Suggested 4-ETF Portfolio Either As Standalone Or In Combination With A Dividend Growth Portfolio

Includes: VNQ, VOE, VPU, VWO, VYM
by: Kent Candee


Outlines a 4-ETF portfolio based on returns, standard deviations, and coefficient of variation.

Coefficient of variation is a statistic used to evaluate volatility versus return for a given investment.

Using VYM, VNQ, VOE, and VPU ETFs create a reasonable yield, decent return, potential dividend growth, and low coefficient of variation portfolio.


I have been encouraged by Dividend Sleuth in his "Blending Individual Equities And ETFs" September 12, 2016, article to share about the 4-ETFs I would use for either a standalone portfolio or in combination with a Dividend Growth portfolio. This all stemmed from comments I submitted to his article "Is 3 Enough? How About 6? Re-Thinking A 10-ETF Portfolio" and an Instablog I wrote in response to my comment.

I am currently reading the book titled "Investing at Level3" by James Cloonan and I share many of his thoughts and comments on investing. In full disclosure, I am a lifetime member of the American Association of Individual Investors, which Mr. Cloonan founded back in 1978. This association has been instrumental in my investment knowledge, but I have not taken full advantage of the wealth of information available. I share this not as an endorsement, but only to provide some background to some of the comments that I made.

Return, Standard Deviation, and Coefficient of Variation

These three investment characteristics are ones that I pay attention to and find value in primarily when researching individual dividend growth stocks and I transferred them to looking at ETFs. My Instablog was submitted because I promised I would research out Dividend Sleuth's proposed 6-ETFs to Mr. Cloonan's Level3 3-ETF suggestion relative to these three characteristics.

Ideally I believe that when looking at ETFs, it is best to have at least 5-years worth of returns and preferably 10-years. My reasoning for this is that economic cycles tend to happen in 10-year increments although this is not always the case. This rationale is also a carryover from investing in dividend growth stocks.

Return and standard deviation are investing terms I believe readers fully understand. This may not be the case for coefficient of variation and I offer the following definition for reader clarity:

Investopedia: A coefficient of variation is a statistical measure of the dispersion of data points in a data series around the mean. It is calculated as follows: (standard deviation) / (expected value). The coefficient of variation represents the ratio of the standard deviation to the mean, and it is a useful statistic for comparing the degree of variation from one data series to another, even if the means are drastically different from one another.

I like to use CV or CoV to compare various investments. Volatility can have a huge impact, both positive and negative, on your investment returns. I strive for the lowest amount of volatility for the greatest amount of return and the CoV assists me in completing this analysis. The lower the CoV ratio (return standard deviation divided by the return mean for a given period of time) means the greater the return with lower volatility. My yardstick when using the CoV has been a ratio of 2 or less when I look at a dividend growth stock. This translates to having the standard deviation be only twice the mean return. When the standard deviation is more than twice the mean return it indicates to me there is volatility in the price of the stock greater than I like to see. When I look at the CoV for the ETFs referenced in my Instablog they are greater than 2 for the 10-year data with the exception of one ETF while the 5-year data only had two ETFs over 2. This suggests that when looking at ETFs the price volatility for 10-year data is likely to be greater than 2 and I might need to rethink the acceptable CoV ratio.

The CoV is not without some downside. This downside stems from the use of standard deviation and the mean. When these two statistics are used there is the assumption that the returns will follow what is known as a normal distribution. In reality, return data for a given investment may have a greater propensity to fit a lognormal distribution or neither type of distribution. Geometric mean and geometric standard deviation are two statistical measures used when the data fits a lognormal distribution. However, these two statistics are not readily available when researching an investment from my experience. If the return raw data is available there are statistical tests that can be used to validate the distribution fit. In some respects it does not matter whether or not the raw data fits a particular distribution as long as you are applying the CoV to each investment.

The use of the CoV is simply a tool (one of many) that can be applied to evaluating volatility and return. I do want to disclose that I am not a statistician and only have a very basic knowledge of statistics.

My 4-ETF Portfolio

My proposed a 4-ETF model:

  • Vanguard High Dividend Yield ETF (NYSEARCA:VYM)
  • Vanguard REIT Index ETF (NYSEARCA:VNQ)
  • Vanguard Mid-Cap Value ETF (NYSEARCA:VOE)
  • Vanguard Utilities ETF (NYSEARCA:VPU).

My suggested ratios are: VYM (50%), VNQ (20%), VOE (20%) and VPU (10%). I limited my selection of ETFs to only those referenced in my Instablog and Dividend Sleuth's 6-ETF suggested portfolio and the ETF had to have at least 5-years of data.

I eliminated any international exposure, including emerging markets. This is strictly a personal choice based on my own investment philosophy and not an endorsement. In my review of world markets I believe we are dealing with one world economy and I expect the U.S. economy to be one of the top economies in this world economy during my lifetime. Many of the U.S. companies have sales in multiple countries which does provide some foreign investment exposure. My review of foreign investments revealed that their volatility is generally greater than the U.S. economy.

This hypothetical portfolio is proposed based on the analysis that I completed in my Instablog and is based only on 5-year returns because VYM lacks 10-year data. The premise for this ETF portfolio is to create an ETF portfolio that could either standalone in lieu of a dividend growth portfolio or in combination. If I understood Dividend Sleuth's objective, the portfolio is to generate income during the retirement years. I kept this objective in mind when considering volatility, returns, standard deviation, and CoV.

Considering these four characteristics my proposed ETF offers the greatest return (14.4%), highest dividend yield (2.95%), and lowest CoV (0.86) when compared to the Level3 (12.8%, 2.14%, 0.98 respectively) and Dividend Sleuth's (8.94%, 1.79%, 4.36 respectively). The CoV for Dividend Sleuth's proposed 6-ETF needs some clarification. The 4.36 CoV is a result of the high (47.13 CoV for the Vanguard FTSE Emerging Markets ETF (NYSEARCA:VWO)). If you substitute the 10-year CoV, 6.21, into the calculation the portfolio CoV would be reduced to 1.91, a more respectable CoV and in-line with the other two CoVs.

How did I determine the ETFs in the portfolio and the percentages? The selected ETFs are based on their characteristics and the ability to generate income with potential dividend growth, low volatility and above a 5-6% expected return. (I use a 5-6% return based on the S&P 500 dividend yield and the inflation rate.) VYM delivers the lowest CoV, provides an opportunity for dividend growth, and delivers a greater than S&P 500 yield (SPY: 2.07% on 9/26/16). This ETF has some resemblance to a dividend growth strategy. The use of VNQ and VOE are holdovers from the Level3 portfolio. The introduction of VPU provides diversification and increases the portfolio yield. In regards to percentages, I simply adopted the Level3 percentages adjusting for the inclusion of VPU.


I believe the review of returns, standard deviation, and coefficient of variation values make a compelling case for my 4-ETF proposal (VYM-50%, VNQ-20%, VOE-20%, VPU-10%).

I want to thank Dividend Sleuth for even suggesting a possible move to ETFs from a dividend growth portfolio as I would not have considered this alternative. The comments and discussion among SA readers provided incentive to take a closer look at the potential for creating an ETF portfolio that could be easily managed. One reason for my interest is the endeavor is because I have a spouse that has no interest in managing our investments and I thought it was a worthwhile endeavor given the Level3 portfolio and Dividend Sleuth's proposal.

Feel free to comment and I will do my best to respond, but responses will likely be in the late evenings.

I do not own any of the ETFs listed in this article nor do I have any immediate plans to purchase shares in any of the ETFs. If you have an interest in stocks that I have currently purchased, I have created Instablogs.

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

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.