以往研究主要在探討財務資訊對於財務分析師的影響，相對的，本研究計劃則在探 討非財務資訊—專利權，與證券分析師之間的關係。本計畫提出兩項議題：(1)專 利權屬性對於分析師決策(盈餘預測與股票推薦)的影響，本文預期：a.專利權數量 愈多或品質(引證與請求權)愈佳的公司，分析師愈會發佈有利的盈餘預測與推薦； b.專利權原創程度(引證以往其他專利中，屬於不同專利權分類碼的比例)愈高，分 析師愈會發佈有利的盈餘預測與推薦；c.產業創新的外溢程度高於競爭性時，或是 企業接受外溢程度的能力愈強，分析師愈會發佈有利的盈餘預測與推薦。(2) 專利 權屬性對於分析師間的預測一致性(指:各分析師之間預測誤差的相關性)、私有訊息 使用程度的影響，本文預期：a.專利權原創程度愈高的公司，分析師的一致性愈小， 此時分析師私有訊息的使用量愈大；b. 專利權技術複雜性愈高的公司，分析師的 一致性愈小，此時分析師私有訊息的使用量愈大。 This project will explore the association between non-financial information (especially patent) and the analysts』 decisions, as compared to prior study which examines the association of financial information with analyst』 decision. I address two primary questions: (1) the effect of properties pf patent on analyst』s earnings forecasts and stock recommendation. I expect that analysts are more likely to issue favorable forecasts with respect to future earnings and stock recommendation for firms with more patent granted and higher quality of patents, as proxy by numbers of claims and citation record. I also hypothesize that the associated between patent and the likelihood of issuing more favorable forecast and recommendation is more significant for firms investing in more original technologies, measured as percentage of citations made by firms』 patents that belong to a broader set of technologies, and for firms in the setting with more inter-firms spillovers, firms with the better ability of to receive effect of spillovers. (2) the effect of properties of patent on analyst』s consensus, measured as the correlation in analysts』 forecast errors. I expect that the degree to which the mean forecast aggregates private (idiosyncratic) information and is more accurate than an individual analyst』s forecast increases with the originality of patented inventions and diversity of firms』 technology of patents portfolio.