Workshops

The Applied Statistics Workshop 2008

※統計数理研究所リスク解析戦略研究センター金融・保険リスク研究プログラム との共催ワークショップ

※ 2008年3月30日現在の予定です。




<本年度終了分>

日時

2008年4月11日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Kosuke Oya/大屋幸輔 (Osaka University / 大阪大学経済学部)

A Test for Cross-sectional Dependence of Microstructure Noises and their Cross-Covariance Estimator
要旨(Abstract)

高頻度観測される約定データにもとづく Integrated Volatility や Integrated Covariance の推定量は Bid-Ask Bounce に代表される Market Microstructure Noise の存在により、バイアスをもち、その分散も過大なものになっている。さまざまな推定量の改良が提案されているが、それらの多くは Microstructure Noise の dependence の構造を既知としたものである。この従属性の構造を明らかにするために、本報告では直接観測できない Microstructure Noise の相互自己共分散がゼロであるかどうかを検定する統計量と相互自己共分散関数の推定量を提案する。

日時

2008年4月18日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Shin Kanaya/金谷信 (University of Wisconsin-Madison)

Non-Parametric Specification Testing for Continuous-Time Markov Processes: Do the Processes Follow Diffusions ?
要旨(Abstract)

I propose a new non-parametric testing procedure to determine whether an underlying continuous-time process is a diffusion. While many papers in economics and finance presuppose that the dynamics of economic variables are described by diffusion processes, an empirical validation of the diffusion hypothesis is rarely found. I develop a new theorem which non-parametrically and fully identifies diffusion processes within a class of univariate stationary Markov processes through their infinitesimal generators - functional operators computed via derivatives of the conditional expectations with respect to time. I construct test statistics based on this theorem and derive their asymptotic distributions. I also propose a simulation-based technique to approximate the asymptotic distributions, since the distributions of the original statistics depend upon a large number of unknown parameters and functions. Monte-Carlo simulations are conducted to study the finite-sample size and power properties of the test. I apply the proposed method to short-term interest rates and foreign exchange rates to examine the validity of the diffusion hypothesis.

日時

2008年5月9日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Hiroki Masuda/増田弘毅(九州大学大学院数理学研究院)

Covariation Estimation Using Realized Multipower Variations: Synchronous-Sampling Case
要旨(Abstract)

Realized multipower variation (MPV) は実現ボラティリティの一種の拡張であり, セミマルチンゲールの第二特性量行列 (連続部分の累積ボラティリティ) の推定量と して高い汎用性を持つ.近年MPVに関する漸近理論が整備されたが,非対角要素まで 込めた第二特性量行列の推定への適用にはあまり注意が払われていない.本報告で は,異なる複数のMPVを組み合わせることで第二特性量行列の推定量を構成し,適当 な規格化を施した下での一次漸近挙動を示す.これは異なる確率過程間の共分散構造 や共ジャンプ構造の解析に関する基本的な道具となる.

日時

2008年5月21日(水 Wednesday) 12:00 - 1:00※臨時ワークショップ※

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Joel Horowitz(Department of Economics, Northwestern University)

Oracle Efficient Nonparametric Estimation of an Additive Model with an Unknown Link Function [PDF]

日時

※特別講義のお知らせ※

第1回: 2008年5月21日(水 Wednesday) 9:00 - 11:00

第2回: 2008年5月22日(木 Thursday) 9:00-11:00
場所

東京大学大学院経済学研究科棟 2階 203号室
at the Room No.203 on the 2rd floor of the Economics Research Building

報告

Joel Horowitz(Department of Economics, Northwestern University)

Bootstrapに関する特別講義
Handbook of Econometricsの執筆部分を元に、わかり易く説明して頂きます。 大学院生の参加を歓迎いたします。

日時

2008年5月23日(金 Friday) 3:30 - 6:10

※時間と場所にご注意下さい。

場所

東京大学大学院経済学研究科棟 12階 第3共同研究室
at the Conference Room No.3 on the 12th floor of the Economics Research Building

報告

1) 3:30 - 4:40 Masayuki Hirukawa/蛭川雅之(Northern Illinois University)

Time Series Nonparametric Regression Using Asymmetric Kernels with an Application to Estimation of Scalar Diffusion Processes

2) 5:00 - 6:10 Seung Ahn(Arizona State University)

Eigenvalue Ratio Test for the Number of Factors

要旨(Abstract)

1) This paper considers a nonstandard kernel regression for strongly mixing processes when the regressor is nonnegative. The nonparametric regression is implemented using asymmetric kernels [Gamma (Chen, 2000b), Inverse Gaussian and Reciprocal Inverse Gaussian (Scaillet, 2004) kernels] that possess some appealling properties such as lack of boundary bias and adaptability in the amount of smoothing. The paper investigates the asymptotic and .nite-sample properties of the asymmetric kernel Nadaraya-Watson, local linear, and re-weighted Nadaraya-Watson estimators. Pointwise weak consistency, rates of convergence and asymptotic normality are established for each of these estimators. As an important economic application of asymmetric kernel regression estimators, we reexamine the problem of estimating scalar di?usion processes.
2) This paper proposes two new estimators for determining the number of factors in approximate factor models. We exploit the well known fact that the r eigenvalues of the variance-covariance matrix of N response variables, where r is the number of comment factors in the variables, grow unboundedly as N increases. The criterion functions used for the two estimators are related to the ratio of two adjacent eigenvalues. An important advantage of the estimators is that they do not require the use of penalty functions. The estimators can be viewed as a reformulation of the well known scree test. We show that the estimators are consistent under the general conditions of Bai and Ng (2002). Our simulation results show that the estimators have good finite sample properties unless the signal-to-noise-ratio of each factor is too low. They perform much better than the Bai-Ng estimators do when either the number of the response variables analyzed or the number of time series observations, T, is small.

日時

2008年6月2日(月 Monday) 12:00-13:10 ※臨時ワークショップ※
※Micro WorkshopEmpirical Micro Brown-Bag Lunch Seminarと共催※

※日時と場所にご注意ください※

場所

東京大学大学院経済学研究科棟 12階 第2共同研究室
at the Conference Room No.2 on the 12th floor of the Economics Building

報告

Hiroyuki Kasahara/笠原博幸 (The University of Western Ontario)

Sequential Estimation of Structural Models with Fixed Point Constraint [PDF]

要旨(Abstract)


This paper considers the estimation problem of structural models of which empirical restrictions are characterized in terms of fixed point constraint, such as a structural dynamic discrete choice model and a model of dynamic games. We analyze the conditions under which the nested pseudo-likelihood (NPL) algorithm achieves convergence and derive its convergence rate. We find that the NPL algorithm may not necessarily converge when the fixed point mapping does not have a local contraction property. To address the issue of non-convergence, we propose alternative sequential estimation procedures that can achieve convergence even when the NPL algorithm does not and, upon convergence, some of our proposed estimation algorithms produce more efficient estimators than the NPL estimator. We also show that the similar convergence results hold for models with (time-varying) unobserved heterogeneity, where the EM algorithm is incorporated into the NPL algorithm. Furthermore, we extend the idea behind the NPL algorithm to the moment estimators, developing a recursive extension of popular two-step moment methods called the sequential generalized method of moments (GMM) algorithm. The sequential GMM algorithm has the convergence properties similar to those of the NPL algorithm.

日時

2008年6月13日(金 Friday) 4:50 - 6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Michael McAleer(University of Western Australia)

The Ten Commandments for Optimizing Value-at-Risk

要旨(Abstract)

I have two papers relating to forecasting VaR, both of which have just appeared in the Journal of Forecasting. Paper 1 examines very practical methods for evaluating VaR and minimizing daily capital charges: (1) McAleer, M. and B. da Veiga (2008), "Single index and portfolio models for forecasting value-at-risk thresholds", Journal of Forecasting, 27, 217-235. Paper 2 presents a new theoretical model for evaluating VaR, and evaluates the usefulness of the new model in practical VaR examples: (2) McAleer, M. and B. da Veiga (2008), "Forecasting value-at-risk with a parsimonious portfolio spillover GARCH (PS-GARCH) model", Journal of Forecasting, 27, 1-19.

日時

※統計数理研究所・リスク解析戦略研究センター/
東京大学大学院経済学研究科・応用統計ワークショップ 共同セミナー※

2008年6月20日(金 Friday)
1:30 - 5:15

場所

統計数理研究所・講堂

報告

統計数理研究所・リスク解析戦略研究センター主催ワークショップ

内容

「信用リスクの統計解析:CDS, CDO, サブプライム問題」
詳細はこちらをご参照下さい。 

日時

2008年6月27日(金 Friday) 4:50 - 6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Taisuke Otsu/大津泰介(Yale University)

On Bahadur efficiency of empirical likelihood

要旨(Abstract)

This paper studies the Bahadur efficiency of empirical likelihood for testing moment condition models. It is shown that under mild regularity conditions, the empirical likelihood overidentifying restriction test is Bahadur efficient, i.e., its p-value attains the fastest convergence rate under each fixed alternative hypothesis. Analogous results are derived for parameter hypothesis testing and set inference problems.

日時

2008年7月1日(火 Tuesday) 12:00-13:00 ※臨時ワークショップ※

※日時と場所にご注意ください※

場所

東京大学大学院経済学研究科棟 12階 第1共同研究室
at the Conference Room No.1 on the 12th floor of the Economics Building

報告

Edward Vytlacil (Yale University)

Monotonicity, Instrumental Variables and the Sign of the Average Treatment Effect (joint with Cecilia Machado and Azeem Shaikh)

要旨(Abstract)


日時

※特別講義のお知らせ※
第1回: 2008年7月10日(木 Thursday)
9:00 - 11:00
第2回: 2008年7月11日(金 Friday) 9:00-11:00

場所  ※会場が異なりますのでご注意下さい※
第1回:東京大学大学院経済学研究科棟 2階 203号室
at the Room No.203 on the 2rd floor of the Economics Research Building

第2回:東京大学大学院経済学研究科棟 3階 第4教室
at the Lecture Hall No.4 on the 3rd floor of the Economics Research Building

報告

Richard Blundell(University College London)

Lectures on the Distributional Dynamics of Income, Earnings and Consumptoin
[paper1] [paper2] [lecture note]
日時

2008年7月11日(金 Friday) 4:50 - 6:10
ミクロワークショップおよびマクロワークショップと共催

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Richard Blundell(University College London)

Employment, Hours of Work and the Optimal Design of Earned Income Tax Credits

要旨(Abstract)

This paper examines the optimal schedule of marginal tax rates and the design of earned income tax credits. The analysis is based on a structural labour supply model which incorporates unobserved heterogeneity, fixed costs of work and the detailed non-convexities of the tax and transfer system. An analytical framework is developed that allows explicitly for an extensive margin in work choices and also the partial observability of hours of work. This is contrasted to the standard case in which only earnings (and non-labour income) are observable to the government. The empirical motivation is the earned income tax credit reforms in Britain which include a minimum hours requirement at 16 hours per week and a further bonus at 30 hours. Our analysis examines the case for the use of hours-contingent payments and lends support for the overall structure of the British tax credit reforms. However, we also provide a strong case for a further reduction of marginal rates for lower earners but only those with school age children.

  

連続講義「統計的リスク管理と保険」のお知らせ

研究会「金融と保険の統計学2008」では,「応用統計ワークショップ」及び「統計数理研究所 リスク解析センター」と共同して東京大学大学院経済学研究科(CIRJE)において, 統計的リスク管理論の理論と応用において国際的に著名な研究者であるAlexander J. McNeil教授 による連続講義を計画中です。
滅多にない連続講義なので,統計的リスク管理論, ファイナンス・保険や統計学の応用に興味を持つ院生は特に見逃すことなくご聴講ください。

<参加をご希望の方へ>
教室の収容人数と配布資料準備等の関係で,参加人数を把握しておきたいと思いますので, 参加される方は9/9(火曜)までに,電子メールで
成城大学経済学部 塚原英敦 (tsukahar@seijo.ac.jp) 
まで必ずご連絡ください
(収容人数の関係で実務界に御所属の方々の中で聴講が認められる 人数には制限があることにご注意ください)。なお,お申込みの際には電子メールの件名に 「McNeil集中講義参加申込」とお書きください。

日時

第1回 2008年9月16日(火 Tuesday)  4:30 - 6:30
第2回 2008年9月18日(木 Thursday) 4:30 - 6:30
第3回 2008年9月19日(金 Friday)      4:30 - 6:30

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Alexander J. McNeil (Heriot-Watt University, Edinburgh)

Statistical Risk Managements and its Applications
(統計的リスク管理とその応用)

概要

内容:    前半は近年発展が目覚ましい統計的リスク管理論と保険業などへの応用についての 入門的講義(2回)を行う。最後の1回は最近の話題に関するセミナーとなる予定である。
予備知識:    予備知識として学部レベルの確率論・数理統計学の基礎を仮定する。
参考文献:    A.J. McNeil, R. Frey and P. Embrechts (2005),
"Quantitative Risk Management: Concepts, Techniques and Tools," Princeton UP
(翻訳「定量的リスク管理:基礎概念と数理技法」塚原他,2008, 共立出版)

日時

2008年10月17日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Tsunao Okumura/奥村綱雄 (横浜国立大学経済学部)

Concave-Monotone Treatment Response and Monotone Treatment Selection: With Returns to Schooling Application(共著者:臼井恵美子(名古屋大学大学院経済学研究科)
要旨(Abstract)


This paper identifies sharp bounds on the mean treatment response under the concave monotone treatment response (concave-MTR) and monotone treatment selection (MTS) assumptions. Empirical application to the mean returns to schooling shows that the estimates of our bounds are substantially narrower than (1) the estimates using only the concave-MTR assumption of Manski (1997) and (2) the estimates using only MTR and MTS assumptions of Manski and Pepper (2000). Our upper-bound estimates on the returns to schooling fall in the lower range of those point estimates given in previous studies which assume linear functions.

日時

2008年10月24日(金 Friday) 12:00-1:00

場所

東京大学大学院経済学研究科棟 12階 第2共同研究室
in Conference Room No. 2 on the 12th floor of the Economics Building

※日時と場所が通常と異なりますのでご注意ください

報告

Ming-Tien Tsai (Institute of Statistical Science, Academia Sinica)

The optimal rank tests for modal directions on spheres
要旨(Abstract)

For a general class of unipolar, rotationally symmetric distributions on the multidimensional unit spherical surface, a characterization of locally best rotation-invariant test statistics is exploited in the construction of locally best rotation-invariant rank tests for modal location. Finite sample permutational distributional perspectives are appraised. Allied statistical distributional problems are appraised, and in the light of these assessments, asymptotic relative efficiency of a class of rotation-invariant rank tests (with respect to some of their parametric counterparts) is studied. Two-sample problem is also investigated.

日時

2008年10月29日(水 Wednesday) 4:00-6:10
※一橋大学グローバルCOEプログラム
「社会科学の高度統計・実証分析拠点構築」 (代表者:深尾京司)/
科学研究費補助金(基盤研究(A))
「高頻度データを用いた日本の証券市場の計量分析」(代表者:渡部敏明) と共催※

場所

東京大学大学院経済学研究科棟 12階 第1共同研究室
in Conference Room No. 1 on the 12th floor of the Economics Building

※日時と場所が通常と異なりますのでご注意ください

報告

(1) 4:00 - 5:00 Fulvio Corsi (University of Siena)

Volatility forecasting: the jumps do matter

(2) 5:10 - 6:10 Torben Andersen (Northwestern University)

Model-Free Implied Volatility
要旨(Abstract)


(1) This study reconsiders the role of jumps for volatility forecasting by showing that jumps have a positive and mostly significant impact on future volatility. This result becomes apparent once volatility is correctly separated into its continuous and discontinuous component. To this purpose, we introduce the concept of threshold multipower variation (TMPV), which is based on the joint use of bipower variation and threshold estimation. With respect to alternative methods, our TMPV estimator provides less biased and robust estimates of the continuous quadratic variation and jumps. This technique also provides a new test for jump detection which has substantially more power than traditional tests. We use this separation to forecast volatility by employing an heterogeneous autoregressive (HAR) model which is suitable to parsimoniously model long memory in realized volatility time series. Empirical analysis shows that the proposed techniques improve significantly the accuracy of volatility forecasts for the S&P500 index, single stocks and US bond yields, especially in periods following the occurrence of a jump.

(2) The notion of model-free implied volatility (MFIV), constituting the basis for the highly publicized VIX volatility index, can be hard to measure with accuracy due to the lack of precise prices for options with strikes in the tails of the return distribution. This is reflected in practice as the VIX index is computed through a tail-truncation which renders it more compatible with the related concept of corridor implied volatility (CIV). We provide a comprehensive derivation of the CIV measure and relate it to MFIV under general assumptions. In addition, we price the various volatility contracts, and hence estimate the corresponding volatility measures, under the standard Black-Scholes model. Finally, we undertake the first empirical exploration of the CIV measures in the literature. Our results indicate that the measure can help us refine and systematize the information embedded in the derivatives markets. As such, the CIV measure may serve as a tool to facilitate empirical analysis of both volatility forecasting and volatility risk pricing across distinct future states of the world for diverse asset categories and time horizons.

日時

2008年11月14日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Seiya Imoto/井元清哉 (東京大学医科学研究所)

Temporal and Spatial Modeling in Systems Biology and Pharmacogenomics
要旨(Abstract)

In this talk, we present a statsitical method for extracting information about gene networks from genome-wide transcriptome data such as microarray data and proteome data like protein-protein interactions; those data are typicallly composed of several thousands or more feature variables. Our aims of the analysis are to gain better understanding of cellular response against a chemical compound like drug and to accelarate the development of genome-based medicine. To understand response of the cells for external stimulus, we should know the following information: when genes are transcribed and where proteins work in the cell. We create a new short time-course microarray dataset of human umbilical vein endothelial cell (HUVEC) treated with fenofibrate that is a drug for hyperlipidemia and apply our method to elucidate how cells respond to the stimulus of fenofibrate in terms of time and space; this response is summarized as a gene network that could show the drug efficacy of fenofibrate. Based on the results obtained by our approach, we would like to discuss the possibility to find novel drug targets.

日時

2008年11月19日 (水 Wednesday) 4:50-6:10
※曜日と場所が通常と異なりますのでご注意ください

場所

東京大学大学院経済学研究科棟 12階 第1共同研究室
at the Conference Room No.1 on the 12th floor of the Economics Research Building

報告

William Dunsmuir (University of New South Wales, Australia)

Parsimonious Garch Modelling of Australian Yield Curves with Multivariate Asymmetric t-Distribution
要旨(Abstract)


In this talk returns of generic interest rate yields with maturities of 1, 2, 3, 5 and 10 years, produced by the Reserve Bank of Australia based on bonds on issue, are modelled using multivariate GARCH models. The model presented has two novel features. The first innovation is the incorporation of a multivariate asymmetric t-distribution in which the degrees of freedom for each series varies with maturity. The second innovation is to describe the parameters specifying the GARCH model and the degrees of freedom using simple smooth functions of time until maturity of the component series. This approach allows an empirical description of the complete interest rate yield curve returns therefore allowing the model to be used for interpolation to additional maturities beyond those used to construct the model. Application is made to fitting the model to daily data from the period 1996-2001 with maturities of 1, 2, 3, 5 and 10 years and applying this to out-of-sample Australian Treasury yield returns for the period 2000-2004 and all annual maturities from 1 to 10 years.

日時

2008年12月5日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Fumikazu Miwakeichi/三分一史和 (千葉大学工学部)

時空間脳信号データの統計解析
要旨(Abstract)


脳信号は様々な計測装置によりミクロからマクロなレベルにおいて, 時空間データとして計測される。 解析方法としてはデータの動力学的情報を直接的に考慮しない相関解析, 回帰分析が広く用いられており,本講演ではその例として機能的磁気共鳴画像 法(fMRI)で標準的な解析方法として確立されているStatistical Parametric Mapping(SPM)の原理と行動経済学,神経行動学における応用例を解説する。 動力学的情報を考慮する多変量時系列解析は脳信号解析の分野ではあまり一般 的ではないが,それでもいくつか研究例があり,その例として赤池による脳波解析の 例,ならびに,筆者らが開発した時空間フィルターと統計検定を組み合わせた膜電 位イメージングデータの解析法を紹介する。

日時

2008年12月12日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Atsushi Inoue/井上篤 (North Carolina State University)

Testing for Identification in Possibly Nonlinear Models
要旨(Abstract)

In this paper we propose a chi-square test for identification. Our proposed test statistic is based on the distance between two bias-corrected shrinkage extremum estimators. The two estimators converge in probability to the same limit when identification is strong, and they converge weakly to different random variables when identification is weak. The proposed test is consistent not only for the alternative hypothesis of no identification but also for the alternative of weak identification, which is confirmed by our Monte Carlo experiment results. We apply the proposed technique to test whether the structural parameters of a representative Taylor-rule monetary policy reaction function are identified.

日時

2009年1月9日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Paul Embrechts (ETH, Switzerland)

Extremes from meta distributions and the shape of the sample clouds
要旨(Abstract)

In several recent applications in financial risk management, models used build on so-called meta distributions. Examples include the Li Gaussian copula model and the t-based Merton-KMV model for the pricing of credit risk. The credit crisis has shown that several models used have serious problems in correctly describing the extremal behaviour of underlying events. In this talk we investigate the behaviour of (multivariate) extremes in meta distributions based on the recent, more geometric theory of extremes as presented in [1]. It turns out that the asymptotic properties of the level sets of the underlying distribution functions play a crucial role. The results of this talk are derived in [2] which can be found on www.math.ethz.ch/~embrechts. References:
[1] G. Balkema and P. Embrechts. High Risk Scenarios and Multivariate Extremes: A Geometric Approach. European Mathematical Society, Zurich, 2007.
[2] G. Balkema, P. Embrechts and N. Lysenko. Meta densities and the shape of their sample clouds, 2008, Preprint ETH Zurich.

日時

2009年1月23日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Nabil Maghrebi (和歌山大学)

Estimation of Japanese equity risk with the model-free Nikkei 225 implied volatility index
要旨(Abstract)

This study addresses some numerical issues in calculating the model-free Nikkei225 implied volatility index, as well as its stochastic properties and forecasting power. It examines the significance of truncation errors associated with the calculation methodology underlying the VIX New index and the limited spectrum of strike prices. It addresses also the long- term relationship between implied volatility and realized volatility in the Japanese and US markets. There is evidence of higher out-of-sample forecasting performance for the implied volatility index relative to alternative GARCH specifications. The results indicate that despite its upward bias, the model-free implied volatility index is able to capture the dynamics of market volatility better than volatility models based on historical returns, even upon accounting for asymmetric effects. This empirical evidence lends support to drawing inference about economic uncertainty from options markets.

日時

2009年1月30日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

下津克己(Queen's University)

Sequential Estimation of Structural Models with a Fixed Point Constraint
要旨(Abstract)

This paper considers the estimation problem of structural models for which empirical restrictions are characterized by a fixed point constraint, such as structural dynamic discrete choice models or models of dynamic games. We analyze the conditions under which the nested pseudo-likelihood (NPL) algorithm achieves convergence and derive its convergence rate. We find that the NPL algorithm may not necessarily converge when the fixed point mapping does not have a local contraction property. To address the issue of non-convergence, we propose alternative sequential estimation procedures that can achieve convergence even when the NPL algorithm does not. Upon convergence, some of our proposed estimation algorithms produce more efficient estimators than the NPL estimator.

日時

2009年2月6日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Yang Feng(Princeton University)

Local Quasi-Likelihood with a parametric guide
要旨(Abstract)

Generalized linear models and quasi-likelihood method extend the ordinary regression models to accommodate more general conditional distributions of the response. Nonparametric methods need no explicit parametric specification and the resulting model is completely determined by the data themselves. However nonparametric estimation schemes generally have a slower convergence rate such as the local polynomial smoothing estimation of nonparametric generalized linear models studied in Fan, Heckman and Wand (1995). In this work, we propose a unified family of parametrically-guided nonparametric estimation schemes. Asymptotic results and numerical simulations demonstrate the improvement of our new estimation schemes over the original nonparametric counterpart. This is joint work with Prof. Jianqing Fan and Prof. Yichao Wu.

日時

2009年3月27日(金 Friday) 4:50-6:10

場所

東京大学大学院経済学研究科棟 3階 第3教室
at the Lecture Hall No.3 on the 3rd floor of the Economics Research Building

報告

Walter Beckert(University of London)

Maximal Uniform Convergence Rates in Parametric Estimation Problems (joint with Daniel L. McFadden)
要旨(Abstract)

This paper considers parametric estimation problems with independent, identically, non-regularly distributed data. It focuses on rate-efficiency, in the sense of maximal possible convergence rates of stochastically bounded estimators, as an optimality criterion, largely unexplored in parametric estimation. Under mild conditions, the Hellinger metric, defined on the space of parametric probability measures, is shown to be an essentially universally applicable tool to determine maximal possible convergence rates. These rates are shown to be attainable in general classes of parametric estimation problems.