在財務金融的領域上，如何準確的預測隔日的股票價格是很重要且具挑戰性的問題。傳統的預測方法，是使用低頻數據進行時間序列的建模與進行相對應的預測。此預測方法的優點是利用長時間的走勢來對未來數值進行評估；然缺點是需取長時間的資料來進行建模，故在波動產生劇烈變化時將無法準確地得到預測值。另一方面，可以使用高頻資料來進行預測，高頻資料能在一天內提供大量的交易資料來對波動有更好的估計，但卻無法展示出股票價格的長期走勢。本研究試圖將兩種方法結合在一起。首先，提出一個演算法來模擬出同時帶有高、低頻數據的金融交易數據，然後提出新的估計方法，同時利用到高低頻的數據。實證結果顯示，新提出來的方法，其預測能力可修正傳統最小平方法高達25%。

In the field of finance, how to accurately predict the stock price
is an important and challenging problem. The traditional prediction method is to use low-frequency data to model time series and make corresponding predictions. The advantage of this prediction method is to use long-term trends to evaluate future values; however, the disadvantage is that it takes a long time to model, so it will not be able to accurately obtain the predicted value when the fluctuation changes sharply. On the other hand, high-frequency data can be used to forecast. High-frequency data can provide a large amount of transaction data in one day to better estimate the volatility, but it cannot show the long-term trend of stock prices.This study attempts to combine the two methods. First, an algorithm is proposed to simulate the financial transaction data with both high and low frequency data. then a new estimation method is proposed, which uses high and low frequency data at the same time. The empirical results show that the prediction ability of the newly proposed method can modify the traditional least square method by up to 25%.

摘要 i
Extended Abstract ii
英文延伸摘要 ii
誌謝 v
目錄 vi
表格 vii
圖片 viii
第 1 章 緒論 1
第 2 章 生成模擬資料 3
第 3 章 模擬參數估計 6
3.1 低頻 AR(1)模型參數估計方法 6
3.2 高頻數據波動率的估計方法 6
3.3 高低頻結合後新的估計方法 9
第 4 章 模擬資料分析 11
第 5 章 實證研究 13
第 6 章 總結及未來研究方向 16
參考文獻 17
附錄 A.圖片 19
附錄 B.表格 25

[1] Anat R Admati and Paul Pfleiderer. A theory of intraday patterns: Volume and price variability. The Review of Financial Studies, 1(1):3–40, 1988.
[2] Yacine Ait-Sahalia, Per A Mykland, and Lan Zhang. How often to sample a continuous-time process in the presence of market microstructure noise. The Review of Financial Studies, 18(2):351–416, 2005.
[3] Torben G Andersen and Bollerslev. Deutsche mark–dollar volatility: intraday activity patterns, macroeconomic announcements, and longer run dependencies. The Journal of Finance, 53(1):219–265, 1998.
[4] Torben G Andersen and Tim Bollerslev. Answering the critics: Yes, arch models do provide good volatility forecasts. 1997.
[5] Torben G Andersen and Tim Bollerslev. Answering the skeptics: Yes, standard volatility models do provide accurate forecasts. International economic review, 39(4):885–905, 1998.
[6] Torben G Andersen, Tim Bollerslev, and Jun Cai. Intraday and interday volatility in the japanese stock market. Journal of International Financial Markets, Institutions and Money, 10(2):107–130, 2000.
[7] Federico M Bandi and Jeffrey R Russell. Separating microstructure noise from volatil-ity. Journal of Financial Economics, 79(3):655–692, 2006.
[8] Ole E Barndorff-Nielsen, P Reinhard Hansen, Asger Lunde, and Neil Shephard. Realized kernels in practice: Trades and quotes. The Econometrics Journal, 12(3):C1–C32, 2009.
[9] Ole E Barndorff-Nielsen, Peter Reinhard Hansen, Asger Lunde, and Neil Shephard. Designing realized kernels to measure the ex post variation of equity prices in the presence of noise. Econometrica, 76(6):1481–1536, 2008.
[10] Recep Bildik. Intra-day seasonalities on stock returns: evidence from the turkish stock market. Emerging Markets Review, 2(4):387–417, 2001.
[11] Tim Bollerslev. Generalized autoregressive conditional heteroskedasticity. Journal of econometrics, 31(3):307–327, 1986.
[12] Stephen Boyd and Lieven Vandenberghe. Convex optimization. Cambridge university press, 2004.
[13] David K Ding and Sie Ting Lau. An analysis of transactions data for the stock exchange of singapore: Patterns, absolute price change, trade size and number of transactions. Journal of Business Finance ＆ Accounting, 28(1-2):151–174, 2001.
[14] C. A. E. Goodhart and L. Figliuoli. Every minute counts in financial markets. Journal of International Money and Finance, 10(1):23–52, 1991.
17 [15] D Graupe, DJ Krause, and J Moore. Identification of autoregressive moving-average parameters of time series. IEEE Transactions on Automatic Control, 20(1):104–107, 1975.
[16] Steven L Heston. A closed-form solution for options with stochastic volatility with applications to bond and currency options. The Review of Financial Studies, 6(2):327–343,1993.
[17] Jean Jacod, Yingying Li, Per A Mykland, Mark Podolskij, and Mathias Vetter. Microstructure noise in the continuous case: the pre-averaging approach. Stochastic Processes and Their Applications, 119(7):2249–2276, 2009.
[18] Prem C Jain and Gun-Ho Joh. The dependence between hourly prices and trading vol-ume. Journal of Financial and Quantitative Analysis, 23(3):269–283, 1988.
[19] Liang-Ching Lin and Meihui Guo. Optimal restricted quadratic estimator of integrated volatility. Annals of the Institute of Statistical Mathematics, 68(3):673–703, 2016.
[20] A Low and Jayaram Muthuswamy. Information flows in high frequency exchange rates.
C. Dunis, 28:3–32, 1996.
[21] Thomas H McInish and Robert A Wood. Intraday and overnight returns and day-of-the-week effects. Journal of Financial Research, 8(2):119–126, 1985.
[22] Gilbert Thomas Walker. On periodicity in series of related terms. Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character, 131(818):518–532, 1931.
[23] Jonathan H Wright. A new estimator of the fractionally integrated stochastic volatility model. Economics Letters, 63(3):295–303, 1999.
[24] George Udny Yule. Vii. on a method of investigating periodicities disturbed series, with special reference to wolfer’s sunspot numbers. Philosophical Transactions of the Royal Society of London. Series A, Containing Papers of a Mathematical or Physical Character, 226(636-646):267–298, 1927.
[25] Lan Zhang. Efficient estimation of stochastic volatility using noisy observations: A multi-scale approach. Bernoulli, 12(6):1019–1043, 2006.
[26] Lan Zhang, Per A Mykland, and Yacine Aït-Sahalia. A tale of two time scales: Determining integrated volatility with noisy high-frequency data. Journal of the American Statistical Association, 100(472):1394–1411, 2005.