His main thesis is to narrowly compare the absolute performance of high-dividend-paying stocks to value stocks. This ignores academic and practitioner research documenting empirical results and theoretical insights that support sustainable high-dividend payers.
While often thought of as similar, value and dividend strategies are distinct. Value stocks are often good investments, but sometimes they are cheap for a reason and consistently exhibit higher volatility. Ignoring their superior risk-adjusted returns, and the empirical and theoretical backing for lower drawdowns, does high-dividend payers a disservice. In fact, my firm’s analysis shows that:
Screening out the highest – the most risky and unsustainable – dividend payers can lead to the best results, consistent with existing research (Patel et al., 2006). Research shows that high-dividend paying stocks’ management teams are more conservative, less driven to empire building, and tend to steer safely clear of negative surprises (Arnott and Asness, 2003).
Consistently achieving earnings and cash flows strong enough to generate high, sustained dividends requires a management team with commitment and competence. High-dividend payers are spread across industries and sectors, allowing for portfolio diversification as well. We focused our analysis on the Russell 3000 universe from January 1979 through October 2017. This high-dividend stock basket was constructed by, first, screening out the most expensive decile of stocks in a manner consistent with many academic papers and, second, selecting the second-highest 10% of dividend-paying stocks (skipping the highest 10% as potentially unsustainable, consistent with Patel et al., 2006). Then we equal-weighted this basket to create a diversified basket of 300 stocks.
For value stocks, we ranked the universe by price/earnings and price/book, to identify the cheapest stocks, and then created an equal-weight basket of the cheapest decile of stocks, also a diversified basket of 300 stocks.
Our main finding is that a more diversified high-dividend basket has lower drawdowns compared to a diversified value basket. Good companies with the financial discipline to keep paying dividends often draw down less when markets suddenly fall, due to their flight-to-safety nature.
What we also found is that, over this broad universe of the largest 3,000 stocks, companies that are the most likely to pay sustainable high dividends had similar returns, 30% lower volatility and a 30% lower maximum drawdown than value stocks.
The largest drawdowns occurred during the global financial crisis, with high-dividend stocks falling 49.1% compared to 70.7%.
The maximum-drawdown difference between Swedroe’s cited research and ours was caused by our use of a more diversified basket of 300 stocks, for both value and high-dividend payers, rather than the more concentrated 100 stocks used by the authors in Swedroe’s analysis, and employing the broader Russell 3000 index rather than a more limited larger cap universe. During the financial crisis, knowing these numbers would have made it an easier choice to stay invested in the market. When offered smaller drawdowns and better risk-adjusted returns, most investors would take those characteristics every single day and say, “Thank you.” History suggests we may all be reintroduced to these concepts, soon enough.
Contrary to the research Swedroe cites, dividends are valuable. By reinvesting their dividends during down markets, investors can increase the number of shares they own, which will increase in value as the stock price rises.
As an alternative, investors can sell stocks to produce required income, as Swedroe points out. But in down markets that can lead to dollar-cost “ravaging,” as investors have fewer shares left to participate in market recoveries because they sold at a low. Vehicles such as ETFs allow an easy way to reinvest dividends and take advantage of the compounding effect of dividend reinvestment.
So picking dividends over value may be, as Swedroe suggests, a stubborn human choice. Yet it is neither illogical nor irrational, and that choice is conclusively supported by facts and data.