在財務理論中，Ross(1976)提出“套利定價理論“ (Arbitrage Pricing Theory，簡稱APT)，試著以較為細膩的角度去詮釋股票報酬率的行為。APT認為會影響股票報酬率的行為是由多個因素所決定，而非如資本資產定價模型(CAPM)所表示的，只有市場風險因素造成。統計多變量分析的因子分析中提供了一個方向，希望能夠藉此找出適合的因子來詮釋與分析，以對股票報酬率的行為與特性有更多了解。
傳統因子分析都假設因子皆獨立且具有 分布。但是股票的超額報酬率往往具有強烈的非均齊變異(Heteroscedasticity)。本文探討因子在具有條件非均齊變異情況下，介紹參數的估計方法，以及在估計因子風險價格上所遇到的問題，並利用EM演算法改善概似估計法在風險價格上的缺失。最後將模型配適在台灣證券市場中各類股票與三高電子股的超額報酬率上，以實證結果說明股票超額報酬率之行為與特性。

1-1 研究背景與動機 …………………………….……1

1-2 研究目的……………………………….……….….2

1-3 論文內容概述………………………….……….….2

第二章 文獻回顧

2-1 套利定價模式………………………..….…………3

2-2 非均齊變異過程……………………….……….….4

2-3 FACTOR-ARCH模型…………………..…………5

2-4 條件非均齊變異之因子模型

2-4-1 模型架構……………………………………….7

2-4-2模型探討………………………………………..8

2-4-3風險價格討論…………………………………..9

第三章 研究方法與討論

3-1 最大概似估計…………………………………….10

3-2 因子之估計…………. ..………………………….11

3-3 EM 演算法……………………………………….12

3-4 模擬試驗結果與討論……………………….…..15

第四章 實證結果與分析

4-1 台灣證券市場股票之實證研究…………….…..17

4-2 三高電子股之實證研究…………………….…..21

第五章 結論與建議 …………………………………..35

參考文獻……………………………………………………36

Antonios, A. & Ian, G. & Richard, P. (1998):”Macroeconomic variables as common pervasive risk factors and the empirical content of the arbitrage pricing theory,” Journal of Empirical Finance ,5,221-240
Bollerslev, T. (1986) :”Generalized Autoregressive Conditional Heteroskedasticity,” Journal of Econometrics ,51,307-327
Bollerslev, T. & Chou, R.Y. & Kroner, K.F. (1992):”ARCH modeling in Finance:A review of the theory and empirical evidence,” Journal of Econometrics ,52, 5-59
Demos, A. & Sentana, E. (1998):”An EM Algorithm for Conditionally Hetroscedastic Factor Models,” Journal of Business and economic statistics ,16, 357-361
Diebold, F. & Nerlove, M. (1989):”The Dynamics of Exchange Rate Volatility:A Multivariate Latent Factor ARCH Model,” Journal of Applied Econometrics,4,1-21
Engle, R. F. (1982):”Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation,” Econometrica , 987-1006
Engle, R. F. & Ng, V. K. & Rothschild, M.(1990) :“Asset Pricing With a Factor-ARCH Structure :Empirical Estimates for Treasury Bills,” Journal of Econometrics ,45, 213-254
Harvey, A.C. (1989):”Forecasting,structural models and the Kalman filter” Combridge University Press , Combridge
Harvey, A.C. & Ruiz, E. & Sentana, E. (1992): “Unobvervable Component Time Series Models With ARCH Disturbances,” Journal of Econmetrics,52,129-158
King, M. & Sentana, E. & Wadhwani, S. (1994) :“Volatility And Links Between National Stock Markets,” Econometrica ,62, 901-933
Sentana, E. (1998):”The Relation between Conditionally Heteroskedastic Factor Models and Factor-GARCH Models,” Econometrics Journal ,1,1-9
Sentana, E. & Fiorentini, G. (1999):”Identification,Estimation and Testing of Conditionally Heteroskedastic Factor Models,” CEMFI unpublished
Sentana, E. (1999):”The Likelihood Function of Conditionally Heteroskedastic Factor Models,” CEMFI unpublished
Sentana, E. (2000):”Factor Representing Portfolios in Large Asset Markets,” CEMFI unpublished
Shephard, N. (1996):”Statistical aspects of ARCH and stochastic volatility,”
Time Series Models in Econometrics,Finance and uther Fields
金崇賢 (1998) : “Arch 效應對因子分析的影響” 清華大學統計所碩士論文