Thesis Jacques Bontje
6013B0326Y 10663509
Economie & Bedrijfskunde Finance & Organisation Supervisor: Razvan Vlahu 2nd February 2016
2
Table of contents
Introduction Page 3
Literature review Page 4
1. Data description Page 8
2. Methodology Page 9
3. Results and empirical analysis Page 10
4. Conclusion and discussion Page 16
5. Reference list Page 17
6. Appendix Page 19
Statement of Originality
This document is written by Jacques Bontje who declares to take full responsibility for the contents of this document.
I declare that the text and work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Introduction
It has been a long discussion whether value investing delivers higher returns than growth investing. On the one hand, Fama and French (1997) demonstrated that value stocks have higher returns than growth stocks, when stocks were selected on certain characteristics in the period from 1975 to 1995. One of these characteristics to pick stocks is the dividend yield.
On the other hand Tom Lauricella claims in an article published on the 3rd of May 2015 in The
Wall Street Journal, that US growth stocks have outperformed US value stocks in a 10 year period from 2003 to 2013.
In 1991, Michael O’ Higgins and John Downes published a book called ‘Beating the Dow’. They discuss selecting the high dividend yield stocks of the Down Jones Industrial Average. These stocks where called the ‘Dogs’. They consequently make a portfolio of these high dividend yield stocks and compare these stocks to the market return of the DJIA. They found that this portfolio outperformed the market from 1957 to 2003.
In this thesis it is tested whether this investment strategy is successful using stocks from the Eurozone. A portfolio of high dividend yield stocks from the Stoxx Europe 600 index from January 2001 till December 2015 is compared with the market index. The research question in this thesis is: “Does the dividend yield investment strategy on the Stoxx Europe 600 index outperform the Stoxx Europe 600 on a monthly basis from 1 January 2001 to 31 December 2015, adjusting for risk?”. If that is the case, there is a superior investment strategy which can be applied to markets around the world.
Section 1 reviews the literature on dividend yield strategy. Data is presented in section 2. In section 3 the methodology is outlined and then the results are described in section 4. Section 5 concludes.
4
Literature review
In this chapter the existing literature on dividend yield investing is reviewed. Fama and French (1998) show that value stocks have better returns than growth stocks in the period of 1975 to 1995 in markets around the world. They identify value stocks on four characteristics. The first characteristic is book to market value ratio, which is the book value of a stock divided by the market value. The second characteristic is the earnings to price ratio; that ratio is explained by dividing earnings per share with the shares’ market price. The third characteristic is the cash flow to price ratio and is calculated by dividing the cash flow by the shares outstanding. The last characteristic is dividend yield. The dividend yield is calculated by dividing the annual dividend per share by the market price per share. They state that value stocks are characterized by high ratios of the first three
characteristics and that growth stocks are identified by low ratios. In different countries over the world Fama and French sorted stocks on these characteristics and ranked them on a list. When they compared the top 30% of the list against the bottom 30% they found that stocks in the upper 30% had above average returns in comparison with the bottom 30%. When a portfolio of stocks was formed with stocks having high ratios of book to market value, earnings to price or cash-flow to price, this portfolio had higher returns than a portfolio consisting of stocks having low values of these characteristics in 12 out of 13 countries used. When a portfolio was formed with dividend yield as characteristic, it was only higher in 10 out of 13 countries. In this thesis the dividend yield
characteristic is considered.
Black and Scholes (1974) question what effect dividend policy has on the price of stocks. They argue that investors prefer stock with a high dividend yield, because a dollar in the hand is more certain than a dollar capital gain increase. Consider two stocks which are similar in all
characteristics, except the dividend yield. Then the stock with a higher dividend yield than the other similar stock is worth more to an investor, because the dividends are higher. They show that there are different groups who prefer a different kind of dividend pay-out ratio. An example is
corporations who pay higher tax on capital gains than on dividend. These corporations prefer a high dividend pay-out ratio. On the other hand there are groups that are tax exempt and therefore are indifferent between capital gain and dividend. So they are indifferent about what dividend pay-out ratio the company has for the shares they hold. This is the tax effect of dividends.
Subsequently Black and Scholes (1974) say there is a diversification effect on portfolio’s containing high dividend stocks or low dividend stocks. This is because high dividend portfolios and low dividend portfolios don’t have a perfect correlation because there are systematic differences between these stocks. Therefore it isn’t possible to create a high dividend yield portfolio or low dividend yield portfolio with equal diversification to a portfolio consisting of high dividend and low
5 dividend stocks. Thus a portfolio consisting of a mix between high and low dividend yield stocks has a better diversification then a portfolio formed with only high or low dividend yield stocks. Because of less diversification, the high or low dividend portfolios of stocks are more risky than a market
portfolio.
Black and Scholes (1974, p. 4) also mention the uncertainty effects, in which they state “that it is not possible to demonstrate, using the best empirical methods, that returns of high yield securities are different from low yield securities, either before or after tax.” They imply that investors are uninformed about the effects of dividend yield on stock returns. Maintaining a portfolio with high or low dividend yield stocks causes less diversification of a portfolio and more transaction costs, hence an investor can decide to avoid using dividend yield completely.
Miller and Modigliani (1961) explain that if the dividend pay-out ratio corresponds with the investor’s preferences, then these preferences will lead a company towards a specific group of investors, according to their dividend pay-out ratio. As a consequence the valuation of the firms will be the same, according to Black and Scholes (1974). This is because corporations change their dividend policy to meet the highest demand of a certain kind of dividend policy. This maximizes the value of their shares, because a certain kind of dividend policy is demanded the most. If every firm tries to maximize the value of their shares by changing their dividend policy, an equilibrium is reached. So no firm can change its dividend policy to change the price of their shares. Black and Scholes call this the supply effect of shares.
According to Miller and Modigliani (1961) the value of a firm is calculated following the fundamental principle of valuation, using the dividend and capital gains. The current dividend policy affects the price of stocks, and therefore firm value, in three different ways. The first way is that dividend affects total firm value, the second way is that dividend conveys otherwise unavailable information about a stock, like information about probable future dividend policy. The last way is the dividend problem: that reducing dividend increases retained earnings for investing in projects, or increase dividend so floating new shares will increase capital for investments, but that dilutes the share price. Miller and Modigliani find that these two effects cancel each other out. As consequence they conclude that the dividend pay-out policy of a firm has no effect on the current price of firm’s shares, and has no effect on the total return of the shareholders.
Black and Scholes (1974) tested the effect of dividend on share prices. They researched if an increase of dividend will raise the price of shares. They did the research in the following form: Will a dividend increase, reduce the expected return of the shares of the company? They were unable to find that changes in dividend pay-out would change stock prices or the expected return, according to them this implies that changes in dividend yield policy doesn’t affect share prices. Also they couldn’t
6 validate that returns of high dividend portfolios are significantly different from returns of low
dividend portfolios.
Harkavy (1953) presented an analysis of common stock prices and dividend policies. He demonstrated that stock prices vary directly with the percentage of the retained earnings
distributed to stock holders. Harkavy sets forth that if two stocks are identical in all characteristics, but one stock has a higher dividend pay-out ratio than the other stock, then the stock with the higher dividend pay-out ratio will be valued higher than the other stock. He also demonstrated that stocks of large US corporations, who keep a greater proportion of earnings, have the tendency to increase in share value over the years. This includes corporations which are retaining earnings and not paying dividends to invest in growth opportunities for the company. A vital condition is that the company uses investors’ money profitable. This doesn’t mean that companies who have a low dividend pay-out ratio are retaining their earnings to expand the company and that the prices of shares of companies with a low dividend pay-out ratio appreciate a lot over time. It implies that the earning power must be increased by increasing the book value from retained earnings.
Walter (1956) shows that retained earnings influence stocks prices with the possibility of future dividends. He also states that ‘growth’ stocks are stocks which carry characteristics that are preferable for long-term price growth. According to Walter growth stocks carry the following characteristics: low dividend pay-out ratio, low capitalization rates and rapid growth over time. Growth stocks especially include low dividend pay-out stocks since a dollar invested in these stocks is worth more than a dollar invested somewhere else at a lower growing rate. Growth stocks are affected by preferential treatment of capital gains over dividend revenue because the value of the retained earnings is enhanced and therefore the value of the growth stocks is augmented. Walter also mentions that retained earnings can say something about dividend stability, however this doesn’t promise dividends in the future.
DeAngelo, DeAngelo and Skinner (2004) study the nominal amount and real amount of dividends paid between 1978 and 2000. They found that since 1978 fewer firms paid dividend. This decrease of the number of dividend paying firms was largely caused by firms who paid little dividend, but don’t pay any dividend anymore because they got acquired or got into financial distress. DeAngelo et al. also found that the dividend pay-out ratio of large industrial firms increased over the period. This was caused by an increase in earnings of these firms. The top 25 industrial firms pay 54.9% of all dividends. As a result dividend payments are much more concentrated in 2000 than in 1978.
To compare the returns of the high dividend portfolio and the market portfolio, a
7 are used to compare a portfolio’s performance in some time period relative to another period or to compare different portfolios in the same time period.” In this thesis the Sharpe ratio is used. Sharpe (1975) describes that measuring performance is irrelevant when risk isn’t taken into account. This is because there is a trade-off between risk and return. Stocks with higher risk will, on average, have a higher return. The Sharpe ratio measures the reward to the variability, and the larger the Sharpe ratio, the better the performance of a stock or portfolio.
To control for tax and transactions costs, the efficient market theory is used to set some assumptions. According to Fama and French (1970) an efficient market fully reflects all available information about securities and stocks. So all securities and stocks prices reflect all information that is available about these securities and stocks. To get capital market efficiency, we need to take conditions which help or hinder adjustments to prices when information gets available. The first condition is that there are no transaction costs in trading stocks or securities. The second condition Fama and French mention is that all information is available free of cost to all market participants. The last condition is that all market participants agree on the assumptions the information gives about the current price and future prices of stocks and securities. With these conditions the prevailing price of a security reflects all information that is available to each market participant.
Filbeck and Visscher (1997) researched the high dividend yield strategy on the British stock market and found that is wasn’t effective from March 1984 to February 1994. Visscher and Filbeck (2003) did the same research on the Canadian stock market, the Toronto 35, and found that the high dividend yield strategy had superior returns, even the risk-adjusted performance measure indicated that this strategy was performing better than the market. The same research was done in Latin America, where Da Silva (2003) found that the dividend portfolio had larger returns, but there were no significant results. On the contrary, Rinne and Vähämaa (2001) found empirical evidence that the
high dividend yield strategy outperformed the Finnish market, even adjusting for risk. Brzeszczynski
and Gajdka (2008) tested the same strategy on the Polish stock market and found that the high dividend yield strategy beat the market portfolio.
8
1. Data
The data is obtained from DataStream. For the research, monthly returns of the Stoxx Europe 600 are used, which is a market index of 600 stocks based on free float market capitalization. The index tracker of this market is SXXP. The index represents companies from 18 different countries in Europe. Table 4 is provided in the appendix with all securities used in this study from the Stoxx Europe 600. Because the research is about stocks in the Eurozone only, stocks denominated in euros will be used. The Eurozone countries include Belgium, France, Luxembourg, Germany, Austria, Greece, Spain, Italy, The Netherlands, Portugal, Denmark, Ireland, Finland and one company from the United Kingdom. This company was denominated in euros. The data obtained is from the first of January 2000 to the last day of December 2015.
The interest rates are obtained from DataStream as well, and I use the 3-month EURIBOR interest rates, since these rates represent the risk-free rate in the Eurozone. The monthly risk-free rate is calculated by taking the twelfth root of the number plus one.
The data used is the Total Return per month, this means that the dividends are re-invested to purchase additional units of equity. The same applies to the market portfolio. Dividend yields of all stocks were collected on a per year basis starting on the first of January 2000. The twenty companies with the highest dividend yield of the last year form the high dividend yield portfolio. Therefore the stocks with the highest dividend yield over the year 2000 will form the high dividend yield portfolio in the year 2001. The stocks in the high dividend yield portfolio are equally weighted. The dividend portfolios are presented in Table 1. In total, there are 118 different companies in the timespan from 2000 to 2015. Wereldhave, a Dutch real-estate investment company, appears 12 times. Further, an Italian electricity company, Enel, and a French real-estate investment company, Fonciere des Regions, appear 10 times in this timespan. 42 companies appear only once in the dividend portfolio. Each year, on average 11.2 companies are replaced by other companies and on average 8.8 companies stay in the dividend portfolio. Therefore the turnover is 11.2.
In Table 1, provided in the appendix, the dividend yield portfolio for the year 2000 is presented. However, this portfolio isn’t taken into account, as the dividend data and market return aren’t fully available. It is presented because it gives more information on the turnover rate of the dividend portfolio.
The Stoxx Europe 600 is calculated on the free float market mechanism, which is the share of a stocks’ full market capitalization that is available for trading. The Dow Jones Industrial Average is calculated on the same mechanism (Stoxx Index guide, p.21).
Companies with a ‘superdividend’ or abnormally high dividend in comparison with previous year aren’t deleted from the sample.
9
2. Methodology
First the excess return of the market is calculated. According to Berk & DeMarzo (2014) this is captured in equation 1. Then the return of the dividend portfolio is calculated as in equation 2:
Equation 1
=
Where
- Rm is the market return
- Mt is the price of the market portfolio at time t,
- Mt-1 is the price of the market portfolio at time t-1
- DIVt is the dividend at time t.
Equation 2
=
- Rdp is the dividend portfolio return
- DPt is the value of the dividend portfolio at time t
- DPt-1 is the value of the dividend portfolio at time t-1
- DIVt is the dividend in time t
Berk and DeMarzo (2014) show that the volatility of a portfolio declines when other stocks are added, still there always remains systematic risk. The dividend portfolio is of equally weights, so €1000 is invested in each stock at the beginning of the year, adding up to €20.000 euro. €20.000 euro is invested in the market portfolio as well. When doing the analysis, the yearly returns of the portfolios are calculated to see whether the dividend yield portfolio will outperform the market
portfolio.The hypotheses of this thesis are as follows:
- H0: Dividend yield strategy does not outperform the index, adjusted for risk.
- H1: Dividend yield strategy outperforms the market index, adjusted for risk.
With this hypothesis the question “Does the dividend yield investment strategy on the Stoxx Europe 600 index outperform the Stoxx Europe 600 on a monthly basis from 1 January 2001 to 31 December 2015, adjusting for risk?” will be analysed.
10
3. Results and empirical analysis
The results are obtained using a paired difference t-test with n-1 degrees of freedom and are presented in Table 2. The yearly return of the market index and the high dividend yield portfolio were calculated using monthly data. The results are from a one year investment horizon. The t-test is introduced by Visscher and Filbeck (1997). Equation 3 describes this t-statistic:
Equation 3
=
∗ √
Where
- is the mean difference from the market and the high dividend yield portfolio per month
- is the standard deviation of the excess return between the market and high dividend
yield portfolio per month
- n is the amount of observations, in this analysis 12 months, so n=12.
Table 2. Eurozone top-20 high dividend yield strategy versus market index for the period of 2001 to 2015 Year Yearly return market index Yearly return high dividend yield portfolio Market adjusted return1 Mean Difference between dividend portfolio and market return2 Standard deviation difference dividend portfolio and market return2 t-test 2001 -14.969% 17.075% 32.043% -0.02413 0.04364 -1.9158* 2002 -30.783% -0.0027% 30.780% -0.02917 0.03769 -2.6808** 2003 17.158% 30.864% 13.706% -0.01086 0.02319 -1.6230 2004 13.343% 29.771% 16.428% -0.01140 0.01785 -2.2123** 2005 26.979% 41.877% 14.898% -0.00889 0.01523 -2.0217* 2006 20.845% 32.405% 11.560% -0.00764 0.01118 -2.3671** 2007 2.829% -3.177% -6.006% 0.00481 0.01487 1.1205 2008 -43.413% -31.587% 11.827% -0.01318 0.02089 -2.1858* 2009 33.348% 53.847% 20.5% -0.02722 0.08508 -1.1084 2010 13.147% 6.289% -6.858% 0.00523 0.06107 0.2964 2011 -7.891% -13.702% -5.883% 0.00520 0.02803 0.6431 2012 17.726% -10.736% -28.462% 0.01916 0.04455 1.4900 2013 21.502% 39.644% 18.142% -0.01495 0.04651 -1.1139 2014 7.792% 20.297% 12.505% -0.01006 0.02275 -1.5319 2015 10.163% 7.163% -3.000% 0.00271 0.02118 0.4437
1The market-adjusted return = yearly return high dividend-yield portfolio minus yearly return market index
2The values are based on monthly returns
*Significant result at 10% level, critical value with 11 degrees of freedom is over 1.796 or under -1.796 **Significant result at 5% level, critical value with 11 degrees of freedom is over 2.201 or under -2.201
11 As can be seen in Table 2, the dividend-yield portfolio outperforms the market index on 10 years out of 15 years, especially from 2001 to 2006. In 2001, the dividend portfolio had 17.075% return, where the market index has -14.969% return. The market-adjusted return is 32.043%. Factors of influence on the results in 2001 were the attacks on the World Trade center on 11 September, the collapse of Enron and the United States and allies invading Afghanistan. One of the other factors is that one stock of the dividend portfolio, Glanbia, increased in value about 2.5 times in that year. The next year, in 2002, the market index had a negative return of 30.783%. The dividend portfolio only suffered a small loss of 0.0027%. When the market had big downturns following the Dotcom crash, the dividend portfolio didn’t follow the market closely. When the market suffered big losses in June and July of 7.7% and 14.4%, the dividend portfolio suffered fewer losses of 2.6% and 8.0%. This resulted in the market index losing value and the dividend portfolio staying almost identical at the end of the year. In 2003 both the market and the dividend portfolio had positive yearly returns, the standard deviation of the difference decreased in comparison with 2001 and 2002, resulting in a less volatile market. From 2002 to 2006 the volatility of the differences between the market index and the dividend portfolio decreased, this implicates that the markets were less volatile. The yearly returns from 2003 were positive till 2007. In 2004, 2005 and 2006 market portfolio had returns ranging from 13.34% to 20.85% and the dividend portfolio had returns ranging from 29.77% to 41.88%. In these years the dividend portfolio had substantial bigger returns than the market
portfolio differing 11.56% to 16.43%. In 2005 this increase was largely caused by two companies. The companies Boskalis Westminster, a Dutch dredging and marine expert, who rose over 128% in value and Metso, a Finnish company specialized in mineral processing and machinery, rose over 100% in value.
In 2007, for the first time in this analysis, the market index performed better than the dividend portfolio. The market had a return of 2.83% whereas the dividend portfolio endured a loss of 3.18%. In 2008 the market index and the high dividend portfolio suffered big losses. The market index lost 43.41% and the dividend portfolio 31.59%. The main explanation for the losses is that the housing bubble in the U.S. burst and because of that, financial institutions all over the world came into trouble and national governments had to bail-out many financial institutions. In September 2008, Lehman Brothers filed for bankruptcy, which is seen in the results of the market index and dividend portfolio which reported losses of 10.1% and 8.5% in the same month. Both portfolios had even bigger losses with 13.2% and 11.7% in October. The market index and the dividend portfolio recovered from the losses of 2008 in 2009 and had an increase of 33.39% and 53.85%, which is the largest increase in the time-series measured. One of the reasons for this, is that the ECB lowered the Euribor interest rate drastically over 2009 as can be seen in the data. The market portfolio realized a
12 bigger gain in 2010 than the dividend portfolio. The gain was 13.15%, where the dividend portfolio only gained 6.29%. Both portfolios had a negative return in 2011, although the market portfolio had a less negative return than the dividend portfolio. Then in 2012, the results are striking, since the market index has a positive return of 17.73% for the year, but the dividend portfolio dropped in value with 10.74%. This was large due to three companies: KPN, Unipol Gruppo Finanziari and Unipolsai lost more than 50% of their value. After three consecutive wins of the market portfolio, the dividend portfolio had the highest return in 2013, which is 18.14% higher than the market index return. The yearly return of the market portfolio and the dividend portfolio are 21.5% and 39.6%. As well as in 2013, the dividend portfolio reported bigger gains than the market index in 2014. The dividend portfolio increased 20.3% in value compared to an increase of 7.8% of the market portfolio. In 2015 the market index outperformed the high dividend yield portfolio by 3.0%. The better result of the market can be explained by losses of 30% of two companies in the high dividend portfolio, E-ON and Banco Santander.
The results in 2001, 2005 and 2008 are significant at 10% level. In 2002, 2004 and 2006 the results are significant at 5% level.
In this section a risk adjusted performance measure is used to compare the results obtained from the previous test. With the Sharpe ratio (1975) the excess return is measured, adjusting for risk. A higher Sharpe ratio means that the reward to volatility is better. The Sharpe ratio is calculated as in equation 4:
Equation 4
Sharpe ratio =
&'()*'+,' -./-00 (-)1(2&'()*'+,' 34(,45,+,)6
=
789:; (< =>98:;
Where
- ?9 @; − B = mean of monthly difference between the market index return or dividend
portfolio return and the Euribor interest rate over 12 months.
- SD(Rp) = Standard deviation of the monthly differences between market index return or
dividend portfolio return and the Euribobor interest rate over 12 months
Table 3 shows the risk adjusted performance of the market index and the high dividend-yield portfolio using the Sharpe ratio. According to Berk and Demarzo (2014) the standard deviation of a return is the volatility and it measures the variability of the returns. According to them the volatility measures firm-specific risk and systematic risk.
13 Out of the 15 years this analysis is performed, the Top-20 high yield dividend portfolio has a better Sharpe ratio 8 times, as can be seen in table 3. The Stoxx Europe 600 market index has a better Sharpe ratio than the dividend portfolio 7 times. It is notable that the dividend portfolio has a better Sharpe ratio in the first 6 years of this analysis and that the market index had a consecutive streak of better Sharpe ratios from 2009 to 2013.
There are two possible explanations for a higher Sharpe ratio of the dividend portfolio in the first 6 years. The first one is that the mean difference of the high dividend yield portfolio and the Euribor interest rate is bigger than the mean difference of the market index and the Euribor interest rate and because of that the Sharpe ratio increases. Briefly this means that the returns of the dividend portfolio are greater than the market index. The second explanation is that the standard deviation of the difference between the high dividend yield portfolio and the Euribor interest rate is lower than the standard deviation of the differences of the market index and the Euribor rate. This means that the volatility of the returns is lower.
The yearly returns of the dividend portfolio were higher in the years 2001 to 2006, 2008, 2009, 2013 and 2014. But the reward-to-risk ratio was better in only 8 years, represented in the third column of Table 3. The Sharpe ratio of the market was higher in 2009 and 2013, this can be explained by the fact that the standard deviation of the differences of the high dividend yield portfolio and the Euribor interest rate was larger than the standard deviation of the differences of the market index and the Euribor interest rate. In 2009 the standard deviation was 14.61% for the dividend portfolio, while the market index only had a standard deviation of 6.94%. The standard deviation of the dividend portfolio was almost three times as big as the standard deviation of the market index in 2013, 6.04% against 2.10%.
To test if the Sharpe ratio is distinguishable between the portfolios, the z-test of Memmel (2003) is used. He corrected the original test of Jobson and Korkie(2008). Memmel suggests:
- H0= SharpeDY Portfolio = Sharpeindex Portfolio
- H1= SharpeDY Portfolio ≠ SharpeIndex Portfolio
The test statistic is calculated as in equation 5. Under H0 the z-value is normally distributed. Kim
(2012) applies this test in practise, and for this analysis the same test is used. Equation 5
C =DℎFG@HIJ K− DℎFG@H L √M
14 Where
- SharpeDY = Sample mean of the excess return of dividend yield portfolio divided by the
sample standard deviation of the excess return of the dividend yield portfolio.
- Sharpeindex = Sample mean of the excess return of market index portfolio divided by the
sample standard deviation of the excess return of the market index portfolio.
- M = N O[2 − 2RS + N U9DℎFG@HU L+DℎFG@HUJ K− 2DℎFG@H LDℎFG@HIJ KRSU - V WX XFY@ZH [Y\HG ]B ]\XHG^F W] X - RS WX ℎH XFY@ZH _]GGHZF W] \H `HH Ha_HXX GH [G X ]B ℎH bW^WbH b @]G B]ZW] F b YFGcH @]G B]ZW]
A positive z result will infer that the market index portfolio outperforms the dividend yield portfolio and a negative z-result will infer that the dividend yield portfolio outperforms the market index portfolio. The z-values are represented in the last column of Table 3. The Sharpe ratio between the portfolios are statistically different in 2001 and 2002 in favour of the dividend portfolio. In 2012 and 2013 the Sharpe ratio is statistically different in favour of the market portfolio.
The second way to test whether the dividend yield portfolio outperforms the market index portfolio is the test of averages over the years. Keller (2012) describes this test as in equation 6. Because the variances of the portfolios aren’t equal this is done as follows:
- H0 = Mean return DY portfolio = Mean return index portfolio
- H1 = Mean return DY portfolio ≠ Mean return index porQolio
Equation 6 t= defg hgijg jkl m
no l pqdefg hgijg
kdefg hgijg o l pqjkl mkjkl m
- N is the sample size
- R is the mean return of the dividend portfolio or market index portfolio
- Stdev is the average standard deviation of the dividend or market index portfolio
The result of this test is a value of 0.3379, so H0 is not rejected. Therefore the claim that the dividend
15
Table 3. Sharpe ratio of high dividend yield portfolio versus market index from 2001 to 2015.
Sharpe ratio marke t index Sharpe ratio high dividen d yield portfoli o Winner Mean differenc e market index and Euribor1 Mean differenc e high-dividend yield portfolio and Euribor1 Standard deviation of difference s of market index and Euribor Standard deviation of difference s of high dividend yield portfolio and Euribor Z-test score of Memme l (2003) 200 1 -0.2548 0.2866 Dividen d -0.01524 0.00899 0.05979 0.03105 -2.26** 200 2 -0.4483 -0.0285 Dividen d -0.03069 -0.00152 0.06844 0.05321 -2.25** 200 3 0.2643 0.3817 Dividen d 0.01234 0.02321 0.04670 0.06080 -1.09 200 4 0.4423 0.5764 Dividen d 0.00894 0.02034 0.02021 0.03528 -1.13 200 5 0.6729 0.9537 Dividen d 0.01865 0.02754 0.02771 0.02887 -1.38 200 6 0.5192 0.8293 Dividen d 0.01369 0.02133 0.02637 0.02572 -1.94 200 7 -0.0217 -0.1588 Market -0.00070 -0.00551 0.03212 0.03468 1.05 200 8 -0.7396 -0.5417 Dividen d -0.04808 -0.03489 0.06500 0.06441 -1.69 200 9 0.6540 0.4973 Market 0.04542 0.07264 0.06944 0.14607 1.25 201 0 0.2317 0.0536 Market 0.01060 0.00538 0.04576 0.10038 1.45 201 1 -0.1727 -0.2296 Market -0.00711 -0.01232 0.04120 0.05367 0.36 201 2 0.3447 -0.0744 Market 0.01397 -0.00519 0.04053 0.06974 2.10** 201 3 1.5769 0.7957 Market 0.03311 0.04806 0.02100 0.06040 2.36** 201 4 0.2417 0.4337 Dividen d 0.00642 0.01648 0.02656 0.03801 -1.00 201 5 0.1685 0.1327 Market 0.00964 0.00692 0.05718 0.05217 0.33
1These values are based on a monthly basis
16
4. Conclusion and discussion
This thesis has examined whether the top-20 high dividend yield portfolio of the Stoxx Europe 600 performed better than the market index from 2001 to 2015. The results show that the high dividend yield portfolio return exceeded the market return in 10 years and the Sharpe ratio of the dividend portfolio yearly returns exceeded the Sharpe ratio of the market index yearly returns on 8 out of 15 years. The Sharpe ratio was statistically different in 4 years. The claim that the high dividend yield portfolio outperformed the market index portfolio cannot be supported because a test of averages didn’t give a significant result.
In 2001, 2002, 2004, 2005, 2006 and 2008 the market adjusted return was significant, but in the years afterwards there were non-significant results. This can be explained by an increase in standard deviation of the difference between the dividend yield portfolio return and the market index return from 2007 and onwards. In the research on the British stock market Filbeck and Visscher (1997) demonstrated that the dividend portfolio was not effective from 1984 to 1994.
One of the possible explanations why the dividend portfolio does not outperform the market index is that high yield dividend investing is known for a long period. Because investors are aware of this strategy, it isn’t possible to outperform the market anymore.
No tax and transaction costs were taken into account. With the yearly rebalancing costs of the high dividend yield portfolio, this portfolio had higher costs to maintain. Therefore it is less profitable then stated in Table 2. Also, choosing a high dividend portfolio with less securities could give another answer than the answer in this analysis. Furthermore this research was limited due to a data range from 2000 to 2015. A more conclusive answer can be made when data would have been available in the years before 2000, which could have given a more empirical analysis from a longer time-series.
Nevertheless this research have shown that in this time-series the high dividend yield portfolio doesn’t outperform the market index, but more research is needed to conclude why. Future research can give an answer to why the dividend portfolio outperforms the market index on several years. Also, more research on high dividend yield investing is needed in other markets than the Eurozone area. And if the high dividend yield strategy performs better in bear or bull markets or in certain industries.
17
5. References
Berk, J., DeMarzo, P. (2014). Corporate Finance, 3rd edition. Essex, England. Pearson Education
Limited.
Black and Scholes (1974). The effects of dividend yield and dividend policy on common stock prices and returns. Journal of Financial Economics, vol. 1, pp. 1-22.
http://doi.org/10.1016/0304-405X(74)90006-3
Brzeszczynski, J., Gajdka, J. (2008). Performance of high dividend yield investment strategy on the Polish Stock Market 1997-2007. Investment Management and Financial Innovations. 5(2) pp. 86-92. Da Silva, A. L. (2001). Empirical tests of the Dogs of the Dow strategy in Latin American stock
markets. International Review of Financial Analysis, 10(2), 187-199.
DeAngelo, H., DeAngelo, L., Skinner, D.J. (2004). Are dividends disappearing? Dividend concentration and the consolidation of earnings. Journal of Financial Economics, 72(3), 425-456.
Fama, E. F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal of Finance, 25(2), 383–417. http://doi.org/10.2307/2325486
Fama, E. F., & French, K. R. (1998). Value versus Growth: The International Evidence. The Journal of Finance, 53(6), 1975–1999. Retrieved from http://www.jstor.org/stable/117458
Filbeck, G., & Visscher, S. (1997). Dividend yield strategies in the British stock market. The European Journal of Finance, 3(4), 277-289.
Harkavy, O. (1953). The Relation Between Retained Earnings and Common Stock Prices for Large, Listed Corporations. The Journal of Finance, 8(3), 283–297. http://doi.org/10.2307/2976396 Jobson, J. D., & Korkie, B. M. (1981). Performance hypothesis testing with the Sharpe and Treynor measures. Journal of Finance, 889-908.
Keller, G. (2012). Managerial Statistics. 9th edition. South-Western Cengage Learning.
Kim, D. (2012). Is currency hedging necessary for emerging-market equity investment? Economic Letters, 116(1), 67-71.
Lauricella, T. (2015, 3rd of May). Growth beats value. (If investors care.). Retrieved from:
18 Memmel, C. (2003). Performance hypothesis testing with the Sharpe ratio. Finance Letters 1. 21- 23. Miller, M. H., & Modigliani, F. (1961). Dividend Policy, Growth, and the Valuation of Shares. The Journal of Business, 34(4), 411–433. Retrieved from http://www.jstor.org/stable/2351143 Miller, M. H., & Scholes, M. S. (1982). Dividends and Taxes: Some Empirical Evidence. Journal of Political Economy, 90(6), 1118–1141. Retrieved from http://www.jstor.org/stable/1830941 O’Higgins, M., Downes, J. (1991). Beating the Dow. HarperPerennial.
Rinne, E., Vähämaa, S. (2011). The ‘Dogs of the Dow’ strategy revisited: Finnish evidence. The European Journal of Finance. 17(5-6), 451-469.
Sharpe, W. F. (1975). Adjusting for risk in portfolio performance measurement. The Journal of Portfolio Management, 1(2), 29-34.
Stoxx index methodology guide (Portfolio indices). November 2015. Retrieved from: https://www.stoxx.com/document/Indices/Common/Indexguide/stoxx_indexguide.pdf Stoxx constituents: Retrieved from: https://www.stoxx.com/index-details?symbol=SXXP
Visscher, S., & Filbeck, G. (2003). Dividend-yield strategies in the Canadian stock market. Financial Analysts Journal, 59(1), 99-106.
Walter, J. E. (1956). Dividend Policies and Common Stock Prices. The Journal of Finance, 11(1), 29– 41. http://doi.org/10.2307/2976527
19
6. Appendix
Table 4. List of stocks of Stoxx Europe 600 denominated in Euros. Table 4 Total number
of companies
292
ANHEUSER-BUSCH TOTAL SANOFI INDITEX BAYER ROYAL DUTCH
SHELL A
L’Oreal SAP DAIMLER LVMH SIEMENS BANCO
SANTANDER DEUTCHE TELEKOM ALLIANZ(XET) BNP PARIBAS BASF UNILEVER CERTS. AXA
TELEFONICA MBW ENI AIRBUS GROUP ING GROEP INTESA SANPAOLO
BBV.ARGENTARIA HEINEKEN DANONE ORANCE AIR LIQUIDE IBERDROLA
CONTONETAL ENGIE SOCIETE
GENERALE
ENEL ASML HOLDING VINCI UNICREDIT SCHNEIDER
ELECTRICE SE
HERMES ENTL. CREDIT AGRICOLE DEUTSCHE BANK DEUTSCHE POST NOKIA RENAULT CHRISTIAN DIOR ESSILOR INTL. LUXOTTICA SAFRAN ASSICURAZIONI
GENIRALI
LINDE MUENCHENER
RUCK
PERNOD-RICARD VIVENDI FRESENIUS PHILIPS ELTN.
KONINKLIJKE
EDF
UNIBAIL-RODAMCO
KBC GROUP SAINT GOBAIN SAMPO ‘A’ CARREFOUR FIAT CHRYSLER
AUTOS
KERING FRESENIUS MED. CARE
RYANAIR HOLDINGS
ATLANTIA BEIERSDORF CAIXABANK ENDESA GAS NATURAL SDG HEINEKEN
HOLDING
REPSOL YPF TELECOM ITALIA MICHELIN NUMERICABLE
SFR
ADIDAS AHOLD KON. AKZO NOBEL
DASSAULT SYSTEMES E ON FERROVIAL NATIXIS RELX SNAM
UCB AMADEUS IT
HOLDING
KONE ‘B’ KPN KON. PEUGEOT SODEXO
CAP GEMINI INFINEON TECHS. LEGRAND PUBLICIS GROUPE AENA SHS KLEPIERRE
THALES ABERTIS
INFRAESTURAS
ACCOR AEGON ALCATEL-LUCENT ALTICE A SHARES
BANKIA BOUYGUES DEUTSCHE
BOERSE EVONIK INDUSTRIES TELEFONICA DTL. TENARIS VEOLIA ENVIRONNEMENT
ERSTE GROUP BANK ILIAD KERRY GROUP ‘A’ NN GROUP RANDSTAD HOLDING
VALEO VONOVIA BANK OF
IRELAND
BOLLORE COMMERZBANK FORTUM HANNOVER RUCK. HEIDELBERGCEMENT MAN MERCK KGAA PROXIMUS THYSSENKRUPP
ADP AGEAS(EX-FORTIS) ARCELORMITTAL EXOR ORD GBL NEW OMV
PROSIEBEN SAT 1 PF. RED ELCTRICA CORPN.
SOLVAY ACS ACTIV. CONSTR.Y SERV
BANCO DE SABADELL
BANCO POPULAR ESPANOL CNH INDUSTRIAL EDP ENERGIAS DE
PORTUGAL
ENEL GREEN POWER
GRIFOLS ORD CL A SUEZ ENVIRONMENT
TERNA RETE ELETTRICA UNITED INTERNET WOLTERS KLUWER ALSTOM BUREAU VERITAS CNP
ASSURANCES
DELAIZE GROUP DSM KONINKLIJKE INGENICO GROUP KABEL
DEUTSCHLAND
METRO UPM-KYMMENE AIR FRANCE-KLM ATOS DASSAULT AVIATION DEUTSCH
WOHNEN
ENAGAS FINMECCANICA GALP ENERGIA SGPS JCDECAUX JERONIMO MARTIN MEDIOBANCA
BC.FIN
TECHNIP ZODIAC AEROSPACE
BANCO POPLARE
BANKINTER ‘R’ BIC COLRUYT GECINA GEMALTO MAPFRE
MEDIASET PADDY POWER SCOR SE SYMRISE TELENET GROUP
HOLDING
UNIONE DI BANCHE ITALIAN
WARTSILA ARKEMA BANCA MONTE
DEI PASCHI BRENNTAG CASINO GUICHARD-P DAVIDE CAMPARI MILANO DEUTSCHE LUFTHANSA EIFFAGE EURFINS SCIENTIFIC EUTELSAT CIMMUNICATIONS GAMESA CORPN. TEGC. GEA GROUP GROUPE EUROTUNNEL
KINGSPAN GROUP NESTE PORSCHE AML.HOLDG.
RWE UNIPOLSAI
ANDRITZ AZIMUT HOLDING BANCA POPOLARE DI MILANO BANCA PPO. EMILIA ROMAGNA BANCO COMR. PORTUGUES ‘R’ BPSS (HUGO)
ELISA EURAZEO FAURECIA FONCIERE DES
REGIONS
20 IMERYS K+S MEDIASET ESPANA COMUNICACION MERLIN PROPERTIES MTU AERO ENGINES OCI
RAIFFEISEN BANK INTL REMY COINTREAU SAIPEM SMURFIT KAPPA GROUP
STORA ENSO ‘R’ UMICORE
VOESTALPINE WENDEL WIRECARD AALBERTS
INDUSTRIES
ACKERMANS & VAN HAAREN
ASM
INTERNATIONAL AXEL SPRINGER BOSKALIS
WESTMINSTER
BPOST CELLNEX TELECOM DELTA LLOYD GROUP
DISTRIBUIDORA INTNAC.DE ALIMENTACION
EDENRED FRAPORT HELLENIC
TELECOM ORG. HUHTAMAKI LAGARDERE GROUPE LANXESS NATIONAL BANK OF GREECE
NOKIAN RENKAAT ORION ‘B’ ORPEA OSRAM LICHT PRYSMIAN QIAGEN RECORDATI INDUA
CHIMICA
REXEL SEB TECNICOLOR TELEPERFORMANCE
TF1 (TV.FSE.1) TNT EXPRESS UNIPOL GRUPPO FINANZIARI
VOPAK ZARDOYA OTIS AMER SPORTS BANCA PP. DI
SONDRIO
BOLSAS Y MERCADOS ESPANOLES
CONFINOMMO FREENET KESKO ‘B’ KION GROUP
LEG IMMOBILIEN METSO OPAP POSTNL RUBIS SBM OFFSHORE
UBISOFTENTM. WERELDHAVE AAREAL BANK BILFINGER BERGER DEUTSCHE EUROSHOP
DIALOG SEMICON.
DUERR GERRESHEIMER IMMOFINANZ RHEINMETALL STADA
ARZNEIMITTEL
VALLOUREC
VISCOFAN LEONI MORPHOSYS CRH RLT GROUP SES FDR
STMICROELECTRONICS (MIL)
VOLKSWAGEN PREF. HENKEL PREF. FUCHS PETROLUB PF.
22
Table 1 Stocks which comprise the high dividend yield portfolio. Company name
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Number of
Appearances
ABERTIS INFRAESTRUCTURAS X X 2
ACS ACTIV.CONSTR.Y SERV. X X X 3
AEGON X X 2 AGEAS (EX-FORTIS) X X X X X X 6 AIRBUS GROUP X X 2 ALSTOM X 1 AMER SPORTS X 1 ARCELORMITTAL X 1 AXA X 1
BANCA POPULARE DI MILANO X X 2
BANCA MONTE DEI PASCHI X 1
BANCO COMR.PORTUGUES 'R' X X X 3
BANCO DE SABADELL X X X 3
BANCO POPOLARE X X 2
BANCO POPULARE ESPAGNOL X X 2
BANCO SANTANDER X X X X 4
BANK OF IRELAND X X 2
BBV.ARGENTARIA X 1
BOLSAS Y MERCADOS ESPANOLES X X X 3
BOSKALIS WESTMINSTER X 1
BOSS (HUGO) (XET) X X 2
BOUYGUES X 1 CAIXABANK X X 2 CHRISTIAN DIOR X 1 CNP ASSURANCES X 1 COMMERZBANK X 1 COFINIMMO X X X X X X X X 8 CREDIT AGRICOLE X 1 DAIMLER (XET) X X X 3 DASSAULT AVIATION X X 2
DEUTSCHE BANK (XET) X 1
DEUTSCHE EUROSHOP (XET) X X X X 4
DEUTSCHE TELEKOM (XET) X X X X X 5
DEUTSCHE WOHNEN (XET) BR.SHS X X 2
DSM KONINKLIJKE X 1
DUERR (XET) X 1
EDF X 1
EDP - ENERGIA DU PORTUGAL X X X X X 5
EIFFAGE X X X 3
ELISA X X 2
ENAGAS X 1
23 ENEL X X X X X X X X X X 10 ENGIE X X 2 ENI X X X X 4 E ON (XET) X X X 3 FINMECCANICA X 1
FONCIERE DES REGIONS X X X X X X X X X X 10
FORTUM X X 2
FREENET (XET) X X 2
FUCHS PETROLUB PF.(XET) X X X X 4
GAS NATURAL SDG X 1 GECINA X X X X X X 6 GLANBIA X X 2 HELLENIC TELECOM.ORG X X X 3 IBERDROLA X 1 ICADE X X X X 4 IMERYS X X 2 ING GROEP X X X X X X 6 INTESA SANPAOLO X X 2 K + S (XET) X 1 KBC GROUP X 1 KESKO 'B' X X X X X X X 7 KONE 'B' X 1 KPN KON X X X X 4 MAPFRE X 1 MEDIASET X X X X 4
MEDIASET ESPANA COMUNICACION X X X X 4
METSO X X X 3 NATIXIS X X X 3 NESTE X 1 NOKIA X 1 NOKIAN RENKAAT X 1 OPAP X X X X X X X X 8 ORANGE X X X X X X X 7 ORION 'B' X X X X X X X X X 9 PERNOD-RICARD X X 2 PEUGEOT X X 2 POST NL X 1
PROSIEBEN SAT 1 PF.(XET) X X X X 4
PROXIMUS X X X X X 5
RENAULT X 1
RHEINMETALL (XET) X 1
ROYAL DUTCH SHELL A X X 2
RTL GROUP X 1
RUBIS X X X X X X 6
24
SAMPO 'A' X X X X 4
SEB X 1
SCOR SE X X X X 4
SMURFIT KAPPA GROUP X 1
SNAM X X X 3 SOCIETE GENERALE X X X 3 STORA ENSO 'R' X 1 TECHNIP X X 2 TELECOM ITALIA X X X 3 TELEFONICA X X X 3 TELEFONICA DTL. (XET) X 1
TELENET GROUP HOLDING X X 2
TERNA RETE ELETTRICA NAZ X X 2
THYSSENKRUPP (XET) X X 2
UNIBAIL-RODAMCO X X 2
UNICREDIT X X X 3
UNIONE DI BANCHE ITALIAN X X X 3
UNIPOL GRUPPO FINANZIARI X X X 3
UNIPOLSAI X 1 UPM-KYMMEME X X X X X 5 VALEO X 1 VEOLIA ENVIRONMENT X 1 VINCI X 1 VIVENDI X X X 3 VOESTALPINE X X 2
VOLKSWAGEN PREF. (XET) X 1
VOPAK X X 2
WARTSILA X 1
WENDEL X 1
WERELDHAVE X X X X X X X X X X X X 12
ZARDOYA OTIS X X 2
Number of stocks in portfolio Total number of companies: 118
20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
Number of stocks staying in portfolio 12 12 10 7 9 9 9 13 9 8 2 9 9 6 8
Turnover1 (Average is 11.2) 8 8 10 13 11 11 11 7 11 12 18 11 11 14 12
