MMDS大阪大学 数理・データ科学教育研究センター
Center for Mathematical Modeling and Data Science,Osaka University

Mixed-frequency Regression Modelling with Some Financial Applications

Virmantas Kvedaras (Vilnius University)

大証寄附研究部門セミナーシリーズ 第33回

Mixed-frequency Regression Modelling with Some Financial Applications

Virmantas Kvedaras (Vilnius University)

Regression modelling alternatives are considered whenever the regularly sampled explanatory variables are observed at a higher frequency than the dependent one. The ordered weighted aggregation (OWA) and the mixed data sampling (MIDAS) alternatives are discussed in more detail. An LM-type test is proposed to detect the non-linearity induced by the OWA-type effects. Its small sample properties are compared with the established tests for omitted non-linearity. Next, the statistical origin is defined of the non-linear restriction on parameters imposed in the MIDAS regressions, and the tests for its empirical adequacy are characterized in regressions with finite and infinite number of higher-frequency lags of explanatory variables. The relevance of OWA-type effects and the application of the MIDAS testing are illustrated using the stock exchange indices data.

講師: Virmantas Kvedaras (Vilnius University)
テーマ: 大証寄附研究部門セミナーシリーズ 第33回
日時: 2012年06月01日(金) 16:20-17:50
場所: 大阪大学大学院基礎工学研究科 (豊中キャンパス)I 棟 204号室
参加費: 無料
アクセス: 会場までのアクセスは下記URLをご参照ください。
http://www.es.osaka-u.ac.jp/access/
お問い合せ: 本ウェブサイトの「お問い合せ」のページをご参照ください。