教育部人文社科重点研究基地中央财经大学中国精算研究院学术活动
精算论坛第209期讲座
(2022年11月11日)
2022年11月11日中央财经大学中国精算研究院与麦考瑞大学精算研究和商业分析系联合举办线上研讨会。此研讨会涉及保险、养老、投资、消费等诸多领域，重在追踪国内外最新的研究成果，并加强国际合作。欢迎各位老师和同学参加和交流。
研讨会日程安排
时间：2022年11月11日（周五）10:00-15:15 （12：15-13：00休息）
地点：ZOOM会议（会议ID：862 8590 8558，密码：503971）
报告主题一：Worst-case Moments under Partial Ambiguity
报告时间：2022年11月11日10:00am–10:45am （北京时间）
报告人：Qihe Tang(University of New South Wales)
报告主题二：The expected length of stay at aged care facilities in Australia: Current and future
报告时间：2022年11月11日10:45am–11:30am （北京时间）
报告人：Colin Zhang (Macquarie University)
报告主题三：Claims reserving with a robust generalized additive model
报告时间：2022年11月11日11:30am–12:15pm （北京时间）
报告人：Yanlin Shi(Macquarie University)
Lunch/Coffee Break 12:15 pm -13:00 pm
报告主题四：A comparative analysis of several multivariate zero-inflated and zero-modified models with applications in insurance
报告时间：2022年11月11日13:00pm–13:45pm （北京时间）
报告人：Xueyuan Wu (University of Melbourne)
报告主题五：Optimal Asset Allocation For Households With Habit Formation
报告时间：2022年11月11日13:45pm–14:30pm （北京时间）
报告人：Jingzhen Liu (Central University of Finance and Economics)
报告主题六：Investment-consumption Decision with a Quasi-hyperbolic Discount Function and Dynamic Evaluations of Exit Probabilities
报告时间：2022年11月11日14:30pm–15:15pm （北京时间）
报告人：Huiling Wu (Central University of Finance and Economics)
报 告 详 情
报告主题一：Worst-case Moments under Partial Ambiguity
报告时间：2022年11月11日10:00am–10:45am （北京时间）
报告人：Qihe Tang (University of New South Wales)
摘要：The model uncertainty issue is pervasive in virtually all applied fields but especially critical in insurance and finance. To hedge against the uncertainty of the underlying probability distribution, which we refer to as ambiguity, a worst-case framework has been well developed during recent years. However, approaches in this framework often yield results that are overly conservative. We argue that in most practical situations a generic risk is realized from multiple scenarios. In some ordinary scenarios, the risk may be subject to negligible ambiguity so that it is safe to trust the reference distributions, and hence we can apply the worst-case approach only to the other scenarios where ambiguity is significant. We implement this idea to study the worst-case moments of a risk in the hope to alleviate the over-conservativeness issue. Note that under this consideration ambiguity exists in both the scenario indicator and the risk realization in the corresponding scenario, leading to a twofold ambiguity issue. We employ the Wasserstein distance to construct an ambiguity ball and then carefully disentangle the ambiguity along the two folds so as to link our optimization problem to established results in the worst-case framework. Our main result is a closed-form worst-case estimate for the moments. Our numerical studies illustrate that the consideration of partial ambiguity indeed greatly alleviates the over-conservativeness issue.
报告人简介：Professor Qihe Tang joined the UNSW Business School under the SHARP (Strategic Hires and Retention Pathways) scheme in July 2017. Prior to joining UNSW, he worked at the University of Iowa, where he was promoted to Full Professor in 2012 and conferred an Endowed Chair in 2014.
His expertise centres on extreme value theory for insurance, finance, and risk management. Recently, he has been working on various topics from the interdisciplinary field of insurance, finance, applied probability, and operations research. He has published over 100 papers resulting in an H index of 42 according to Google Scholar. His research has been constantly supported by major external grants including two recently awarded discovery grants from the Australian Research Council.
Currently, he is an Editor of the journal Insurance: Mathematics and Economics, and an Associate Editor of the journals Applied Stochastic Models in Business and Industry, Statistics & Probability Letters, and Science China Mathematics. He is an Elected Member of the International Statistical Institute.
报告主题二：The expected length of stay at aged care facilities in Australia: Current and future
报告时间：2022年11月11日10:45am–11:30am （北京时间）
报告人：Colin Zhang (Macquarie University)
摘要：This paper analyzes the hazard functions of exiting from an aged care facility in Australia. Using a comprehensive dataset ranging over 2008– 2018, we find that those functions are heterogeneous across the age, sex and year-of-leaving. The modelling results lead to in-sample estimated expected length of stay (LOS) for residents differed by age (in general, longer for older groups) and sex (longer for females). The estimated LOS declines gradually from 2008 to 2014 and then steadily increase afterwards for all heterogeneous age and sex groups. Out-of-sample predictions up to 2100 suggest that the longest LOS belongs to females aged 100 and older, with the estimated/predicted LOS increasing from 58.6 months in 2018 to 68.9 months in 2100. Relative uncertainty measures are also provided in this paper. Those results can be used to explore the nature of and aspects to improve service quality of Australian aged care facilities for policy makers.
报告人简介：Dr. Colin Zhang is a Lecturer in Department of Actuarial studies and Business Analytics. He has years of experience in delivering lectures for a broad range of Actuarial and Applied Finance units. Dr. Zhang holds a Bachelor of Commerce (Actuarial Studies), a Bachelor of Commerce (First class Honours in Actuarial studies) as well as a Degree of Doctor of Philosophy (Applied Finance and Actuarial Studies) from Macquarie University. His research area includes, life cycle model, optimal control, financial strategy and derivative pricing.
报告主题三：Claims reserving with a robust generalized additive model
报告时间：2022年11月11日11:30am–12:15pm （北京时间）
报告人：Yanlin Shi (Macquarie University)
摘要：In the actuarial literature, most existing stochastic claims reserving methods ignore the excessive effects of outliers. In practice, however, these extreme observations may occur in the upper triangle and can have a non-trivial and undesirable influence on the existing reserving models. In this paper, we consider the situation when outliers of incurred claims are present in the upper triangle. We demonstrate that the model fitting and prediction results of the classical chain ladder method can be substantially affected by these outliers. To mitigate this negative effect, we propose a robust generalized additive model (GAM). An associated robust bootstrap based on strati fi ed sampling is also developed to obtain a more reliable predictive bootstrap distribution of outstanding claims. Using both simulation examples and real-life data, we compare our proposed robust GAM with non-robust counterparts (and robust GLM). We demonstrate that the robust GAM provides comparable results with those of other models when outliers are not present, and that the robust GAM demonstrates significant improvements in estimation accuracy and efficiency when outliers are present.
报告人简介：Dr Shi is a Senior Lecturer at Macquarie University. His research interests include mortality modelling, investment risk modelling, non-life insurance rate making/reserving, and financial econometrics. He has published 60 journal articles in a wide range of actuarial, statistics, finance and related journals, such as Insurance: Mathematics and Economics, ASTIN Bulletin, Scandinavian Actuarial Journal, Demography, Journal of Banking and Finance, and Journal of Financial Econometrics.
Lunch/Coffee Break 12:15 pm -13:00 pm
报告主题四：A comparative analysis of several multivariate zero-inflated and zero-modified models with applications in insurance
报告时间：2022年11月11日13:00pm–13:45pm （北京时间）
报告人：Xueyuan Wu (University of Melbourne)
摘要：Claim frequency data in insurance records the number of claims on insurance policies during a finite period of time. Given that insurance companies operate with multiple lines of insurance business where the claim frequencies on different lines of business are often correlated, multivariate count modeling with dependence for claim frequency is therefore essential. Due in part to the operation of bonus-malus systems, claims data in automobile insurance are often characterized by an excess of common zeros. This feature is referred to as multivariate zero-inflation. In this paper, we establish two ways of dealing with this feature. The first is to use a multivariate zero-inflated model, where we artificially augment the probability of common zeros based on standard multivariate count distributions. The other is to apply a multivariate zero-modified model, which deals with the common zeros and the number of claims incurred in each line given that at least one claim occurs separately. A comprehensive comparative analysis of several models under these two frameworks is conducted using the data of an automobile insurance portfolio from a major insurance company in Spain. A less common situation in insurance is the absence of some common zeros resulting from incomplete records. This feature of these data is known as multivariate zero-deviation. In this case, our proposed multivariate zero-modified model still works, as shown by the second empirical study.
报告人简介：Dr. Xueyuan Wu obtained a B.Sc and an M.Sc from Nankai University, China. He completed his Ph.D. in Actuarial Science from the University of Hong Kong in 2004. Xueyuan joined the Centre for Actuarial Studies (CAS) at the University of Melbourne as a lecturer in January 2006 and was appointed as a senior lecturer in 2010 and an associate professor in 2020.
Xueyuan has a broad range of research interests. His early research field is discrete-time risk models and ruin theory with more than 10 peer-reviewed publications on discrete-time risk models with correlation, dividends, and so on. Xueyuan is also interested in discrete Phase-type distributions and recursive calculations in various aggregate claim models. Xueyuan's most recent research interests include public health and actuarial statistical modeling, in particular, applications of data analytics and machine learning techniques in non-life insurance problems.
报告主题五：Optimal Asset Allocation For Households With Habit Formation
报告时间：2022年11月11日13:45pm–14:30pm （北京时间）
报告人：Jingzhen Liu (Central University of Finance and Economics)
摘要：In this paper, we introduce habit formation into a family’s optimal strategy with investment, consumption and life insurance. We divide the original problem into two stages’ problem. In the first stage the couple are both alive and in the second stage only one of them is alive. By dynamic programming approach, the analytical solution can be obtained with CRRA utility function. Finally, the numerical experiments from the Chinese data illustrate the effect of habit formation on consumption, investment and insurance.
报告人简介：Jingzhen Liu is a Professor at China Institute for Actuarial Science (CIAS), Central University of Finance and Economics (CUFE), China. Professor Liu’s research interests include Actuarial Science, Financial Mathematics, Risk Management, Deep Learning. Jingzhen obtained a BSc in Mathematics from South China Normal University Guangzhou, China and an MPhil in Stochastic Process and its Application in Finance and Insurance at Nankai University, China. Professor Liu got a PhD in Financial Engineering from The Hong Kong Polytechnic University, Hong Kong, China.
报告主题六：Investment-consumption Decision with a Quasi-hyperbolic Discount Function and Dynamic Evaluations of Exit Probabilities
报告时间：2022年11月11日14:30pm–15:15pm （北京时间）
报告人：Huiling Wu (Central University of Finance and Economics)
摘要：This paper studies a multi-period investment-consumption problem where the discount function is quasi-hyperbolic and the exit probabilities are dynamically evaluated over time. That is, the decision maker at any time can adjust the exit probabilities from her current time to the terminal time. The Nash equilibrium investment-consumption strategy is obtained. The essential effects of the quasi-hyperbolic discount function and the dynamically evaluated exit probabilities on the investment-consumption strategy and the expected wealth are analyzed for different decision makers with different risk aversion and bequest motivation. It turns out that the quasi-hyperbolic discount function sometimes can change the classical properties of the investment-consumption proportion and the “hump saving'' phenomenon. Moreover, we find that in most cases the decision maker will consume more proportion of wealth if she or her successors will increase the probabilities of leaving the market before the terminal time. When the current decision maker does not allow the successors to adjust their exit probabilities, this autocratic behavior can reduce the value functions of the successors.
报告人简介：Dr. Huiling Wu is a professor of actuarial science and quantitative finance at Central University of Finance and Economics. She obtained her doctoral degree from School of Mathematics at Sun Yat-sen University. Her recent research interest covers portfolio selection optimization, risk management and the optimal decision in the decumulation phase. She has actively published in some international journals, including Insurance: Mathematics and Economics, Economic Modelling, OR Spectrum, Journal of Optimization Theory and Applications, International Review of Financial Analysis, etc., or in some leading journals in China, including Chinese Journal of Management Science, Systems Engineering-Theory and Practice, Management Review.
邀 请 人：郑敏
