在時間序列資料中，已經有非常多的學者證實有結構性變化的問題。此問題對於預測上來說，是一項非常重要的議題，因為使用結構性變化前的資料去預測，會降低其預測的準確度。針對上述問題，本文採用依時變動參數模型(time-varying parameters model)，還有Inoue et al.(2017)所提出的計量方法，試圖找出最適的樣本數以用來預測，此方法是藉由極小化損失函數(loss function)，以得到最接近實際狀況的預測值。在實證分析上，本文使用上述計量方法，以預測臺灣經濟成長率，分別加入24種外生變數，每次只放入一種外生變數，嘗試去找出哪些變數是適合用來預測經濟成長率，藉由與主計處公告之預測經濟成長率，去做預測表現的比較。實證結果顯示，加入下述任一外生變數，其預測狀況是優於主計處的預測情形，包含了各類貨品外銷訂單金額、製造業銷售量指數、工業生產指數-總指數、M1B(月底數)、同時指標與領先指標。分別放入以上六種外生變數，其預測狀況會優於主計處。

It’s been confirmed by many scholars that there is an issue regarding structural variations in time-series data, and it has shown significance in terms of prediction, due to the fact that it would compromise the predictive accuracy if the research was conducted with pre-structural variation data. As the issue mentioned above, the paper has applied time-varying parameters model and the econometric measurement brought by Inoue et al.(2017), trying to find out the most suitable window size for the purpose of prediction by the loss function approach in order to get a close predictand to reality as possible. On empirical analysis, the paper used the same econometric measurement to predict Taiwan’s economic growth rate, adding one exogenous variable at a time for a total of 24 in conclusion, to determine which ones were relatively suitable, as taking the predictive economic growth rate announced by the Directorate-General of Budget as comparison. The result has indicated that if anyone of the following exogenous variable were included, such as the different products’ Value of Export Orders, the Manufacturing Sales Index, the Industrial Production-Total Index, the MIB, Coincident Indicator and Leading Indicators, then the predictive circumstance would have an advantage over the one that's releasing by the Directorate-General of Budget.

論文審定書. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
誌謝. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ii
摘要. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .iii
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
圖目錄. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .vii
表目錄. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .viii
第一章緒論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 研究動機與目的. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 研究架構. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
第二章文獻回顧. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1 結構性變化模型之回顧. . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 結構性變化檢定文獻之回顧. . . . . . . . . . . . . . . . . . . . . . . 7
2.3 結構性變化實證文獻之回顧. . . . . . . . . . . . . . . . . . . . . . . 8
第三章研究方法. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1 依時變動參數模型. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2 依時變動參數模型的假設. . . . . . . . . . . . . . . . . . . . . . . . .11
3.3 依時變動參數模型的優點. . . . . . . . . . . . . . . . . . . . . . . . .12
3.4 依時變動參數模型的推導. . . . . . . . . . . . . . . . . . . . . . . . .13
3.4.1 運用局部線性法估計參數. . . . . . . . . . . . . . . . . . . . . 14
3.4.2 尋找最適樣本數. . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.5 結構性變化檢定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.5.1 sup F 檢定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..16
3.5.2 Double Maximum 檢定. . . . . . . . . . . . . . . . . . . . . . 17
3.6 單根檢定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 18
3.6.1 DF 單根檢定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.6.2 ADF 單根檢定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.7 落後期數的決定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
第四章實證分析與結果. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.1 資料來源與處理. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..23
4.2 預測範圍. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.3 實證方法. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.4 結構型變化檢定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..24
4.5 實證分析與比較. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..27
第五章結論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
參考文獻. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
附錄A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 37
A.1 資料處理. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 37
A.2 Dmax 檢定結果. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 37
A.3 WDmax 檢定結果. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
A.4 DM test 檢定結果. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40

何金巡、林建甫、于宗先，2008。《主計處總體經濟計量模型的運用》。
嘉義； 國立中正大學。《國立中正大學經濟系含國際經濟研究所》，
<http://econ.ccu.edu.tw/academic/master—paper/080531seminar.doc>。
李慶男，2005。《時間序列講義》。高雄；國立中山大學。《國立中山大學經
濟學研究所》，<http://econ.nsysu.edu.tw/files/11-1124-1342.php>。
林輝銘，2005。《卡爾曼濾波器結合類神經網路在時間序列預測之研究》。
桃園；元智大學工業工程與管理學系碩士論文。
徐士勛、管中閔、羅雅惠，< 以擴散指標為基礎之總體經濟預測>，《臺灣
經濟與政策》，第36 卷第1 期，2005 年10 月，頁1-28。
陳旭昇，2013。《時間序列分析：總體經濟與財務金融之應用二版》。臺北:
東華書局。
康銘顯，2001。《台灣地區貨幣定義新詮釋Time -Varying Parameter 計量模
型之應用》。臺北；東吳大學經濟學系博士論文。
張職揚，2006。《檢驗結構性轉變在網路產品之影響-Chow，CUSUM，
STAR 方法之比較》。高雄；國立中山大學經濟學研究所碩士在職專班碩士
論文。
黃于珍，2007。《實質匯率之結構改變；亞太地區之實證研究》。高雄；國
立中山大學經濟學研究所碩士在職專班碩士論文。
Andrews, D.W.K. (1993).‘Test for parameter instability and structural change with
unknown change point’, Econometrica, Vol. 61, pp. 821–856.
Baudin, A., Nadeau, S. and Westlund, A. (1984).‘Estimation and prediction under
structural instability: The case of the U.S. pulp and paper market’, Journal of Forecasting,
Vol. 3, pp. 63–78.
Bai, J., Perron, P. (1998).‘Estimating and testing linear models with multiple structural
changes’, Econometrica, Vol. 66, pp. 47–78.
Brown, R.L., Durbin, J. and Evans, J.M. (1975).‘Techniques for testing the constancy
of regression relationships over time with comments’, Journal of the Royal
Statistical Society, Vol. 37, pp. 149–192.
Breusch, T.S., Pagan, A.R. (1979).‘A simple test for heteroscedasticity and random
coefficient variation’, Econometrica, Vol. 47, pp. 1287–1294.
Chow, G.C. (1960).‘Test of equality between sets of coefficients in two linear regressions’,
Econometrica, Vol. 28, pp. 591–605.
Cooley, T.F., Prescott, E.C. (1973a).‘An adaptive regression model’, International
Economic Review, Vol. 14 , pp. 364–371.
Cooley, T.F., Prescott, E.C. (1973b).‘Test of an adaptive regression model’, Review
of Economics and Statistics, Vol. 55 , pp. 248–256. 364–371.
Cooley, T.F., Prescott, E.C. (1976).‘Estimation in the presence of stochastic parameter
variation’, Econometrica, Vol. 44 , pp. 167–184.
Estrella, A., Rodrigues, A.P. and Schich, S. (2003).‘How stable is the predictive
power of the yield curve? Evidence from Germany and the United States’, Review
of Economics and Statistics, Vol. 85, pp. 629–644.
Ghysels, E., Alain, G. and Alastair, H. (1997)‘.Predictive tests for structural change
with unknown breakpoint’, Journal of Econometrics, Vol. 82, pp. 209–233.
Giraitis, L., Kapetanios, G. and Price, S. (2013).‘Adaptive forecasting in the presence
of recent and ongoing structural change’, Journal of Econometrics, Vol. 177,
pp. 153–170.
Hamilton, J.D. (1994). Time Series Analysis, Princeton:Princeton University Press.
Inoue, A., Jin, L. and Rossi, B. (2017).‘Rolling window selection for out-of-sample
forecasting with time-varying parameters’, Journal of Econometrics, Vol. 196, pp.
55–67.
Kalman, R.E. (1960).‘A new approach to linear filtering and prediction problems
’, Journal of Basic Engineering, Transactions of the ASME Series D, Vol. 82, pp.
35–45.
Lucas, Jr. R.E. (1976).‘Econometric policy evalution: a critique’, Carnegie-Rochester
Conference Series on Public Policy, Vol. 1, pp. 19–46.
Nyblom, J. (1989).‘Testing for the constancy pf parameters over time’, Journal of
the American Statistical Association, Vol. 84, pp. 223–230.
Pesaran, M.H., Timmermann, A. (2007).‘Selection of estimation window in the
presence of breaks’, Journal of Econometrics, Vol. 137, pp. 134–161.
Pesaran, M.H., Pick, A. and Pranovich, M. (2013).‘Optimal forecasts in the presence
of structural breaks’, Journal of Econometrics, Vol. 177, pp. 134–152.
Ploberger, W., Kramer, W. (1992).‘The CUSUM test with OLS residuals’, Econometrica,
Vol. 60, pp. 271–286.
Quandt, R.E. (1958).‘The estimation of the parameters of a linear regression system
obeying two separate regimes’, Journal of the American Statistical Association,
Vol. 53, pp. 873–880.
Quandt, R.E. (1960).‘Test of the hypothesis that a linear regression system obeys
two separate regimes’, Journal of American Statistical Association, Vol. 55, pp.
324–330.
Rosenberg, B. (1972).‘The estimation of stationary stochastic regression parameters
re-examined’, Journal of the American Statistical Association, Vol. 67, pp. 650–
654.
Rosenberg, B. (1973).‘Random coefficients models: the analysis of a cross section
of time series by stochastically convergent parameter regression’, Annals of Economic
and Social Measurement, Vol. 2, pp. 399–428.
Schrimpf, A., Wang, Q. (2010).‘A reappraisal of the leading indicator properties of
the yield curve under structural instability’, International Journal of Forecasting,
Vol. 26, pp. 836–857.
Sessions, D.N., Chatterjee, S. (1989).‘The combining of forecasts using recursive
techniques with non-stationary weights’, Journal of Forecasting, Vol. 8, pp. 239–
251.
Stock, J.H., Watson, M.W. (1996).‘Evidence on structural instability in macroeconomic
time series relations’, Journal of Business & Economic Statistics, Vol. 14,
pp. 11–30.
Stock, J.H., Watson, M.W. (2003). ‘Forecasting output and inflation: the role of
asset prices’, Journal of Economic Literature, Vol. 41, pp. 788–829.
Young, P.C., Ng, C.N., Lane K., and Parker, D. (1991). ‘Recursive forecasting,
smoothing and Seasonal adjustment of non-stationary environmental data’, Journal
of Forecasting, Vol. 10 , pp. 57–89.