The Brandes Institute extends a landmark academic study on the comparative performance of value and glamour stocks in the United States to cover a 38-year period. Extending the results of the seminal 1994 study conducted by Professors Josef Lakonishok, Andrei Shleifer, and Robert Vishny, we compare the performance of value stocks to glamour securities through April 2006.
Data Source:
The Brandeis Institute uses the same data as the original LSV paper entitled "Contrarian Investment, Extrapolation, and Risk." They examine stocks trading on the NYSE and the AMEX from April 30, 1994 to April 30, 2006. Their datasource is FACTSET, which uses the Compustat database.
Data Specification:
This paper replicates the famous study of LSV. Each April 30th of a given year the researchers sort all the stocks on the NYSE/AMEX into deciles based on their recent fiscal year price/book, price/cash flow, and price/earnings (three rough measures that can separate value stocks and growth stocks). The idea is that companies that are selling for high p/b, p/cf, or p/e represent stocks that are in favor with the market and may be overvalued, whereas, low p/b, p/cf, and p/e companies are stocks that are out of favor with the market and may be undervalued.
Investment Strategy:
Each year on April 30 take all the appropriate data on the universe of NYSE and AMEX stocks (or your favorite universe) and rank them on your favorite 'value' measure. You can try p/b, p/cf, or p/e, a combination of all three, or some derivation of these--the point is to capture the 'value' spread. Once you have ranked all the stocks on your criterion, you then go long the portfolio of stocks with the lowest value measure and hold for five years.
Results:
LSV found that for the P/B sort the average annualized return (for holding a 5 year portfolio) for the low P/B decile was 21.6% and the highest P/B was 11.5%--a substantial outperformance! These returns were based on portfolios formed in 1969-1989 and cover an investment period of 1969 to 1994. Brandeis extends these results and gets the average annualized returns for the more recent portfolios formed in April 30, 1990 through April 30, 2001. Unfortunately, they do not directly show the results of the portfolios in the 1990-2006 period, but instead show the entire times series from 1968 to 2006. In this extended series they find the value stocks earn an average of 17.9% and the glamour stocks earn 10.4%. It seems the tech bubble years put a dent in the value/glamour trading strategy, but the results are still strong. Even so, the rebound from the tech bubble bursting gave value another notch in the win column, which makes the results mixed. It still seems that value is STILL king!
Implementation Issues and Remarks
This is probably one of the easiest strategies to implement. Simply go to the Yahoo Finance Screener or any number of free websites and screen stocks based on book value. Once you have identified your stocks, you buy a portfolio of the cheapest stocks and hold it for 5 years. If the past 30+ years of data are any guide, you should outperform growth investors fairly reliably. The main problem with this strategy is that it takes guts and staying power, which most people do not have. If you are someone who is looking for a nice, smooth return stream, and you get utility from BSing about stocks with your buddies, this is probably not the strategy for you.
Overall, this strategy is easy to implement and cost effective, however, the returns (unless you form a long/short) are still risky at some level--you may outperform growth, but if growth is down 50% and value is down 20%, you still lost 20%!!! Investment potential ranking: 7/10
[URL=http://empiricalfinanceresearch.blogspot.com/]Source: Empirical Finance Research Blog[/url]
Data Source:
The Brandeis Institute uses the same data as the original LSV paper entitled "Contrarian Investment, Extrapolation, and Risk." They examine stocks trading on the NYSE and the AMEX from April 30, 1994 to April 30, 2006. Their datasource is FACTSET, which uses the Compustat database.
Data Specification:
This paper replicates the famous study of LSV. Each April 30th of a given year the researchers sort all the stocks on the NYSE/AMEX into deciles based on their recent fiscal year price/book, price/cash flow, and price/earnings (three rough measures that can separate value stocks and growth stocks). The idea is that companies that are selling for high p/b, p/cf, or p/e represent stocks that are in favor with the market and may be overvalued, whereas, low p/b, p/cf, and p/e companies are stocks that are out of favor with the market and may be undervalued.
Investment Strategy:
Each year on April 30 take all the appropriate data on the universe of NYSE and AMEX stocks (or your favorite universe) and rank them on your favorite 'value' measure. You can try p/b, p/cf, or p/e, a combination of all three, or some derivation of these--the point is to capture the 'value' spread. Once you have ranked all the stocks on your criterion, you then go long the portfolio of stocks with the lowest value measure and hold for five years.
Results:
LSV found that for the P/B sort the average annualized return (for holding a 5 year portfolio) for the low P/B decile was 21.6% and the highest P/B was 11.5%--a substantial outperformance! These returns were based on portfolios formed in 1969-1989 and cover an investment period of 1969 to 1994. Brandeis extends these results and gets the average annualized returns for the more recent portfolios formed in April 30, 1990 through April 30, 2001. Unfortunately, they do not directly show the results of the portfolios in the 1990-2006 period, but instead show the entire times series from 1968 to 2006. In this extended series they find the value stocks earn an average of 17.9% and the glamour stocks earn 10.4%. It seems the tech bubble years put a dent in the value/glamour trading strategy, but the results are still strong. Even so, the rebound from the tech bubble bursting gave value another notch in the win column, which makes the results mixed. It still seems that value is STILL king!
Implementation Issues and Remarks
This is probably one of the easiest strategies to implement. Simply go to the Yahoo Finance Screener or any number of free websites and screen stocks based on book value. Once you have identified your stocks, you buy a portfolio of the cheapest stocks and hold it for 5 years. If the past 30+ years of data are any guide, you should outperform growth investors fairly reliably. The main problem with this strategy is that it takes guts and staying power, which most people do not have. If you are someone who is looking for a nice, smooth return stream, and you get utility from BSing about stocks with your buddies, this is probably not the strategy for you.
Overall, this strategy is easy to implement and cost effective, however, the returns (unless you form a long/short) are still risky at some level--you may outperform growth, but if growth is down 50% and value is down 20%, you still lost 20%!!! Investment potential ranking: 7/10
[URL=http://empiricalfinanceresearch.blogspot.com/]Source: Empirical Finance Research Blog[/url]
