard’s Utilities ETF, Fund page, Barchart Summary. I looked at VPU briefly in my series on REITs and found that VPU comfortably outperformed VNQ, Vanguard’s REIT ETF. VPU is currently in one RP test portfolio, the RPC Income portfolio where it has a 7.5% allocation
PFF
iShares Preferred and Income Securities ETF. Fund page, Barchart Summary. This fund holds preferred shares in a range of companies, but predominantly utilities (90%), with smaller allocations to industrials and healthcare companies. Preferred Shares are a class of stock in which investors generally sacrifice capital appreciation for higher dividends and a priority claim in case of liquidation. I think of them a bit as equities that behave a bit like bonds. I currently hold PFF in three portfolios, the RPC Stability (10% allocation), the RPC Income (15%), and the Levered Butterfly portfolio (10%).
USMV
iShares MSCI USA Minimum Volatility Factor ETF. Fund page, Barchart Summary. Tracks an index of equities with lower volatility than the broader equity market. It’s around 73% large-cap and 27% mid-cap, with about a 2:1 value/growth split. This one I do not currently have in my test portfolios, nor do I own it in real life. Been hearing about it here and there for a while, so I wanted to check it out..
As for the basics of each fund:
How Will I Decide?
Let’s establish the criteria to make a decision:
- The lower correlation with the stock market the better, though I can’t expect the correlations will be as low as gold or commodities. Less correlation with the S&P and even small-cap value will lead to diversification benefits, namely a stronger perpetual withdrawal rate.
- As long as returns are decent, of course. Using a total return perspective, I really want to see how RP portfolios with each of these will fare.
First Criterion: Correlation
I wanted to compare these three funds with a variety of different equity funds, so I chose VTI, VUG, VIOV, VXUS. VTI is my current core US equity fund. VUG and VIOV are my example of a two-part “barbell” approach to US equities, though in real life, I’d replace VIOV with AVUV. VXUS is used to represent international equities, though as I wrote about recently, DFAX would be my actual choice.
Here is the full data from the correlation tool on Portfolio Visualizer with a timeframe back to November, 2011 :
Since USMV is a younger fund, I did a second test to get a comparison back to PFF’s birth in 2007 (note: this also meant changing VXUS for VGTSX, and VIOV for VBR). Results were directionally similar, with VPU less correlated to the other equity funds than PFF, though this time by smaller amounts. Instead of a .3 gap between VPU and PFF in terms of correlation with VTI, the gap was just .05.
From these comparisons, VPU is clearly the least correlated with other equity funds, by a decent amount compared to PFF, and then by a significant amount compared to USMV. Only VPU is in the range of what I was thinking about when I first started the search (.5 or so). PFF might better be termed “marginally lower beta.” Mental note: a second round casting a wider net would be in order.
Meanwhile, USMV probably can’t even really be called “low beta.” With correlations at .89 with VTI, USMV actually is much closer to small-cap value (,87) than it is to either VPU or even PFF. This could be an important factoid for the future. USMV should probably be judged alongside small-cap value instead of other “low(er) beta” funds.
Second Criterion: Total Performance
To run the backtest, I used the skeleton of the Golden Butterfly portfolio, but substituted the 20% normally allotted to small-cap value to go to each of the three funds:
- 20% VFINX: Vanguard S&P 500 Mutual Fund
- 20% TLT: iShares Long-term Treasuries
- 20% SHV: iShares Short-term Treasuries
- 20% IAU: iShares Gold Trust
- Then 20% in either VPU, PFF, or USMV
- I also ran one iteration of the original Golden Butterfly, with VBR (Vanguard’s Small-cap Value fund) in this spot.
The primary backtest goes back to November, 2011 (here is the link for VPU, PFF, and USMV; here is the link with VBR):
You can see her a slight edge to the fund with USMV over VBR and VPU, with all a healthy distance above PFF. You get slightly more return with USMV, with slightly less volatility, but a larger maximum drawdown (which has been January to September of this year, incidentally). A real eye-opener is just how bad the portfolio with PFF has been. It’s hard to believe that one substitution could make that much of a difference, when it’s only 20% of the portfolio.
I then ran a backtest with the VPU, PFF and original VBR versions with a longer timeframe, back to April, 2007. Since VBR and USMV were pretty similar in the first test, I would bet that the longer results for USMV would be in the ballpark of VBR.
Again, the portfolio with Utilities is just behind Small-cap Value (and likely USMV if it had existed), and then both are comfortably ahead of PFF. This query is not looking great for PFF, for sure.
Looking at the funds individually through Portfolio Visualizer’s Fund Analysis tool, you see the superiority of USMV over the past twelve years.
USMV has higher CAGR and Sharpe Ratios than VPU by a good amount and PFF by a lot, with a standard deviation in between the two. On performance alone, USMV is the clear choice.
Other Considerations
I usually track a few nuts-and-bolts elements of prospective funds, namely expense ratios, holdings, and market weight. Not much to say for these three, though. Both VPU and USMV have very low expense ratios, .1 and .15% respectively, while PFF’s is higher (.45%) but not that high. AUM’s are all decent, as well, so you won’t have many problems with liquidity or bid/ask spreads. All three are fairly narrow funds, though: VPU has just 68 holdings, PFF has 501, and USMV has 178. No major issues here.
So…What’s the Preferred Low(er) Beta Equity Fund?
It's complicated. VPU was the better choice in terms of correlation, and the verdict there was that USMV might just be in the wrong category. Then USMV was the clear winner in terms of performance, ahead of VPU. But here, the higher performance leads me to think it should be compared to small-cap value, with which it had a lot in common when running the backtests.
So, even though USMV might be better suited for the investor in the accumulation stage (though whether it is better than small-cap value remains to be seen), VPU is my choice for now as the preferred asset in the category of low(er) beta equity funds.
This doesn’t mean it is better than USMV in general, just that its lower correlation with other equity funds in your portfolio (between .30 and .44 in the correlation matrix above) means it is the better choice as a diversifying asset when you already have allocations to other equity funds. As mentioned, this is a tentative conclusion; I’ll need another search for low(er) beta funds for comparison.
Regardless of that debate between USMV and VPU, we can for sure say that PFF doesn’t look like it has much to offer. I’d rule out PFF unless you have a particular desire for dividends or else you really want a very stable fund - the one category that PFF “won” in the various tests was standard deviation. It is, after all, invested 90% in Utilities through preferred shares, so really, comparing it to VPU is a test of whether you are better off holding common stock or preferred stock, and having capital appreciation or dividends. It looks like common shareholders are getting the better end of the deal.