"It can't really be that easy." Dividend or dividend growth investors sometimes sound too confident and assured: sooner or later those of us who follow other methods simply fall back on the easiest counter-argument. This article presents the results of backtesting a portfolio of low beta, high quality dividend payers. After poking and prodding the results, the answer is: "It would have been just that easy for the past decade or more." Very possibly it will be that easy going forward.
A Ten Year Backtest
Using StockScreen123, the following screen was backtested:
- S&P 500 (SPY) member
- Equal investment in top 20 candidates, as determined by the site's Greenblatt ranking system, which focuses on quality
- Reconstituted every 3 months, 0.5% slippage
Here's a chart of the most recent ten year period:
The results are impressive: IRR for the 10 years is 13%, compared to 7% for the S&P 500 including dividends. The screen outperformed by 4% (every 3 months) in down markets and underperformed by 0.5% in up markets. Turnover averaged 25% for 3 months.
Critiquing the Results
The data source provided by the website prevents survivorship bias. Reconstituting the portfolio every 6 months or even annually improved performance. Results were comparable when applied to the Russell 1000 or the stock screen large cap universe.
There is some large cap bias here. The screen outperformed the S&P 500 when applied to small or medium cap universes, but not by as wide a margin.
The Greenblatt ranking system relies on two inputs: Return on Capital (50%) and Earnings Yield (50%). Impressive results were also realized using a Basic Quality ranking system, which is more complex but still focused on quality.
The maximum period available goes back to 1999, and results were just as good for the longer period.
Shorter time frames provided generally consistent results. The exception was the 5 years period from 2002 to 2008. In this interval between the two major bear markets, the screen's advantage in down markets was less important, and performance was pretty much the same as the index.
Naming Some Names
The current output includes some familiar names:
While there is a concentration in Pharmaceuticals and Tobacco, excluding both industries reduced but did not eliminate the performance advantage.
What About The Financials?
Critics of dividend-oriented investing often cite the financials (XLF) problem. Briefly, prior to the financial crisis, banks were common in portfolios aimed at dividend income. This screen doesn't directly address the issue of commercial or investment banks and they will be included if they meet the basic criteria. Excluding financials didn't affect the outcome by a meaningful amount.
Subsequent to the financial crisis there are no banks that meet the basic criteria. Prior to the crisis, there were low beta dividend paying banks, but they didn't rank high on quality. The portfolio will include P&C insurance companies from time to time.
First one thing works, then another. That being said, a strategy along the lines suggested by the screen can't do much harm, and will do quite a bit of good if any of the current macro concerns develop unfavorably.
The covered call strategy is subject to the criticism that it gives away the upside while accepting the full downside. Logically, covered calls over a portfolio of the type produced by this screen would resolve some of the criticism, since the quality of the stocks reduces the downside and the upside is more muted.
Disclosure: I am long SPY.