本文延伸傳統策略群組與移動障礙的觀點，透過動態策略群組分析法探討其進入與擴張，提出了「策略移動路徑」的觀點以呈現我國IC產業廠商長期的策略發展的動態歷程，並探討策略移動路徑對廠商績效的影響。過去文獻中探討進入策略對績效的影響，所使用的資料通常侷限於廠商進入產業的初期，而分析移動障礙及其影響時，多數文獻則僅就廠商目前的績效進行分析。本文同時考慮進入與擴張，連結了進入初期到最終期策略與績效的完整資訊，可彌補只顧頭或只顧尾的缺憾，對不同廠商間績效差異亦能有較佳的解釋力。 研究結果發現，在1994-2003年這段期間內，台灣IC產業共可區分出八期策略穩定時期，這驗證了高科技產業快速變動的特性。整體而言，所有廠商的策略可被區分成六個主要群組，早期進入的廠商多選擇移動障礙較低的群組作為初期進入產業的路徑，這與Bogner et al.(1996)的研究結果一致。本研究 亦發現，採取不同類型的策略移動路徑，其績效有顯著差異。採取間接策略的廠商，績效表現優於採取直接策略的廠商。若就策略變動的頻率、幅度分別分析其對廠商績效的影響，本研究發現策略變動頻率較高的廠商，其總資產報酬成長率的表現較佳，若策略變動幅度較大，則其營收成長率的表現相對較差。綜合本研究之結果可發現，處於激烈變動的產業環境，採取間接、快速反應但微幅調整的策略移動路徑，應是我國IC廠商長期策略的較佳選擇。

This article extends the viewpoint of traditional strategic groups and mobility barriers, and discusses entrance and expansion using dynamic strategic group analysis. It also proposes the "strategic moving path" viewpoint to present the dynamic process of long-term strategy development of semiconductor companies in Taiwan and discusses the impacts of strategic moving path on performance. Articles in the past that discuss the impact of entrance strategy on performance used only data in the initial period when the company first entered the industry. And, most articles only analyzed the current performance when studying dynamic barriers and their impacts. This article encompasses both entrance and expansion, and connects comprehensive information including strategy and performance from the initial period to the final period. In this way, the article makes up for the lack of information by focusing on the earlier or later period and provides better explanation for the difference in performance between different companies. The study shows that Taiwan semiconductor industry experienced 8 strategic stable time periods (SSTPs) from 1994 to 2003. This proves the fast changing nature of high technology industry. On the whole, the strategies of all companies can be divided into 6 main groups. Companies that entered the industry in the earlier period chose groups with fewer mobility barriers as the initial path to enter the industry. This finding is consistent with the result of Bogner et al. study (1996). This study also ascertains that there is significant difference in performance when different types of strategic moving paths are adopted. The performances of companies that adopt indirect strategies are better than those that adopt direct strategies. By analyzing the impacts of frequency and scale of strategy change company performance, the study discovers that companies with higher frequency of strategy change perform better in terms of asset return rate. If the scale of strategy change is great, the revenue growth rate will be comparatively poorer. Integrating the findings, this study shows semiconductor companies in Taiwan shall adopt an indirect, quick response, but slightly adjusted strategic moving path as their long-term strategy in an industry that experiences intense changes.