On March 16, David Van Knapp authored the article Has Dividend Growth Kept Up With Inflation?. As is the case with most of David's excellent articles, a lively and informative comment stream ensued. However, there was one specific series of comments relating to back testing that inspired me to write this article. There is a dialogue between the author, David Van Knapp, and two readers, David Fish and Chowder, regarding the complexity, validity and even the desirability of conducting a back test as follows:
Chowder commented as follows:
"In my opinion, back testing doesn't prove much as far as I'm concerned. Back testing means you had to hold through feast or famine. Who does that when your objective is hold and monitor?
Back tested results are worthless if you don't follow the system to a T.
I was reading a book last night, The Dividend Growth Strategy. It was written in 2001. At that time, the author listed their top 30 typical dividend growth companies.
I suppose if someone wanted to back test a program, based on what was thought to be the top 30 at that time, I suppose that might be a good way to test a strategy. There certainly wouldn't be any survivor bias, but some might say they were selective at the time. I don't know. What I do know is that I can see where a dividend growth investor would have held some of those companies at that time."
Then Chowder went on to provide the list from the book he was reading that cited its view of the top 30 dividend growth stocks at that time (2001). However, as we will see later, there actually was a certain amount of survivor bias to the 30 names cited in the 2001 book titled "The Dividend Growth Strategy." Chowder's comment was followed up by one from David Van Knapp as follows:
"It would take a lot of work, but that list could be "back-tested" in a semi-realistic fashion. Personally, I would require that a version my own porfolio management rules be followed, including selling stocks (usually) when they cut their dividend. (That's the point you were making about "monitoring" your portfolio.) Reinvesting the proceeds would be problematic, but a rule could be set up for how to do that, as well as how to handle M&A activities. Reinvesting dividends would be a real headache, but again a rule could be set up (like automatic reinvestment).
There are many companies on that list that are in a lot of dividend growth portfolios today. I'm virtually certain that it kept up with inflation since 2001. Wonder what the yields were back then?
Although I was unable to set up a back test that would comprehensively address all of David Van Knapp's requirements, I do feel that I was able to set up a back test that was in many ways more comprehensive and that provided more actionable knowledge. The back test I generated will be provided in great detail later in this article.
Next, David Fish added his commentary that I felt was profoundly important on two fronts. First he pointed out the overvaluation bias that existed in 2001, and then he introduced the important principle of a buy-and-monitor approach.
David Fish's quote is as follows:
"Considering the widespread overvaluations back in 2001 (even for good companies) and the fate of certain companies like AIG, Fannie Mae, etc., it almost looks like it could be a "worst case" scenario for DGI. But even then, it might not have been a total disaster. The argument, as you mention, would be that DG investors would have taken a buy-and-monitor approach, so back-testing a buy-and-hold might have been instructive, but not necessarily realistic."
A Comprehensive Review of the top 30 Thoroughbred Dividend Growth Stocks from the book-"The Dividend Growth Investment Strategy" by Roxann Klugman, J.D., L.L.M.
In my own contributions to the excellent comment stream in David Van Knapp's article, I discussed my aversion to studies in general. My big problem with studies is that they must rely on combining a significant amount of data that is all mixed together in order to provide average results. To my way of thinking, by combining the good, the bad and the ugly into one lump, you are thereby failing to make the distinctions that exist among the individual components. Therefore, since the universe was only 30 names, I felt that a study that specifically revealed each of the 30 names was doable and would be capable of providing clearer insights.
In order to create my mini study, I simply went to the create portfolio section of the graph research tool and typed the 30 names that Chowder provided into the edit tickers box and then clicked update. Then I creatively named the portfolio the Chowder Backtest.
Once my Chowder Backtest portfolio was created, I generated the following portfolio review that listed the 30 companies in alphabetical order by symbol. Note that there are three companies at the top of the portfolio review that are no longer in existence, and there are also companies that were spun-off from Chowder's original list of 30 that were added as a method of compensating for survivor bias. For example, Philip Morris (nka Altria) now includes Kraft Foods (KFT) and Philip Morris International (NYSE:PM).
Once the portfolio review was created, I could look at each individual selection by simply clicking on the button containing the stock symbol. I prefer this more specific analysis in lieu of a study based on a generic combining of huge data sets mixing the good, the bad and the ugly. Therefore, I'm going to show each of these companies by providing an earnings and price correlated F.A.S.T. Graphs™ on each that includes the accompanying performance over the period 2001 to now. I will provide limited commentary on the three best performing companies in this group, followed with some commentary on the three worst. For the remainder of the examples, I will briefly comment while letting the graphs mostly speak for themselves.
When reviewing each of the earnings and price correlated graphs, and for the sake of simplicity, there are a few areas that should be focused on. First, the orange line represents the earnings justified valuation. Also, the earnings growth rate is noted to the right of each graph. The black line is the monthly closing stock price. When the black line touches the orange line, fair value exists. When the black line is above the orange line, overvaluation is evident, and will be a major contributing factor to poor returns. The reader should further note that most of these companies, as David Fish pointed out, were overvalued in 2001. Therefore, it is one of the worst possible time frames from which to evaluate dividend growth stocks on.
The focal point for this analysis should be on the performance tables. Especially on the dividend record since the primary thesis here is whether dividend growth kept up with inflation or not. I've highlighted the 2001 dividend yield to answer David Van Knapp's question regarding what the yields were back then on each company.
The Three Best Dividend Growth Stocks from the Book "The Dividend Growth Investment Strategy"
The three dividend growth stocks from the list that had the best historical performance are Pall Corporation (NYSE:PLL), Illinois Tool Works (NYSE:ITW) and Eaton Vance (NYSE:EV). The best performer, Pall Corporation, has strong earnings growth in excess of 8%, started the time frame only modestly overvalued, but has a very spotty dividend record. Nevertheless, Pall produced the best track record of the entire group.
Illinois Tool Works
Although Illinois Tool Works started with an overvaluation headwind, strong earnings growth allows the company to produce a decent return for shareholders. Dividend growth was strong throughout the period and easily outstripped inflation.
Eaton Vance Corp
Price followed earnings.
The Three Worst Dividend Growth Stocks from the Book "The Dividend Growth Investment Strategy"
Interestingly, the three worst performing growth stocks from the list are each a collapsing earnings story. American International Group (NYSE:AIG) and General Electric (NYSE:GE) clearly suffered from the financial debacle. While Merck (NYSE:MRK) suffered from a combination of excessive valuation and faltering earnings in later years. Each of these three companies underperformed the S&P 500 because of high starting valuations exasperated by falling earnings in the later part of this time frame.
American International Group
General Electric Co.
Merck & Co.
Beginning overvaluation and earnings headwinds had a large impact on this company's performance since 2001. Weaker earnings also caused a frozen dividend for the six-year period 2005 to 2010.
In Alphabetical Order by Symbol the Remaining Dividend Growth Stocks from the Book "The Dividend Growth Investment Strategy"
As previously stated, I will let the graphs tell the story on each of the remaining companies from the original list. However, I will add the occasional comment regarding spinoffs. Remember, the performance tables, especially the dividend cash flow table, tell the story about whether dividend growth kept up with inflation. On the other hand, the earnings and price correlated graphs vividly illustrate how well each company performed on an operating basis, and the impact valuation had on their performance.
Aflac started out slightly overvalued, and even though it is now undervalued, in my opinion, the company still produced an S&P 500 beating track record. Furthermore, dividend growth has been very strong and certainly has exceeded the inflation rate.
Avery Dennison Corp. (NYSE:AVY)
This company fell victim to the great recession of 2008 with both earnings and dividends falling.
ConAgra Foods Inc. (NYSE:CAG)
Flat earnings led to flat performance.
Cintas Corp (NASDAQ:CTAS)
Beginning overvaluation and flat earnings are the story here.
Home Depot Inc. (NYSE:HD)
Overvaluation took its toll more than earnings.
HJ Heinz Company (NYSE:HNZ)
A strong correlation between earnings and stock price.
Hildebrand Industries (NYSE:HI)
Hildebrand Inc. was spun-off and the name was changed to Hill-Rom Holdings Inc. (NYSE:HRC). Both companies are offered for survivorship bias perspective.
Johnson & Johnson (JNJ)
Overvaluation overcame good earnings growth. However, dividend growth correlated to strong earnings growth.
Mostly a valuation issue, however, dividend growth correlated to earnings growth.
Medtronic Inc. (NYSE:MDT)
Overvaluation hurt capital appreciation, but dividend growth followed strong earnings growth.
Altria Group Inc. (NYSE:MO)
Philip Morris , now known as Altria , is one of the companies that did not get past survivor bias. It is commonly known that they spun-off Kraft and Philip Morris International since 2001. Therefore, I've included all three companies in order to provide a perspective of how shareholders in the original Philip Morris would have fared. As you can see, the spin-offs have done fairly well, especially Philip Morris International.
M&T Bank Corp (NYSE:MTB)
In May of 2011, Wilmington Trust joined M&T Bank Corp.
PepsiCo Inc. (NYSE:PEP)
Higher than normal beginning valuation of PepsiCo shares in 2001 lowered the capital appreciation component. However, dividend growth followed earnings growth.
Pfizer Inc. (NYSE:PFE)
Both high valuation and reduced earnings growth have a major impact on performance, including dividend cuts.
Procter & Gamble Co. (NYSE:PG)
Modest overvaluation, but an above-average record of earnings growth allowed Procter & Gamble to generate S&P 500 beating returns. Furthermore, inflation-beating dividend growth followed strong earnings.
State Street Corp. (NYSE:STT)
A high beginning valuation and a low ending valuation produced anemic returns for shareholders. Also, weaker earnings resulting from the financial crisis caused the dividend to be slashed by 96%.
Sysco Corp (NYSE:SYY)
Overvaluation hurt performance, but dividend growth bettered inflation thanks to strong earnings growth.
Tootsie Roll Industries Inc. (NYSE:TR)
No earnings growth at all is the story here.
Wal-Mart Stores Inc. (NYSE:WMT)
High beginning valuation totally destroys exceptional operating results. But strong dividend growth follows strong earnings growth.
Summary and Conclusions
I want readers to understand that I acknowledge that what I've produced here is not in the strictest sense a scientific study. However, I would argue that the information presented here is far superior to what a study based on crunching a bunch of unrelated data points could produce. Even though all the companies in this report are considered dividend growth stocks, it is clear that all similarities stop there. Yes, the companies do pay dividends, and yes, many of them have actually increased their dividends, but each represents a very unique company.
Therefore, I believe that looking at each company on an individual basis provides far more insight and understanding into what individual dividend growth stocks offer individual investors, and why, than any study could ever hope to accomplish. Some of these companies did great, others generated horrible results, and the reasons why are varied from one company to the next. Overvaluation was a major culprit that affected the performance of most of these names. However, individual operating results (earnings growth rates) also varied widely from one company to the next. Finally, although survivor bias was not precisely dealt with, I thought I provided enough information to state that it is an overblown concern.
Disclaimer: The opinions in this document are for informational and educational purposes only and should not be construed as a recommendation to buy or sell the stocks mentioned or to solicit transactions or clients. Past performance of the companies discussed may not continue and the companies may not achieve the earnings growth as predicted. The information in this document is believed to be accurate, but under no circumstances should a person act upon the information contained within. We do not recommend that anyone act upon any investment information without first consulting an investment advisor as to the suitability of such investments for his specific situation.