行為財務學的三篇文章:證券分析師之實證研究

Abstract

在經濟體系中，人際關係的互動是不可或缺的一部分。人與人之間的相處，存在著互惠、合作等影響經濟運作的重要行為。當政府推行某些政策或實施新法規來限制某些行為時，人們的變通時常出人意外。本論文以證券分析師為研究對象，探討其與公司高管的互動關係。證券分析師的個別資訊在社交平台上普及易得，且分析師與企業管理者的互動關係複雜，以致出現相關法規對兩者的互動進行規範。分析師所出具之報告，對證券市場往往極具影響力，亦使之成為值得探討的議題。因此本人收集並整合財務分析師之資料，針對其行為進行實證分析，盼能幫助未來研究更進一步了解人際互動對經濟和社會體系之可能影響。本篇博士論文包含三篇實證文章，主要探討分析師的互惠策略、資訊收集與合作行為。\n\n第一篇文章探討分析師與企業管理者的互惠行為，此文章研究分析師是否會發表迎合討好企業的分析師報告，以及企業管理者如何透過給予資訊優勢來回饋這些分析師。本文發現，未實行互助的分析師，即便與企業高管有社交連結，也得不到企業高管的資訊反饋；而沒有社交連結的分析師，即便討好企業高管，也得不到資訊反饋。此外，在Regulation Fair Disclosure (Reg FD)法規實行前，互惠分析師只能在沒有提供管理階層盈餘預測的公司中，得到帶來優勢的資訊反饋。然而，在Reg FD法規實行後，互惠分析師所得的資訊優勢轉而僅存在有提供管理階層盈餘預測的公司中。在更進一步的Difference-in-differences-in-differences (DDD)分析中，我們發現在Reg FD實行後，這些互惠且有社交連結的分析師會在企業提供(停止)管理階層的盈餘預測後，得到(失去)資訊優勢。這樣的成果指出在Reg FD實行後，企業高管從私下提供特定資訊，轉為公開提供盈餘預測資訊，來協助有連結且互惠的分析師結合自身已收集的資訊馬賽克，來進一步拚奏出資訊優勢 (Cheynel and Levine, 2020)。\n\n第二篇文章更進一步探討，具備社交連結的分析師究竟從企業高管得到哪類型的資訊。研究證據顯示，與企業高管有社交連結的分析師，能夠在有連結之公司的產業皆提出更精確的盈餘預測與更高獲利的股票推薦。這樣的成果指出，分析師從社交連結中得到來自企業管理階層所提供的產業資訊。本文章亦探討，分析師的報告主要是仰賴與公司相關的特定資訊，亦或是涵蓋產業層面的通用資訊。本文章發現企業管理階層在Reg FD之後，從過去提供公司特定資訊給有社交連結的分析師，轉而提供產業資訊。這樣的結果也暗示了產業資訊的提供屬於未被Reg FD規範之灰色地帶，而這灰色地帶值得受到未來監管政策規劃時的重視。\n\n最後，第三篇文章探討分析師同僚間的合作行為。此文章延伸第二篇文章，探討分析師是否會將從社交關係中得到的產業資訊分享給同證券經紀商的分析師同僚。這樣的合作行為提高了分析師的研究報告品質，包含更精確的盈餘預測與獲利度更高的股票推薦。透過企業新聘一個與分析師有社交連結的高管作為分析師合作的外生事件，本研究揭露了一個清晰的知識傳遞鍊。進一步的分析發現，證券經紀商內若存在更高的協作成本與內部競爭，會弱化分析師同僚的合作關係。最後，本文發現從同僚得到的資訊，特別能幫助分析師解析資訊不確性高的企業。\n\n總結而言，本博士論文首先分析存在潛在互惠行為的兩種族群(分析師與企業經營階層)。研究證據指出互惠行為提高分析師報告的精確度，對市場帶來正面效益。再者，即便監管政策部分限制分析師與企業高管的私下資訊交流，企業高管依舊找到新的方法(例如:資訊馬賽克、產業資訊)來繞過規範，以便提供有社交連結的分析師資訊優勢。最後，本論文亦發現同儕之間的資訊分享與合作行為，能為分析師帶來更高的生產力。

People’s interaction is an indispensable part in the economic system. Among people’s interactions, there exist several important behaviors (e.g., reciprocity and cooperation) that could be influential for the economic system. It is interesting to note that when the government enacts and implements new laws, the reaction of people is sometimes unexpected. Since analyst reports are influential in the equity market and analysts’ interactions with firm manager are complicated, which even induce regulation, it becomes a topic worth studying. Therefore, I collect the analysts’ personal data and conduct empirical analyses. This dissertation aims to improve our understanding of how people’s interactions could influence society and the economic system. Specifically, this dissertation examines the favoring strategy, information acquirement, and cooperation of equity analysts in three essays.\n\nThe first essay studies the reciprocal relationship in which sell-side analysts curry favors with firm managers by speaking in two tongues and firm managers reciprocate with information edge. I demonstrate that relationships without reciprocity are not relationships at all, and that social connections keep such a reciprocal relationship viable. Strikingly, social-connected reciprocal analysts obtain information advantage only in stocks without management guidance before Reg FD. In sharp contrast, they only do so in stocks with management guidance after Reg FD. My difference-in-difference-in-difference (DDD) tests based on the initiation and cessation of quarterly management guidance in the post-FD period reveal that social-connected reciprocal analysts` information advantage increases (decreases) after firms initiate (stop) quarterly management guidance. These results suggest that after Reg FD, managers rely on public earnings guidance instead of private disclosure to reciprocate social-connected analysts by helping them complete information mosaic (Cheynel and Levine, 2020).\n\nThe second essay further investigates what kind of information is acquired by analysts through their social connection with the firm managers. The finding shows that relative to analysts without school ties, analysts with ties make more accurate forecasts and more profitable recommendations on non-tied firms in the same industry. This suggests that analysts acquire industry information through school ties. Furthermore, this essay also assesses the relative importance of firm-specific vs. industry information reflected in connected analysts’ forecasts and recommendations. Moreover, the evidence shows that after Reg FD, firm managers shift from privately disclosing firm-specific information to industry information, which is responsible for connected analysts’ continued (albeit smaller) information advantage. These results suggest selective disclosure of industry information is a legal gray area that deserves further regulatory attention.\n\nLast, the third essay studies cooperation among analysts. This essay’s evidence suggests that analysts cooperate through the sharing of industry information and knowledge with colleagues in their brokerage houses. Such shared industry knowledge significantly improves colleagues’ forecast accuracy and recommendation profitability. Quasi-experiment tests based on the change of exogenous information sources reveal a clear information sharing (knowledge spillover) channel, which indicates the cooperative behavior among analysts. Greater in-house competition and coordination costs will hamper cooperation and knowledge spillover. Knowledge spillover is more beneficial to analysts covering firms subject to greater information uncertainty.\n\nOverall, this dissertation first examines the potential reciprocal behaviors between two groups (i.e., sell-side analysts and firm managers). The evidence shows that reciprocity helps improve the information accuracy of analysts’ research reports, generally benefitting the market. Furthermore, although the Reg FD restricts the private communication between analysts and firm managers, firm managers still come up with new methods (e.g., information mosaic, industry information) to circumvent the regulation in order to provide information advantage to social-tied analysts. Last, cooperation and information sharing between colleagues could help improve the productivity of analysts.