Date of Degree


Document Type


Degree Name





Jun Wang

Committee Members

Armen Hovakimian

Rong Huang

Lin Peng

Subject Categories



Earning Forecast Revision; Last Minute; Security Analyst; Stock Recommendation


This dissertation consists of 3 chapters.

Chapter 1: Last Minute Earning Forecast Revision

Chapter 2: Security Analysts' Double Down Behavior in Stock Recommendation

Chapter 3: Cost of Speaking in Two Different Tongues

Chapter 1: I study financial analysts who revise their earnings forecasts just a few days before firms' earnings announcement. Analysts who apply this strategy are more accurate in their earnings forecasts, and they are more likely to move to a larger brokerage firm. All-star analysts are more likely to take this last minute revision strategy than non-all-star analysts. All-star analysts using the strategy are likely to maintain their all-star status and get a better career path, while non-all-star analysts are less likely to get promotion when they apply this strategy.

Chapter 2: When a stock experiences a significant loss after a favorable recommendation from an analyst, the analyst can either continue to issue a favorable recommendation or reverse the course. I define the behavior of continuing favorable coverage as the analyst's doubling down behavior. I find analysts who double down are likely to get demoted and less likely to become an all-American analyst in the next year. Analyst makes biased recommendation in the double down behavior because of overconfidence instead of defending the firm. Stocks recommended by doubling down analysts have worse performance than stocks recommended by other analysts. Investors perform worse if they follow the buy recommendations of doubling down analysts.

Chapter 3: Malmendier and Shanthikumar (2014) find that some analysts prefer to issue relatively higher stock recommendation ratings and relatively lower earnings forecast on the same firm at the same time. They describe this behavior as speaking in two different tongues. Following Malmendier and Shanthikumar (2014)'s definition of the two-tongue strategy, I explore the external influence of this strategy on financial analysts. I show that when analysts take the two-tongue strategy, they are suffering a cost by sacrificing their accuracy on their target firms. The more frequent an analyst takes the two-tongue strategy in the previous year, the less likely this analyst would be promoted to a top 10 brokerage house and be nominated as an All-American analyst in the current year. Moreover, investors respond more positively to the higher stock recommendation ratings but ignore the lower earnings forecast on the firms where analysts apply the two-tongue strategy by showing no significant negative reaction.

Included in

Business Commons