In a comment David Van Knapp's article seekingalpha.com/article/805611-open-let..., I brought up the idea of weighting a portfolio of dividend growth stocks by income as opposed to market value. I thought this would be something intriguing to dividend growth investors as so often I read how many of the investors that follow this investment approach have little or no interest in the underlying market value of the stocks other than at the point of purchase.
Surprisingly, the idea received a rather cold reception by a few of the frequent dividend growth contributors (the exception being Robert Schwartz who seemed interested). Perhaps this was a function of me (Counterpoint) bringing this up as I know I am generally viewed as an "outsider" so my comments tend to be viewed with skepticism regardless of their content. On the other hand, perhaps there is a subset of dividend growth investors who call themselves this but are really just plain old value investors trying to wear dividend growth clothing…even though they violate one of its key rules…letting market values materially affect their post-investment decisions.
Let me state, contrary to the implications of the comments responding to this idea, I would not view either approach, income or market value weighting, as something that should be rigidly applied to portfolio management any more than an MPT practitioner would rebalance his or her ratio of stocks and bonds with each movement of the markets. While it would actually be possible to do this with a dividend income weighted portfolio due to dividend changes happening, typically, once per year as opposed to second-by-second with prices, it would still result too frequent rebalancing in a portfolio of 20 or 30 stocks in my opinion. Much like I manage my asset allocation portfolio, I would not reallocate more frequently than once a year (of course, one of the nice things about dividend growth investing is that you have periodic mini-rebalancing opportunities if you reinvest your dividends).
More importantly, I would not rigidly say that each holding has to produce the exact same amount of income after a rebalancing. Again, as I do with my asset allocation portfolio, I create "curbs" such that each category stays with a range of a few percent. So, let's say that in a portfolio of 20 stocks for example, where in a completely equal situation each stock would provide 5% of the income, any stock with more than 6.5% or less than 3.5% of the total dividends would be adjusted downwards or upwards respectively at the time of a rebalancing. Let me note too that special dividends would, of course, not be part of such a calculation (can't believe I had to add that).
I am not making any assertion that one will achieve a superior return with such an approach, but that is not why I raised this idea. I raised it because the purpose of diversifying your holdings is risk management. As an asset allocator, I view "risk" as the possibility of a deterioration in the market value of my holdings so I diversify my holdings based on market value. Dividend growth investors view risk as the possibility of a deterioration of their income stream. If this is the case, their holdings should be diversified based upon the income not the market value.