發(fā)布時(shí)間：2019-07-08 10:38

【摘要】：本文針對(duì)匯率問(wèn)題的復(fù)雜性,根據(jù)匯率與宏觀經(jīng)濟(jì)變量的非線(xiàn)性關(guān)系和匯率數(shù)據(jù)自身的非線(xiàn)性特征,以神經(jīng)網(wǎng)絡(luò)技術(shù)為主要工具,研究了人民幣匯率基于購(gòu)買(mǎi)力平價(jià)理論的預(yù)測(cè)和時(shí)間序列預(yù)測(cè)問(wèn)題,并建立了一個(gè)基于神經(jīng)網(wǎng)絡(luò)的人民幣匯率風(fēng)險(xiǎn)管理的VaR 模型。 本文應(yīng)用神經(jīng)網(wǎng)絡(luò)考察匯率與物價(jià)指數(shù)之間的非線(xiàn)性關(guān)系。在應(yīng)用協(xié)整技術(shù)檢驗(yàn)發(fā)現(xiàn)購(gòu)買(mǎi)力平價(jià)理論對(duì)人民幣匯率不成立的情況下,論文建立了一種基于神經(jīng)網(wǎng)絡(luò)的非線(xiàn)性協(xié)整檢驗(yàn)方法,對(duì)匯率與物價(jià)指數(shù)進(jìn)行了非線(xiàn)性協(xié)整檢驗(yàn)。并在此基礎(chǔ)上建立了預(yù)測(cè)模型。 論文在分析人民幣匯率收益序列長(zhǎng)記憶性的基礎(chǔ)上,建立了神經(jīng)網(wǎng)絡(luò)預(yù)測(cè)模型。引入方向準(zhǔn)確率作為預(yù)測(cè)模型的一個(gè)評(píng)價(jià)指標(biāo),對(duì)模型的預(yù)測(cè)結(jié)果進(jìn)行了統(tǒng)計(jì)檢驗(yàn)。并考察了計(jì)量經(jīng)濟(jì)模型的選擇標(biāo)準(zhǔn)中的信息準(zhǔn)則法,在建立神經(jīng)網(wǎng)絡(luò)預(yù)測(cè)模型過(guò)程中的作用。 論文根據(jù)神經(jīng)網(wǎng)絡(luò)技術(shù)中的混合密度網(wǎng)絡(luò)能夠預(yù)測(cè)數(shù)據(jù)之間的條件概率密度函數(shù)這一特點(diǎn),對(duì)外匯資產(chǎn)組合收益的條件密度函數(shù)進(jìn)行了預(yù)測(cè),并在預(yù)測(cè)的基礎(chǔ)上建立了一個(gè)VaR 模型。 論文比較系統(tǒng)地分析了人民幣匯率的特征,考察了應(yīng)用神經(jīng)網(wǎng)絡(luò)技術(shù)進(jìn)行匯率預(yù)測(cè)的建模過(guò)程和模型選擇技術(shù),討論了對(duì)人民幣匯率進(jìn)行有效預(yù)測(cè)和風(fēng)險(xiǎn)管理的可行性。這些工作,將有助于拓寬對(duì)人民幣匯率問(wèn)題的研究思路,為解決人民幣匯率預(yù)測(cè)和風(fēng)險(xiǎn)管理問(wèn)題提供了一些可供選擇的方法。
文內(nèi)圖片：
圖片說(shuō)明：-1資本市場(chǎng)的均衡圖中ME表示貨幣市場(chǎng)的均衡曲線(xiàn)，，它是向上傾斜的
[Abstract]:In view of the complexity of exchange rate problem, according to the nonlinear relationship between exchange rate and macroeconomic variables and the nonlinear characteristics of exchange rate data itself, this paper studies the prediction and time series prediction of RMB exchange rate based on purchasing power parity theory with neural network technology as the main tool, and establishes a VaR model of RMB exchange rate risk management based on neural network. In this paper, neural network is used to investigate the nonlinear relationship between exchange rate and price index. When it is found that purchasing power parity theory is not valid for RMB exchange rate by using cointegration technique, a nonlinear cointegration test method based on neural network is established in this paper, and the nonlinear cointegration test of exchange rate and price index is carried out. On this basis, the prediction model is established. Based on the analysis of the long memory of RMB exchange rate return series, a neural network prediction model is established in this paper. The direction accuracy is introduced as an evaluation index of the prediction model, and the prediction results of the model are statistically tested. The role of the information criterion method in the selection criteria of econometric models in the process of establishing neural network prediction model is also investigated. According to the characteristic that the mixed density network in neural network technology can predict the conditional probability density function between data, this paper forecasts the conditional density function of foreign exchange asset portfolio income, and establishes a VaR model on the basis of prediction. This paper systematically analyzes the characteristics of RMB exchange rate, investigates the modeling process and model selection technology of applying neural network technology to exchange rate prediction, and discusses the feasibility of effective prediction and risk management of RMB exchange rate. These work will help to broaden the research ideas of RMB exchange rate and provide some alternative methods to solve the problems of RMB exchange rate forecasting and risk management.
【學(xué)位授予單位】：吉林大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2005
【分類(lèi)號(hào)】：F832.6

【引證文獻(xiàn)】

相關(guān)期刊論文 前1條

1 張晨曦;楊一文;;基于混合密度網(wǎng)絡(luò)測(cè)度股市流動(dòng)性“周內(nèi)效應(yīng)”[J];中國(guó)證券期貨;2010年09期

相關(guān)碩士學(xué)位論文 前9條

1 鄭東;我國(guó)商業(yè)銀行匯率風(fēng)險(xiǎn)形成機(jī)制及其管理研究[D];廣東外語(yǔ)外貿(mào)大學(xué);2006年

2 張靜;開(kāi)放條件下我國(guó)商業(yè)銀行匯率風(fēng)險(xiǎn)管理問(wèn)題研究[D];湖南大學(xué);2006年

3 鄧雄;我國(guó)商業(yè)銀行匯率風(fēng)險(xiǎn)管理研究[D];西南財(cái)經(jīng)大學(xué);2007年

4 王映喬;基于神經(jīng)網(wǎng)絡(luò)的匯率預(yù)測(cè)及系統(tǒng)設(shè)計(jì)[D];西南財(cái)經(jīng)大學(xué);2007年

5 周文;基于密度預(yù)測(cè)的人民幣匯率組合預(yù)測(cè)研究[D];中南大學(xué);2006年

6 姜瑩;匯率體制轉(zhuǎn)軌時(shí)期我國(guó)商業(yè)銀行外匯風(fēng)險(xiǎn)管理研究[D];南京師范大學(xué);2008年

7 馬廣元;我國(guó)商業(yè)銀行匯率風(fēng)險(xiǎn)管理研究[D];西南財(cái)經(jīng)大學(xué);2009年

8 盧寧;企業(yè)外匯風(fēng)險(xiǎn)防范決策支持系統(tǒng)[D];河南科技大學(xué);2012年

9 侯英超;基于VaR的中國(guó)股指期貨風(fēng)險(xiǎn)實(shí)證研究[D];北方工業(yè)大學(xué);2013年

本文編號(hào)：2511530