Testing the Monday Effect in U.S. Stock Return: A Spatial Dominance Approach 1

• 영문명: Testing the Monday Effect in U.S. Stock Return: A Spatial Dominance Approach 1
• 발행기관: 한국계량경제학회
• 저자명: Sungro Lee Chang Sik Kim In-Moo Kim
• 간행물 정보: 『한국계량경제학회 학술대회 논문집』2009년 공동학술대회, 1~28쪽, 전체 28쪽
• 주제분류: 경제경영 > 경제학
• 파일형태: PDF
• 발행일자: 2008.11.30

영문 초록

This paper presents tests of Monday effect in U.S. stock market using the concept of spatial dominance that was developed in Park (2007). The spatial dominance provides a new tool to test the distributional dominance of one process over the other even under nonstationarity. Extending the notion of well-developed stochastic dominance into the nonstationary time series whose distributions are time varying, the existence of the Monday effect in U.S. stock market is investigated. Test procedures in the previous literature including stochastic dominance compare expected utilities only at a specific time, usually market closing time. Spatial dominance, however, enables us to analyze the expected sum of instantaneous utilities during trading time. We, therefore, can consider intraday patterns of returns using high frequency data for the testing of the Monday effect. When we apply spatial dominance approach to S&P 500 return over the period 1988-2004, significant Monday effect is not found whereas Cho, Y.-H. et al. (2007) reports inconclusive results for the same period with stochastic dominance. Furthermore, we find dominance relationships among groups different weekdays. The group consists of Tuesday and Wednesday returns spatially dominates the other groups with three other weekdays. We also find that returns on Wednesday also spatially dominates returns on Friday. Our results are consistent with recent findings that Monday effect in U.S. stock market has been disappeared since 1987.

목차

1 Introduction

2 Testing Procedures

3 Empirical results

4 Conclusion

키워드

spatial dominance local time spatial distribution subsampling