In February 2017, I wrote what has turned out to be, in terms of page views, my second most popular article ever for this platform. That article? A Dividend Portfolio Built From The World's Best Dividend ETFs.
In short, that article took advantage of the combined filtering criteria of 5 world-class, dividend-focused ETFs in an attempt to find between 12-15 stocks which an investor desirous of starting their own dividend-focused portfolio could seriously consider as a starting point.
Five months later, in July 2017, I followed that with my most popular article, 20 Top Stocks For A Monthly Dividend Portfolio, which built on the work of the first article but selected a few additional stocks such that one would receive at least some dividends each and every month. At the conclusion of the article, I revealed that I had put my money where my mouth was: selling all dividend-focused ETFs I held in my retirement portfolio in favor of the 20 featured stocks plus an additional 2 "bonus stocks" I suggested.
Lastly, in May of this year, I provided a 2-Year Update on the portfolio.
Segue To 2019 - Part Un
With the exception of two small moves, as revealed in the '2-Year Update' article, these 22 stocks are still the only single-stock holdings in my personal portfolio (sorry, it's my Q1 2019 update, life was a little complicated for me during Q2 and so I hope to next publish an update for Q3).
A couple of weeks back, I thought to myself: Based on all the theory I laid out in that original article, it's really about time I re-ran the exercise. After all, time does not stand still. Much has changed since then. I mentioned the two small moves I already had to do to adapt to that fact. Further, since these are the stocks I hold in my personal portfolio, I have a very direct vested interest in whatever answer the results of such an analysis would reveal!
Oh yeah, and I could also write an article about it for Seeking Alpha. Score!
Back To The Beginning
I won't reproduce the entire original article for you here. This thing would get so far beyond unwieldy that readers would either have to read the article at their favorite coffee shop, wired on cold-brew coffee, or perhaps with a couple of hits of No-Doz. Tell you what, let me quickly give you the basics of the idea and method and let any data geeks who so desire go back to the original article for the minutiae. Deal?
If an investor was desirous of building a small, but yet reasonably diversified, dividend-focused portfolio, they could do research on their own, from scratch. If done well, that is probably the best approach. However, it is not without its challenges, and I outlined a few that I could foresee.
Might there, however, be an easier way? I realized that, in the course of my research and writing as ETF Monkey, I had already reviewed the following ETFs:
- iShares Core Dividend Growth ETF (DGRO)
- iShares Core High Dividend ETF (HDV)
- Schwab US Dividend Equity ETF (SCHD)
- Vanguard Dividend Appreciation ETF (VIG)
- Vanguard High Dividend Yield ETF (VYM)
Of these, DGRO and VIG focus on dividend growth, while HDV, SCHD and VYM focus on high current dividend yields. What I learned, though, in the course of my research is that all of these ETFs have stringent selection processes. In that original article, I went on to list each, in detail. For our synopsized purposes here, let's summarize it as 'different variants of sort of the same end goal.'
One beautiful feature of ETFs is that you can typically visit their official websites and download a spreadsheet containing their complete portfolio holdings. So I started by doing that.
Next, I copied the Top 25 stocks from each file into a common spreadsheet. As you will quickly grasp, while each ETF contains somewhere between roughly 80-400 holdings, the Top 25 holdings constitute a substantial portion of that weighting. As it turned out, it mathematically aggregated to 63.84% of the 5 selected ETFs. Lastly, using the 125 rows copied into my common spreadsheet, I simply used Excel's sorting and summing formulas to come up with the stocks that had the highest cumulative weighting, and also a count of how many ETFs contained each stock.
Here are the Top 12 stocks from that original February 2017 analysis.
|Symbol||Name||Sector||Rank||Cumulative %||ETF Count|
|JNJ||Johnson & Johnson||Health Care||1||20.9366||5|
|PG||Procter & Gamble||Consumer Staples||5||14.6237||4|
|IBM||International Business Machines||Technology||10||10.8346||4|
Finally, I concluded the article by featuring the concept (after all, I am ETF Monkey) that, building from this base, one could use ETFs to round out the portfolio; from exposure to the U.S. total market to international stocks, REITs and bonds.
Segue To 2019 - Part Deux
With that grounding, it's time to get down to the 2019 exercise.
Using Google Sheets, I built an updated version of the same spreadsheet. Here's a view-only link for readers interested in taking a closer look. More precisely, it is a workbook. In the workbook, you will find separate tabs containing the data dump of the full holdings of each of the 5 ETFs, the "parent" analysis, along with a LOOKUPS tab that helped me immensely in standardizing what were quite different "dump" formats from the various providers (NOTE: In my opinion, iShares provided the best dump, Vanguard 2nd best, and Schwab sorta brought up the rear, in terms of quality).
Without any further ado, here are the results. Make sure you take advantage of the 'Click to enlarge' option to take a closer look. I had to shrink the spreadsheet to 90% normal size to include all the helpful information I wanted to share. Actually, even then I missed one detail I need to feature, but I will address that later. Anyway, have a look and then I will follow up with comments.
First, one little detail that may save you a trip back to the original article (you data geeks have already been there, haven't you?).
If you look carefully, you will see two little inverted triangles on several of the rows. In Google Sheets, those indicate hidden rows (notice that you see Row 2 and then Row 7 next). Take a closer look at Procter & Gamble. The spreadsheet shows a cumulative weighting of 20.00%, but you only see the DGRO line and 2.46%. But you will also notice that it is in all 5 ETFs.
Sure enough, here's the 'exploded view' of Procter & Gamble, with the hidden rows exposed. Add up the five '% of fund' values and you arrive at 20.00%. And so it goes with the rest of the stocks.
Back to the main picture. As you can quickly see, I expanded the view to 29 companies. Compared to 2017, the 3 companies in bold green are the new 'winners' in the exercise. These are:
|Symbol||Name||Sector||Rank||Cumulative %||ETF Count|
|MRK||Merck & Co.||Health Care||12||7.67||3|
|HD||Home Depot||Consumer Discretionary||14||7.18||2|
Conversely, the two stocks featured in bold red are the 'losers' in the exercise.
Note: I officially stopped at 12 stocks in the original article, so Altria Group (MO) was not on my list. Nevertheless, I included it in my Top 15 here, for reconciliation purposes.
"Wait a second," you're thinking. "I see three additions but only two subtractions. What gives?" You may remember that, above, I alluded to the fact that I zoomed the spreadsheet down to 90% normal size and still missed one detail.
That detail is AT&T. In the latest pass, AT&T slipped all the way to 37th place, with a 2.43% cumulative weighting, and only making it past the filters of one ETF; VYM.
Basically, then, we are talking about:
- Dropping IBM (IBM), Altria (MO), and AT&T (T)
- Adding Texas Instruments (NASDAQ:TXN), Merck & Co. (NYSE:MRK), and Home Depot (NYSE:HD).
Again, that result is purely based on the mathematical results from this analysis. I literally did nothing further to get to this point. When I started this article, I had not necessarily planned on linking these pages. However, I did so because these were the first places I looked to ask myself the question: "Would I actually replace the stocks, as proposed above, in my own portfolio?" What I found was an amazing amount of truly helpful information. Here's a hint. Without giving away the farm, I will simply note that what I saw on those 6 pages convinced me that, at the very least, I need to think very seriously about doing so. I was sufficiently impressed by the total picture those pages offered me that I found myself motivated to at least recommend them to readers.
Portfolio Visualizer Test
I concluded my 2017 article by running 3, 5, and 10-year backtests of the two variants I offered. I decided that I would once again do something similar for this article, except that I have made a couple of adjustments in the period covered.
For documentation purposes, the following 5 graphics were all linked from this Portfolio Visualizer backtest. If you go into Portfolio Visualizer to play around for yourself, you will just need to change the 'Start Year' setting to reproduce the two sets of results I will feature. In each backtest, the initial balance was $10,000, all dividends were included and reinvested, and the portfolio was rebalanced annually.
First, here is the construction of Portfolio 1. This represents the new, or updated set of 15 stocks, with all stocks equal-weighted (the 3 new holdings are at the end).
Portfolio 3 represents my chosen benchmark. It is a 100% allocation to SPDR S&P 500 ETF Trust (SPY), a proxy for the S&P 500 index.
For the first comparative backtest, I deliberately chose 2007 as my starting point. I did so because, in each case, I wanted to capture the effects of the 2007-2009 downturn on each portfolio.
Here are the results.
Certainly, it would appear that Portfolio 1 is the clear winner over this time period. This variant managed to outperform Portfolio 2 by roughly 1.4% per year, with an almost $6,000 variance in the final balance. Its standard deviation and max drawdown are higher, but not significantly so, and ultimately the risk/reward profile is reflected in slightly higher (better) Sharpe and Sortino ratios.
Now, as this article is written, with U.S. markets hovering right around all-time highs, a question that may be of interest is: How did Portfolios 1 & 2 compare through the 2007-2009 downturn? This detailed supporting graphic from the backtest provides the answer. Please allow your eyes to be drawn to the two areas I've highlighted in yellow.
First, the results through the 2007-2009 downturn. In short, as suggested by the overall results, Portfolio 1 was slightly more volatile than Portfolio 2, but not significantly so. And both held up far better than the S&P 500 index!
Second, you will also note that I highlighted the 'income' section of the graphic. Follow the data for Portfolio 1 and Portfolio 2, and you will note that Portfolio 1 generates slightly less income. Well, we did drop IBM, MO and T, three notoriously high dividend payers.
Lastly, here is the same top-level results graphic for all 3 portfolios, but this time taken all the way back to 1994, as far back as the backtest would allow per the inception date of SPY. I won't bore you with any further commentary save to note that the general themes borne out in our first comparison appear to hold true over a longer period; 13 years longer in this case.
Really, I don't have much more to add, save to perhaps briefly summarize.
Two years ago, I proposed a dividend-focused portfolio, the stocks for which were selected by doing nothing more than aggregating, in effect, the results of the filtering processes used by 5 world-class dividend ETFs. One "problem" with this approach is that things change over time. The stocks selected two years ago might not be the same as those chosen today. Further, this could put your portfolio at some risk, as you might now hold stocks that had been eliminated for very valid reasons.
And thus, this exercise. To re-perform the analysis and try to determine what we could learn.
What am I going to do with my own portfolio? Honestly, I don't know yet. As any long-term investor should do, I'm going to sit back and reflect. I'll probably do a little more research on the stocks in question. And then I will make a decision.
You may have noticed that I am now posting all updates to my personal portfolio on my personal site. So, you'll have to keep an eye out. Likely, I will also link any updates from a future Seeking Alpha article as well.
As always, until next time, I wish you happy investing!
Author's note: At the top of this article, next to my name, you will see a "Follow" button. If you like my work, I would be profoundly grateful if you would take a minute to do this, as well as feature my work to friends, colleagues and/or relatives who may be interested in the subject matter. Growing one's readership base is critical to any author and I am no exception. Your support will enable me to continue my efforts.
Disclosure: I am/we are long SEE PERSONAL PORTFOLIO LINK. 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.
Additional disclosure: I am not a registered investment advisor or broker/dealer. Readers are advised that the material contained herein should be used solely for informational purposes, and to consult with their personal tax or financial advisors as to its applicability to their circumstances. Investing involves risk, including the loss of principal.