آیا تقاضای سرمایه گذار خارجی برای اطلاعات بر دقت پیش بینی تاثیر می گذارد ؟ شواهدی از بازار بورس چینی

Prior international accounting studies have examined mainly the supply side of analyst research. Specific trading rights and financial reporting systems in China provide a unique opportunity to test both demand and supply factors of analyst information. For shares traded only by less-informed foreign investors, the increased information demand may create incentives for analysts to exert greater efforts than for shares traded by local investors. This study provides evidence that analyst effort (proxied by analyst coverage) and expertise (proxied by consensus cross-analyst forecast variability) affect forecast accuracy more significantly for shares traded by foreign investors than for shares traded by local investors. Additionally, I find that the relation between analyst characteristics and relative forecast accuracy is stronger for shares traded by the foreign investors. The findings are consistent with analysts responding to demands for information by less-informed investors.

The Chinese financial markets have become increasingly important for investors around the world.1 Issues such as financial analysts’ earnings forecasts and analysts’ activities in China, however, have received little attention in the academic literature. Foreign investors had only two ways to invest in equity securities in China between 1991 and 2000. They could either buy B shares traded only in China or buy H shares of Chinese companies traded in Hong Kong. In contrast, local Chinese investors could buy only A shares of companies traded in China. Financial analysts followed almost all B shares or H shares but only a handful of A shares. This study is motivated by the extensive analyst coverage of shares traded by foreign investors and leads to the question of why analysts mostly followed shares issued only for foreign investors and followed few shares traded only by local investors. I am interested in exploring two research questions on the phenomenon. First: Do analyst effort and expertise affect earnings forecast accuracy differently for shares traded only by foreign investors than for shares traded only by local Chinese investors? Second: Are the findings consistent with analysts responding to demands for information by foreign investors? I document and discuss information related to the motivation question and later combine it with the two research questions. Next, I empirically test the ability of analyst coverage and forecast variability to explain consensus forecast accuracy and proxies of individual analyst effort and expertise to explain relative forecast accuracy across the different types of shares. The specific trading rights and financial reporting standards in China provide a unique opportunity to examine these associations. Finally, for the second research question, I examine whether analysts respond to increased demand from foreign investors by exerting greater effort and gravitate to fulfill the increased demand for information by the less-informed foreign investors.For shares traded only by foreign investors, the increased demand for earnings information among foreign investor clienteles may create incentives for analysts to intensify their efforts. Analyst following and consensus volatility may have a stronger impact on forecast accuracy, and analysts with superior characteristics (Barniv et al., 2005, Bonner et al., 2003, Clement, 1999, Clement and Tse, 2003, Jacob et al., 1999 and Mikhail et al., 1997) are more likely to issue a superior forecast relative to their peers. For the A shares traded by local investors, weaker investor demand for analysts’ information may reduce the economic incentives of analysts to extend their efforts and would have a smaller effect on analysts’ consensus forecast accuracy and forecast accuracy relative to their peers. Most studies focus on the supply side of analyst forecast accuracy in the United States (e.g., Brown et al., 1987 and Brown and Mohd, 2003) and across countries (e.g., Barniv et al., 2005 and Hope, 2003). A few studies report that analysts tend to respond to investors’ information demands (e.g., Frankel et al., 2006, Lang et al., 2004 and Moyer et al., 1989).2 Prior studies find that local investors in China are better informed than foreign investors and that information asymmetry plays a significant role in the pricing of shares of companies traded only by foreign investors versus shares traded by local investors (Chakravarty et al., 1998, Chan et al., 2003 and Chan et al., 2008). One inference from these articles is that less-informed foreign investors experience special needs for additional information and, in response, analysts have a greater incentive to exert effort. They become more active and provide more coverage for shares traded by foreign investors. The extra analyst effort and expertise should have a greater effect on forecast accuracy and on analysts’ performance relative to their peers. This study examines whether analyst effort and expertise explain consensus and relative forecast accuracy better for B and H shares than A shares. Specifically at the consensus level, the number of analysts and consensus cross-analyst forecast variability are used as proxies for analyst effort and expertise. Additionally, at the individual analyst level, I use analysts’ frequency of forecasts and the length of the time interval between the forecast issue date and the earnings announcement date as proxies for effort, and each analyst's experience and the percentage of companies followed by the analyst in the same industry as proxies for expertise. I expect these proxies and additional control variables to better explain accuracy for shares traded by foreign investors than for shares traded by local investors. In additional analyses, I examine whether these analyst characteristics explain the relative forecast accuracy better for firms with higher foreign ownership. Findings supporting this expectation would indicate that analysts respond to the specific demand for information by foreign investors. Foreign investors could invest in B and/or H shares; therefore, I also examine and compare the relations between accuracy and analyst characteristics using shares traded by foreign investors in China and shares of Chinese companies traded by foreign investors in Hong Kong. Consistent with the research expectations, I find analyst coverage and consensus cross-analyst forecast variability affect forecast accuracy more significantly for shares traded by foreign investors than for shares traded by local investors after controlling for firm size and earnings effects. The relations between the proxies for analyst's effort and expertise and the relative forecast accuracy are also stronger for shares traded by the foreign investors. Finally, I find that, higher foreign-investor ownership, a proxy for investor demand, increases forecast accuracy, and the proxies for analyst's effort and expertise better explain the relative forecast accuracy for companies with higher foreign ownership.3 Overall, this study contributes to the literature in three ways. First, by examining the association of forecast accuracy with specific characteristics across different types of shares, it complements prior studies that examine the impact of (a) listing location on forecast accuracy (Lang, Lins, & Miller, 2003); (b) differences in forecast accuracy associations between U.S. and non-U.S. firms (Das and Saudagaran, 1998 and Das and Saudagaran, 2002); (c) the association between accounting disclosures and forecast accuracy (Hope, 2003); (d) and the association between the legal and financial reporting environments and relative forecast accuracy of individual analysts (Barniv et al., 2005). The specific associations (c) and (d) have not been examined in prior research in China. Second, with regard to specific and unique attributes of the Chinese equity markets, the study complements prior studies that examine (a) the information asymmetry between different types of shares (Chan et al., 2008); (b) the value relevance of different accounting standards used for financial reporting by types of shares (Abdel-khalik et al., 1999, Chen et al., 2001, Lin and Chen, 2005, Liu and Liu, 2007 and Sami and Zhou, 2004); (c) and various Chinese capital market, accounting, and regulatory issues (Allen et al., 2005, Ball et al., 2000a, Ball et al., 2000b, Chen and Yuan, 2004, DeFond et al., 2000, Haw et al., 2000, Haw et al., 2005 and Sun and Tong, 2003). Third, given the idiosyncratic investor rights in China, I examine whether foreign ownership affects consensus forecast accuracy and relative forecast accuracy of individual analysts. With regard to information demand, the study complements prior research on investor demand and analyst informativeness (e.g., DeFond and Hung, 2007 and Frankel et al., 2006), particularly by indicating that analysts respond to demand for information by less-informed foreign investors. The remainder of this article is organized as follows. The next section provides a background discussion within the unique context of the segmented Chinese stock market. The third section provides the rationale and research expectations, and the fourth section provides the methodology. The fifth section presents the data and sample, and the sixth section sets forth the main results. The seventh section provides additional analyses and robustness tests, and the eighth section concludes.

Using a unique setting of differential trading rights for foreign and local investors in China, this research investigates an interesting and challenging question of analysts responding to increased information demands by less-informed foreign investors. Analysts followed almost all shares issued only for foreign investors but followed few shares traded only by local investors. The study examines whether analyst effort and expertise affect earnings forecast accuracy differently for shares traded only by foreign investors than for shares traded only by local Chinese investors. I also analyzed the effect of foreign ownership on forecast accuracy, which helps interpret whether the findings on factors affecting forecast accuracy are consistent with analysts responding to demands for information by foreign investors and, therefore, follow mainly shares traded by foreign investors. I hypothesize that increasing analyst coverage and smaller cross-analyst forecast dispersion, proxies for analyst effort and expertise, would improve forecast accuracy for shares traded by foreign investors more than for shares traded by local investors. Further, I hypothesize a greater effect of proxies for individual analyst effort and expertise on the relative forecast accuracy for analysts following shares traded by foreign investors than for analysts following shares traded by local investors. In addition, I expect greater forecast accuracy as foreign-investor ownership, a proxy for investor demand for analyst information, increases. The empirical results support these predictions. A main conclusion is that analysts’ effort and expertise affect forecast accuracy more significantly for shares traded only by foreign investors than for shares traded only by local investors. The findings are consistent with analysts responding to demands by the less-informed foreign investors. The study focuses on testing analyst response, that is, the supply side, and provides only indirect examination of investors’ demand. I consider significant foreign-ownership coefficients in the predicted direction to constitute evidence consistent with analysts responding to demand for more accurate forecasts, but the unavailability of data to construct alternate proxies definitely limits testing the supply and demand effects. The unavailability of some financial reporting and analyst activities data used in prior global studies poses a similar constraint. The differences in information environments have been only partially controlled in the study. Finally, some changes in political, economic, and environmental factors may also influence the results, and the research design provides only partial control for these potentially confounding effects. Less-informed investors need more information and may choose to invest in companies for which analysts may issue more accurate forecasts. Analysts followed almost all B shares traded only by foreign investors and only a few A shares traded by local investors. Analysts cease coverage for some B shares when local investors were allowed to trade in B shares in 2001 and initiate coverage for relatively many A shares when China permitted trading in these shares to selected foreign institutional investors in 2003. Taken together, these issues suggest that foreign investor demand motivates analysts, and they present a unique opportunity to test how analysts respond to foreign-investor demand for analyst information. I investigate this environment where the demand for analyst information is provided by foreign investors who were the only group allowed to trade in B and H shares. Future cross-country studies may test investor demand for analyst information and analyst response where foreign and local investors trade in cross-listing companies and, subject to data availability, for shares traded simultaneously by local and foreign investors in other countries.