Swedroe: Highest Expected Returns Not Always Best

Don’t forget about the importance of diversification.

Reviewed by: Larry Swedroe
Edited by: Larry Swedroe

A regular reader of my articles contacted me recently to discuss current valuations and a value-oriented strategy. He observed: “It doesn’t matter which approach you like: a value investor doesn’t prefer U.S. stocks now.”

He also pointed out that, while the MSCI World Index currently contains 58.6 percent U. S. stocks, the iShares MSCI World Value Factor UCITS ETF, based on the MSCI World Enhanced Value Index, contains just 38.8 percent U.S. stocks. He then quoted Meb Faber: “A value approach works not just by investing in the cheapest markets, but also by avoiding the most expensive.”

You can see evidence supporting his assertion by looking at the value metrics of value funds managed by Dimensional Fund Advisors (DFA). (Full disclosure: My firm, Buckingham, recommends DFA funds in constructing client portfolios.) The table below compares the price-to-book (P/B) and price-to-earnings (P/E) ratios of the firm’s domestic value funds with those of their international value funds.

FundPrice-to-Book (P/B)Price to Earnings (P/E)
DFA U.S. Large Value III (DFUVX)1.415.0
DFA International Value III (DFVIX)0.912.6
DFA Emerging Markets Value (DFEVX)0.89.0
DFA U.S. Small Value (DFSVX)1.215.7
DFA International Small Value (DISVX)0.912.0

Source: Morningstar, as of Aug. 31, 2015

As the table demonstrates, international and emerging market value stocks have significantly lower prices relative both to book value and earnings than do domestic value stocks. That implies they also have significantly higher expected returns. One measure of expected real returns used by financial economists employs the inverse of the P/E ratio, which gives you an earnings yield.

Thus, the real expected returns for the three large value funds (DFUVX, DFVIX and DFEVX) are 6.7 percent (domestic large value), 7.9 percent (international developed-market large value) and 11.1 percent (emerging market large value), respectively. For the pair of small value funds (DFSVX and DISVX), the expected real returns are 6.4 percent (domestic small value) and 8.5 percent (international developed market small value), respectively.

Returns Not The Only Goal

Clearly, if investors want the higher expected returns, they should consider tilting their portfolios (have a higher allocation) to international developed-market value stocks and emerging markets value stocks. However, earning the highest expected returns isn’t generally an investor’s only objective, or sole consideration.

If earning the highest expected returns possible was, in fact, an investor’s sole consideration, we would likely concentrate portfolios to a greater degree than is prudent. Consider the following. The majority of passively managed value funds, such as index funds, generally categorize a stock as value or growth in one of two ways.

The first common approach is to split the market into halves. The stocks with the lowest prices relative to some metric (such as book value, earnings, cash flow, dividends or sales) are considered value stocks. The stocks with the highest prices relative to that metric become growth stocks.

The second common convention is to split the market three ways. The lowest 30 percent of stocks based on a given metric are classified as value, stocks in the middle are classified as core, and the highest 30 percent are classified as growth. Doing so provides investors with a great deal of diversification, sufficient to minimize the idiosyncratic risk of individual stocks.

Increasing Concentration

Given these two “schemes,” the methodology using the 30 percent of stocks with the lowest prices will have higher expected returns than the one using the 50 percent rule. The reason is that the value premium is basically monotonic, increasing as you go down deciles. Stocks with the highest prices have the lowest returns, and stocks with the lowest prices having the highest returns. Thus, if an investor sought higher expected returns than those available from the bottom 30 percent of stocks, they could limit their holdings to, say, the bottom 20 percent.

Taking it a step further, you could own just the bottom 10 percent. And taking it even further, you could own the bottom 1 percent, or perhaps even the single stock with the lowest price. Of course, most investors would not want to take such a concentrated risk.

In other words, there’s a trade-off between expected returns and diversification. When it comes to large value stocks, a mutual fund probably should hold, at a minimum, 200 stocks to be effectively diversified.

With smaller stocks, which tend to have more idiosyncratic risk, a mutual fund would probably want to hold several times that figure. As you increase the number of stocks, you reduce idiosyncratic risk, but you also lower expected returns. The key is to find the right balance between the two.

Idiosyncratic Risks

The same principles apply to owning U.S. stocks or international developed and emerging market stocks. Yes, you could increase your expected return today by lowering your allocation to U.S. value stocks and more heavily weighting international developed, and especially emerging market, value stocks. However, doing this would also decrease your level of diversification, increasing idiosyncratic risks. Again, we have the same trade-off.

My recommendation is to create an investment policy statement that clearly defines the amount of exposure you believe provides a prudent amount of diversification among U.S., international developed and the emerging markets. And then stay the course, rebalancing along the way.

When one asset class outperforms (and its value metrics likely increase, lowering its relative expected returns), you will sell some of it to buy the asset class that underperformed (and now likely has higher expected returns). In that way, you’ll stay sufficiently diversified and still have a value strategy.

There’s one more important point to emphasize. While low prices relative to a value metric do predict higher returns, those higher returns are expected, not guaranteed. Lower prices are also an indicator, at least from the conventional finance point of view, that more risk present is present. In other words, higher expected returns aren’t a free lunch.

Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.

Larry Swedroe is a principal and the director of research for Buckingham Strategic Wealth, an independent member of the BAM Alliance. Previously, he was vice chairman of Prudential Home Mortgage.