For the past couple of years, an active discussion or debate has been going on between proponents of dividend growth investing [DGI] and proponents of modern portfolio theory [MPT]. I have been in the middle of some of those debates. I often wonder whether the two strategies for portfolio construction can be reconciled. I will state my preference up front: My primary investing strategy is a DGI strategy. I recognize that many readers will see this as biasing this article. I am sure that any mistakes, stereotypes, or misconceptions can be rectified in the comments section.
MPT intrigues me, and I have made many efforts to compare the two paradigms of investing. MPT's main principle - that it is possible both to decrease the volatility and increase the total returns of a portfolio by investing in uncorrelated asset classes - strikes me as both non-intuitive and very much supported by most of the data and many studies. I think that the original insights and research border on brilliant.
That said, I am troubled by the absolutist nature of what MPT has become, which is dogma in many pockets of the investment advice industry. Most troubling is that MPT is focused on a single objective - maximizing total return for a given amount of "risk" - when many investors have other objectives or are looking for solutions to specific problems. In particular, many investors desire to build reliable growing income streams for retirement, income streams that they can live off without selling the underlying assets. MPT does not address this goal directly.
Instead, MPT targets the building of maximum wealth. Funding retirement then becomes a matter of combining whatever income the MPT portfolio may provide organically with the selling of assets. The idea of selling assets, or "decumulating," strikes many individual investors as scary, because there are clearly scenarios in which a person might outlive their assets. While everyone recognizes the necessity to sell assets sometimes, for example, to make a large capital purchase or to cope with an emergency, the idea of making it the routine process for funding retirement strikes many as imprudent.
Sometimes it seems as if the two investment strategies exist in parallel universes. Even though in the end both are aimed at maximizing the effectiveness of investing, the two "schools" have different objectives, utilize different strategies and tactics, and even use nomenclature differently. None of this helps proponents from either school understand the other. For example, it is not uncommon for MPT proponents to become confused at DGI's focus on income rather than total return. The other side of that coin is that many DGI investors do not understand why MPT proponents are so focused on acquiring maximum wealth if that wealth is subject to market risks that may strike just when they need the money.
While I am continually trying to learn more, currently I do not think that the two approaches can be fully reconciled. In the sections that follow, I will explain my thinking on topics that are important to both MPT and DGI.
One main reason I believe that MPT cannot be fully reconciled with DGI is that DGI, by its very nature, is focused on stock investing. Dividends come from stocks. MPT, on the other hand, is focused on multiple asset classes - not only stocks, but also bonds, cash, commodities, "alternative investments," and really anything that can be defined as an asset class. Many MPT proponents consider DGI's focus on stocks to be imbalanced, undiversified and imprudent.
I have come to view dividend growth stocks as an asset class in their own right. This was originally suggested to me by David Jackson, SA's founder and an MPT proponent. However, that is not an idea that has gained much traction, especially with MPT proponents who fail to see how dividend growth stocks are uncorrelated with the asset class stocks themselves or such recognized stock categories as large-cap-value. To be a true asset class in MPT, an investment type, must be minimally correlated with other asset classes.
I understand that, but I do wonder why one of the most obvious differentiators between stocks, namely, whether they pay dividends or not, has not gained more general recognition as being of paramount importance. The yield of every stock is either zero or not. That seems to be an easy and obvious first cut at cleaving the stock universe into two broad categories that are fundamentally different from each other.
I would take it one step further. To dividend growth investors, it is not just the issuance of a dividend from time to time that is important, but rather the reliable distribution of increasing dividends. The Dividend Champions document compiled and updated monthly by David Fish is the best compilation in existence of stocks that have increased their dividend payouts for five years or more. I would suggest that this document represents an asset class in common sense if not in MPT parlance.
DGI involves picking individual stocks to construct a portfolio. Investors look for the particular elements that will advance their goals. Typically, those elements include, but are not limited to, dividend history, current yield, dividend growth history, business model, financial fundamentals, dividend sustainability, and valuation.
MPT, on the other hand, turns away from individual stock picking. Asset classes are defined, and it is the various asset classes that are "picked" to construct the portfolio. More specifically, investment products such as mutual funds or ETFs are picked, with different products representing the various asset classes. The overall portfolio is constructed from these investment products, which form the building blocks of the portfolio.
Minimizing Negative Market Impacts
Both MPT and DGI attempt to minimize the impact of market gyrations on results, but they do so in very different ways. DGI attempts to remove the market from results by focusing on that portion of total return that is not controlled by markets: Dividends.
MPT focuses on total return, which includes dividends. The equation is simple: Total Return = Price Return + Dividends. But dividends are independent from the market. Dividends are direct transactions between companies and their owners. The market is not involved in the declaration or payment of dividends. This is very important to most dividend growth investors, in fact it is a principal reason that they utilize a DGI strategy. They do not want markets messing with their income stream.
MPT comes at the impacts of markets differently. Rather than rendering market actions irrelevant by focusing on dividend transactions that take place outside the market, MPT attempts to minimize the damage from market fluctuations by participating in many markets simultaneously. The idea is that different markets (different asset classes) zig and zag at different times, with the end result that overall volatility is reduced at the same time that overall returns are improved. That is why MPT leads to portfolios constructed of many different asset classes. The thesis is that owning several uncorrelated asset classes results in smoother and better results.
Another principal reason why DGI and MPT cannot be fully reconciled is that they are focused on different goals. MPT focuses on achieving total return. It is presumed that greater wealth is superior to lesser wealth, and that "wealth" is measured by the total market price of all assets at any given time.
When it comes time for retirement, the financial advice industry has developed strategies for converting that wealth from capital into the monthly or quarterly income streams that retirees need to live on. Two of the more well-known retirement funding strategies are the "4% rule" and the "bucket strategy." Each depends on "decumulating" (selling) assets to convert capital assets into spendable money. Each also depends on the value of the remaining assets (after some are sold) increasing in value at a rate at least equal to inflation so that assets are not spent down too fast. Under ideal conditions (growth exceeding inflation), the decumulation strategies actually can result in significantly more wealth piling up, despite the periodic withdrawals. The equation for that would be Fewer Assets x Higher Prices = More Wealth.
DGI, on the other hand, does not see retirement as a finish line for achieving the most wealth possible. DGI instead sees retirement as an inflection point in how dividends are used. Prior to retirement, dividends are reinvested to accelerate the process of accumulating shares that generate income organically. The increasing number of shares generates increasing dividends. The compounding effect over many years can build up an impressive income stream. There is actually a second layer of compounding, namely the annual dividend increases that the DGI investor receives. The DGI investor may actually have a different concept of "wealth" itself. It may be seen more as the ownership of assets with the inherent ability to produce increasing income rather than the total market value of the assets.
Upon the retirement inflection point, the dividends are no longer reinvested, but instead are taken as spendable income. Because DGI involves ownership of companies that have a history and culture of raising dividends annually, DGI depends on the income stream from the portfolio rising enough each year to at least equal the rate of inflation. In the ideal case, no assets need to be sold. Over long periods (decades), the DGI investor expects the market prices of shares to rise too, resulting in considerable conventional wealth from the unsold assets.
One of MPT's weaknesses, in my opinion, is its reduction of "risk" to a single metric, volatility. In most discussions of MPT, the volatility of total return is used to define risk. Volatility is usually measured by the standard deviation of total returns. A secondary measure might be "maximum drawdown," showing how much the portfolio's value dropped during a market crisis such as the dot-com bubble or the financial meltdown in 2008. Risk tolerance is assessed by how much an investor thinks he or she could stand to see their nest egg drop.
Portfolio risk was discussed in an article in the most recent AAII Journal. The author, Jason Brady, is classically trained and a CFA charterholder. He points out that the traditional single axis of risk does not take into account a number of risks that investors face, particularly inflation.
I agree with Brady. In my view, MPT's one-size-fits-all definition of risk is inadequate. It ignores the risks that are most important in DGI, namely (1) the risk to each stock's dividend, and (2) the risk of inflation. I know of no easy way to quantify dividend risk. Clearly it exists, as any dividend growth company bears some danger of freezing or cutting its dividend. But without study, you don't know which companies are in dividend danger nor when. DGI investors use various metrics - such as payout ratios and fundamental financial measures - to assess dividend risk in advance, and they also use portfolio management guidelines - such as selling if there is a dividend freeze or cut - to ameliorate the risk. It seems obvious that the risk to any dividend growth company's dividend is far smaller than the risk to its price, as prices rise and fall all the time whereas dividends do not.
Few DGI investors are uninterested in their portfolio's total value, but many tend to view it as a secondary consideration, rather than the only consideration as is done in MPT. Typically the dividend growth investor accepts portfolio volatility as part of market reality, and because they are more interested in growing income streams than in overall portfolio values, they are less interested in short-term volatility.
Inflation is a risk borne by all no matter what investments schemes you use. While MPT does not ignore inflation - indeed, MPT expects that portfolio values will rise faster than inflation over the long haul - it does not explicitly include inflation in its definition of "risk." DGI, on the other hand, explicitly considers inflation as a risk to be overcome. DGI investors strive to construct portfolios whose dividend streams rise faster than inflation.
Can DGI Serve as the Stock Portion of MPT?
One way in which the two investing strategies might be reconciled would be to use DGI for the stock portion of MPT. Under a "typical" 60/40 stock/bond basic allocation, one could build the 60% stock portion using DGI principles. I believe that in reality many investors do this to a greater or lesser degree even if it is not part of a formal strategy.
However, I believe that most MPT proponents would not consider a DGI stock portfolio to be sufficiently diversified even among stocks. Most pie-chart allocation schemes that I have seen recommend holding a variety of stock types. For example, one scheme that has gained popularity is the 7Twelve portfolio, offered as a "guideline for achieving optimal portfolio diversification." In that model, five stock and three "diversifying" categories are recommended, all to be purchased via mutual funds:
- Large US Stock
- Midcap US Stock
- Small US Stock
- Developed Non-US Stock
- Emerging Non-US Stock
- Real Estate
I should point out that many DGI portfolios contain stocks that fit into most or all of the categories above. The commonly held belief that "dividend growth stocks" are all large-cap US-based stocks is incorrect. The only requirement is that the stocks have dividends and raise them regularly. Here are examples of stocks in the above categories from my Top 40 Dividend Growth Stocks series:
- Large US Stock-AT&T (T), McDonald's (MCD)
- Midcap US Stock-Alliant Energy (LNT), Darden Restaurants (DRI)
- Small US Stock-Avista (AVA)
- Developed Non-US Stock-AstraZeneca (AZN), Rogers Communication (RCI)
- Emerging Non-US Stock-Endesa Chile (EOC)
- Real Estate-Realty Income (O)
- Materials-BHP Billiton (BBL)
- Commodities-Alliance Holdings (AHGP)
This is probably the area where there is the most agreement between MPT and DGI. Both use the principles of diversification to promote safety.
MPT diversifies both among asset classes and within asset classes. As to the latter, some MPT adherents believe that once an asset class is defined, the investor should own every individual asset within that asset class-all large US stocks, for example. This is accomplished by the purchase of funds or ETFs.
Most DGI investors diversify as well. I have seen arguments for portfolios of as few as 10 stocks and as many as 100, with every variation between. Beyond the number of stocks, as illustrated above it is easy within DGI to diversify across industries, country of origin, cap size, and the other traditional stock categories utilized in MPT. In addition, most DGI investors hold cash-which is an MPT asset class-in some measure, if for no other reason than to cover ongoing expenses or as a designated "two year spending needs" fund.