Date of Award

Spring 5-2-2018

Document Type


First Advisor

Samuel W. Woolford

Second Advisor

Dhaval M. Dave

Third Advisor

Jahangir Sultan


Financial research analysts are experts who analyze the financial markets and company fundamentals to make investment recommendations. Based on the US analysts’ stock recommendations, I examine the issue of financial analysts’ over-optimism, the investment value of analysts’ stock recommendations, and analysts’ leader-follower herding behavior. The goal of this thesis is to enhance the understanding of the roles that financial analysts play in promoting information transmission in the financial market. The study of analysts’ over-optimism focuses on the market reaction asymmetry. Consistent with analysts being over-optimistic, the financial market responds more strongly to analysts’ unfavorable recommendations than to their corresponding favorable recommendations. This market reaction asymmetry reflects how investors perceive analysts’ over-optimism and discount the potential bias. The study measures the effect of regulatory efforts to reduce the market reaction asymmetry through the adoption of NASD Rule 2711, NYSE Rule 472 and the Global Analyst Research Settlement. However, after considering the “Self-Correction” mechanism of the financial market, the analysis indicates that the actual over-optimism mitigation due to the regulatory intervention was only a 0.7% reduction in the market reaction asymmetry. viii The study of the investment value of analysts’ recommendations is performed through a portfolio construction approach. I use both a passive asset allocation strategy and an active asset allocation strategy to create recommendation-based portfolios. The passive strategy utilizes a market-value weighting to rebalance the portfolio and the active strategy uses a Black-Litterman model to determine the optimal stock weight in rebalancing. I find that both strategies generate positive abnormal returns. The active asset allocation strategy outperforms the passive strategy as it allows investors to gain incremental values by overweighting outperforming industries and underweighting underperforming industries. In studying analysts’ herding behavior, herding behavior is defined as issuing leader-follower recommendations in the presence of clustered recommendations. Based on a network analysis approach, I find that analysts’ herding network structure and analysts’ centrality can explain their performance on non-herding recommendations. This finding is consistent with a learning by herding hypothesis, which states that analysts can acquire knowledge from other analysts when learning is the motivation to herd.