Keeping It In Perspective

by Craig L. Israelsen
December 2005
Reprinted by permission from Financial Planning Magazine

The bull market of the 90s is as fresh in everyone's memory as the drama of its collapse. That memory might distort the real value of growth as a style category. It seemed to be so overwhelmingly in the ascendancy during much of that decade that it virtually wiped out the historical notion of a value premium. But if you look back over time, the big bull looks a little more like a calf.

We begin our investigation of growth versus value by looking at a unusual risk-return frontier (see Classic Curves). It is unusual because it incorporates more than simply the S&P 500 Index as the equity component. Additional equity surrogates include large value, large growth, mid value, mid growth, small value, and small growth. The S&P 500 Index is a large cap index that combines both value and growth styles. The data in "Classic Curves" demonstrate the varying slopes of a two-asset, risk-return frontier when different equity indexes are utilized in combination with cash (i.e. T-Bills).

The yellow frontier line represents 11 different combinations of cash & the S&P 500 Index (from 100% cash/0% stock, 90% cash/10% stock,…to 0% cash/100% stock). The leftmost yellow dot is a 100% cash portfolio and the rightmost yellow dot represents a 100% S&P 500 Index portfolio. Over the 25 year period from 1980 and 2004 the annualized return of 3 month Treasury Bills was 6.38%, with an annualized standard deviation of return of 3.53% (leftmost yellow dot). The 25 year annualized return of the Standard & Poor's 500 Index was 13.50% with a standard deviation of 16.31% (rightmost yellow dot). A 50% cash/50% S&P 500 Index portfolio (annually rebalanced) had a 25 year annualized return of 10.24% with a standard deviation of return of 8.55% (as represented by the middle dot on the yellow line). Thus, compared to a 100% S&P 500 portfolio, a 50/50 portfolio had 24% lower return and 48% less volatility of return.

The blue lines represent portfolios utilizing three different U.S. value equity indexes (large value, mid value, and small value) in various combinations with cash. For example, the 25 year annualized returns and 25 year standard deviation of annual returns for "large value" represents the equal-weighted performance of the Dow Jones Large Value Index and the Wilshire Large Value Index. The same methodology applies to the Mid Value and Small Value data. The cash component is represented by the 3 month T-Bill.

The red lines represent portfolios utilizing three different U.S. growth equity indexes (Large Growth, Mid Growth, and Small Growth) in various combinations with cash. As with the value indexes, the returns and risk of each growth index represents the average of the respective Dow Jones Growth Indexes and Wilshire Growth Indexes. The annual returns of the Dow Jones and Wilshire style-specific indexes were averaged so as to obtain aggregated performance figures, rather than rely on one specific index family.

In "Classic Curves", you can see a pronounced "value premium" over the 25 year period from 1980-2004, particularly in mid cap value and small cap value portfolios. This is noted by the fact that the LV, MV, and SV frontiers are to the left and above of the S&P 500 frontier. The value indexes generated higher returns with equal or less volatility of return than the U.S. equity market in general (as measured by the S&P 500 Index). Moreover, value indexes (in each market cap category) significantly outperformed their growth index counterparts.

The data table ("Historically Speaking") shows that the 25-year annualized return of U.S. small value (as an asset by itself without any cash component) was 16.33%, or nearly 300 basis points higher than the S&P 500 Index and nearly 600 basis points higher than U.S. small growth. In dollar terms, the performance of small value over the full 25 year period (as a specific style category) turned $10,000 into $438,827 before taxes and inflation. A $10,000 investment into the S&P 500 Index account became $237,174, and the small growth account had a future value of $121,081. Value-based indexes produced staggering differences in ending account value, and with lower return volatility.

So, how can anyone argue against value as an investing style? The primary argument is obvious: growth outperforms value periodically. How often and by how much?

One approach to answer that question is to examine returns over rolling time periods, rather than from one point in time to a distant point in time. During the 25 year period from 1980-2004 there were 21 five-year rolling periods. The first 5-year rolling period was from 1980-1984, the second from 1981-1985, and so on. The real issue here is to determine how persistent a value premium is through time.

The average five-year rolling return of the S&P 500 Index (by itself) was 14.24%, for T-Bills alone 6.26% (see "Rolling Fives"). The average five-year rolling return for large growth was 12.71% and 14.37% for large value. The value premium widened among mid caps, and widened even more among small caps.

Rolling Fives

As shown in "Time Passages", during the first five-year period (1980-1984), large cap value had a 432 basis point higher five-year annualized return than large cap growth, or a 432 bps premium. Mid value had a 422 bps premium, and small value a 1,110 bps premium.

Over this 25-year period, there was a value premium among large caps 62% of the time with an average premium was 471 basis points. Large growth outperformed large value in eight of the 21 five-year periods (or 38% of the time). The average "growth premium" among large caps was 329 basis points, which is a considerably smaller premium than the average value premium of 471 basis points.

U.S. mid cap equity indexes experienced a value premium 71% of the time, and the average value premium (512 bps) was 214 bps higher than the average growth premium (298 bps). Among small cap U.S. equity indexes, there was a value premium in 19 out of the 21 five-year rolling periods, or 90% of the time. The average value premium was 659 basis points.

The five-year rolling periods ending in 1991 and 1999 were the only two periods in which there was a growth premium across all three market cap categories. The five-year growth premium in 1999 was in excess of 1,000 basis points for large, mid and small cap U.S. equity indexes. Though the growth premium during that particular time period was dramatic, it was relatively short-lived.

In 2001, James Davis noted that "Perhaps the biggest disappointment in the past three decades is the inability (or unwillingness) of [mutual] funds to capture the value premium that has been observed in common stock returns during the period" (Mutual Fund Performance and Manager Style, Financial Analysts Journal, Jan/Feb 2001). One possible explanation is that many fund managers remain committed to growth stocks because they fear missing out on the growth premium-when it occurs-because their "underperformance" will be so noticeable. Interestingly, over the past 25 years there has been a growth premium among large cap U.S. equity indexes only 38% of the time, among mid cap indexes only 29% of the time, and among small cap indexes only 10% of the time (as measured by five-year rolling returns). The notion of seasons may well be the guiding principle here. Among large and mid cap U.S. equity indexes, value outperformed (using five-year rolling performance) from 1980-1989. From 1990-1994, growth had the advantage. Value outperformed from 1995-1997 and then growth took over again from 1998-2000. From 2001 to 2004, U.S. value equity indexes have performed considerably better than U.S. growth equity indexes. It will surprise no one when growth enjoys its season again.

In summary, over the past 25 years a performance "premium"-both in frequency and in magnitude-has been observed among value-based U.S. equity indexes. Among small cap U.S. equity indexes, the value premium is particularly evident. Nevertheless, when growth premiums occur, they can be significant in their magnitude. There is a season for value and a season for growth. Making prudent use of the seasonal shifts is the real key.

Classic Curves

Historically Speaking

¹  Average of Dow Jones Large Growth Index and Wilshire Large Growth Index

²  Average of Dow Jones Large Value Index and Wilshire Large Value Index

³  Average of Dow Jones Mid Growth Index and Wilshire Mid Growth Index

⁴  Average of Dow Jones Mid Value Index and Wilshire Mid Value Index

⁵  Average of Dow Jones Small Growth Index and Wilshire Small Growth Index

⁶  Average of Dow Jones Small Value Index and Wilshire Small Value Index

Time Passages

¹  Average of Dow Jones Large Growth Index and Wilshire Large Growth Index

²  Average of Dow Jones Large Value Index and Wilshire Large Value Index

³  Average of Dow Jones Mid Growth Index and Wilshire Mid Growth Index

⁴  Average of Dow Jones Mid Value Index and Wilshire Mid Value Index

⁵  Average of Dow Jones Small Growth Index and Wilshire Small Growth Index

⁶  Average of Dow Jones Small Value Index and Wilshire Small Value Index