| |
 November 2004 In the Eye of the Beholder
Of all the various equity styles, growth and value are perhaps the most hotly contested. Proponents of each are as steadfast as Yankee or Red Sox fans, and equally as unwavering in their disdain for the other. Growth investors look for stocks of firms that are growing earnings at or above the market average (earnings momentum) or at a steady rate over time. They look for quick gains based on the expectation that earnings growth will translate into share price appreciation. Value investors are bargain hunters. They seek stocks that are mispriced to the downside and have the potential to appreciate when the market finally brings them back to fair value. Since they're already undervalued, they have less to fall should stocks correct, and are often seen as "safer" than their growth counterparts. We've pointed out before that styles -- particularly value and growth -- are not as different as black and white, but rather fall along an investing spectrum. Deep value and price momentum anchor the ends, but all other shades of value and growth fall somewhere in between. Since the Growth vs. Value Debate won't go away, it's time to give it another look, not based an emotional attachment like a sports fan but from a quantitative perspective. Starting BigSo which is better, growth or value? As odd as it may seem, this question isn't as straight forward as it appears.When financial pundits speak of growth vs. value, it's often framed in terms of the large cap stocks of the S&P 500. These are household names that everyone knows, so it's understandable that this is what you'd see in Barron's or CNBC. Using this benchmark the debate looks pretty cut and dried. The S&P/Barra 500 Growth and S&P/Barra 500 Value Indexes split the market capitalization of the large cap S&P 500 evenly between growth and value stocks. Book value, essentially what's left on a company's balance sheet once all debt has been removed, is the determining factor. Companies with high book value are usually considered "value" while those with low book value are "growth". The S&P/Barra style indexes are rebalanced semi-annually on January 1 and June 1 using the book-to-price ratios and market capitalizations as of November 30 and May 31, respectively.
Chart 1 shows the returns of the S&P/Barra Growth and Value Indexes from their inception (January 1, 1975) through December 2003. Based on that, there's not much question about which approach wins out: Value clearly dominates returns. Even after growth's run in the 1990s, $1 invested in the S&P/Barra 500 Value Index on January 1, 1975 grew to $52.91 while that same dollar invested in the growth index only came to $31.46, a significant difference. So there you have it. Value is the clear winner. Or is it? Have we really addressed the growth vs. value issue to its fullest extent? Probably not, for at least three reasons: First, twenty-eight years of data is considerable, but more would be even more compelling if it covered a longer period and more market cycles. The S&P/Barra 500 style indexes don't go back beyond 1975, but other growth/value series do. Secondly, the S&P/Barra method of defining growth and value based on book-to-price and overall market capitalization is but one of many ways to construct style indexes. Others may actually yield a truer picture of the two styles. Thirdly, as investors painfully learned in the recent bear market, large cap stocks do not define the market. Small and mid cap stocks held up much better during large caps' recent correction from the excesses of the 1990s. Just as growth and value characteristics may help determine performance, market capitalization may play a role as well. The folks at S&P and Barra know this, too. That's why they also have style indexes for the S&P 400 Midcap Index and S&P 600 Smallcap Index. They use the same methodology of dividing stocks into growth and value camps, and both have the same inception as the large cap style indexes. There's certainly nothing wrong with this methodology, but it's worth considering others, especially those that offer a look at the broader market. Fortunately, there are alternatives. A Different ApproachThe Russell 2000 is arguably the most well known measure of small cap stocks, but the Frank Russell Company offers a full complement of equity indexes including large and broad market measures as well as style indexes. Until 1994, Russell, like S&P/Barra, used book value as the means of distinguishing between growth and value stocks.Currently, Russell uses I/B/E/S forecasts of individual stocks' long-term growth mean and a probability methodology to make the distinction. As opposed to S&P/Barra's approach where stocks are classified as either growth or value, about 30% have mixed characteristics and are placed in both growth and value indexes. As a result, they have an impact on both series. This is an odd result and from a quantitative perspective, diminishes the value of the indexes. In addition, the available Russell data series only date back to January 1979, rendering them more limited than the S&P/Barra series.
There are, however, other style indexes which, although less well known to most investors, are more comprehensive and quantitatively complete. The Fama-French style indexes are based on all stocks traded on the NYSE, AMEX, and Nasdaq. Unlike the S&P/Barra and Russell indexes, they extend all the way back to July 1927. Like the other style indexes, these use book value of equity (BVE) to market value (MV) as the determining factor. Stocks with high ratios are considered to be value stocks while those with low ratios are growth. The indexes' creators, Eugene Fama and Ken French, consider all stocks traded on the NYSE, AMEX, and Nasdaq in order to set breakpoints for style and capitalization. Unlike the other indexes, not all stocks are included in the style indexes. Instead, the 30% with the highest BVE/MV ratio are value while those in the lowest 30% are growth. The remaining 40% are considered to be "blends". Results are quite similar to those for the S&P/Barra style series, in fact they're even more striking, particularly for small caps. Each appears on Chart 2 with returns plotted on a logarithmic scale. From a return standpoint, value clearly dominates for both large and small caps. All series were relatively even until the mid-1970s when value pulled ahead. Even though large cap growth outperformed in the bull market of the 1990s, value's lead was hardly diminished. There are three other Fama-French indexes that further illustrate the value/growth relationship. Two, simply called "growth" and "value", are created by again taking the top 30% and bottom of all stocks traded on the NYSE, AMEX, and Nasdaq based solely on BV/MV without regard to capitalization. The third is a derived series resulting from the geometric difference between respective growth and value returns. (For more details on these calculations, E-mail your request.)
Chart 3 has no surprises. Again value handily outperforms growth. The value premium is a little more interesting. It was actually negative -- meaning growth outperformed value -- from 1927 through 1940. It then climbed in an almost linear fashion until the late 1980s and has been relatively flat ever since. The fact that the premium has changed over time demonstrates that the relationship between growth and value has not remained constant over time. If it had, the premium line would be roughly horizontal -- as it has been over the past 15 years. This may also suggest that the historical value premium is not a reliable prediction for the future. What About Risk?Up until now, we've looked exclusively at return, but what about risk? In an efficient market, there should be a direct relation between risk and return. In other words, the riskier the asset or series, the greater the expected return. A rational investor is willing to bear more risk as long as the expected return rises at least proportionally. How do the Fama-French Style Indexes fare when risk is also a concern?Chart 4 plots each series' risk (as measured by standard deviation) in the horizontal axis and average annual return on the vertical axis. The diagonal line is the Security Market Line (SML) created by drawing the line passing through the risk-free rate (here the 30-day T-Bill, not shown on the graph) and the S&P 500. Series plotting higher and to the left are preferable to those plotting lower and to the right.
There's nothing wrong with series being riskier (plotting further to the right) as long as their returns are at least proportionally higher (plotting further up the graph) than their more conservative counterparts. A quick and dirty way to make this comparison is to see where various series fall with respect to the SML: Those plotting above it are preferable to those falling below. Consider, for example, the Small Cap Value Index. It falls the furthest to the right on the graph indicating it has the most risk. Nevertheless, it plots above the SML while its counterpart the Small Growth Index falls well below. Although it plots to the left of the Small Value Index (31.65% vs 35.45%), the Small Value Index is more efficient from a risk/return standpoint. The same can be said for the Value and Growth Indexes. But what about Large Cap Growth and Value? Here the determination isn't as easy since both fall below the SML. It's clear that Large Cap Value has a higher return since it plots above Large Cap Growth, but it's equally clear that it's much more risky, falling much further to the right. The Sharpe Ratio provides a way to make this comparison. It's a measure of risk-adjusted return calculated by subtracting the risk-free rate from a series' return, and then dividing that result by its standard deviation. The greater the standard deviation (risk), the lower the ratio at any given level of return. There's that direct relation between risk and return again. Chart 5 shows the Sharpe Ratios for the Fama-French Indexes. It comes as no surprise that the ratios for the Small Cap Value and Value series are higher than those of the Small Cap Growth and Growth series. We knew that from their relative positions on Chart 4. The large cap series are a different story. Large Cap Growth actually has a higher Sharpe Ratio that Large Cap Value, .1653 vs. .1613, respectively. In other words, despite having a lower annualized nominal return from July 1927 through December 2003, Large Cap Growth offers greater annualized risk-adjusted return.
This is surprising for two reasons: First, it's like winning the war after losing every battle. When you only consider Chart 4's historical annual nominal returns or Chart 2's cumulative results, Large Cap Value seems to far outdistance its growth complement. But that's only half of the equation since over that same time, Large Cap Value actually dominated on a risk-adjusted basis. Despite its lower return, it's been more efficient -- not what you'd normally expect. Secondly, investors typically expect growth stocks to be inherently more risky that value issues. As mentioned earlier, the presumption is that value stocks are already underpriced and have less to lose in a down market while growth shares are often priced to perfection (or higher) by momentum investors. For the latter, even minor disappointments can lead to major declines. But that's not the case; and not just for large caps but for all three Fama-French series. Look back at Chart 4 where you'll notice each value series plots to the right of the corresponding growth index. They fall to the right of their counterparts because they're riskier. While this makes sense in regard to the Efficient Market Hypothesis (higher returning series should be expected to involve more risk) it flies in the face of the common belief that value stocks are "safer" than growth stocks and value investors are more conservative than growth investors. Based on the Fama-French data, the exact opposite is true. "The Best"Far from providing the definitive conclusions, this data suggests the Value vs. Growth Debate -- like beauty -- is in the eye of the beholder. A simple analysis of index returns doesn't tell the whole story. Whichever approach you deem "the best" will not just depend on returns, but on your preferred means of comparison.If nominal returns are your only concern, the Fama-French series do favor value over growth. But that assumes you focus solely on return without regard to risk. Once risk considerations are involved, the situation become murkier. Each value series has a higher standard deviation than the corresponding growth series, making the former "riskier" in the traditional sense of the term. If you have a different definition of risk -- and there certainly are others -- the results may be different. If you're interested in risk-adjusted returns, there's a mixed bag. The Fama-French Small and Value indexes have higher results than their counterparts, but that's not the case for the Large Cap series. So which is "better"? Or are they both? Or do they fluctuate? Or do you even measure "value" and "growth" like Fama and French? Again this might seem to question the obvious, but it's really quite critical to the overall value of the Fama-French results. While it's true that book value plays a role in the S&P/Barra, Russell, and Fama-French style indexes, it plays a slightly different role in each. In the latter, Fama and French have a specific definition of book value. (If you're interested, contact us and we'll send you the details.) It may not be the way you measure book value. Even more basically, the Fama-French style series may not truly reflect your concept of "value". For instance, investors typically associate "value" stocks with high book values, low P/Es, or similar factors. How do these definitions square with the stocks in the Fama-French Value series?
We couldn't run an exact test since we don't have access to fundamental data on all stocks traded on the NYSE, AMEX, and Nasdaq. The broadest universe we had was the Russell 3000, the combination of the Russell 2000 (small stocks) and Russell 1000 (large stocks). While this clearly falls short of the number used in creating the actual indexes, it can provide some insight. To simulate the Fama-French methodology, we first calculated BVE for each stock. Next, we ranked them based on their BVE/MV ratios. Finally, we took the top 30% -- those with the highest BVE/MV -- to represent the approximation of Fama-French value stocks. To see how these aligned with the typical definitions of value stocks, we ordered the stocks of the Russell 3000 by book value and took the top 30% -- those with the highest BV -- to represent value stocks. We also made similar orderings based on trailing four-quarter P/E and projected four-quarter P/E, this time taking the lowest 30% to represent value. The final step was to compare the "typical" value stocks to the Fama-French results. As you'd expect, the book value group had the highest number in common, but it was only 50%. The crossover based on P/E was 42% for trailing and 39% for projected. Clearly, there's not a great deal of overlap between the different methodologies. Again, this was far from the perfect test since our universe was much smaller than that used by Fama and French, but the results are still something to think about. If, as this test implies, the Fama-French approach doesn't align well with what many investors commonly think of as value, then their results may not be as informative as they first appear. As an investor, your concern in the value/growth debate probably stems from a desire to apply the right investing strategy. It does you no good if there's evidence -- even overwhelming evidence -- that one style is better than the other if your definitions of key terms (e.g. return, risk, and value) differ. It's up to you, the investor, to put these findings in context -- however you define it. Search this site! Just enter you key word or words:
Get current quotes or follow your own custom portfolio,
courtesy of E-Line Financials:
|
|||||||||||||||||||||||||||||||
EFORE THEIR POPULARITY EXPLOSION in the 1990s, the goal for most mutual funds was simply to beat the market -- however "the market" was defined. As their distribution increased, funds have become so specialized that investors now expect them to be. Much of this is atrributable to Morningstar and Lipper's rating systems that provide not only specific categories, but benchmarks as well. "Beating the market" has now become "beating the benchmark".
