If Dividend Growth Investing can beat Index investing, where is the portfolio that shows this?
In the various debates about whether or not Dividend Growth Investing is a viable strategy and how it stacks up to Index Investing, one question is often raised by proponents of Index Investing. And that question is, where are all the DG portfolios that beat Index Investing?
While the answer is complex, it is a fair question. I know with my own portfolio ( here is the latest update) is not currently beating the indexes I have compared it to. Part of that is due to the short time frame it has been in existence. Part of that is due to my directing cash flows into sectors and stocks where prices were falling while the indexes are doing the opposite. I bought all good companies and by purchasing them when prices were low I was able to grow dividend income faster. I fully expect this to pay off down the road. But I don’t expect my portfolio to beat the comparison indexes each year in total return. While I don't worry about it very much, I do expect that over the long term dividend growth will result in price growth too.
Understand as well that actual investors who use index funds under-perform their indexes. That may sound counter-intuitive. But actual investors tend to buy more shares of funds that are going up and sell shares of funds that are going down in price. This can and often does lead to investors buying high and selling low. So it’s not just investors who buy individual stocks that act in ways that reduce their long term performance.
Individual actions and reactions can make a big difference in outcomes but I would like to see how the basic strategies compare. To do that, I will use Portfolio Visualizer to track how 3 different model portfolios would have performed over the last 15 years. I wanted to go 20 years, but the ETFs I used were not all in existence for a full year until 2003.
Let’s look at the model portfolios and why I chose them
Let’s first look at the stocks I selected for my model DGI portfolio. I actually created this model portfolio for my second article on Seeking Alpha. I was refuting an article by an author who claimed that dividend growth investing wasn’t very good. Much of his claim rested on the poor performance of his DGI model portfolio and I argued that this poor performance was due not to a lack in the strategy but in the fund he selected to model it.
Figure 1 Source
As I explained in my original article I took 20 stocks from Mike Nadel’s DG50 list (latest article here). I chose only 20 in part because that seemed a good number of positions and in part because that keeps the display of the portfolio positions at a good size. I have written articles on every one of these companies. I don’t think any of the companies would have been an unreasonable pick for DG investor in 2003 (or 1999 as the start date for the original article). Several have had various issues. I don’t think anyone can claim that I cherry picked General Electric (GE) for instance. My aim was to pick stocks that an actual dividend growth investor at the time would have picked and to avoid picking any stocks that I now know to have a performance that would ensure this portfolio was a winner.
My DGI model portfolio is composed of shares of the following: Coke (KO), Johnson & Johnson (JNJ), Exxon Mobil (XOM), Realty Income (O), Procter & Gamble (PG), Dominion (D), McDonald’s (MCD), Walmart (WMT), IBM (IBM), Microsoft (MSFT), AT&T (T), Clorox (CLX), J.M. Smucker (SJM), Emerson Electric (EMT), General Electric, 3M (MMM), Lockheed Martin (LMT), Colgate Palmolive (CL), WEC Energy (WEC) and Wells Fargo (WFC). With a few exceptions in 2003 each had increased dividends each year since 1999 according to David Fish’s CCC List(which contains data on companies that have raised their dividend each year for 5 or more years). Dominion Energy had a frozen dividend but as a utility and a good one, it’s a reasonable pick anyway. Microsoft had just begun paying a dividend in 2003, but as a technology leader it’s a solid growth company and had the potential to turn into a dividend growth company. Lockheed Martin had frozen its dividend but had resumed increases. As a defense contractor it certainly had the potential to resume being a good dividend growth company.
Next up, I created a model portfolio with the common 60% stocks and 40% bonds configuration. For this portfolio I used the ETF (SPY) for the stock portion. This is a fairly common and well know S&P500 ETF and has low fees as well. For the bonds I looked at both (AGG) based on a corporate bond index and (TLT) which is based on a long term treasury bond index. I ended up using TLT because that produced better results since January 2003. Since I want to show that DG investors can beat basic index strategies, I am okay with picking a bond fund that does better over the studied time period.
Figure 2 Source
Finally, sticking with the 60% allocation to stocks and 40% allocation to bonds I created a model portfolio with a wider selection of stocks. First, as the NASDAQ exchange has some pretty impressive growth stocks, I added (QQQ) to capture some of that growth. Next Mid-cap and Small Cap stocks also tend to have higher total return over long periods so I added (MDY) and (IJR) to capture that class of stocks as well. I looked into using (SDY), since that uses an index from Standard and Poor’s (the S&P Small Cap 600), but it doesn’t have a full year of data until 2006. The performance of IJR and SDY are pretty close as well.
Figure 3 Source
So how did the 3 portfolios do in comparison to each other?
I originally wanted to do a 20-year comparison, but unfortunately the ETFs I used have been around quite that long. 2003 is the first year where my tool has full year results for all the ETFs, and 15 years is still a long time. I think showing performance over the period from January of 2003 to the end of September 2018 will show how each portfolio performs over the long term.
A limitation, and in this case a feature, of Portfolio Visualizer is that each portfolio is managed in essentially the same way. For the purpose of this study this is very much a feature because the comparisons I make will not be influenced by differences in actions taken. So for instance, if Microsoft shares take off in price, the DGI portfolio won’t buy a bunch more (or sell them at the higher price). When GE announced a dividend cut, the DGI portfolio won’t sell those shares. Just like the Simple 60/40 portfolio won’t sell bonds when the Fed announces a rate hike (or cut).
For the first case study, we will model how the portfolios would have performed if each portfolio was purchased with $1 million at the beginning of January 2003. Any distributions will be reinvested by buying more of the security that paid them. Positions will be allowed to grow or shrink freely. No re-balancing will be done.
Figure 4 Source
Above are the results. Notice that the DGI portfolio does in fact beat both of the index-based 60/40 portfolios. Over just 15 years, the DGI portfolio has around 8% higher market value than the best of the other 2. More importantly, look at the values for the Sharpe and Sortino ratios. These ratios measure traditional risk and the higher the number the lower the traditional risk. So by these metrics the DGI portfolio delivers 58 basis points more return per year for essentially the same risk as the Complex 60/40 portfolio and 237 basis points more return per year than the Simple 60/40 portfolio at less risk.
It’s great that all 3 portfolios are growing nicely and should provide a nice nest egg for the owners, but it’s also important to take inflation into account. It has been called the hidden tax and it clearly erodes the purchasing power of one’s savings. Portfolio Visualizer does allow one to show inflation adjust returns on the performance graph, so below I included that graph.
Figure 5 Source
All the portfolios do well more than countering the effects of inflation over the long term. However, 2009 looks to be a bad year for all of them. I don’t think it’s unexpected that the Great Recession had a big negative impact.
I did run the numbers for a scenario where $5000 was added to each portfolio each year. However, its results were very similar to the first case. The big difference was that each portfolio had a higher market value. Every portfolio had roughly $200K more but the risk metrics were the same.
The third case is based on our intrepid investors having already saved $1 million when they decide it is time to retire. Each of the 3 will put their $1 million into 1 of the 3 portfolios starting at the beginning of January. Each has decided that in order to meet expenses that they will withdraw 5% of the balance each year. Now this is more than the yield of the DGI portfolio, but this was a scenario suggested by a poster who is a proponent of index investing. And it’s fair enough, while most DG investors aim to cover their retirement expenses entirely with dividends, plenty of folks find themselves short of this goal.
Figure 6 Source
In this scenario all the portfolios grow less than they did in the previous one. Which, given that the investor is withdrawing money to live on shouldn’t be unexpected. And with a 5% withdrawal rate, the investor gets more money each year. Except around 2009. This nicely highlights the problem with withdrawing a fixed percentage each year. When the portfolio is growing faster than average the investor could end up taking out too much money, so that when growth slows (or the portfolio even shrinks) there might not be enough money to cover expenses. Notice that the Sharpe and Sortino ratios are the same as in the first case. This is because cash is withdrawn proportionally from all positions.
So as case #2 showed us, withdrawing a fixed percentage each year can lead one to withdraw too much money in the boom times leaving not enough to cover the bust times. Can withdrawing a fixed dollar amount, adjusted for inflation correct for that? Let’s take a look.
Figure 7 Source
The good news is that the drop in market value for all the portfolios is less, as was to be expected. The DGI portfolio didn’t even drop below $1 million (like it did with the fixed percentage). While the Great Recession was pretty deep it didn’t last all that long. And it didn’t last long enough to damage any of our model portfolios.
These results do, however, show an issue with sequence of return. Sequence of return becomes a problem when yearly results are significantly lower than the average. This can become a problem when losses (or slow growth) happen early in the performance window. This can be a big problem for those who plan to harvest capital gains, but it is not entirely absent from the portfolios of those who only collect dividends. While DG investors try to diversify their portfolios so that they are unlikely to contain a large number of companies cut dividend payments at the same time, it can happen.
Let’s check out the casino
One tool used to assess the safety of a portfolio and its ability to support generating cash over long periods of time is called Monte Carlo simulation. Basically this method uses past price movements to simulate future price movements pseudo-randomly. Typically, 10 thousand or so simulated periods are run, and the idea is that the portfolio is a good one if 95% or more of the simulated portfolios will still have money left at the end of the time frame chosen.
The Portfolio Visualizer site also has a tool for doing Monte Carlo simulations, so let’s compare the 3 portfolios with that tool as well. For each portfolio I will pull $50,000 each year adjusted for inflation over a 40-year period. To account for sequence of return risks, I will place the worst 2 years of performance first.
Here are the results of the DGI portfolio.
Figure 8 DGI Portfolio Monte Carlo Results
The 9802 results that succeeded out of 10,000 shows that this portfolio should be okay. And the 13.44% rate of return for the 50th percentile is pretty good.
Here are the results for the Simple 60/40 portfolio.
Figure 9 Simple 60/40 Portfolio Monte Carlo Results
With only 9641 outcomes that don’t run out of money, the Simple 60/40 portfolio is slightly more likely to fail but it is still well above the 9500 threshold. With the provision that the 2 worst years happen at the start, this is not likely to happen, so investors should be okay with this portfolio as well. That it has more failures and lower returns does show that sequence of return risk has a greater impact on it because it requires more cash flow from share sales than the DGI portfolio.
Here are the results for the Complex 60/40 Portfolio.
Figure 10 Complex 60/40 Portfolio Monte Carlo Results
This portfolio had the most outcomes that didn’t run out of money at 9970, making it slightly safer than the DGI portfolio. However, its yield was much lower as well.
These results further reinforce the idea that a dividend growth based portfolio can beat the market (as modeled by broad based index funds) without any more risk. In fact, the DGI portfolio here beat the market with less risk.
In subsequent comments a slightly different 60/40 stock and bond portfolio was suggested using (MDY) as the stock component. In back testing it do do better than the Quick DGI portfolio I put together here. But it doesn't fair as well in the Monte Carlo simulation.
Figure 11 MDY 60/40 Portfolio Monte Carlo Results
I often here the claim that it is very hard if not impossible to beat the market, so investors shouldn’t try and just use index funds. Well, I picked a portfolio of dividend growth stocks that are not particularly outstanding in their performance. Some might argue that plenty of them are in fact mediocre. And one, GE, is doing very badly, one might even say disastrously currently.
And yet this portfolio, on a total return basis even, beat a fairly standard 60/40 stock and bond portfolio. And an even more complex portfolio that included not only an S&P500 index, but a mid-cap and small cap index funds. While this in no way guarantees that every investor can put together a portfolio that will beat the market (especially if they aren’t trying to do so), it does say it is possible.
Note: I hope you all got something out of this article. I do appreciate the time you took reading it. If you are one of those who follow me here, I appreciate it; if you'd like to include yourself amongst those individuals, please hit the "Follow" button next to my name as well as following other contributors whose work you enjoy. As always, please leave any feedback and questions you may have in the comments below. SA has also added a like button at the bottom of articles.
Disclaimer: This article is intended to provide information to interested parties. As I have no knowledge of individual investor circumstances, goals, and/or portfolio concentration or diversification, readers are expected to complete their own due diligence before purchasing any stocks mentioned or recommended. The price I call fair valued is not a prediction of future price but only the price at which I consider the stock to be of value for its dividends.
Disclosure: I am/we are long KO, JNJ, XOM, O, PG, D, T, EMR. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.