Leaving Investment Style Boxes Behind

Recently I was interviewed by a financial magazine on the topic of Exchange Traded Funds (ETFs). I pointed out that ETFs are an excellent investment tool to implement a sector rotation strategy, and that sector rotation was completely different from the traditional Morningstar Style-Box approach to equity selection. Morningstar is perhaps the best-known research firm specializing in independent mutual fund analysis and its star-rating system is widely used in the investment industry. In 1992 Morningstar introduced a system for categorizing mutual funds by investment style. The Morningstar Style Box divides the equity mutual fund universe by two major characteristics: market capitalization (or the market value of the companies in a fund) and the valuation of the stocks in a fund based on the P/E and P/B ratios. So a mutual fund that owns large cap stocks with high P/E and P/B ratios is likely to be found in the large growth, or upper right hand quadrant of the box.

Over the years this approach to categorizing investment managers has been quite useful for investors who want to make “apples to apples” comparisons of fund managers. You can easily go to the Morningstar database and list all the funds in the large growth universe and rank and sort them by any number of metrics. In the strategic, buy and hold paradigm of asset allocation, where presumably an investor would own a fixed percentage of large cap, mid cap, and small cap funds, as well as growth and value funds, finding a manager that outperforms the average fund in the universe is the ticket to generating above market returns. I call this search for winning mutual funds in each category the “manager search” model for the investment industry. Not surprisingly the search for the best manager in each style category dominates the conversation of portfolio analysts at every level of the profession, from those who are working at the consumer level, to those who consult with the country’s largest pension plans. For mutual fund managers, the style box becomes a sort of jail cell where they are absolutely constrained to only invest in stocks that fall within a neatly defined style box. Doing otherwise, called “style drift,” is almost guaranteed to get a fund manager fired, because consultants want a large growth manager to remain large growth. Shifting to a different style means the overall asset allocation of the portfolio is now off model and needs to be rebalanced.

All of this works fine until you change your view of how you want to categorize the market of publicly traded stocks. For example, MSCI and Standard & Poor’s has developed the GICS method for classifying stocks, (Global Industry Classification System), which groups all publicly traded stocks into 10 sectors, 24 industry groups, and 68 industries. If you look at the stock market by sectors instead of by market cap and value characteristics, you see that the 10 major stock sectors perform quite differently at different points in the economic cycle. During economic contractions investors want to own those sectors that have the most predictable and least cyclical earnings — utilities, consumer staples, and health care. As the economy shows signs of early recovery investors should favor those sectors that do best early in the cycle — financials, technology, and consumer discretionary. And as the economic expansion enters its later stages investors should favor industrials, materials, energy, and then telecom, as they prepare to shift once again to the defensive sectors that will outperform at the next downturn. ETFs (exchange traded funds) are wonderful securities to use to execute a sector rotation strategy because most ETFs are designed to own a GICS-defined index and investors can easily buy and sell sectors and industries that will outperform depending on your market view.

I thought it would be interesting to go back to the Morningstar Style Boxes and see what happens when you view them in terms of sectors and industries. The results are actually very strange. Let’s start with the premise that if you’re a defensive investor you would want a mutual fund that owns large-cap value stocks, which presumably are more defensive than either small cap funds (generally) or large cap growth stocks. I screened the 320 mutual funds that comprise the universe of large cap value funds in the Morningstar database and found that in aggregate, 66.4% of the funds were invested in cyclical stocks, including 3.5% basic materials, 7.8% consumer cyclical, 19.6% financial services, 12.1% energy, 11.4% industrials, and 10.4% technology. In other words, if you thought that owning a large value fund was a defensive investment, you’re in for a surprise. I consider all of the above “high octane” sectors of the market that are typically more volatile than the market as a whole. Only 33.7% of the average large value fund is actually invested in defensive sectors like utilities, consumer defensive, healthcare, and communication services.

Large cap growth funds seem more useful to investors who believe we’re in an economic expansion, but there are still some major issues. The first problem is that 22.4% of these funds are invested in defensive stocks which are almost guaranteed to underperform in an economic expansion. The second problem is that there’s no way to differentiate between early and late cyclical holdings. For example, if we’re early in the cycle then the average fund invests 24.1% of the fund in late cyclicals (like energy, industrials, and materials). Investors are left to hope that fund managers are actively rotating their mutual fund stock portfolio to the proper sector as the cycle progresses.

As the investment community becomes more sophisticated, the style box approach to categorizing mutual fund managers is going to be left behind. Investors will become more interested in discussing sector rotation and a manager’s ability to add excess returns by investing in those sectors and industries that outperform as the market relentlessly grinds through its cycle.

Leave a Reply

Your email address will not be published.