1) Manager of the TCW, MetWest, and TCW Alternative Fund Families INSIGHT ALTERNATIVE VIEWPOINT High Dividend Equities – Fertile Grounds, Just Beware the Landmines IMAN H. BRIVANLOU, PHD | MARCH 2016 Summary Historically, high dividend stocks have outperformed the broader equity markets and produced better risk-adjusted returns, even during periods of rising interest rates. However, the volatility associated with the highest dividend payers increases the probability of loss and lowers the risk-adjusted returns: (i) the top 5% of dividend payers in the S&P are over-represented in the bottom decile of performance, similar to stocks that do not pay any dividend, and (ii) dividend cutters considerably underperform. Thus, the high dividend universe is clearly “enriched” both with regard to stocks with potential for high returns and those primed for significant underperformance. Therefore, both long and short investors would be well-served to look carefully at high dividend companies. Introduction Iman H. Brivanlou, PhD Managing Director High Income Equities Dr. Brivanlou heads TCW’s High Income Equities group and is the Lead Portfolio Manager for the TCW High Income Equities and TCW Global REITs funds. He joined TCW in 2006 as an Analyst in the U.S. Equity Research group. His research coverage has spanned the real estate, insurance, business services, transportation, and consumer staples sectors. Prior to TCW, he developed and tested quantitative models aimed at predicting the sizes of catalyst “events” using option pricing for Kayne Anderson, a hedge fund in Century City. Dr. Brivanlou was a Howard Hughes post-doctoral fellow in Molecular Neurobiology at the Salk Institute in La Jolla. He holds a BS in Physics from MIT, a PhD in Neuroscience from Harvard, and an MBA in Finance and Strategy from the UCLA Anderson School of Management. The demographic trend of the aging of the population – a pronounced phenomenon prevailing across most developed markets – means income investing will likely remain a primary area of focus for decades to come. Traditionally, investors seeking income also desire safety and stability, and have usually looked to the credit markets to meet their requirements. However, with yields having been suppressed by unprecedented central banking policies of accommodation and quantitative easing, and with the credit cycle showing clear signs of aging, investors should pause and re-assess their options going forward. One of the compelling cases for investing in high dividend stocks is that they can provide a significant source of income while delivering attractive long-term returns through some degree of upside capture. Furthermore, by the time a company is in a position to initiate (or increase) a dividend, its business model has generally been tested and vetted by the marketplace and its cash flows and competitive position are usually stable. Dividend-paying companies have management teams that have demonstrated they take their fiduciary responsibility of proper capital allocation seriously, ranking a return of capital to shareholders above more speculative investments. As such, dividend paying companies tend to be the larger, higher quality, more established and better run companies. Fertile Grounds It should therefore not come as a surprise that dividends tend to correlate positively with total returns and that high dividend stocks have historically outperformed the broader equity markets. Exhibit 1 below documents this phenomenon using data where
2) ALTERNATIVE VIEWPOINT High Dividend Equities all listed stocks have been considered and segregated into 10 portfolios based on dividend yield. The figure shows that the top decile dividend portfolio has historically produced an average annual yield of 6.42% and has outperformed the S&P 500 by ~3.5% per year. The high level of annual yield has made it an attractive asset class across the full time frame, with nearly half (49%) of the total return of the high dividend portfolio coming from the dividend payments. Despite higher total return volatility (more on this later), the high dividend portfolio also boasts a higher Sharpe Ratio, and low correlation and beta relative to the S&P – overall a meaningfully better risk-adjusted return profile. Exhibit 1: Historical Performance of Top Decile Dividend Portfolio Versus S&P 500 Average Dividend Yield of Decile Portfolios (1960-2014) High Dividend Stock Portfolio 7% 6.42% 6% 5.14% 5% 4% 3% 2% 1% 1.62% 2.55% 2.08% 3.41% 2.92% 3.83% 4.34% 0.95% 0% Returns of High Dividend Portfolio* Versus S&P 500 Index 1200 1000 800 600 400 200 0 1960 1964 1968 1972 1976 1980 1984 1988 High Dividend Portfolio 1992 1996 2000 2004 2008 2012 S&P 500 Index Annual Sharpe Return* Volatility Ratio Correlation Beta Top Decile Dividend Portfolio 13.8% 15.3% 0.48 0.57 0.71 S&P 500 10.3% 12.3% 0.31 *Data from 1960 to 2015 Source: Global X Research. Data sourced from: http:/mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html 2 1 1
3) ALTERNATIVE VIEWPOINT High Dividend Equities The outperformance of the high dividend portfolio persists, on average, even during periods of rising interest rates. This may seem counter-intuitive; the space is often viewed as consisting primarily of “bond proxies,” which would be expected to wither with rising rates. The historical data indicates otherwise. Exhibit 2 below shows every instance of a sustained rising rate environment, as measured by changes in the U.S. 10-year Treasury yields, dating back to 1965. The table shows that across rising interest rate regimes, top decile dividend stocks outperformed the S&P Index in seven out of the ten observed periods, and by an aggregate annualized average of 0.8%. The three instances of underperformance occurred during periods with the most rapidly increasing rates – approximately 15-20 basis points per month. In today’s macro-economic backdrop, most prognosticators would agree that the likelihood of a rapid spike in 10-year Treasuries is low, so perhaps the more meaningful comparisons would be the instances where the rates rise more slowly. In the five periods in which the increase in 10-year rates is below 10 basis points per month, the relative outperformance of the high dividend portfolio was 8.8%. Exhibit 2: Performance in Rising Rate Environments Period Months U.S. 10-Year Treasury Increase (bps) Avg Rate of Increase Per Month (bps) Top Decile Dividend Portfolio S&P 500 Difference 6.3 11.2% 0.9% 10.3% Jul-65 to May-70 59 371 Oct-71 to Aug-75 47 247 5.3 3.8% -0.2% 4.0% Jan-77 to Sep-81 57 811 14.2 11.1% 7.8% 3.3% May-83 to Jun-84 14 318 22.7 8.5% 1.8% 6.7% Dec-86 to Oct-87 11 241 21.9 -1.1% 19.4% -20.5% Oct-93 to Nov-94 14 263 18.8 -8.4% 3.1% -11.5% Jan-99 to Jan-00 13 194 14.9 5.4% 19.6% -14.2% Jun-03 to May-06 36 178 4.9 17.2% 13.2% 4.0% Dec-08 to Apr-10 17 143 8.4 49.4% 27.2% 22.2% Jun-12 to Jan-14 19 133 7.0 28.4% 24.9% 3.5% 12.4 12.6% 11.8% 0.8% Average 28.7 289.9 Source: Global X Research, TCW Research There are several plausible reasons for the outperformance. First, high dividend stocks are not merely interest rate-sensitive bond proxies. Many are stable, high quality, and well managed businesses with meaningful economic upside capture. If interest rates are going up at a reasonable pace and for the “right reasons” (i.e., robust employment metrics, better than expected GDP growth, etc…), these companies stand to benefit. Second, higher interest rates imply higher discount rates for future cash flows. High dividend companies are usually established and mature, and their valuations are more resilient than those of high-flying growth companies to higher discount rates. Finally, high dividend companies typically allocate capital more efficiently due to the discipline imposed on management teams to make do with less. The fruits of this discipline are reaped as higher interest rates drive up the cost of capital and increase hurdle rates, exposing the wasteful projects of many peers. 3
4) ALTERNATIVE VIEWPOINT High Dividend Equities Landmines What about the group’s aforementioned higher volatility? Dividends from the highest decile group are usually high for a reason, and it is easy to imagine that many of the highest dividend-paying stocks are companies with deteriorating fundamentals, high leverage, or marginal dividend coverage. For at least a fraction of this segment, the high yield is likely a function of higher risk or a declining share price – where the market is expressing a concern that the current dividend level may not be sustainable. every other bucket. Also, the top 5% of dividend payers, while posting the highest returns, have considerably greater risk across both probability of loss and volatility. In fact, the substantial increase in volatility renders the risk-adjusted return (returns/volatility) of the top 5% worse than any other bucket, including non-dividend stocks. As outlined above, this result makes intuitive sense; a very high dividend yield is probably reflective of a riskier or at least more controversial stock. Market participants demand higher compensation as they await the “verdict.” Therefore, proper analysis and stock selection is especially critical with stocks in the top 5% as they are ripe for profitable trades on both the long and short side. Exhibit 3 addresses this point. The figure shows that there is a clear historical relationship between the highest dividend yields and measures of risk such as probability of loss and volatility. The Russell 1000 was divided into seven buckets (top 5% of dividend payers, dividend payers by quintiles, and non-dividend stocks), and average 12-month rolling total returns were plotted against two measures of risk (left: the probability of loss [frequency of negative returns]; right: volatility in 12-month rolling returns). Using the Russell 1000 Index rather than the S&P expands the breadth and reduces size effects from the results. Perhaps most interestingly, the table shows that stocks in Quartile 2 (top 20% to 40%) of dividend payers have the best risk-adjusted returns, posting a high average return profile with considerably lower probability of loss or volatility metrics. These stocks are likely fundamentally sound companies for which the high dividend yield may be the result of share price appreciation lagging dividend growth, or more benign reasons. While still above average, the dividend yields of this cohort (currently ranging from ~2.6% to ~3.5%) are not alarmingly high and do not by themselves signal market concern. First, note that the earlier finding that high dividend stocks outperform is replicated. Quintiles 1 and 2 (top 40% of dividend payers; green box) have average returns that exceed Exhibit 3: Russell 1000 Dividend Yield Quintiles Performance (1984-Present) 17% 17% Quintile 2 Average 12-Month Rolling Returns Average 12-Month Rolling Returns Top 5% 16% Quintile 1 15% Quintile 3 No Dividend 14% Quintile 5 13% 12% 15% Quintile 4 16% 17% 18% 19% 20% 21% Risk - Probability of Loss 22% 23% Top 5% 24% 16% Quintile 2 Quintile 1 15% Quintile 3 14% Quintile 5 13% Quintile 4 12% 16% 25% No Dividend 18% 20% 22% Quintile 1 (High) Quintile 2 Quintile 3 Quintile 4 24% 26% Risk - Volatility 28% Quintile 5 (Low) No Dividend Avg. Rolling 12M Return 16.6% 15.3% 15.7% 14.9% 12.8% 13.5% 14.6% Prob of Loss 20.3% 19.7% 15.3% 19.2% 20.3% 20.8% 23.6% Volatility 33.4% 21.9% 17.9% 18.7% 17.3% Risk-Adj. Return (Av/Vol) 0.50 0.70 0.88 Source: BofA Research., TCW Research 4 0.80 0.74 18.9% 27.1% 0.71 0.54 30% 32% 34%
5) ALTERNATIVE VIEWPOINT High Dividend Equities Exhibit 4 further illustrates the point that very high dividends can portend poor performance. The left panel shows that over the past ten years, the top 5% of dividend payers account for over 8% of the bottom decile of performance. In other words, the bottom decile in terms of performance is over-represented by stocks whose dividend yields were very high. Similarly, we also see that non-dividend payers are also over-represented in the bottom decile of performance. They have historically accounted for 21% of the index as a whole, but make up over 31% of the bottom performance decile. This effect was even more pronounced in 2015. At the start of the year, the top 5% dividend yielders of the S&P 500 had an average indicated yield of 6.4%. These 25 names ended the year with an average return of -9.4% (returns ranging from -47.4% to +35.6%). Twelve of the names (48% of the 25 companies) outperformed the S&P index, a larger fraction than the entire set of constituents. However, four of the names (16% of the 25) were in the bottom 5th percentile in terms of performance, and eight (32% of the 25) had total returns of -25% or worse (vs. 16% of companies with a loss of greater than 25% for the entire index). Landmines indeed… Exhibit 4: Historical Performance of Top 5% Dividend Stocks and Non-Dividend Payers (left), Annualized Excess Returns of Dividend Cutters and Growers (right) Average Concentration of Stocks By Dividend S&P 500 (2005-2015) Annualized Excess Return S&P 500 (2005-2015) 50% 45% 40% 35% 2% % of Index 1% % of Bottom Decile of Performance 0% 31.3% -1% 30% 25% -2% 21.3% 20% -3% 15% 10% 5% 5.0% -4% 8.1% -5% 0% -4.8% -6% High Dividend Payers No Dividend Dividend Cutters Source: Wells Fargo Research, TCW Research 1.4% 5 Top 20% by Dividend Growth
6) ALTERNATIVE VIEWPOINT High Dividend Equities Conclusion While the high dividend universe is fertile ground with regard to stocks with potential for high returns, the space also contains an over-abundance of landmines, which could detract very meaningfully from returns. Income-oriented investors would be well-served to separate the wheat from the chaff and perform fundamental research, especially on stocks with abnormally high yields, in order to avoid those with deteriorating fundamentals and to ensure that dividends are indeed well supported. Also, we believe the high dividend equity space is clearly well-suited for fundamental analysis/research-based long-short strategies. A simple starting point would be to look for longs in the top 20% to 40% of dividend payers, and for shorts in non-dividend stocks and those with abnormally high yields. As a parting thought, consider the right panel of Exhibit 4 on the previous page. Companies that cut dividends have historically underperformed the S&P by 4.8% per year. Identifying these names can potentially be an interesting approach for investors who employ short strategies. n This material is for general information purposes only and does not constitute an offer to sell, or a solicitation of an offer to buy, any security. TCW, its officers, directors, employees or clients may have positions in securities or investments mentioned in this publication, which positions may change at any time, without notice. While the information and statistical data contained herein are based on sources believed to be reliable, we do not represent that it is accurate and should not be relied on as such or be the basis for an investment decision. The information contained herein may include preliminary information and/or “forward-looking statements.” Due to numerous factors, actual events may differ substantially from those presented. TCW assumes no duty to update any forward-looking statements or opinions in this document. Any opinions expressed herein are current only as of the time made and are subject to change without notice. Past performance is no guarantee of future results. © 2016 TCW 865 South Figueroa Street | Los Angeles, California 90017 | 213 244 0000 | @TCWGroup 6
