Should Investors Follow the Prophets or the Bears? Evidence on the Use of Public Information by Analysts and Short Sellers

Transcription

1 Should Investors Follow the Prophets or the Bears? Evidence on the Use of Public Information by Analysts and Short Sellers Michael S. Drake Lynn Rees Edward P. Swanson The Accounting Review January 2011 (pp ) Financial Planning Workshop: Strategic Issues for Financial Planners October 2011

2 Mean six-month abnormal returns to long-short trading strategy ( ) (firms sorted into quintiles by Recommendations and Short Interest) Portfolio Recommend Levels Short Interest 1 = Buy Signal = Sell Signal Return Spread ** 0.043**

3 Six-month size-adjusted returns to portfolios based on analyst strong buy and strong sell recommendations

4 Mean six-month returns to long-short trading strategy using both recommendation levels and short interest ( ) Short Interest Strong Sell Consensus Stock Recommendation Strong Buy Low High Return spread when signals concur: 2.0% Return spread when signals conflict: 9.6%***

5 Does information use explain why trading with the shorts against the analysts produces high returns? Fundamental signals with predictive information (from prior research): 1. Market value of equity 2. Stock price momentum 3. Analyst earnings forecast revisions 4. Standardized unexpected earnings 5. Earnings-to-price 6. Book-to-market 7. Daily volume turnover 8. Analyst long-term growth forecasts 9. Sales growth 10. Accruals 11. Capital expenditures

6 Ordered logistic regressions of analyst recommendation level quintile on fundamentals Variable Pred. Coefficient t-stat Market Capitalization (Ln) neg Momentum Variables (Price or Earnings) Stock Price Momentum (6 months) pos Analyst Forecast Revision pos Quarterly Earnings Change (seasonal) pos Contrarian Variables (Fundamental or Growth) Earnings-to-Price (4 quarter average) pos Book-to-Market pos Stock Turnover neg Analyst Consensus LT Growth Forecast neg Sales Growth (actual) neg Total Accruals neg Capital Mays Business Expenditures School neg

7 Ordered logistic regression of short interest quintile on fundamentals Variable Pred. Coefficient t-stat Market Capitalization (Ln) pos Momentum Variables (Price or Earnings) Stock Price Momentum (6 months) neg Analyst Forecast Revision neg Quarterly Earnings Change (seasonal) neg Contrarian Variables (Fundamental or Growth) Earnings-to-Price (4 quarter average) neg Book-to-Market neg Stock Turnover pos Analyst Consensus LT Growth Forecast pos Sales Growth (actual) pos Total Accruals pos Capital Mays Business Expenditures School pos

8 Conclusion About Use of Information Shorts Short interest captures the predictive information in fundamentals to a remarkable extent Analysts Recommendation levels are based primarily on stock momentum and earnings growth (glamor stocks) Recommendation revisions only partially correct information in recommendation levels

9 Conclusions & Implications 1. Investors can profit by combining signals from analysts and shorts. 2. The most profitable strategy is to follow the shorts when analyst recommendations and short interest disagree.

10 Conclusions & Implications 3. Shorts play a role in the price formation process by taking positions in firms whose fundamentals are consistent with overvaluation. 4. Regulators should therefore exercise caution in adding further restrictions to reduce short selling.

11 Thank you for the invitation to visit today!