An Out-of-Sample Test for Nonlinearity in Financial Time Series: An Empirical Application

• Published in: Computational economics Vol. 36; no. 2; pp. 121 - 132
• Main Author: Panagiotidis, Theodore
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01.08.2010; Springer; Society for Computational Economics; Springer Nature B.V
• Series: Computational Economics
• Subjects: Algorithms; Behavioral/Experimental Economics; C22; C53; Cointegration; Computer Appl. in Social and Behavioral Sciences; Datasets; Econometric models; Econometrics; Economic models; Economic theory; Economic Theory/Quantitative Economics/Mathematical Methods; Economics; Economics and Finance; Forecasting; Forecasting techniques; G10; Interdisciplinary subjects; Math Applications in Computer Science; Nearest neighbour; Non-linear models; Nonlinearity; Operations Research/Decision Theory; Securities markets; Stock exchanges; Studies; Time series; Unit root; Greece; Athens Greece; C53; Nearest neighbour; Nonlinearity; C22; G10
• ISSN: 0927-7099; 1572-9974
• DOI: 10.1007/s10614-010-9225-z

This paper employs a local information, nearest neighbour forecasting methodology to test for evidence of nonlinearity in financial time series. Evidence from well-known data generating process are provided and compared with returns from the Athens stock exchange given the in-sample evidence of nonlinear dynamics that has appeared in the literature. Nearest neighbour forecasts fail to produce more accurate forecasts from a simple AR model. This does not substantiate the presence of in-sample nonlinearity in the series.