Methodology for EMH
1. Methodology
1.1 Impact of Buy and Sell Recommendations
We examined the effect of Barrons’ buy, sell and hold recommendations on stock prices, using the Brown and Warner (1985) standard event-study method to compute the daily abnormal returns. We used a two-step procedure to compute the average daily abnormal returns with stock price data from CRSP.
First, we estimated the alpha and beta coefficients of each firm by using a single-factor market model. We used the days from –301 to –46 as the estimation window for this purpose. Second, we computed the abnormal return by subtracting a firm’s expected daily return from its observed return. We calculated the cumulative abnormal returns by adding up the abnormal returns over the periods from days -30 to 0, days -5 to 0, days 0 to 1, days 0 to 5, and days 2 to 30, where day 0 represents the next trading day following the Saturday when the Barrons’ recommendation was published. These abnormal returns are estimated for buy, sell and hold recommendations.
1.2 Portfolio Performance
In order to evaluate the average performance of recommendations published by Barrons we assembled a Barrons portfolio by equally weighting their recommendations for 374 days starting January 3, 2012 till June 28, 2013. We used two approaches to do this.
First, we added a new security into the portfolio whenever there was a buy recommendation published for it in Barrons. The security remained in the portfolio for 6 months if there was no subsequent buy or hold recommendation during this period. If Barrons published another buy or hold recommendation on this security during the 6 month period then the stock remained in the portfolio for an additional 6 month period starting on the date when the recommendation was published. A security was removed from the portfolio if it had remained in the portfolio for 6 months and a new buy or hold recommendation was not made during this time. A security was also removed from the portfolio
1.1 Impact of Buy and Sell Recommendations
We examined the effect of Barrons’ buy, sell and hold recommendations on stock prices, using the Brown and Warner (1985) standard event-study method to compute the daily abnormal returns. We used a two-step procedure to compute the average daily abnormal returns with stock price data from CRSP.
First, we estimated the alpha and beta coefficients of each firm by using a single-factor market model. We used the days from –301 to –46 as the estimation window for this purpose. Second, we computed the abnormal return by subtracting a firm’s expected daily return from its observed return. We calculated the cumulative abnormal returns by adding up the abnormal returns over the periods from days -30 to 0, days -5 to 0, days 0 to 1, days 0 to 5, and days 2 to 30, where day 0 represents the next trading day following the Saturday when the Barrons’ recommendation was published. These abnormal returns are estimated for buy, sell and hold recommendations.
1.2 Portfolio Performance
In order to evaluate the average performance of recommendations published by Barrons we assembled a Barrons portfolio by equally weighting their recommendations for 374 days starting January 3, 2012 till June 28, 2013. We used two approaches to do this.
First, we added a new security into the portfolio whenever there was a buy recommendation published for it in Barrons. The security remained in the portfolio for 6 months if there was no subsequent buy or hold recommendation during this period. If Barrons published another buy or hold recommendation on this security during the 6 month period then the stock remained in the portfolio for an additional 6 month period starting on the date when the recommendation was published. A security was removed from the portfolio if it had remained in the portfolio for 6 months and a new buy or hold recommendation was not made during this time. A security was also removed from the portfolio