85. December 15, 2017, Friday
84. November 21, 2017
83. November 17, 2017
82. November 13, 2017
81. November 10, 2017
80. October 24, 2017
79. September 11, 2017
- Speaker: Prof. Zhang, Bing (Nanjing University)
- Title: Transparency and Local Accountability: A National-Scale Field Experiment in China on the Disclosure of Information about Pollution
- Time: 14:30--16:00
- Venue: B127
- Chair: Dr. Cui, Jingbo
- Abstract: Central governments face compliance problems when they rely on local governments to implement policy. In authoritarian political systems, these challenges are pronounced because local governments do not face citizens at the polls. We designed and implemented a national-scale, randomized field experiment in China to test whether the public dissemination of performance ratings of municipal governments by non-state actors caused municipal governments to release more information about the management of pollution to the public as mandated. We find significant and positive treatment effects on the release of information to the public about the management of pollution after only one year. These results reveal important roles that non-state actors can play in enhancing accountability of local governments in authoritarian political systems.
84. November 21, 2017
- Speaker: Dr. Tang, Cheng Yong (Temple University)
- Title: Disentangling and Assessing Uncertainties in Multiperiod Corporate Default Risk Predictions
- Time: 10:00--11:30
- Venue: B127
- Abstract: Abstract: Measuring credit risks for individual companies, industrial segments, and market systems is fundamentally and broadly important in economics, finance and beyond. For such a purpose, various quantitative methods have been developed to predictively assess the probabilities of companies going default in future. However, as a more difficult yet crucial problem, evaluating the uncertainties associated with the default predictions remains little explored. In this paper, we develop, for the first time in the scenario of default predictions, a procedure for quantifying the level of associated uncertainties by carefully disentangling multiple contributing sources. Our framework effectively incorporates broad information from historical default data, financial records, and economic environmental conditions by a) characterizing the default mechanism, and b) capturing the future dynamics of various features contributing to the default mechanism. Our development of the framework overcomes major challenges in this tremendously large scale statistical inference problem and makes it practically feasible by using parsimonious models, innovative methods, and modern computational facilities. By appropriately predicting the market-wise total number of defaults and assessing the associated uncertainties, our method can effectively evaluate the aggregated market credit risk level. Upon analyzing a US market data set with our method, we demonstrate that the level of uncertainties associated with default risk assessments is indeed substantial. More importantly and informatively, we also find that the level of uncertainties associated with the default risk predictions is correlated with the level of default risks, indicating potential for benefiting practical applications including improving the accuracy of default risk assessments. This is a joint work with Miao Yuan, Yili Hong, and Jian Yang.
83. November 17, 2017
- Speaker: Dr. Zhang, Chuanyong (East China Normal University)
- Title: Capitalization of Transaction Taxes in Housing Markets: A Natural Experiment in Shanghai
- Time: 10:00--11:30
- Venue: B129
- Abstract: This paper contributes to the literature by identifying the causal effect of residential tax reform on housing price. Although economic theory implies that differences in residential transaction tax rates should be fully capitalized, empirical support for this proposition is uneven. This paper builds a search model that links the bargain power and reduction of transaction tax in a partial equilibrium model. The model indicates that the tax reduction will increase the transaction price, and to what extent the increase benefit goes to the buyer or seller is determined by the bargain power. Then we treat this tax reduction as one natural experiment and find that after March 1st, 2012, the threshold date, transaction prices of the NRP increased significantly compared with those of Non-NRP. Specifically, transaction prices of NRP increased by 2.5%, moreover, prices of smaller-than-90-m2 NRP increased much higher than those of larger-than-90-m2 NRP. We document some heterogeneity of the effect across geographic areas and various floor areas and find that this effect is stronger from the inner belt to the outer belt than other areas with smaller floor areas.
82. November 13, 2017
- Speaker: Prof. Kou, Steven (National University of Singapore)
- Title: The Science of Money
- Time: 10:00--11:00
- Venue: B251
- Abstract: We will discuss three aspects of the application of scientific methods to finance: (1) Investment (2) Derivatives (3) Risk Management. Although more advanced tools, such as stochastic differential equations, Monte Carlo simulation, psychology, statistical inference, optimization, and functional analysis may be needed to study these topics, only high school mathematics will be used in the talk. We aim at giving some concrete examples in these topics, to inspire interests among the general public. In particular, we will focus on three examples: (1) Optimal portfolio choices, e.g. Kelly and Merton criteria. (2) The binomial model for option pricing. (3) Axioms for risk measures.
81. November 10, 2017
- Speaker: Dr. Zhang, Rong (Tencent)
- Title: 强化学习简介
- Time: 14:00--15:00
- Venue: B127
- 摘要:本次讲座将会介绍 (1)强化学习:Markov决策过程,如何求解强化学习问题。 (2)深度学习:卷积神经网络,循环神经网络。 (3)强化学习与深度学习的结合:如何使用 DQN 来玩 Flappy Bird 等简单游戏。
- Chair: Dr. Sun, Xiang
80. October 24, 2017
- Speaker: Prof. Sun, Yixiao (UCSD)
- Title: Three in One: a Unified Estimator of Average Treatment Effects
- Time: 10:00--11:30
- Venue: A321
- Abstract: We consider estimating average treatment effects under the unconfoundedness and overlap assumptions. We introduce a method to integrate the three conventional estimators --- the regression adjustment estimator, the inverse probability weighting (IPW) estimator, and the doubly robust (DR) estimator --- into a single three-in-one (TIO) estimator. The TIO estimator is all three of them: it is a regression estimator, an IPW estimator, and a DR estimator. From an empirical point of view, this removes the burden from having to choose one of the three estimators. In finite samples, the TIO estimator outperforms the three conventional estimators.
79. September 11, 2017
- Speaker: Prof. Kuan, Chung-ming (National Taiwan University)
- Title: Large-Scale Multiple Inequality Testing without Data-Snooping Bias
- Time: 14:30--16:00
- Venue: A511
- Abstract: We will present a comprehensive review of the testing procedures for a large number of multiple inequalities, with properly controlled familywise error rates. These tests include the stepwise procedures of White's Reality Check and Hansen's Superior Predictive Ability (SPA) test. The asymptotic properties and simulation results of the stepwise SPA tests will also be discussed.