SPHD: Misleading Name Hides The True Risk
The Invesco S&P 500 High Dividend Low Volatility Portfolio ETF(NYSEARCA:SPHD) appears to be a simple way for investors to gain exposure to higher yields with blue chip large-cap stocks while reducing volatility risk. At least that would be my assumption based on the name of the fund. But what you read is not always what you get. Let’s dig a little deeper into this mislabeled fund.
Methodology of SPHD
SPHD tracks the S&P 500 Low Volatility High Dividend index. The index is constructed using two simple filters:
- Select the top 75 highest-yielding stocks in the S&P 500
- Of those 75, keep the 50 which have the lowest realized volatility over the past year
Finally, weight the remaining stocks by dividend yield. Both the fund and index are rebalanced twice a year.
Critical Analysis of Methodology
There are a few questions I have regarding SPHD. Here they are in no particular order.
Why do they use trailing yield instead of indicated yield?
The index selects stocks with the highest trailing 12-month dividend yield. That’s in the past. But unless investors have a time machine, they will get the future dividend. Other funds such as SPDR Portfolio S&P 500 High Dividend ETF (SPYD) take the most recent dividend payment and then annualize it. This weights towards the most recent dividend. You could also weight a portfolio towards announced dividends. Would this make much of a difference in realized dividend yield?
- The simple average of the 75 highest yielding stocks using indicated annual yield in the S&P 500 is around 4.56% (this is also the forward-looking indicated annual dividend yield).
- The simple average of the 75 highest yielding stocks using trailing 12-month dividend yield is 4.45% (reported using indicated dividend yield).
There is not a big discrepancy at this moment in time. In general, there will be a strong correlation to high yielding stocks over the past 12-months and over the following 12-months. But not always. The stock PPL Corp (PPL) is just one example of a stock where the dividend was slashed making the future yield much lower than the trailing yield. Due to the high 12-month trailing dividend yield it is still included in the fund. I would argue that it shouldn’t be present because realized dividend yield will be much lower than trailing.
I am not sure why they use the trailing 12-month dividend yield. Is it easier to calculate? I doubt that’s the reason. My personal belief is related to back-testing bias. The historical simulated total returns are slightly higher (a fraction of a percent) using trailing dividends in the calculation. But maybe there is another reason such as smoothing out irregular dividend payments. I don’t know.
Why do they brand this as a low volatility portfolio?
I find it a stretch to label this as a low volatility portfolio. Keep in mind the screening process. They retain 15% of the S&P 500 index (75 stocks) based on trailing dividend yield. Of that 15% of the index, they remove 5% of the index (25 stocks) which have higher volatility than the other 10% (50 stocks) they decide to keep. I feel that calling this a low volatility strategy is a bit misleading.
- If I screened the highest volatility stocks in an index, then removed those without a dividend yield, would it be appropriate to label this as a high-yield strategy? Not likely.
- If I screened 15% of the S&P 500 index for the deepest value stocks and then removed the 5% with the lowest momentum, could I really call this a deep value – high momentum strategy? Not likely.
So why is this labeled as a low volatility strategy when the volatility component is more like an after-thought?
For investors that are truly looking for low volatility, I find the label attached to this fund to be grossly misleading. This is not a low volatility fund. The trailing average 252-day volatility is in the upper half of the S&P 500 index for some of SPHD’s holdings such as the following:
The chart above shows the standard deviation of daily returns over the past year. Look how much higher the volatility is of a few select holdings as opposed to the SPY. Yes, I cherry-picked them but a fund that claims to be low volatility should think twice before having stocks with this level of volatility.
There are other ways to mix low volatility and high dividend yield such as:
- Having a composite ranking score. Rank volatility and yield independently and then blend the two scores.
- Create a rule that removes any stock in the upper 50% of volatility of the S&P 500 index.
- Or possibly take in the top 100 yielding stocks and keep the 50 that are lower volatility.
The counter argument may be that the dividend yield would be too low if you constructed a portfolio that wasn’t so aggressive in how it filtered high-yielding stocks. But I would argue that the ‘low volatility’ component is too low and it may mislead investors into thinking that they are getting something which they are not. The low volatility factor is a seasoning, and not the meat, in this fund.
This next chart shows the drawdown of SPHD vs SPY ETF in early 2020.
You might expect a low volatility fund to have less drawdown in market crashes. But this didn’t happen in 2020. Higher yielding portfolios carry higher drawdown risk. The ‘low volatility’ filter they use isn’t wrong – but it is likely misleading for many investors. It isn’t low volatility compared to the average large-cap stock.
That being said, I ran my own backtests on the 25 discarded ‘higher volatility’ dividend stocks which were excluded from this fund. Those stocks would have experienced 10%+ more drawdown in 2020. The elevated drawdown risk is true of most bear markets. So while I agree that removing some of the highest volatility holdings is good for risk, it is a long way away from being a ‘low volatility’ hybrid fund.
Why do they weight the fund with trailing dividend yield?
Again, I come back to the same argument with the weighting scheme. Why weight towards something in the past which does not represent the future? The fund tilts heavily toward PPL Corp when it has one of the lowest future dividend yields in the fund.
My second concern is that this weighting scheme may inadvertently weight towards higher volatility stocks. While the ‘highest volatility’ holdings (25 out of 75 initially screened stocks) may be removed, there are still holdings with price action that are not ‘low volatility’. VNO is in the top 5 holdings because of its trailing yield but the volatility is still elevated. IRM makes up a good portion of the fund and it too has higher volatility.
It feels like they try to lean away from volatility by a small amount but then lean right back towards it in other ways.
Final Thoughts On SPHD
I don’t find the performance of SPHD that much different than other funds like SPYD. This is especially true when compared against the SPY.
I believe the ‘low volatility’ addition of SPHD to be little more than a marketing gimmick. The expense ratio of 0.30% is considered high when I compare this to the gross expense ratio of SPYD which is 0.07%. This is especially true when the performance of the two funds is so suspiciously similar.
My final thought for SPHD is that it is an average fund performing an average service for the average investor at an above-average cost. I see no real reason to invest in this fund over the cheaper alternative of SPYD or any other large-cap high-yield fund.