IC設計業財務五力績效評估及股價預測模型之實證研究

Exploring the Performance Evaluation of Five Financial Constructs and Empirical Stock Prices Forecasting Model on IC Design Industry

劉怡君(Yi-Jiun Liu)

指導教授 ： 古永嘉

國立臺北大學/商學院/企業管理學系/碩士(2008年)

摘要

台灣的IC設計業產值已成為僅次於美國的全球第二大重鎮，就產業面而言，IC設計業是台灣最具發展條件的重點產業之ㄧ，因此期能以財務五力評估IC設計業之經營績效，並建構以財務五力績效評估結果對股價之預測模型。以往學術研究財務比率對經營績效的關聯性，多以少數指標或部分財務五力構面的關係分析；實務與學術界對股市資訊揭露的財務五力指標和股價報酬的關係，均尚未見完全的實證(蔡聰源，民國95年)。本研究以財務五力模式評估IC設計業之經營績效，並以財務五力之績效評估結果進行股價走勢預測。
本研究透過財務報表及及公開資訊觀測站取得55家上市櫃IC設計業者五年(2002年~2006 年 )之34項財務比率，進行財務比率之變數標準化程序，以計算財務五力F1~F5，再以集群分析及區別分析檢定分群準確度，於區分為四群時之預測準確度可達94.25%，故以分四群進行變異數分析(包含GLM平均數檢定及DUNCAN多重比較檢定)，判定各群組之財務五力特性，最後歸納各IC設計業者於研究期間之群組變化，以評估各IC設計業者之經營績效及績效群組之變化趨勢。另一方面，藉由迴歸分析建構以本益比為財務五力之依變數的股價預測模型，比對財務五力之績效評估結果是否可預測次年度股價走勢。
本研究實證結果彙總如下：
1、以財務五力評估IC設計業之經營績效是有效的，區別分析檢定分
群於區分為四群時之預測準確度可達94.25%。上市櫃IC設計業之
績效可分四群為：生產力中上餘四構面最佳組、生產力最佳餘四構面中上組、生產力最差餘四構面中下組、生產力中下餘四構面最差組。
2、 IC設計業者當年度財務五力分群為生產力中上餘四構面最佳組者，其對次年度股價走勢預測之中率接近七成。將IC設計業於研究期間五年之財務五力績效變化進行評價，評價結果分為四等級：優、中上、中下及差，其中績效評價為優之IC設計業者，其財務五力績效評估結果可預測次年度股價走勢之命中率約為七成，其他績效評估等級的企業於預測股價走勢則較不顯著。

關鍵字

財務五力 ； 績效評估 ； IC設計業 ； 股價預測模型

並列摘要

The output value of Taiwan’s IC design industry has enabled Taiwan as the second most important IC nation worldwide, next only to the United States of America. From the industrial perspective, the IC design industry is one of the key industries in Taiwan that possesses the best condition for development. Therefore, the five financial constructs is used in order to evaluate the IC design industry’s operating performance. A stock price forecast model is also constructed based on the evaluation results of the five financial constructs. In the past, when academic research attempted to examine the correlation between financial ratios to operating performance, usually only a few indicators or portions of the five financial constructs aspect for relation analysis are utilized. A complete five financial constructs about the relationships with stock returns of information disclosure in the stock market has not yet been worked empirically by practitioners and academics (Tsai, Tsung-Yuan 2006). This research utilizes the Five Financial Constructs Analysis Model to evaluate the operating performance of the IC design houses and uses the evaluation result of the Five Financial Constructs Analysis to forecast the fluctuation of the stock price.
Through financial reports and public market observation information, this research obtained 34 sections of financial ratios out of 55 publicly offered IC design houses covering a range of 5 years (2002 to 2006). The process of standardization of variables of financial ratios was performed to calculate the Five Financial Constructs F1~F5. Cluster analysis and discriminant analysis were then conducted to determine the accuracy of the cluster. When categorized into 4 clusters, the accuracy of the forecast reaches 94.25%. Therefore, 4 clusters were used to conduct the variance analysis (including GLM mean test and DUNCAN multiple comparisons test) to determine each cluster’s Five Financial Constructs trait. Lastly, each IC design houses was summed up into cluster changes during the research period to evaluate each IC design house’s operating performance and the changing trend of the performance cluster. On the other hand, through aspect of the regression analysis, the price earning ratio is used as the dependent variable in the Five Financial Constructs stock price forecast model to verify whether the result from the Five Financial Constructs performance evaluation can be used to forecast this year’s stock price movement.
The result collected by this study based on factual evaluation is illustrated below:
1. The use of Five Financial Constructs to evaluate the operating performance of the IC design houses is effective. When the discriminant analysis test separates the clusters into 4 clusters, the accuracy of the forecast reaches 94.25%. The performance of the publicly offered IC design houses can be classified into the following 4 clusters: those with the best 4 perspectives but upper middle production capability , those with upper middle 4 perspectives but best production capability, those with lower middle 4 perspectives but worst production capability, and finally those with the worst 4 perspectives but lower middle production capability.
2. For the IC design houses with the best 4 perspectives but upper middle production capability for its financial affairs of that year, the accuracy on the forecast of their stock prices for the following year is close to seventy percent. When evaluation is performed on the changes of the five financial constructs performance of the IC design houses during the five years period examined by this research, the evaluation result can be classified into four grades: excellent, above average, below average and poor. Amongst those, IC design houses with a performance evaluation of “excellent,” the evaluation result of their five financial constructs can forecast the price fluctuation of their stock for the following year with an accuracy of approximately seventy percent. The forecast of stock price fluctuation of other enterprises with other evaluation ratings were not as significant.

並列關鍵字

Five Financial Constructs ； Performance Evaluation ； IC Design Industry, ； Stock Prices Forecasting Model

參考文獻

8.施思佳，電子業財務危機預警模式之研究－以現金流量觀點，國立台北大學 企業管理研究所碩士在職專班碩士論文，民國91年。9.章長原，全球半導體市場趨勢與台灣IC設計產業經營績效之關聯性研究，中原大學企業管理研究所碩士論文，民國93年7月。

12.許政宏，研究發展、財務指標與經營績效之關聯性研究－以台灣上市上櫃IC設計產業為例，中原大學國際貿易研究所碩士論文，民國96年。

19.劉以德，企業經營績效、財務槓桿與研究發展費用之關聯性研究，中原大學會計研究所碩士論文，民國90年。

1. Baginski, S. P., “Intra-industry Information Transfers Associated with Management Forecasts of Earnings,” Journal of Accounting Research, Chicago：Vol.25, Iss.2；Autumn 1987. p. 161.

2. Diakoulaki, D., G. Mavrotas, and L. Papayannakis, “Determining Objective Weight in Multiple Criteria Problems：The Critic Method, ” Computers & Operations Research, New York：Vol.22, Iss.7；Aug 1995. pp. 763-770 .