In a recent article Eli Inkrot showed the results of his study on the return of all the stocks presently on the Dividend Champions list. It showed that the Champions significantly outperformed the rest of the market over the past 15 years. But one of the main criticisms of the article was the issue of survivorship bias, and that any back test of this sort will be skewed towards the stock that have done well over whatever time period is being studied. If you study stocks that have done well, then obviously the study will show that they did well! There certainly is validity to this argument, and every back test must take survivorship bias into account. And yet I still think Eli's article was excellent, and was a strong demonstration of the merits of a dividend growth investing (DGI) philosophy.
But the critics do have a point. Back tests, although giving some valuable information, are inherently flawed. The best study is a forward looking, prospective study in which a group of stocks is selected, based on certain criteria, and then followed into the future to see how well they do. In this way the bias of back tests, in which only the surviving companies are selected, cannot skew the results.
Fortunately this kind of study is possible. In the comment section of Eli's article Chowder (a highly respected SA contributor) gave the names of 20 stocks that were highlighted in the book "The 100 best stocks to own in America" fourth edition by Gene Walden. This book was published in 1996. So we have a list of stocks that was recommended back in 1996, and we have the ability to look at how these stocks have done over the past 17 years. It's a perfect opportunity to do a prospective study, and I was happy to undertake the project.
Here is the list of stocks as presented by Chowder.
- The Gillette Company ((NYSE:G))
- Wm. Wrigley Jr. Company (WWY)
- Franklin Resources (NYSE:BEN)
- The Coca-Cola Company (NYSE:KO)
- Albertson's, Inc. (ABS)
- UST, Inc. (UST)
- Medtronic, Inc. (NYSE:MDT)
- Merck & Company (NYSE:MRK)
- Schering-Plough Corp. (SGP)
- H&R Block, Inc. (NYSE:HRB)
- Philip Morris Companies, Inc. (NYSE:MO)
- PepsiCo, Inc. (NYSE:PEP)
- Northwest Corp. (NOB)
- Wal-Mart Stores, Inc. (NYSE:WMT)
- Abbott Laboratories (NYSE:ABT)
- Walgreen Company (WA)G
- General Mills, Inc. (NYSE:GIS)
- The Home Depot (NYSE:HD)
- Crompton & Knowles Corp. (NYSE:CNK)
- Federal Signal Corp. (NYSE:FSS)
Since I am a dividend growth investor, and this discussion was part of an article promoting the benefits of DGI-ing, I felt that the first thing to do would be to figure out which of these stocks, back in 1996, would in fact have been considered to be a dividend growth stock, and could reasonably have been expected to be selected for a DGI portfolio. It is the returns of the stocks that would have been included in a DGI portfolio that are of real interest to us. The criteria I used to determine if they would have been DGI stocks was very simple. In part this was mandated by the relative lack of information I could find about the stocks dating back to 1996. For example I did not include a certain payout ratio as a criteria, even though many DGIers (including me) use a payout ratio cut off as part of their screen, because I could not find information about their payout ratios from 1996. Secondly, since many DGIers use many different criteria for inclusion into their portfolios it would be hard to decide which should or should not be used. It seems that the only universal criteria is a history of at least 5 years of dividend growth. Therefore, the only criteria I used to determine if they would have been suitable for a DGI portfolio was a record of increasing their dividend each and every year from 1991 through 1995. I ignored valuation, payout ratio, metrics of financial strength (such as S&P rating), or anything else investors use, because the information was too hard to find, especially for the stocks that no longer existed due to buy outs and mergers, and because not everybody agrees on which of these criteria to use.
Out of the 20 stocks presented I could not find any price or dividend information about two of them, and so they were not included in the study. They are Albertson's and Crompton and Knowles. Albertson's was bought by Supervalu (NYSE:SVU) and CVS Corp. (NYSE:CVS) in 2006. Crompton and Knowles merged with Witco Corp. in 1999 to form Crompton Corp, and then Crompton Corp. merged with Great Lakes Chemical Corp in 2005 to become Chemtura Corp (NYSE:CHMT). The historical quotes for CHMT only go back to 2010. Without historical quote back to 1996 I could not include the company in the study.
General Mills was also removed from the prospective DGI portfolio because it froze its dividend in 1994 and 1995, only beginning to increase it again in 1996. Although some people probably would have continued to hold GIS when it froze its dividend, especially considering its long dividend history, I think most DGIer would agree that once it had frozen its dividend they would not have considered it as a new purchase for a DGI portfolio until it began raising them again, and perhaps for at least 5 years. So although I did calculate GIS' return during the time period, and will report it in this article, it will not be included in the performance of the portfolio as a whole.
For all 17 stocks that were included in the DGI portfolio historical prices and dividend data was taken from yahoo.com for those companies still trading, or from the acquiring company's websites if the companies had been bought out. $10,000 of each company was bought on the first trading day of 1996, to the nearest whole share. Therefore the portfolio started with $170,000. All dividends were reinvested in the stock that paid them, including fractional shares. The value of each stock was determined at the end of trading on 9/9/2013, and the annual return was determined for each stock, and for the portfolio as a whole.
If a stock was bought out for cash (Wrigley's, UST Inc.), the cash was split up equally amongst the other remaining original stocks, and more shares of each were bought. If a stock was bought out for shares in the acquiring company (Gillette, Schering Plough) the shares of the acquiring company were sold, and once again the money was invested evenly into the remaining original stocks.
When spin-offs occurred the spun off companies were held in the account until the end of the study, and their return was included in the final results.
An equal value ($170,000) of the Spider S&P 500 ETF (NYSEARCA:SPY) was also bought on the first trading day of 1996 and held through 9/9/2013, with all dividends being reinvested. This was used as the control group to represent the return of the market as a whole.
Although many other mergers and buyouts occurred involving some of the companies on the list, I was still able to get stock prices and dividend data from when they were independent companies. These companies, and the merger activity, are listed below:
- Gillette was bought out by Proctor and Gamble (NYSE:PG) in October of 2005. Gillette stockholders were given 0.975 shares of PG for every share they had of G.
- Wrigley's , was bought out by Mars, inc. in October of 2008 for $80.00 per share.
- UST Inc, was bought out by Altria Group (MO) in January of 2009 for $69.50 in cash.
- Schering Plough was bought out by Merck (MRK) in November of 2009 for .5767 shares of Merck plus $10.50, for each share they had of Schering Plough.
Many of the original companies had spin-offs of other companies. In fact, six new companies were spun off during the study time period. The spin-offs were as follows:
Abbott: Abbott spun off Hospira (NYSE:HSP) in May of 2004. ABT Stock holders received 0.1 share of Hospira for each share of Abbott. Abbott did a second spin-off, this time of AbbVie (NYSE:ABBV) in January of 2013. ABT Stock holders received one share of ABBV for each share of ABT.
Merck: Merck spun off Medco in August of 2003. MRK Shareholders received 0.1206 shares of Medco for each share of MRK. Medco was later bought out by Express Scripts (NASDAQ:ESRX) in April of 2012. Medco shareholders received $28.80 plus 0.81 shares of ESRX for each share of Medco.
Philip Morris: First of all, Philip Morris in 2003 changed its name to Altria. Altria then spun off Kraft (NASDAQ:KRFT) in April of 2007. Shareholders received 0.692024 shares of Kraft for each share of MO. MO then split up into two different companies in March of 2008. MO shareholders received one share of MO and one share of Philip Morris International (NYSE:PM) for each share of MO that they owned. Kraft, which had been spun off in 2007, itself split into two different companies in October of 2012. Kraft changed its name to Mondelez (NASDAQ:MDLZ), and shareholders received one share of Kraft for each three shares of MDLZ. So, in the end, MO, bought in 1996 had turned into four different companies: Altria, Philip Morris International, Mondelez, and Kraft. And don't forget, during that time it also bought out UST inc.
There was also one merger:
Norwest: Norwest Corp merged with Wells Fargo (NYSE:WFC) in November of 1998. The joint company continued to trade under the ticker symbol WFC, but the historical quotes and dividend history going back before the merger are those of Norwest.
Following all of these mergers, acquisitions, and spin-offs, of the original 17 companies only 13 were still in existence as of September of this year. One of these, FSS, cut its dividend in December 2003 so it was sold at that time. Therefore at the end of the time period 12 of the original 17 were still in the portfolio, but due to spin-offs the portfolio actually consisted of 18 companies. The following chart shows the 12 stocks out of the original 17 that remained at the end of the study period. The value and annual return includes the value of all spin-offs that originated with that particular stock. The SOLD section shows the value of FSS when it was sold, and the BUY OUTS section shows the value of each company that was bought out at the time it ceased to trade. These values were not included in the final results because, as mentioned above, when bought out the cash was distributed to the other stocks remaining in the portfolio and more shares of each was purchased.
Federal Signal Corp
The final value of the entire portfolio was $1,015,735.50. This is from a starting value of $170,000. This is equal to an annual return of 10.63%. During this time "The Market", as represented by SPY, showed a return of 7.61%, for a final value of $622,777.54.
The DGI portfolio collected a total of $291,364.86 in dividends during the study period, or 172% of the original investment. The dividends collected in the last year were $24,973.19, for a yield on cost (YOC) of 14.69%.
SPY paid out a total of $114,086.43 in dividends, or 67.1% of the original investment. The dividends collected in the last year were $12,128.62, for a YOC of 7.10%.
As an FYI General Mills, not included in the portfolio, had a final value of $93,568.18 for an annual return of 13.48%. A total of $13,920.24 would have been collected in dividends over the course of the study period, and $1,701.63 would have been collected in the past year, for a YOC of 17%.
The DGI portfolio easily outperformed "the market". It beat it by 39.6% on an annual basis, and the total return in the end was 54% higher. I feel that this confirms the argument that a portfolio of quality dividend growth stocks will outperform the market over the long term. The DGI portfolio also paid out more than twice as much in dividends over the study period, and in the final year was creating almost twice as much income as SPY. For many income investors this is even more important than the total return.
Most people use valuation as a consideration when selecting criteria for inclusion into a DGI portfolio. I know I do. I think that the results above would have been even better had a maximum PE ratio cut-off been used. As an example, KO, the second worst performer (besides FSS) out of the original 17, started the study period with a PE ratio of 31.89. Starting at such a high PE ratio it is no wonder that its results suffered compared to the other DGI stocks. And yet it still managed to achieve an annual return of 8.83%, beating the market, and showing once again the power of a solid dividend growth stock.
Finally, maybe long-term dividend history is more important than a recent dividend freeze when considering whether or not a stock should be included in a DGI portfolio. GIS had an annual performance that would have placed it 6th out of the original 17 stocks, even though it had frozen its dividend for the two years prior to the study. Maybe the fact that it had never decreased its dividend in the almost 100 years before the study period started was the more important piece of information. I don't know what the right answer is, but I would say It should at least be taken into account.
Thank you for reading my article. I welcome your comments and criticisms.