Building A Diversified, Low-Volatility ETF Portfolio
Summary
- Excel Solver is set up to build a hypothetical million-dollar portfolio to maximize volatility-adjusted returns constrained by correlation to other funds in the portfolio.
- Correlation of each ETF to the Risk Indicator from the Investment Model is added to the ETF Ranking System and to limit investment limits in ETFs.
- The Optimized Portfolio is not a destination. It is a journey.
Introduction
I dusted "Investing With The Trend" by Gregory L. Morris off the shelf as I set up a system to build a diversified, low-volatility ETF Portfolio. Discussing investing by the stage of the business cycle, he says, "... you can use a momentum analysis and always be in the top four sectors and probably do well. Clearly, this is certainly better than buy and hold or index investing." As I skim through the book again, the quote from Jud Doherty, Stadion Money Management, "I am a firm believer in active management to add safety, not return, and that a disciplined process is critical to success." is particularly important.
Christine Benz from Morningstar says in "Assess Your Asset Allocation for Retirement" that “the mix of stocks and bonds that delivers the highest possible return with the least amount of risk over a given time frame - will be apparent only in hindsight.” She goes on to say that portfolios need to be based on investors own situation and should be reviewed periodically. Ms. Benz will be a guest speaker at the American Association of Individual Investors in Las Vegas this October.
There is some merit to using standard deviation to adjust returns for volatility as the S&P 500 (VOO) currently has a standard deviation of 10 while short-term bonds (MINT) has a standard deviation of less than 1; however, standard deviation changes with volatility. I use returns adjusted by standard deviation to influence the ETF Rank for volatility.
I use momentum, fundamentals and five analysts' ratings (such as Morningstar and Market Edge) to narrow about a thousand ETFs down to approximately 100 funds, which include most of Morningstar’s classifications. I have built the Ranking System and Portfolio Optimizer to adjust the portfolio for the Business Cycle including sector rotation.
I use Excel Solver to create a hypothetical million dollar ETF portfolio maximizing the weighted 6 month, volatility adjusted return constrained by diversification (correlation to other funds in the portfolio) and fund investment limit. The most recent month gets 29% of the weight and the sixth month gets a weight of 5%.
The funds are classified as core, support, rotation, and theme and given a maximum allowable investment limit. To account for risk, I use Morningstar's Bear Market Ranking System combined with the fund's correlation to my Risk Indicator from the Investment Model, analyst's ratings, and standard deviation to establish the fund investment limit.
Correlation To Risk
Will McGough, VP Portfolio Management, is quoted in "Investing With The Trend" as saying, "The Ulcer Index attempts to quantify risk around measuring drawdowns. Drawdowns are the true source of risk that investors face, not some normal distribution variance around a mean return." My Risk Indicator is my Ulcer Index. The worse that a fund has performed in past corrections, the lower the amount that I am willing to invest at this time in preparation for the next downturn.
The Risk Indicator consists of the Chicago Fed, St. Louis Fed, and Kansas City Fed Financial Stress Indicators, CBOE S&P 100 Volatility Index, and Economic Policy Uncertainty Index available on the St. Louis Federal Reserve FRED website.
Source: Created by the author with data from the St. Louis Federal Reserve FRED database.
Below are some of the fund categories that I rank (with 1 being the best), grouped by the Morningstar categories. The Rank is composited from over 20 metrics including analyst’s ratings, return on equity, return over different time periods, volatility (beta and standard deviation), risk/reward (Alpha and Sharpe Ratio), net flows, valuation, historical and projected earnings growth, technical indicators, among others.
I have added the correlation to Risk Indicator (0 being the least correlated) to the Ranking System and included it in the table, along with beta and Morningstar’s Bear Market Ranking (1 is better performance and 100 is poor performance). Risk correlation is developed using Excel’s CORREL function.
Source: Created by the author with data from Morningstar. The Rank is calculated by the author with data from Morningstar, Fidelity and Schwab. The risk correlation is calculated from returns from Morningstar and the author's Investment Model Risk Indicator.
Diversification
Correlation measures how closely two indexes move in the same direction. During market downturns, correlation tends to be high and investing in different stock styles provides little protection. In the table below, I have calculated the correlation of large blend funds that I track over the past year. The passively managed funds tend to be in the upper left corner and are more correlated to each other. The Enhanced Strategy (momentum, low volatility, factor) Funds tend to be in the lower right corner and are less correlated.
Source: Created by the author with data from Morningstar.
Optimizing With Correlation And Beta Constraints
I selected nearly a hundred ETFs from different Morningstar categories that had lower correlations to other funds. I used Excel Solver to build a hypothetical million-dollar portfolio that maximizes risk-adjusted monthly returns constrained by maximum allowable investment and correlation to other funds. Below are the correlations of the ETFs selected to be in the hypothetical portfolio.
Source: Created by the author with data from Morningstar.
“Deploying Multi-Factor Index Allocations in Institutional Portfolios” by MSCI Index Research, Inc. describes the benefits of using multifactor funds for reducing turnover and risk. I select both passive and enhanced strategy funds to be optimized, but believe that Smart Beta funds offer advantages.
Portfolio
A good place to start to build a low-volatility portfolio is with the iShares Edge Minimum Volatility ETFs: United States (USMV), International Developed (EFAV), and Global Markets (ACWV).
Employer-sponsored plans often have limited fund options. Vanguard has the US Multifactor ETF (VFMF), US Liguidity Factor ETF (VFLQ), US Momentum Factor ETF (VFMO), Minimum Volatility ETF (VFMV), US Quality Factor ETF (VFQY), and Global Minimum Volatility Mutual Fund (VMVFX), among others, that may be too new to include in the optimization, but may be valid substitutes. These "smart" funds may have the impact of increasing diversification and reducing volatility.
Below is how the selected funds have performed over the past 12 months. Two of the more speculative funds are Vanguard Energy ETF (VDE) and First Trust Global Wind and Energy (FAN).
Source: Created by the author with data from Morningstar.
The hypothetical portfolio is shown below. The asset-weighted correlation of the funds is 0.34 with a beta of 0.63. The percentage of stocks in the portfolio is 50% with 8% of the portfolio invested in foreign stocks. About 43% is in cash and bonds. The funds at the top of the list are more favored by the analysts and the ones with lower favorable consensus are at the bottom of the table.
Source: Created by the author with data from Morningstar.
The distribution of the funds is shown in the investment pyramid below. I classify the funds by standard deviation, but beta results in a similar distribution.
Source: Created by the author with data from Morningstar.
I set the hypothetical portfolio up in Morningstar. Below is a summary of the results. I used the X-Ray Interpreter available in the Premium Service in Morningstar to evaluate the portfolio. The assessment of the portfolio is for an investor with a three to ten-year timeline with moderate concern for volatility is what I was trying to achieve. It is a somewhat defensive, highly diversified, low-volatility portfolio with upside potential.
Source: Morningstar.
The Macro Environment
For a better understanding of exchange rates and how they are impacting international investments, I turn to Larry Summers in “The dollar is getting weaker. That should worry us.” He states, “U.S. 10-year interest rates are about 220 basis points above German rates and about 280 basis points above Japanese rates.
This implies that markets expect depreciation of the dollar by more than 25 percent against its major competitors over the next decade.” He adds, “The pattern of higher interest rates and a weakening currency suggests that on multiple dimensions, U.S. assets have to be put on sale at bargain prices to persuade foreigners to hold them or to induce Americans not to diversify into overseas assets.”
As described in, “Why the U.S. Treasury Likes a Weak Dollar,” the Federal Reserve is issuing short-term debt and shrinking its balance sheet. This appears to have the effect of weakening the US dollar. The weaker dollar has historically lifted the price of gold and helped exporters. .
Indicators
Below are many of the main indicators that I use in the Investment Model. It is consistent with slow growth and moderate risk and inflation.
Source: Created by the author with data from the St. Louis Federal Reserve FRED database.
Investment Model
The Investment Model is constructed to measure conditions for the stock market over the next six months. It shows that the time to be aggressive (shaded area) has faded. The Index is still high, but starting to decline. The economy is continuing to grow at a slow pace and corporations are getting stronger. Tax reform and fiscal stimulus will assist growth. A tight labor market and weaker dollar may increase inflationary pressures. The result is that the Federal Reserve will likely continue to raise interest rates.
Source: Created by the author with data from the St. Louis Federal Reserve FRED database.
Conclusion
I set my personal portfolio up in a similar manner with the funds that I own and the ones that I may be interested in buying. I use Excel Solver to identify the most effective changes to the portfolio to maximize volatility adjusted return constrained by beta and correlation. It identifies where my funds are too concentrated. I prefer to make small changes monthly, because historical returns have no guarantees of trending forward. I am not as defensive as the hypothetical portfolio, but see some benefit to trending in that direction.
This article was written by
Analyst’s Disclosure: I am/we are long PSL BNDX WOOD VDE COMT VMVFX. 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.
I am not a professional investor nor an economist. The information in this article should not be taken as advice. Investors should do their own research or consult with an investing professional.
Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.