- Dividend growth stocks are desirable for income and retirement-focused investors.
- Low-volatility filters can reduce portfolio gyrations, max loss and even boost returns.
- Low-volatility filters combined with value factors can provide further enhancements.
Low Volatility and Dividend Growth is a Winning Combo
Dividend Growth Investing is a popular topic in the Seeking Alpha community and for many good reasons.
- Confidence. A company that can afford to consistently pay increasing dividends year after year exudes confidence. When you hear of a company cutting or failing to raise the dividend, you wonder if management is beginning to worry.
- Simple and logical. An investor can easily create an investment and withdrawal plan based on increasing income payouts. It is sensible and logical.
- Protects against emotional trading. When the markets crash, investors typically sell when the pain is highest after a major loss. Dividend growth investors are able to focus on soaring yields, and their excitement level is often inversely proportionate to the share price decline. As long-term value/income/growth investors -- they will likely accumulate in a bear market while others are selling.
I have written various articles on how to combine dividend growth investing with various other strategies or factors -- but the one area that I have been largely silent about is on combining low volatility with dividend growth. Long-term investors would likely appreciate a reduction in portfolio volatility -- if this can be achieved -- and low volatility investing doesn't carry the high turnover like other strategies, such as momentum or short-term price reversion. While dividend growth investors are no doubt willing to replace holdings when conditions change -- many dislike frequent portfolio adjustments.
The first part of this article is to look at a variety of volatility filters to be implemented on stocks which have at least 5 years of consecutive dividend increases. The second part will add an additional filter onto the most promising low-volatility filter. And the end of this article will consider any additional modifications that may enhance dividend growth investing.
All testing is carried out using the Portfolio123 platform.
To view our results in context, we need a benchmark. Below is an equal-weight custom index that holds all stocks that have at least 5 consecutive years of annual dividend increases. All figures include dividends and assume re-investment of income. The universe is point in time -- which means the 5-year dividend growth minimum is true at every testing point. The universe will include stocks that de-listed or later cut their dividends so as to be free of survivorship-bias.
The next step is to run various volatility filters across this universe. In every example, we will only retain a quartile, or one-fourth, of the total universe. We want to discover which volatility factors work best.
ATRN. The first volatility filter uses ATRN or Average True Range Normalized. We examine the average daily range (daily high minus daily low) over the past 100 trading days, which is normalized to the price of the security. The purpose is to find stocks with relatively low trading ranges instead of large intra-day gyrations.
The weekly turnover using this factor is 5%. This factor may be suitable for those wanting to lessen their daily portfolio swings and possibly reduce the downside in a bear market at the expense of some return. The reduction in standard deviation, and max drawdown is significant and desirable.
3-Year Weekly Beta. This next filter uses the 3-year Beta using weekly returns. The lowest beta quartile is held.
This is an interesting filter that helps the risk-adjusted returns significantly. Standard deviation is quite low. There are some abnormal spikes in turnover though and if this filter was adjusted, I believe that average portfolio turnover would be only 1% or 2% per week instead of 5%.
3-Year Daily Beta. This will be the same as above, but it will look at the daily beta over the past 3 years and keep the stocks in the lowest quartile.
I find this particular volatility filter desirable as turnover is very low -- only 2% per week -- and returns are enhanced slightly. The drawdown is reduced and standard deviation is half of our benchmark. We will re-visit this filter later on.
1-Year Trailing Sharpe. This is a very common ratio that compares returns to the standard deviation of the returns.
Interestingly enough, I find nothing compelling about using the trailing 1-year Sharpe Ratio even though it is widely used by advisors and investment dealers to promote products. In fact, there may actually be a case against using this metric as a screening tool. While a high Sharpe ratio is good when evaluating the past performance of your existing portfolio, I would advise against using the trailing 1-year Sharpe Ratio when deciding which investment products to buy -- at least as it relates to dividend growth stocks.
1-Year Trailing Sortino. The Sortino ratio is closely related to the Sharpe, but it only considers downward volatility instead of both up and down moves.
Again, this ratio is useful in looking at your portfolio ex-post, but I am dubious about using the trailing Sortino ratio when considering which dividend growth stock to purchase for future low volatility.
This concludes our first round of testing and the next section will add a second filter onto our most promising volatility filter -- namely the 3-year Beta using daily returns.
Adding a Second Filter
My favorite volatility filter tested above is the 3-year daily Beta. Can we further enhance portfolio performance by adding a second filter that will keep 50% of the remaining stocks? The table below details the results obtained by taking our low volatility quartile and keeping the best half according to our second filter.
|Total Return||Annualized Return||Max Drawdown||Sharpe Ratio||Sortino Ratio||Standard Deviation||Correlation with Bench||R-Squared||Beta||Alpha||Weekly Turnover|
|Benchmark - Low 3 Year Daily Beta quartile||474.81%||11.89%||-43.67%||0.97||1.21||8.17%||0.58||0.34||0.24||7.42%||2%|
|High 52 - 4 week performance||610.42%||13.35%||-31.99%||1.06||1.33||8.85%||0.59||0.35||0.26||8.79%||9%|
|High 3 Year Dividend Growth||563.64%||12.86%||-44.55%||0.9||1.13||9.90%||0.51||0.26||0.25||8.33%||2%|
|High Sustainable Growth Rate||426.59%||11.20%||-40.35%||0.81||1||8.91%||0.56||0.31||0.25||6.74%||2%|
|High Earnings Yield||1129.87%||17.40%||-37.69%||1.48||1.85||9.11%||0.45||0.2||0.2||12.75%||5%|
|Low 90 Day RSI||1385.94%||18.83%||-35.92%||1.46||1.83||10.19%||0.51||0.26||0.26||14.07%||14%|
|High Working Capital to Price||520.03%||12.37%||-43.67%||1.01||1.26||8.30%||0.62||0.38||0.26||7.86%||5%|
The three best secondary filters would be earnings yield, dividend yield and low 90-day RSI. However, I would exclude the low 90-day RSI due to the higher turnover, which is not desired in this type of investing.
The 10 largest names (marketcap) in the higher yield filter are as follows:
|(NYSE:VZ)||Verizon Communications Inc||Diversified Telecommunication Services|
|(NYSE:T)||AT&T Inc||Diversified Telecommunication Services|
|(NYSE:EPD)||Enterprise Products Partners LP||Oil, Gas & Consumable Fuels|
|(NYSE:DUK)||Duke Energy Corp||Electric Utilities|
|(NYSE:KMP)||Kinder Morgan Energy Partners LP||Oil, Gas & Consumable Fuels|
|(NYSE:SO)||Southern Co (The)||Electric Utilities|
|(NYSE:BCE)||BCE Inc.||Diversified Telecommunication Services|
|(NYSE:WPZ)||Williams Partners LP||Oil, Gas & Consumable Fuels|
|(NYSE:PPL)||PPL Corp||Electric Utilities|
|(BSYBY)||British Sky Broadcasting Group PLC||Media|
What else can be done to customize a dividend growth portfolio so as to be tailor-fit to the investor's needs?
1. Expand the universe. In addition to including dividend growth stocks -- include those that have either raised or maintained the dividend for 5 consecutive years. Perhaps the company kept the dividend the same for two years in a row. These companies are still worth looking at and it will expand the database of investable securities.
2. Combine factors in the secondary process and adjust tolerances. For instance, the first filter keeps the lowest volatility (3-year daily Beta) half of the dividend growth universe. The second process filters out and keeps only the top quartile of earnings yield and dividend yield in the original universe. This creates 9% weekly turnover, an annualized return of 24.35%, max DD of 27%, a Sharpe Ratio of 1.67 and an average of 30 tickers held in the portfolio.
3. Apply the low-volatility logic with dividend yield and earnings yield to other universes, which makes sense to do so. For instance, you may apply this system to a high-yield strategy or to the utility sector.
If there is a particular set-up you would like to see, please write about it in the comments section below and I will consider it for a future article.
Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.