Open Letter To ETF Industry: Create A Better Dividend Growth Product

by: David Van Knapp

Dear ETF Industry,

There is a vibrant market for an ETF or mutual fund that implements a sensible dividend growth strategy. With all due respect, the myriad "dividend" products that you have created until now don't cut it. I spent some time this past spring looking for good dividend growth ETFs (I, II, III), but I finally gave up in frustration.

I am happy enough being an individual dividend growth investor myself, and I don't personally need a good dividend growth ETF at the moment. But many other investors could use such a product. And some day I will die. I would like to be able to tell my wife, who is not very interested in investing, to "just put it into XXX." But a suitable XXX does not exist.

I cannot understand why that is the case. The interest in dividends is at a generational high. That interest only figures to grow as Baby Boomers turn 65 at a rate of about 10,000 per day and will continue to do so for the next 17 years. I would think that you would be using all of your knowledge and expertise to create the best dividend growth products imaginable. Beyond the Baby Boom generation, many younger investors in their accumulation years are seeing the value in using a dividend growth strategy to take advantage of compounding over a long period of time. The market is huge.

Instead of a good dividend growth investment product, however, we get products with weird, unexpected flaws:

  • "Dividend" ETFs that do not appear to have a goal of maximizing and growing dividends, despite their names. Some of them report only on total returns, despite their apparent principal appeal to income-seeking investors.
  • ETFs that do not react to red flags such as dividend cuts.
  • Lousy yields.
  • Sorry dividend growth rates and yields on cost.
  • Distributions that have been cut from the prior year rather than showing steady year-over-year increases.
  • Annual reconstitution of the stocks held, therefore an inability to react to dividend events that happen in the middle of the year.
  • Oddly high volatility.

You guys can do better than that!

I am submitting this modest proposal about how to construct a dividend growth product that can stand up to the tests of time, common sense, and principles that have been developed by individual investors for generations to create stable, growing income streams for years and years.

I will be the first to admit that I do not know all the mechanics of constructing an ETF that is practical, low in cost, and complies with all regulations and legal requirements. But I do know how to select dividend growth stocks and manage a portfolio of them. For technical guidance in preparing this modest proposal, I have relied somewhat on "Construction Rules for Morningstar Wide Moat Focus Index," a Morningstar Methodology Paper dated June 12, 2012. It is that paper that convinced me that a better dividend growth index can be created. The ETF will be based upon such an index.


The primary purpose of the dividend growth ETF will be to provide an investment product that creates growing income streams of meaningful size over time through investments in common stocks, REITs, and MLPs. (I will use "stocks" to refer to all of these dividend- and distribution-paying instruments.) A specific goal will be to have each year's income stream be larger than the previous year's, every year. This will be accomplished by investing in stocks that are expected to grow their dividends regularly.

Individual ETF owners who reinvest their dividends back into the ETF will experience faster dividend growth than those who do not (through the purchase of more shares that generate more dividends), but everyone will receive growing income from their investment whether they reinvest or not, because the ETF's constituents will be growing their distributions as described above.

A secondary goal will be growth in total return, which is expected to occur roughly in concert with the growth of earnings of the constituent companies. If reporting regulations require an emphasis on total return, this ETF will also clearly explain that its primary goal is income growth, and it will report its income performance as prominently and clearly as total return is reported.

A third non-financial goal for this ETF will be to become a model of transparency and clarity in the world of ETFs. Marketing materials for this ETF will emphasize that its primary goal is reliable, growing income, and that total return is a secondary goal. In all other communications and reports, the ETF will strive to be a beacon of education and clear communications regarding goals, strategies, methods, and outcomes. None of the existing "dividend" ETFs that I examined highlighted dividend returns or the growth in their distributions. None spoke at length about the pros and cons of dividend growth investing during accumulation years nor during retirement. My vision is that the literature connected with this ETF can serve educational as well as strictly marketing or reporting functions.

Index/ETF Committee

The index will be created and administered according to rules that are created, maintained, and occasionally refined by an Index/ETF Committee. It is important that the index be rules-based, so that accurate back-testing (free of survivorship bias or distortions) can be performed to produce performance statistics that go back as far as possible given limits of data availability.

The ETF will be managed by the same Committee. The ETF will attempt to track the index as closely as possible. This means that the ETF is passively managed in the sense that it tracks an index, but also that the ETF is "active" in the sense that the index it is tracking is actively managed. Widespread industry practice now often fails to clearly explain that a "passive" fund nevertheless can have ever-changing constituent investments, because the underlying index is itself repeatedly updated. The Committee will do its best to explain these conditions as part of its charge to operate as transparently as possible.

Stock Selection, Step 1: Qualifying Stocks

Dividend growth investing depends on picking appropriate stocks. There are two stages in stock selection: Identifying stocks that are fundamentally eligible, then selecting from among them those that are favorably valued.

By far the best compilation of candidates for a dividend growth index is the Dividend Champions document compiled and maintained by David Fish of the DRiP Investing Resource Center. The document is updated monthly, and it contains a world of information about the stocks of companies that have increased their dividend payouts every year consecutively for strings of five (Challengers), ten (Contenders), and 25 years or more (Champions). I modestly suggest that you only include in your new index (which I hereby dub The Best Damn Dividend Growth Index Ever, or BDDGX) stocks that are on this document, that is, stocks that have increased their dividend payouts for at least five years in a row.

I have some other ideas about stock selection for BDDGX:

  • Minimum initial yield (at time of purchase) of 2.75%. That number represents a "floor" yield of 3% plus stocks that could get to the floor yield with a single 10% dividend increase.
  • Minimum 5-year dividend growth rate of 5%. This minimum might be adjusted (i.e., placed on a sliding scale) for stocks that have higher yields, in the Committee's best judgment. However, the rule must be made clear and consistently followed in keeping with the rules-based requirement.
  • Rules-based fundamental stock selection criteria that are intelligently chosen to identify stocks that are most likely to be able to maintain and to grow their dividend and distribution payouts. These shall be clearly laid out for all to examine and consider. These criteria shall include due diligence on company fundamentals and shall not be restricted to dividend characteristics only.

Stock Selection, Step 2: Valuation

Valuation is a co-equally important pillar of stock selection for BDDGX. The fundamental idea is that BDDGX will only contain stocks that are well-valued. "Well-valued" must be defined by the Committee, but it should be based on transparent, accessible, and widely accepted criteria for determining what is commonly known as "intrinsic value." Examples would be "true worth" valuations as determined by Chuck Carnevale's F.A.S.T Graph system, "fair value" as determined by Morningstar's star rating system, or an average of analysts' price valuations. The range of valuations suitable for BDDGX shall be determined by the Committee. An example would be that the Committee might decide to restrict the BDDGX to companies that have a current price of no more than 120% of their intrinsic value, and to add new companies only if they have a valuation of no more than 85% of their intrinsic value. This clearly means that not every company that passes the Step 1 eligibility criteria will necessarily be a member of BDDGX at any particular time.

BDDGX is not a subset of any other index. For example, there is no requirement that any constituent be a member of the S&P 500 or any other stock index. It would be, as mentioned above, a subset of the DRiP Dividend Champions document.

How Many Stocks in the Index?

As to the number of stocks in BDDGX, my modest suggestion would be that the number of stocks be allowed to float, comprising at any time all stocks that satisfy the selection criteria, and not be restricted to any fixed number such as 20 or 30 or 50 or 100. The flexibility of the floating number will allow you to add new stocks as they become eligible and drop existing stocks as they become ineligible without worrying about the need to adhere to an arbitrary number.

Stock Weighting

Constituents in the BDDGX should be assigned an equal weighting. Because capital preservation and income reliability are very important qualities in a dividend growth strategy, simple equal weighting provides the best balance for this type of index. Weightings shall be determined by market cap unless the Committee decides otherwise.

Inception Date, Base Value, and Index Calculation

The inception date of BDDGX shall be 10 years prior to the ETF's date of introduction, assuming that sufficient data exists to accurately back-test that far. Daily price, dividend, and total returns shall be made available from that date forward. Also to be made available will be the lists of component companies and a complete chronological history (specific as to dates) of every change made to the index from its beginning date forward. The index base market value at inception will be 1,000.

Current index values will be calculated in real-time and updated at 15 second intervals.

Changes to and Reconstitution of the Index

One of the glaring problems with existing dividend ETFs is their slow reaction times to dividend events. BDDGX will be updated monthly, with near-instantaneous changes when unexpected dividend events happen. The regular index reviews will take place and be implemented on the first Monday of each month, while "instant" changes will be made as soon as possible after the event precipitating them, no later than the following Monday.

Since this index must be rules-based, the following rule will be applied mechanically even though "best practice" might be to allow discretion: Any stock will be dropped from BDDGX if it cuts, freezes, or suspends its dividend.

The regular monthly reviews will allow consideration of other reasons that the Committee adopts for dropping a stock. Here is a proposed list of rules that might be developed. A stock shall be dropped if:

  • Its valuation surpasses 120% of intrinsic value (or some other percentage as determined by the Committee).
  • It is announced that the company is going to be acquired or merged.
  • It announces plans to split itself up or spin off a separate company.
  • Its current yield drops below 2.75 percent.
  • It under-performs the market in total returns (price + dividends) for four years running.

When stocks are dropped, they are not replaced, since the index holds every eligible security. As a rough guess at this time, it is expected that the population of BDDGX will be around 150 to 200 stocks. There is no "waiting list." The monthly meeting will also be the time to add stocks that newly became eligible during the preceding month.

Every sixth month (April and October), the index will be rebalanced back to approximate equal weightings. The committee can decide on an acceptable range of "over-weightings" or "under-weightings" of each stock from precise equal weightings. Some small range of departures, perhaps in the 1 percent range, shall be utilized to keep inconsequential changes to a minimum. However, again, the index must be rules-based, so once the Committee determines the acceptable range of departures from equal weighting, those ranges will be enforced mechanically.

Every change to the index components will be announced and explained in a timely press release, and each change will be recorded in the running history of BDDGX. All of this information shall be freely available and easily accessible.

Submitted humbly and sincerely on behalf of dividend growth investors everywhere,

David Van Knapp

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.