This paper examines the relationship between career concerns and herding behavior among security analysts, focusing on the labor market for security analysts. The authors use data from the Institutional Brokers Estimate System (I/B/E/S) to analyze the timing of earnings forecasts and the deviation from consensus forecasts as measures of herding. Key findings include: 1. **Timing of Forecasts**: Older analysts are more likely to produce the first earnings forecast for a firm compared to younger analysts. This suggests that older analysts are less likely to herd. 2. **Deviation from Consensus**: Older analysts also deviate more from the consensus forecast than younger analysts, further supporting the idea that older analysts are less likely to herd. 3. **Sensitivity Analysis**: The results are robust to various methodological changes, including different cutoff dates for forecasts and changes in the sample size. 4. **Auxiliary Findings**: The paper also explores the relationship between forecast accuracy and frequency of revisions with job tenure, finding that older analysts are less accurate and revise their forecasts less frequently. The authors argue that these findings are consistent with reputation-based models of herding, where older analysts, having more to lose if they are wrong, are less likely to deviate from the consensus. The study provides empirical evidence to support the idea that career concerns influence herding behavior in the labor market for security analysts.This paper examines the relationship between career concerns and herding behavior among security analysts, focusing on the labor market for security analysts. The authors use data from the Institutional Brokers Estimate System (I/B/E/S) to analyze the timing of earnings forecasts and the deviation from consensus forecasts as measures of herding. Key findings include: 1. **Timing of Forecasts**: Older analysts are more likely to produce the first earnings forecast for a firm compared to younger analysts. This suggests that older analysts are less likely to herd. 2. **Deviation from Consensus**: Older analysts also deviate more from the consensus forecast than younger analysts, further supporting the idea that older analysts are less likely to herd. 3. **Sensitivity Analysis**: The results are robust to various methodological changes, including different cutoff dates for forecasts and changes in the sample size. 4. **Auxiliary Findings**: The paper also explores the relationship between forecast accuracy and frequency of revisions with job tenure, finding that older analysts are less accurate and revise their forecasts less frequently. The authors argue that these findings are consistent with reputation-based models of herding, where older analysts, having more to lose if they are wrong, are less likely to deviate from the consensus. The study provides empirical evidence to support the idea that career concerns influence herding behavior in the labor market for security analysts.