中文摘要：

关注客户特征甚至同时跟随客户,既能促进公共信息含量及其质量的提高,又能获得某些私人信息,进而改善分析师盈余预测行为。基于企业层面的证据表明：（1）客户越稳定,分析师对企业的盈利预测越准确,预测分歧度与偏差度越小;（2）企业生产非耐用品、处于行业弱势地位、稳定客户销售占比较高以及与客户关系较好时,稳定客户对分析师盈余预测准确性的影响更显著;（3）稳定客户对分析师盈利预测准确性的影响,可能源自稳定客户对企业盈余以及盈利稳定性的提高。这说明稳定客户是分析师的重要信息来源,即使分析师不同时跟随客户,稳定客户对于分析师预测也有信号传递效应,有助于提高其对企业盈余预测的准确性。基于分析师层面的证据表明：相对于未同时跟随客户与企业的分析师,同时跟随客户与企业的分析师的预测偏差更低,而且这种关系主要体现在客户稳定组。此外在同时跟随客户与企业的样本中,稳定客户与分析师预测偏差负相关。这说明分析师同时跟随稳定客户,可获得供应链知识的溢出效应,从而降低分析师对企业盈利预测的偏误。这些研究结论有助于丰富客户和分析师文献。

英文摘要：

Focusing on a firm＇ s customer characteristics and following the customer may improve the quantity and quality of the analysts＇ information on the covered firm, thus may enhance the analyst＇ s forecast accuracy. The empirical evidence based on ＂firm-level＂ analysis, shows that both the analysts＇ forecast error and dispersion of a supplier firm are significant negative with the firm＇ s stable relationship with its customers. Further analysis shows that, the above positive effects are more prominent when the supplier produces nondurable goods, has a weaker position in the industry, sell higher goods to the stable customers or has a good relationship with customers. And the impact of the stable relationship with customers on analyst forecasts behavior may derives from the effect that the stable relationship can bring higher profitability and profitability persistence. All of these results suggest that stable customers are an important information source of the analysts. Even through don＇ t follow a covered firm＇ s customer, the stable relationship with customers may also improve the analyst forecasts, consistent with the ＂signaling effects＂. The empirical evidence based on ＂firm- analyst-level＂ analysis show that compared with the analysts who do not following a covered firm＇ s customer, analysts who followed provide more accurate earnings forecasts for the supplier firm. And this positive relationship exists only in the stable relationship group. What＇ s more, in the subsample which the analysts follow the covered firm＇ s customer, the analysts＇ forecast error is negative with the stable relationship. All of these results imply that following a covered firm＇ s customer does provide some material incremental information, consistent with the ＂supply chain knowledge spillover effects＂. Based on these finds, our study may enrich the customer and analyst literatures.