Dividend strategies tend to have a strong value tilt and lower price-to-book (P/B) and price-to-earnings (P/E) ratios than the market as a whole. But those strategies can become overvalued. When this occurs – we are currently in such a period – building a high-dividend, low-risk portfolio requires extra care.
Dividend-paying stocks can offer investors consistent income, historical outperformance and lower volatility relative to the market as a whole. In a recent article, I reviewed these characteristics as well as arguments by a number of experts that dividend-focused investing may represent a diluted form of value investing. I concluded that selecting stocks on the basis of dividends is a legitimate strategy, with the caveat that one must also consider measures of valuation such as price-to-earnings and price-to-book ratios.
This article takes up that challenge and shows how to construct a high-dividend portfolio that is “cheap” on a value basis.
Larry Swedroe, director of research for the BAM Alliance, a community of more than 130 independent registered investment advisors, and Mebane Faber, portfolio manager at Cambria Investment Management, have expressed concern that U.S. dividend-paying stocks have become expensive in recent years relative to a broader universe of value stocks. Dividend-focused investing is becoming more popular, with the attendant risk that the historical valuation advantages (e.g., that you are buying stocks with low prices relative to fundamentals) will be diluted or entirely erased. The asset flows into dividend-equity strategies – especially domestic stocks – raise the possibility that dividend investing is a crowded trade.
Looking beyond the U.S., the story becomes more nuanced. The yields on a range of sectors in a number of regions exceed those of their U.S. counterparts, but these higher yields come with higher risk. Investors will not be well served by simply chasing yield across borders.
I will review the arguments suggesting that the value advantages of dividend-paying stocks have been diluted. I will then address the crucial question for income investors: Is there a population of stocks that have both attractive yields and valuations comparable to traditional value indexes?
U.S. dividend stocks are relatively expensive compared to a value index
In a recent article, Faber (see Figure 7 in the article) looked at the relative valuation of dividend payers. He found that a portfolio of the 20% top dividend-yielding U.S. large-cap stocks since 1963 consistently traded at lower P/E than the S&P 500. This high-dividend portfolio never traded at higher P/E than the S&P 500 until 2008 but has been consistently trading above it since 2011. This simple but compelling analysis demonstrates that portfolios formed exclusively on the basis of dividend yield have become relatively highly priced, at least since 2008, compared to their historical P/E ratios.
My recent article on dividend investing had a table that showed that the P/E and P/B of two dividend-oriented index funds (SDY and VIG) are higher than the same measures for two value-oriented index funds (IWD and VTV). In fact, those two dividend exchange-traded funds (ETFs) have higher P/Es and P/Bs than the S&P 500. An expanded and updated version of this table is below.
Dividend yield, price-to-earnings and price-to-book for value (in red) and dividend (in blue) ETFs
Source: current dividend yield from Morningstar.com and TTM P/E and P/B from Ameritrade.com
I included an international-dividend fund (IDV) and a dividend-growth fund (VIG) along with two domestic high-dividend funds (DVY, SDY). Dividend growth strategies, embodied by VIG, are less focused on current yield than on dividend increases over time, so it is not surprising that VIG does not have the value tilt associated with other dividend strategies.
The domestic-dividend funds (VIG, DVY and SDY) have a higher P/B and P/E than the value funds. This confirms that dividends are relatively expensive compared to pure value indices. One caveat is important here: The domestic-dividend index funds are also less volatile than the value index funds and have lower betas.
Dividend-paying portfolios typically have a value bias, which at least partially explains their historical outperformance. Now, dividend indices are sufficiently popular that their prices have been bid up to the point where they no longer have a value tilt.
Building a more value-oriented dividend portfolio
One approach to avoid paying too much for dividend-paying stocks is screening for high dividend yield and low value measures. My goal is to balance yield, P/B, P/E and risk measures to create a portfolio with a consistent income stream over time. The question is whether it is possible to create a portfolio with yield at or above that of the dividend-focused ETFs but with P/E and P/B below that of the pure-value ETFs.
Using FINVIZ.com, I searched for mid- and large-cap stocks with yields of at least 3%, P/E less than 15 and P/B less than 2. I removed all real estate investment trusts, business development companies and master limited partnerships from the resulting list. I then calculated risk levels (based on volatility of returns) and removed the companies with the highest risk levels. I restricted total portfolio risk to a level comparable to that of the dividend-index ETFs DVY and SDY.
My goal was to reduce portfolio risk rather than the risk of individual positions. A stock that might look more attractive on a standalone basis may be less attractive in the context of the overall portfolio because of the correlations among positions.
The final step in my selection process was to examine the dividend histories for the stocks. My assumption was that investors prefer consistent income. I favored U.S. stocks, which tend to have steadier dividend streams than non-U.S. stocks, because of the lack of currency risk. Non-U.S. stocks also carry a higher degree of uncertainty with regard to dividends and country-to-country variations in policy. International dividend-paying stocks (as represented by IDV, for example) have both higher yield and higher volatility than their domestic counterparts. I selected international stocks that provide yield, valuation and risk levels that are similar to domestic stocks.
With any type of incremental portfolio-selection process, but especially with one that has so many constraints, a range of portfolios can satisfy the general goals. My ultimate target was to identify a portfolio with higher yield than the well-known dividend index funds, but with P/B less than the Russell 3000 and Russell 1000 index funds and volatility comparable to the dividend index funds. The result is below:
Model low-cost dividend portfolio
Source: Current dividend yield from Morningstar.com and TTM P/E and P/B from Ameritrade.com
The equal weighted portfolio has a yield of 4.1%, P/B of 1.5 and P/E of 12.5. The trailing three-year volatility of this portfolio is 10.8%.
This portfolio is concentrated in four sectors: energy, financial, telecoms and utilities. This type of concentration is typical for dividend-oriented portfolios but may be unacceptable to some investors. By reducing the requirements on dividend yield and valuation, a wider set of sectors could be included.
While non-U.S. stocks have higher yields than U.S. stocks in many sectors, their yields in this portfolio are almost identical, as are the P/E and P/B. This was not a design criterion but a result of the risk constraints in the portfolio-selection process.
The aggregate dividends paid by this portfolio of companies have been stable or rising over the past 10 years. To illustrate the stability of the dividend payments, I divided the annual dividend for each stock by the 2009 dividend, so that the normalized dividend for each company is $1 in 2009. I then averaged the normalized dividends for all of the portfolio holdings in each year. This shows dividend consistency through time for the portfolio holdings, without having to account for differences in yield.
While there is variability in dividend payouts, the average payments have been stable and rising over the past 10 years.
Normalized average dividend history for portfolio components
This shows it is possible to build a high-dividend portfolio with lower P/B and P/E than major value indices, with moderate volatility. I have written about the importance of the ratio of yield and volatility. The model portfolio developed here also has a high yield-to-volatility ratio and low beta.
Model portfolio compared to reference ETFs
The portfolio has higher yield than the domestic dividend funds, higher yield-to-volatility than both the domestic and international dividend funds and lower P/E and P/B than the broader market and value indices. Its volatility being lower than that of the S&P 500 and the three value index benchmarks.
Conclusions
Dividend stocks are expensive in the current market. This poses a challenge for investors who focus on index funds but is less of a concern for those who are willing to build a portfolio of individual stocks. Even with current valuations, it is possible to build a portfolio of dividend-paying stocks that also has price-to-book and price-to-earnings ratios below major value indices.
A common criticism of dividend-focused strategies is that they are under-diversified because they are limited to the sectors that pay the highest dividends. This is evident in the model dividend portfolio presented here. On the other hand, dividend portfolios have historically generated high risk-adjusted returns relative to the S&P 500. Portfolio risk-adjusted return is typically considered the concrete demonstration as to whether a portfolio is effectively diversified or not
As prices rise, the expected future return of an asset tends to fall. In the current bull market, valuations on almost every risk asset are fairly high. Dividend-paying stocks, as a group, have become disproportionately expensive as yield-starved investors seek income. The prices of some well-known dividend indices, and the funds that track them, have increased to a point that investors at current levels are not buying stocks with low costs as compared to fundamentals. The historical outperformance of dividend stocks has been explained by their low P/E and P/B. Yield-seeking investors have reduced or perhaps eliminated that value advantage. The dividend yields available at current market levels are less attractive than they have been for years. At some point, dividend yields will be too low to justify their risks.
But a careful approach to selecting dividend stocks demonstrates that even today it is possible to build portfolios with relatively high dividends and valuations consistent with, or cheaper than, value indices.
Geoff Considine is founder of Quantext (www.quantext.com) and the developer of Quantext Portfolio Planner, a portfolio management tool. Quantext is a strategic adviser to FolioInvesting, a brokerage firm specializing in offering and trading portfolios for advisors and individual investors
Read more articles by Geoff Considine
