By training and by inclination, financial planners typically construct portfolios under the illusion of certainty.  The most popular investment strategy for professional investors, strategic asset allocation (“SAA”), assumes that financial markets are efficient and that risk premiums, the amount that stocks outperform inflation, will be positive and stable over time.  The most cherished assumption for financial planners and investment advisors is that stocks will outperform inflation, cash, and bonds over time with certainty.  Therefore, in order to add to expected portfolio returns, a planner need only add to the equity allocation of a portfolio. 

However, everything we know about managing risk in SAA portfolio construction changes with the realization that risk premiums are not stable and positive over time.  In fact, the data shows that markets are surprisingly inefficient and returns can be much lower than average expected returns for periods as long as twenty years.  Specifically, the data shows that when buying the “market” as represented by the S&P 500 index, either at peak profits or at peak price/earnings (P/E) multiples, the expected 20-year returns for the S&P 500 Index can be much less than the average expected returns assumed in SAA.

Figure 1 shows the 20-year returns of the S&P 500 Index ranked by decile from lowest returns to highest returns.  The chart shows the range of each decile return and the average return for each decile.  Finally the chart shows the average P/E multiple of the S&P 500 at the beginning of each decile and the end of each decile. 

Clearly the “average” historic expected returns of 11% are not realized by investors who own the “market” for periods of 20-years when they buy the market at high P/E multiples.  The chart illustrates that high return periods occur when the multiple is low at purchase and then expands during the holding period, but the returns are well below the expected average return when multiples are high at purchase and then contract during the period.

The data also seems to suggest that historic “average” returns are not realized over twenty year time periods when buying the market at peak earnings.  Figure 2 shows the 20-year nominal and real returns of the S&P 500 index, excluding dividends, from each period of peak profits over the past sixty years.  Again, the only periods where expected average returns are realized are the 1980’s and 1990’s when P/E multiples rapidly expanded.  The remaining time periods show 20-year nominal and real returns that are significantly less than expected.

A traditional view of value investing implies that buying stocks at or beneath their intrinsic value minimizes the downside volatility of the investment since, in theory, prices should not fall too far below the asset class’s fair or intrinsic value.  Value investing also implies that buying below intrinsic value enhances the opportunity for capital gains as investors bid prices up to fair or intrinsic value.  Clearly, buying stocks at high prices versus their earnings, or when the economy operates at peak earnings, does not constitute good value, and investors can expect higher volatility and lower returns when buying stocks at such high prices.  The converse also applies:  buying stocks at low P/E multiples and below the peak earnings of an economic cycle should minimize portfolio volatility and maximize capital gain opportunities.

Depending on how one analyzes the data, today’s market environment is one of either peak profits and fair P/E multiples, or peak profits and high P/E multiples.  In either case, planners should not be certain that average historic returns will be returned by buying and holding stocks for long time periods of up to twenty years.  If this is true, then planners must consider making portfolio construction decisions under conditions of uncertainty.

¹   Source:  Crestmont Research, net total returns including gains, dividends, and transaction costs of 2%.