The efficient market hypothesis (EMH) posits that stock prices reflect all available information, making profit generation challenging. However, studies suggest investor overreactions to new information, creating market inefficiencies. Analyst recommendations, according to signaling theory, provide valuable insights into firm value, potentially impacting stock performance due to information asymmetry. Previous research has shown mixed results regarding the investment value of analyst recommendations and the impact of analyst reputation. Some studies found that recommendations from reputable analysts, often measured by rankings in Institutional Investor Magazine, lead to positive market responses. Others showed that these effects are short-lived and may not translate into profitable investment strategies. This study aims to address these inconsistencies by examining whether star-ranked analysts (using StarMine's proprietary quantitative algorithm) generate abnormal returns and outperform non-star analysts in both short and long portfolios. The study focuses on the NASDAQ market, using StarMine analyst rankings, and avoids the methodological limitations of previous research by considering all years after an analyst is deemed a star, rather than isolating specific years. The key research questions are: 1) Can investors profit from analyst recommendations? and 2) Do reputable analysts (stars) make more profitable recommendations than less reputable analysts (non-stars)? The study utilizes a multi-factor asset pricing model (CAPM, Fama-French 3-factor, Carhart 4-factor) and tests for endogeneity and incorporates investor sentiment using GARCH modeling and frequency-domain causality analysis.

The literature on the economic value of analyst recommendations has expanded significantly since the seminal works of Stickel (1995) and Womack (1996). Womack found that upgrades trigger persistent price drifts, while downgrades can persist for up to six months. Stickel investigated the relationship between analyst reputation (proxied by Institutional Investor Magazine rankings) and short-term price reactions, finding a positive correlation. However, subsequent research revealed that price reactions are often short-lived, not necessarily leading to profitable investment strategies. Studies have explored the relationship between analyst reputation and stock performance, with findings varying widely. While some studies found a positive correlation between analyst reputation and recommendation performance, others raised concerns about potential biases in ranking systems like Institutional Investor Magazine, suggesting they may be influenced by factors beyond skill and that even top-ranked analysts may have achieved their status through fortunate circumstances. The study also examines differences in analyst ranking systems, differentiating between objective (StarMine, Wall Street Journal) and subjective (Institutional Investor Magazine) approaches. Objective rankings rely on quantitative measures of performance, while subjective rankings are based on surveys of institutional investors, potentially introducing biases.

The study uses data from Thomson Reuters' Refinitiv Eikon for daily holding period returns, Thomson Financials Institutional Brokers' Estimate System (I/B/E/S) Detailed Recommendations file for analyst recommendations, and Fama-French factors daily frequency database for market, size, book-to-market, and momentum factors. The sample includes recommendations for 93 NASDAQ 100 companies from 2010 to 2020. Analysts are categorized as "star" or "non-star" based on StarMine's Top Stock Pickers (TSP) and Top Earnings Estimators (TEE) rankings. Only changes in recommendation levels and initial recommendations are considered. A calendar-time portfolio methodology is employed, constructing long and short portfolios for each analyst group. The long portfolio contains stocks with "strong buy" and "buy" recommendations, while the short portfolio includes stocks with "hold," "sell," and "strong sell" recommendations. Portfolio returns are calculated using equal monetary investment, and risk-adjusted returns (alphas) are calculated using CAPM, Fama-French 3-factor model, and Carhart 4-factor model. The Durbin-Wu-Hausman test is used to assess the endogeneity of the momentum factor. The GARCH (1,1) model, with NASDAQ Volatility Index (VXN) as a proxy for investor sentiment, is used to investigate the effect of investor sentiment on stock return volatility. A frequency-domain causality test by Breitung and Candelon (2006) is used to analyze the causal relationship between investor sentiment and stock market volatility across different frequencies.

The study's key findings are: 1) All analyst groups (star, non-star) generate abnormal returns exceeding market averages, rejecting the EMH. 2) Star analysts significantly outperform non-stars in short-term portfolios, with a 0.5565% monthly alpha differential (Carhart 4-factor model). However, no significant difference is observed in long-term portfolios. 3) The top-ranked StarMine analysts ('Star-1') consistently outperform other groups in short portfolios, exhibiting a highly significant monthly alpha differential compared to non-star analysts. 4) The Durbin-Wu-Hausman test confirms the exogeneity of the momentum factor. 5) GARCH analysis shows a positive and statistically significant effect of investor sentiment (proxied by VXN) on stock return volatility. 6) Frequency-domain causality analysis shows that changes in investor sentiment affect stock volatility in both the short and long run, while changes in stock market volatility affect investor sentiment within 5-10 days, without permanent effects. The study further explores the effect of high weight firms and positivity bias on the outcomes, particularly explaining why non-stars seem to perform equally to stars in long portfolios.

The findings support the value of analyst recommendations, especially for short-term investment strategies. Star analysts' superior performance in short portfolios could be attributed to their enhanced access to information and expertise, enabling them to identify and capitalize on short-term market opportunities. The lack of significant difference in long-term portfolios may be due to information quickly being incorporated into prices, reducing the advantage of superior information. The positive relationship between investor sentiment and volatility supports behavioral finance theory. The short-term influence of volatility on investor sentiment may imply that investors rapidly adjust their expectations based on recent market movements. The study's findings highlight the dynamic interplay between analyst recommendations, market sentiment, and volatility, indicating the limitations of simplistic interpretations of market efficiency. This cyclical relationship implies that market participants may be able to leverage systematic patterns in market responses.

This study provides empirical evidence supporting the usefulness of analyst recommendations for stock selection, particularly in short-term portfolios. StarMine's ranking system effectively identifies high-performing analysts. The results highlight the importance of considering both analyst reputation and investment horizon. Future research could explore the impact of different types of news and economic indicators on investor sentiment and the role of top-tier analysts in mitigating information asymmetry in various market conditions.

The study's limitations include the use of a specific dataset (NASDAQ 100 companies from 2010 to 2020). Expanding the dataset to include more markets (NYSE, S&P 500) and a longer time period could enhance the generalizability of the findings. Additionally, the study uses a specific proxy for investor sentiment (VXN). Exploring alternative sentiment measures could provide a more comprehensive view.