买方分析师的盈利预测更准确吗?

Do Buy-side Analysts＇ Earnings Forecasts More Accuracy？

本文利用一家基金管理公司的独特数据，考察了买方分析师与卖方分析师的盈利预测准确度差异及其产生原因。结果发现，与卖方分析师相比，买方分析师的盈利预测准确度显著较差。进一步分析表明，买方分析师准确度较差的原因可能有：（1）买方分析师的预测特征不同（如发布报告的频率较高、跟踪的时间较长）；（2）盈利预测相对保守；（3）做好盈利预测的动力相对不足；（4）与管理层的信息沟通较少。本文首次提供了买方分析师盈利预测的证据，拓展了现有的分析师文献。

There are mainly two kinds of analysl, namely buy-side analysts and sell-side analyst. Buy-side analysts are more independent in earnings forecast and stock rating, while sell-side analysts face several conflicts of interest and then have low independence. Therefore, a question naturally comes up is that whether buy-side analysts＇ earnings forecasts more accu- racy. However, because that data is unavailable, research on this question is vacant in China and is scarce even in developed markets. The Chinese＇setting is different from developed mar- kets in lwo ways. First, the history of stock analysts in China is very short and market con- straint mechanisms such as analyst reputation are underdeveloped. Therefore, the extent of conflict of interest of stock analyst is more pronounced in China. Second, the information en- vironment of stock market in China is relatively poor and thus earnings forecast accuracy is worse than that in developed markets. And analysts in China may be more depend on private information rather than public information because of poor information transparency. There- fore, the results from an emerging market like China may different from those from developed markets. To test our research question, we obtain a proprietary data of research report from a fund management company and process the data on buy sides analysts manually. Using ana- lyst forecast data from 2007--2011 in China, we investigate the difference of earnings forecast accuracy between sell-side analysts and buy-side analysts and the reasons behind the differ ence. Our results show that, compare to sell side analysts, buy side analysts is worse in earnings forecast accuracy. Further findings suggest that, lhe possible reasons why buy-side analyst is worse in earrings forecast are： （1） buy side analysts＇ forecast characteristics （e. g. , firm-specific experience, forecast frequency）. （2） buy side analysts are relatively more conservative in earnings forecast. （3） buy-side analysts＇ incentive to prepare earnings fore- cast is relatively weak than sell-side analysts. （4） buy-side analysts are less likely to access manager＇s private information. This paper contributes the literature in two ways. First, we fill the blank of buy-side analysts literature by firstly examine the earnings forecast of buy- side analysts. Second, by exploring the reasons behind the difference of earnings forecast ac- curacy between buy-side analysts and sell-side analysts, we extend existing literature on ana- lyst earnings forecast and helps to further understand analysts＇ forecast behaviors. Moreo- ver, this paper also has implications for the practitioners by proving that sell-side analyst has comparative advantage in earnings forecast.