When I first started creating ETF-based portfolios, most of the index methodology at that time was market-cap weighted. I decided to build four risk-based model portfolios: conservative, balanced, growth and dynamic growth.
I then selected ETFs to fill the core and satellite positions for the different asset classes and geographic breakdowns of each portfolio, in keeping with each portfolio’s investment policy.
Initially, core positions were broad-based, market-cap, low-cost indices representing the S&P/TSX (XIC), S&P 500 (VUN) and EAFE (ZDM) for stocks, and the ScotiaMcLeod Universal bond index (XBB). Satellite holdings were more sector-focused, especially if I had high conviction that a certain sector, such as technology (XLK), would outperform the S&P 500.
Over the years the ETF space has seen exceptional growth in factor-based products. I was an early adopter of “factor-based” or “rules-based” ETFs, which track indices that are constructed to choose securities based on attributes associated with higher returns. I was impressed with the Nobel Prize-winning research of Eugene Fama and Kenneth French. After analyzing the white paper on First Trust’s AlphaDEX index methodology, which is based on Fama and French’s work, I started using some of their factor-based ETFs in my portfolios as satellite positions for U.S. mid-cap (FNX) and European (EUR) exposure.
The so-called “smart-beta” index strategies usually have higher fees than traditional market-cap indices, so my basic criteria for adding them to a portfolio was that they had to generate better returns and/or lower portfolio volatility compared to the market-cap, low-cost, core portfolio positions mentioned above.
In the early years, factor-based investing using ETFs usually focused on index construction using filters for one or two of the five factors that are persistent and repeatable over time: value, momentum, quality, size and minimum volatility. Without delving into the pros and cons of each index strategy, suffice it to say that there were compelling reasons to consider using different indexing strategies, along with traditional market cap, to diversify portfolio holdings and enhance returns.
Factors will perform differently throughout an economic cycle; for example, the size factor may perform best during the recovery phase. This led to the question: Is there a way to time the use of factors in investing to maximize a portfolio’s alpha?
I personally did not go that route, since I had embraced ETFs based on my conviction that no one can successfully time the market. As Vanguard’s John Bogle said to me: “In my 60 years of investing, I have never met anyone who even knows anyone who can successfully time the market.”
What led me to integrating factor-based ETFs into my models was scientific-based research showing that over time, the index methodology would provide superior returns and, in some cases, reduce portfolio volatility.
My performance results using factor ETFs have been positive overall. However, I have been using factors long enough to see that single factors have periods of outperformance and underperformance. I was happy with the outcome of low volatility factor-based ETFs like ZLB (Canadian equity) and EEMV (emerging market) during periods of high market volatility. Although I am a believer in the Fama and French methodology, it has a value tilt that has not been in favour over the last few years. Growth has been the place to be and my decision to overweight technology stocks since the inception of my portfolios has paid off. I did not do that with a factor-based growth ETF; instead I used a sector ETF (XLK) to overweight technology.
As the industry continues to evolve, so do ETF offerings. For managers concerned about the issue of factors and market timing, multifactor ETFs have now become available. These products can be viewed as solutions, not just for satellite portfolio holdings, but for core positions as well.
I’ll take a look at that in my next article.