受学校国际合作与交流基金资助，应哈工大管理学院惠晓峰教授的邀请，美国康斯维星大学普莱维尔分校经济系Abdol Soofi于2007年8月19日-9月1日在我校访问、讲学。Abdol Soofi教授于8月22日-31日在哈工大科学园国际会议中心106报告厅做8场学术报告。本次讲学内容如下：
第1讲
题目：Time-delay Embedding and Local Linear Prediction Methods: A Tutorial
时间：2007年8月22日上午 9:00-11:30
地点: 哈工大科学园国际会议中心106报告厅
摘要: This lecture is a generally introduction to theory of time-delay embedding and nonlinear dynamical systems. Analyze the Lorenz system’s curve, fractal chaotic system is also discussed.
第2讲
题目:Mutual prediction and testing for nonlinear interdependency of stock markets
时间：2007年8月23日上午 9:00-11:30
地点: 哈工大科学园国际会议中心106报告厅
摘要: Sudden financial crisis that spreads from one country to another is commonly known as contagion. The review of the contagion literature indicates that sound methodologies in testing for dynamical interdependency between the financial markets either in its acute form during financial crisis (contagion) or for more stable periods do not exit. Moreover, the correlation method of testing for interdependence of markets (interdependence is a necessary condition for contagion to occur) is based on the assumption of linearity. The linear models when applied to nonlinear data may produce misleading results. The methodological shortcomings combined with nonlinearity of financial time series require use of nonlinear methods of testing for interdependency of dynamical systems. The work forming the foundation of the presentation takes contagion analysis from the linear domain into nonlinear domain. The talk provided the details of algorithms for:
1. Testing for nonlinearity of stock market indices using surrogate data analysis,
2. Testing for interdependency of the stock markets using nonlinear mutual prediction method.
第3讲
题目: Prediction and Volatility of Black Market Currencies: Evidence from Renminbi and Rail Exchange Rates
时间：2007年8月27日上午 9:00-11:30
地点: 哈工大科学园国际会议中心106报告厅
摘要: Out-of-sample prediction is performed on both fixed and black market Chinese renminbi/US dollar, and black market rial/US dollar exchange rates by using the time-delay embedding technique and the local linear prediction method. An artificially generated chaotic time series is also predicted with and without noise for the purpose of validation of the methods used in this study. In all examples tested the prediction results significantly outperform those by the benchmark mean value predictor based on a statistic defined by Harvey et al. (1997). Another interesting result found is that one may use the embedding dimension as a measure of volatility of a financial asset.
第4讲
题目: Measuring the Complexity of Currency Markets by Fractal Dimension Analysis
时间：2007年8月28日上午 9:00-11:30
地点: 哈工大科学园国际会议中心106报告厅
摘要: The theory of nonlinear dynamical systems is used to measure the complexity of currency markets by estimating the correlation dimension of the returns of the Dollar/Pound and Dollar/Yen daily exchange rates (the spot rates). The significance of the results is tested by comparing them to correlation dimension estimates for surrogate time series, i.e. stochastic linear time series with the same power spectrum and amplitude distribution as given by the original data. The discernible nonlinear structure is found in the returns of the Dollar/Pound daily rate.
第5讲
题目: Capital Account Liberalization, Risks, Financial Crises, and Risk Mitigations
时间：2007年9月1日上午 9:00-11:30
地点: 哈工大科学园国际会议中心106报告厅
摘要: After discussing Washington Consensus which is based on the assumptions of the neoclassical economic theory I argue that recent financial liberalization policies in many emerging economies may have led to financial crisis and massive economic dislocations in many of these liberalizing countries. These financial/economic difficulties mostly emerge because of the problems of asymmetric information. Furthermore, the theory of financial fragility which is inherent in market economies is discussed and methods of financial risk mitigation are introduced.
第6讲
题目: Capital Account Liberalization, Currency Markets’ Volatility and Prediction: Financial Implication of the P.R. of China’s Accession to the WTO
时间：2007年8月30日上午 9:00-11:30
地点: 哈工大科学园国际会议中心106报告厅
摘要: Out-of-sample prediction is performs on both fixed and black market Chinese renminbi/US dollar, and black market rial/US dollar exchange rates by using the time-delay embedding technique and the local linear prediction method. An artificially generated chaotic time series is also predicted with and without noise for the purpose of validation of the methods. In all examples tested, the prediction results significantly outperform those by the benchmark mean value predictor based on a statistic defined by Harvey et al. (1997). Another interesting result found is that one may use the embedding dimension as a measure of volatility of a financial asset.
第7讲
题目: Testing for Nonlinearity of Exchange Rates Using Surrogate Data: An Information Theoretic Approach
时间：2007年8月31日上午 9:00-11:30
地点: 哈工大科学园国际会议中心106报告厅
摘要: Nonlinear modeling of an exchange rate time series offers a clear alternative to modeling the series using a linear stochastic method. However, before fitting a nonlinear model to the observed data, one should have some indications that the data are generated by a nonlinear process. An information theoretic method and surrogate data analysis is used to test for nonlinearity of 4 daily dollar exchange rates. The nonlinear structures are observed in all currencies examined.
第8讲
题目: Testing for Long Memory in the Asian Foreign Exchange Rate
时间：2007年8月29日上午 9:00-11:30
地点: 哈工大科学园国际会议中心106报告厅
摘要: The plug-in and Whittle methods are used based on spectral regression analysis to test for the long memory property in 12 Asian/dollar daily exchange rates. The results according to the plug-in method show that with the exception of Chinese renminbi all series may have long memory properties. The results based on the Whittle method, on the other hand, show that only Japanese yen and Malaysian ringgit may have long memory properties.
It is well known that inference about the differencing parameter, d, in presence of structural break in a series entails considerable difficulties. Therefore, given the financial crisis of 1997-1998 in Asia, further tests for unraveling of the memory property and presence of structural break in the exchange rate series are required.
Abdol Soofi教授简历
Dr. Soofi is Professor, Department of Economics, obtained tenure in 1986, University of Wisconsin-Platteville. He is also the Director, Office of Institutional Research and Data Analysis, University of Wisconsin-Platteville from July 2001 to September 2004. From July 2002 to August 2002: Visiting Scholar, Academy of Mathematics and Systems Sciences, Chinese Academy of Science, Beijing , P.R. of China.
Dr. Soofi曾在中科院系统所与汪寿阳等人进行过合作，并曾多次指导过中国学者进行研究并在国际期刊中发表过论文。曾于2002年暑期在中科院系统所就汇率问题讲学。曾于2005年8月至哈工大进行为期一周的讲学和合作研究。
Dr. Soofi是Global Business and Economics Review期刊的联合编辑，并为National Science Foundation, Contemporary Economics, Economics Letters, International Journal of Forecasting, Mathematical and Computer Modeling, 和 Journal of Applied Econometrics.等多种著名期刊做审稿人。曾审阅和推荐中国作者的论文在国际学术期刊上发表。
国际合作处
2007年8月29日