The Income vs Diversification Trade-Off
Since the early days of security analysis, research has emphasized the potential benefit of buying stocks at a price that is reasonable to the fundamentals. The oldest benchmark for valuation is the dividend yield. Stocks that trade at a high dividend yield are often referred to as value stocks. Stocks that trade at a low dividend yield are often referred to as growth stocks. Value and growth investing have given rise to dramatically different records of long-term performance. A large body of evidence suggests that there has been a higher long term-term return both domestically and internationally from investing in value stocks.
Value companies sell at prices that are low relative to fundamentals and/or have experienced a price decline. Their dividend yields, earnings yields, and market-to-book ratios are often relatively high. However, persistent poor performance can lead to dividends, earnings, or book values that decline to zero or even become negative. For these reasons companies that pay a dividend of zero, for example, may be more similar in their attributes to high yielding value stocks than to low yielding growth stocks. Zero yielding value stocks are also similar to value stocks in terms of their subsequent stock market performance.
To verify the tendency for higher yielding stocks to provide higher long-term returns, Triumph of the Optimists: 101 Years of Global Investment Returns (Princeton University Press 2002, Elroy Dimson, Paul Marsh, and Mike Staunton) examined long-term U.S. stock market returns and documented a marked historical return premium from U.S. stocks with above average dividend yields. Between 1926 and 2000, the annualized return on high dividend yield U.S. companies was 12.2%, compared with 10.4% for the low yield companies. Over the long term, the historical record of high dividend yielding stocks has been positive in the United Kingdom as well. Using their own record of dividends on UK companies since 1900, Dimson, Marsh and Staunton found that the annualized return on the value index containing high dividend yield stocks was 11.5% compared to 8.6% for the growth index containing stocks with low dividend yields.
What explains the long-run historical value premium? This is a matter of debate that may never be fully resolved. But among US companies, the declining propensity to pay dividends has raised questions about the use of dividends as a guide to value (see Fama & French 2001). In fact other measures prove to do a better job at capturing the performance gap between value and growth portfolios. Fama and French (1992) documented the tendency for stocks that trade at a high book-to-market ratio (value stocks) to outperform stocks that trade at low book-to-market ratio (growth stocks) by a significantly wider margin than traditional value variables demonstrated. Fama and French (1998) also examined dividend yields along with price-to-book and price-to-earnings to identity value factors and premia in international markets. The results showed the presence of the value effect that they and others had documented in the U.S. Of the four variables tested, dividend yield produced the smallest value spread.
What about the dividend stream itself? Many investors prefer dividends over capital gains because they generate income without selling shares. What are the costs of investing only in firms that pay dividends or only in firms that pay only high dividends? How predictable are dividend payments? Using global data from 1991 to 2012 provided by Bloomberg, Stanley Black (Research Associate, DFA) addressed these questions.
Global portfolios of dividend payers and non-payers have had similar compound average annual returns. The simple average returns annual returns were higher for the non-payers than for the dividend payers, but the standard deviation of the returns of non-payers was higher than for the dividend payers.
From 1991 to 2012, the dividend yield averaged 2.3% for dividend payers vs. 2.0% for the global market. And as of 2012 the dividend yield for payers was 3.1% vs 2.7% for the global market. By focusing on only dividend payers, however, an investor would exclude 35-40% of firms. The portfolios containing 50% and 25% of aggregate market cap have average dividend yields of 3.3% and 3.9% substantially higher than the market and all payers. This result is related to the tendency for a large fraction of aggregate global dividends to be paid by the top dividend paying firms.
If aggregate dividends become more concentrated, a portfolio constructed to provide a high dividend yield may also become more concentrated. This is an important consideration for investors. For instance, getting to 50% of global dividends required about 320 firms in 1991 but only 220 firms in 2012, a 31% decrease. Half of 2012 global dividends were paid by 2.4% of global firms.
The propensity of firms to pay dividends has shown a global decline. The data show that the percentage of firms paying dividends globally dropped from 71% in 1991 to 61% in 2012, with declines occurring in both US and international markets.
In addition, the propensity to pay dividends has shown a great deal of variation across countries. Although less volatile than the capital gain component of stock returns, the aggregate stream of dividend payments is subject to the same broad, macroeconomic risks that affect capital gains. As the experience of the financial crisis of 2008-2009 demonstrated, companies can and do cut dividends in the face of declining profits and economic conditions. In 2009, for example, 14% of firms around the world eliminated their dividend, and 43% of firms reduced their dividend.
If an investor focuses on dividend paying stocks, it is important to understand that the inherent uncertainty about future tax rates and policy translates into added uncertainty about future dividend income streams. The recent uncertainty surrounding the extension or expiration of the 15% dividend tax rate in the US, which became effective in 2003, is only the most recent example. While the U.S. top dividend tax rate ended up being raised from 15% to 20%, one possible outcome of the policy debate was dividend tax rates reverting to ordinary income tax rates, which would have been an effective marginal rate tax rate of roughly 40% at the top. The tax treatment of dividends changes over time and future dividend tax changes are difficult to predict. For example, if tax rates on dividends are raised relative to taxes on capital gains, it is likely firms will tend to reduce dividend payouts and increase the use of stock buybacks to deliver income to investors.
BOTTOM LINE: many investors seem to have a preference for stocks that pay cash dividends, because such stocks generate income without the need to sell shares. Investors should be aware, however, of the very real trade-offs between diversification and the pursuit of higher dividend yield. Global portfolios that purchase only dividend paying stocks will exclude about 47% of available small cap stocks. So while investors may be able to achieve greater dividend yield and dampen portfolio volatility by investing in higher yielding stock funds, they sacrifice diversification to achieve that goal. When investors exclude firms from their portfolios that pay no dividends today, they deprive themselves of the returns of the biggest dividend generators of tomorrow. A broadly diversified strategy that includes both dividend payers and non―dividend payers will ensure investors enjoy the potential rewards of both. Additionally, high dividend oriented stock funds are still equities and will rise and fall with the broad market. Even though yields on fixed income are at historic lows, high dividend stock funds are not bond substitutes, even though they may yield as much or more than short-term bond index funds. If you are using high yield stock funds to boost your income beyond what current interest rates will bear, you are taking equity-correlated risks whether you realize it or not. Your portfolio will be eventually be punished. Having said that, high dividend oriented stock funds can be introduced into a portfolio as a component within the large-cap asset class. Finally, in spite of low interest rates, fixed income still has an important role in portfolio allocation to dampen market risk as a result of relatively low correlation with more volatile equities. Investors should take a balanced approach, one that accounts for all investment related considerations, when choosing portfolios to achieve their investment goals.