I recently conducted a study with Dr. George Athanassakos. In the coming weeks I'm hoping to reformat it for publication in a journal. In the meantime please see below the abstract as well as an excerpt from the paper. Your thoughts would be very valuable.
ABSTRACT
In this paper we study the performance of value versus growth, and the resultant value premium, from 1951 to present. We look specifically at post-war recessionary periods in the United States. We are especially interested in the performance of value versus growth during a “credit crunch” such as the one in the current recessionary period. We find that (1) the only times during recessionary periods that one of either value or growth had positive returns, while the other had negative, were times when growth was negative and value was positive, (2) during both the current financial crisis, and that of the early 90’s, value suffered but performed no worse than growth and (3) there was a significant negative value premium in each of the 12 months leading up to the four most recent bear markets. This paper could prove useful to other researchers of the value premium as well as to practitioners who may be interested in developing a predictive model for use in the field.
EXCERPT
The value premium is the excess return of value stocks over the return of growth stocks. On the aggregate, and for research purposes, value stocks are usually described as those stocks with a low price-to-earnings ratio (or high earnings-to-price ratio). Value stocks can also be selected based on other metrics such as the price-to-book ratio. In practice, the value premium may have even greater significance when an aggregate basket is not used, but rather when specific equity selections are made. In this paper we set out to find how the value premium has performed during the current recession and what impacts a liquidity crisis may have had on value. In doing so we also determine the value premium for specific recessionary periods in the post-war United States and then draw implications from this for future research on the value premium.
A body of research is available demonstrating the significance of the value premium over long time periods (Lakonishok, Shleifer and Vishny 1994, Capaul, Rowley and Sharpe 1993) and in different parts of the world (Athanassakos 2006, Fama and French 1998). Athanassakos (2006) provides evidence for a consistently strong value premium over a 1985-2002 sample period in the Canadian market; a value premium that persists in both bull and bear markets as well as in recessions and recoveries.
There are two main causes of the value premium proposed in the body of research available today; these are the premium as a compensation for added risk and the premium as a result of errors-in-expectations. The most recent research available suggests that the errors-in-expectations theory is the likely cause of the value premium (Athanassakos 2006, Chan and Lakonishok 2004, Lakonishok, Shleifer and Vishny 1994). It is argued that markets are inefficient, evidence being from observations such as “The January Effect” (Athanassakos 1992), and as such there are times when certain equities are unduly overpriced or underpriced due to unrealistic future expectations based on historical returns.
In addition to research on the value premium over long periods of time, there is research available that has looked specifically at value stocks during periods of recession/expansion and bull/bear markets on the aggregate. Some of this research shows that historically the value premium persists during all aspects of the business cycle (Arshanapalli and Nelson 2007, Kwag and Lee 2006). There is also a body of research, however, which demonstrates that fundamentally value stocks are riskier during bear markets and recessions and that the value premium would be negatively affected during these times (Chen, Petkova and Zhang 2006, Xing and Zhang 2005, Black and Fraser 2003, Asness, Friedman, Krail and Liew 2000).
Kwag and Lee (2006) provide evidence that a value portfolio consistently outperforms a growth portfolio throughout the business cycle and that the benefits of value investing are even greater during periods of contraction than during periods of expansion. Arshanapalli and Nelson (2007) also provide strong evidence in support of this, showing that during the 1962-2005 period value portfolios suffered smaller losses during down markets than did growth portfolios. Their study also shows that investing in a high book-to-market portfolio during this time offered a significant value premium in addition to the downside protection that it offered in bear markets. This value premium was significant for all firm sizes. As a result, Arshanapalli and Nelson believe that the value premium doesn’t appear to result from bearing the additional risk these stocks may inherit during a recession.
While the aforementioned evidence provides a strong case for the performance of value over growth stocks during all aspects of the business cycle, value stocks have also performed poorly for significant periods of time (Asness, Friedman, Krail and Liew 2006). Black and Fraser (2003) provide strong evidence that the future expected value premium is negatively related to recently recorded GDP growth over long periods of time. Their research may contradict that of Kwag and Lee (2006) and Arshanapalli and Nelson (2007) in that it lends support to the view of the value premium as a reward for non-diversifiable risk associated with financial distress. They argue that value stocks are riskier during times of recession when investors least want to hold a distressed stock. The findings of Xing and Zhang (2005) provide support to this research as they find that the fundamentals of value firms decline sharply in recessions. While growth firms also experience a decline in fundamentals, the decline experienced is not as deep as that of value firms. The specific areas of underperformance during these times are in earnings growth, dividend growth, and sales and investment growth. It is proposed that as value firms generally have less flexibility than growth firms, due to a higher proportion of tangible assets, they are more susceptible to cost reversibility. As such, they conclude that value firms have a difficult time in smoothing negative aggregate shocks and are more affected by negative business cycle shocks than are growth firms. Providing further evidence to this research are Chen, Petkova and Zhang (2006) who demonstrate a countercyclical value premium, implying that value is riskier than growth in adverse economic times when the price of risk is high. They, like Xing and Zhang (2005) propose that as cutting capital is more costly than expanding capital, value firms don’t have enough flexibility when scaling down during negative business cycle shocks; value is therefore to be more adversely affected by economic downturns.
The previously mentioned research, regarding the demonstrated and expected value premium during negative business cycle shocks, is not unified in findings or expectations. Also, there has not been any research conducted on the performance of the value premium during specific negative business cycle shocks in history. While a significant body of research is available showing that value has outperformed growth in negative business cycles, much of the research points to evidence that value could be expected to underperform when the price of risk is high. This leads us to believe that during a financial crisis, such as the one present during the most recent recession, a significantly negative value premium could be expected. This paper studies a number of specific post-war recessionary periods in U.S. history, including the current one, in order to establish the periods where value did indeed underperform growth. We also provide a real situational analysis from these findings. Our research looks at value versus growth over a longer period than previously conducted research and delves into specific periods of time. To the best of our knowledge this has not been conducted previously. As practitioners believe that the outperformance of certain investment styles can be traced to economic fundamentals (Kao and Shumaker 1999), this paper also considers the macroeconomic indicators of the time periods that are being studied.
We find that it is difficult to make many connections between the realized value premium and the observed macroeconomic indicators in each period. This is likely due to most instances of the variation in the value premium being due to a standard deviation of the value premium that would be expected. There are a number of interesting and specific findings from our research that should greatly add to the study of the value premium in general. Firstly, all the instances where one of either growth or value had negative returns while the other had positive returns were instances where the returns of the growth portfolio were negative while those of the value portfolio were positive. These instances seem to be around situations involving a bubble in the equity markets. Secondly, it is apparent that value did not underperform growth in the most recent financial crisis. It seems that the returns of a value portfolio, while suffering, merely performed as poorly as those of a growth portfolio. This was likely due to a flight to quality across all financial assets and the same phenomenon can be observed during the 1990-1991 recessionary period, which can also be classified as a “credit crunch”. Lastly, and perhaps most interesting, is a persistent negative value premium observed in each of the 12 months prior to the most recent four bear markets. This trend goes back almost 30 years and, as a potential leading indicator, may prove to be the foundation for a predictive model in the future.
Jonathan Goldberg, MBA
http://www.jonathangoldberg.com/
ABSTRACT
In this paper we study the performance of value versus growth, and the resultant value premium, from 1951 to present. We look specifically at post-war recessionary periods in the United States. We are especially interested in the performance of value versus growth during a “credit crunch” such as the one in the current recessionary period. We find that (1) the only times during recessionary periods that one of either value or growth had positive returns, while the other had negative, were times when growth was negative and value was positive, (2) during both the current financial crisis, and that of the early 90’s, value suffered but performed no worse than growth and (3) there was a significant negative value premium in each of the 12 months leading up to the four most recent bear markets. This paper could prove useful to other researchers of the value premium as well as to practitioners who may be interested in developing a predictive model for use in the field.
EXCERPT
The value premium is the excess return of value stocks over the return of growth stocks. On the aggregate, and for research purposes, value stocks are usually described as those stocks with a low price-to-earnings ratio (or high earnings-to-price ratio). Value stocks can also be selected based on other metrics such as the price-to-book ratio. In practice, the value premium may have even greater significance when an aggregate basket is not used, but rather when specific equity selections are made. In this paper we set out to find how the value premium has performed during the current recession and what impacts a liquidity crisis may have had on value. In doing so we also determine the value premium for specific recessionary periods in the post-war United States and then draw implications from this for future research on the value premium.
A body of research is available demonstrating the significance of the value premium over long time periods (Lakonishok, Shleifer and Vishny 1994, Capaul, Rowley and Sharpe 1993) and in different parts of the world (Athanassakos 2006, Fama and French 1998). Athanassakos (2006) provides evidence for a consistently strong value premium over a 1985-2002 sample period in the Canadian market; a value premium that persists in both bull and bear markets as well as in recessions and recoveries.
There are two main causes of the value premium proposed in the body of research available today; these are the premium as a compensation for added risk and the premium as a result of errors-in-expectations. The most recent research available suggests that the errors-in-expectations theory is the likely cause of the value premium (Athanassakos 2006, Chan and Lakonishok 2004, Lakonishok, Shleifer and Vishny 1994). It is argued that markets are inefficient, evidence being from observations such as “The January Effect” (Athanassakos 1992), and as such there are times when certain equities are unduly overpriced or underpriced due to unrealistic future expectations based on historical returns.
In addition to research on the value premium over long periods of time, there is research available that has looked specifically at value stocks during periods of recession/expansion and bull/bear markets on the aggregate. Some of this research shows that historically the value premium persists during all aspects of the business cycle (Arshanapalli and Nelson 2007, Kwag and Lee 2006). There is also a body of research, however, which demonstrates that fundamentally value stocks are riskier during bear markets and recessions and that the value premium would be negatively affected during these times (Chen, Petkova and Zhang 2006, Xing and Zhang 2005, Black and Fraser 2003, Asness, Friedman, Krail and Liew 2000).
Kwag and Lee (2006) provide evidence that a value portfolio consistently outperforms a growth portfolio throughout the business cycle and that the benefits of value investing are even greater during periods of contraction than during periods of expansion. Arshanapalli and Nelson (2007) also provide strong evidence in support of this, showing that during the 1962-2005 period value portfolios suffered smaller losses during down markets than did growth portfolios. Their study also shows that investing in a high book-to-market portfolio during this time offered a significant value premium in addition to the downside protection that it offered in bear markets. This value premium was significant for all firm sizes. As a result, Arshanapalli and Nelson believe that the value premium doesn’t appear to result from bearing the additional risk these stocks may inherit during a recession.
While the aforementioned evidence provides a strong case for the performance of value over growth stocks during all aspects of the business cycle, value stocks have also performed poorly for significant periods of time (Asness, Friedman, Krail and Liew 2006). Black and Fraser (2003) provide strong evidence that the future expected value premium is negatively related to recently recorded GDP growth over long periods of time. Their research may contradict that of Kwag and Lee (2006) and Arshanapalli and Nelson (2007) in that it lends support to the view of the value premium as a reward for non-diversifiable risk associated with financial distress. They argue that value stocks are riskier during times of recession when investors least want to hold a distressed stock. The findings of Xing and Zhang (2005) provide support to this research as they find that the fundamentals of value firms decline sharply in recessions. While growth firms also experience a decline in fundamentals, the decline experienced is not as deep as that of value firms. The specific areas of underperformance during these times are in earnings growth, dividend growth, and sales and investment growth. It is proposed that as value firms generally have less flexibility than growth firms, due to a higher proportion of tangible assets, they are more susceptible to cost reversibility. As such, they conclude that value firms have a difficult time in smoothing negative aggregate shocks and are more affected by negative business cycle shocks than are growth firms. Providing further evidence to this research are Chen, Petkova and Zhang (2006) who demonstrate a countercyclical value premium, implying that value is riskier than growth in adverse economic times when the price of risk is high. They, like Xing and Zhang (2005) propose that as cutting capital is more costly than expanding capital, value firms don’t have enough flexibility when scaling down during negative business cycle shocks; value is therefore to be more adversely affected by economic downturns.
The previously mentioned research, regarding the demonstrated and expected value premium during negative business cycle shocks, is not unified in findings or expectations. Also, there has not been any research conducted on the performance of the value premium during specific negative business cycle shocks in history. While a significant body of research is available showing that value has outperformed growth in negative business cycles, much of the research points to evidence that value could be expected to underperform when the price of risk is high. This leads us to believe that during a financial crisis, such as the one present during the most recent recession, a significantly negative value premium could be expected. This paper studies a number of specific post-war recessionary periods in U.S. history, including the current one, in order to establish the periods where value did indeed underperform growth. We also provide a real situational analysis from these findings. Our research looks at value versus growth over a longer period than previously conducted research and delves into specific periods of time. To the best of our knowledge this has not been conducted previously. As practitioners believe that the outperformance of certain investment styles can be traced to economic fundamentals (Kao and Shumaker 1999), this paper also considers the macroeconomic indicators of the time periods that are being studied.
We find that it is difficult to make many connections between the realized value premium and the observed macroeconomic indicators in each period. This is likely due to most instances of the variation in the value premium being due to a standard deviation of the value premium that would be expected. There are a number of interesting and specific findings from our research that should greatly add to the study of the value premium in general. Firstly, all the instances where one of either growth or value had negative returns while the other had positive returns were instances where the returns of the growth portfolio were negative while those of the value portfolio were positive. These instances seem to be around situations involving a bubble in the equity markets. Secondly, it is apparent that value did not underperform growth in the most recent financial crisis. It seems that the returns of a value portfolio, while suffering, merely performed as poorly as those of a growth portfolio. This was likely due to a flight to quality across all financial assets and the same phenomenon can be observed during the 1990-1991 recessionary period, which can also be classified as a “credit crunch”. Lastly, and perhaps most interesting, is a persistent negative value premium observed in each of the 12 months prior to the most recent four bear markets. This trend goes back almost 30 years and, as a potential leading indicator, may prove to be the foundation for a predictive model in the future.
Jonathan Goldberg, MBA
http://www.jonathangoldberg.com/
