Previous Posts

Jan 2021 | Q1 Takeaways


Dec 2021 | OCIO Selection


Oct 2021 | Q3 Takeaways

Oct 2021 | Diversification

Jul 2021 | Q2 Takeaways


Jun 2021 | Viewpoint

More ...

Viewpoint: Analyzing the Drivers of Performance

Some investors assume that recent returns will persist when forecasting capital markets. That is, they are tempted to extrapolate past performance into the future. For example, after the stock market returned more than 10% for several years in succession, some investors may be inclined to believe that stock prices will continue to rise at that rate. Or, after growth stocks persistently outperformed value stocks for several years, some investors may assume that pattern will continue. We believe that temptation should be resisted.

The analyses in this Viewpoint disaggregate returns into component drivers. Armed with this understanding of the drivers of past performance, investors will be better able to evaluate if recent performance is likely to repeat.

Total Return Attribution: Equities

Stock market returns can be disaggregated into three components: a) dividends received, b) the growth in expected earnings, and c) the change in the price-to-earnings ratio (P/E).  The combination of the latter two variables lets us understand the drivers of the change in price. 

For the five years from 2017 through 2021, the S&P 500 Index returned 18.5% percent per annum. The last column in Table 1 shows the three components that produced that result: dividends (2.0%), earnings growth (11.1%) and P/E multiple expansion (5.3%). The middle column shows that breakdown for the S&P 500’s three-year annualized return of 26.1%.

Because the index’s return exceeded what we and most other market professionals had forecast,  a natural question is, why was there such a difference? In addition, some may ask why our current projections are not in line with the index’s trailing return. Our answer is that the high earnings growth and the significant expansion in P/E multiple that led to the high return were surprises. Because we do not expect them to repeat, the index’s recent performance should not be projected into the future. 

Further elaborating the rationale for our projected return:

  • For the 18.5% annual return to repeat, annual earnings growth plus P/E expansion would need to sum to approximately 16.5%. We believe that is unlikely.  

  • P/E ratios are quite high by historical standards. If they were to continue to expand by 4.5% annually for another 5 years, the cumulative expansion would be 25%, and the S&P 500’s P/E would rise from 22 times to 27.5 times earnings – possible, but we believe it is unlikely.

  • The S&P 500’s earnings have dramatically outpaced economic growth in recent years. We think this unlikely to continue. 

  • In short, with the dividend yield around 2%, if earnings grow in line with nominal GDP and the P/E multiple does not change, investors should expect an S&P 500 Index return of 5% to 6% per annum. 

Total Return Attribution: Fixed Income  

Applying a similar approach to fixed income returns, we can attribute returns to: a) coupons received and b) price movement from interest rate changes. Table 2 examines the return of the Bloomberg Aggregate for three- and five-year periods. While the index returned 4.79% per annum for the three-year period, 2.88% of it was due to falling interest rates. 


Unless one expects interest rates to fall further, the index’s future return is likely to be in the range of its current yield of 1.75%. It will be even lower if interest rates rise to more normal levels.


Relative Return Attribution: Equities

A similar analysis can be applied to relative index returns such as growth versus value stocks, large- versus small-cap stocks, and US versus international stocks. Such analyses examine differences in dividend yields, and the relative changes in the indices’ earnings growth rates and P/E multiples. 

Growth vs. Value

For the three years ending 12/31/21, the Russell 1000 Growth Index outperformed the Russell 1000 Value Index by 16.4% per annum (Table 3). The bulk of the outperformance came from relative P/E expansion: 22.5%  per annum for the growth index (from 18 to 32) versus 10.1% per annum for the value index (from 13 to 17). 


While it is appropriate to expect growth stocks to have higher P/Es than value stocks, the current average P/E of growth stocks is almost two times that of value (well above the 1.33 average relative P/E since 2010). We believe that the gap is unlikely to widen further.

5.2% of growth stocks’ outperformance came from faster earnings growth. Unless aggregate corporate earnings continue to outpace GDP at the same rate they have over the last several years, the spread in earnings growth rates should shrink. We believe that the rapid earnings growth of recent years is unlikely to repeat.

Large cap vs. small cap

Over the past three years, the large-cap S&P 500 Index outperformed the small-cap Russell 2000 Index by 6.2% per year. Large cap’s P/E expanded 14% per annum (from 15 to 22), outpacing the 10% annual expansion for small caps (from 21 to 27). The current relative P/E ratio of large- versus small-caps is 0.84. Since 2010, it averaged 0.69. We do not see fundamental reasons for further convergence. Looking at relative earnings growth, both indices had similar growth rates, yet in principle smaller stocks’ earnings growth should be faster over the longer-term. 


US vs. International Stocks

Over the last three years US stocks outperformed international by 12.6% per annum. Table 5 shows that it was due to US stocks having both relative P/E multiple expansion and faster earnings growth.


The Russell 3000’s P/E multiple climbed 14% per annum (from 15 to 22) while the average P/E of international stocks grew 5% per annum (12 to 14). US stocks are now trading at 1.87 times the P/E multiple of international stocks, up from 1.40 times at the start of the period. The average relative P/E since 2010 was 1.27 times. We do not believe there is a fundamental reason for further expansion in US stocks’ relative P/E.

US stocks’ earnings grew by 8.7% per annum versus only 5.1% for international stocks. Most of the difference came from top-line growth: 5% per annum revenue growth in the US, while the aggregate revenue of international stocks declined 1% during the same period.  Looking forward, we believe that as the pandemic fades and the global economy heals, US and international stocks should have similar revenue growth.

In terms of return expectations, US stocks start a bit behind international due to their lower dividend yield. To forecast that US stocks will outperform international stocks, investors would need to expect either that US stocks will continue to expand their relative P/E, or that the earnings of US listed stocks will grow significantly faster than international’s. Either may happen, but we believe such expectations represent more of a tactical than a consensus view of the longer-term fundamentals.



While it is tempting to extrapolate past performance when developing capital market expectations, we believe that short- to medium-term performance is often largely determined by unexpected events; i.e., surprises. Unless one expects the surprises to repeat, it is unwise to base return estimates on past performance. Instead, investors can gain a better perspective on longer-term return expectations when they base their forecasts on reasonable expectations of the components such as dividends, earnings growth, interest rates and changes in valuation.


This article is provided for the general information of clients of Alan Biller and Associates and others whom we believe will find it of interest. Alan Biller and Associates is an investment adviser registered with U.S. Securities and Exchange Commission. Registration does not imply a certain level of skill or training. This article does not consider the investment needs, objectives, or circumstances of any person, and does not constitute investment advice to or for any person, or on which any person may rely. Certain articles are based on information provided by third-party sources. While we believe third-party sources are reliable at the time an article is prepared, because our use is limited to articles and similar communications, we do not independently verify the accuracy of the information provided by the third-party sources, or monitor any subsequent changes in such information following preparation of articles. Readers are alerted we cannot and do not guarantee the accuracy of information in the articles we publish, which are provided on an “as is” basis without any warranty whatsoever, and are subject to change without notice. Past performance is no indication or guarantee of future investment results.