Sequential Break-Point Detection in Stationary Time Series: An Application to Monitoring Economic Indicators

Abstract

Monitoring economic conditions and financial stability with an early warning system serves as a prevention mechanism for unexpected economic events. In this paper, we investigate the statistical performance of sequential break-point detectors for stationary time series regression models with extensive simulation experiments. We employ an online sequential scheme for monitoring economic indicators from the European as well as the American financial markets that span the period during the 2008 financial crisis. Our results show that the performance of these tests applied to stationary time series regressions such as the AR(1) as well as the AR(1)-GARCH(1,1) depend on the severity of the break as well as the location of the breakpoint within the out-of-sample period. Consequently, our study provides some useful insights to practitioners for sequential break-point detection in economic and financial conditions.

Partial Sum Processes of Residual-Based and Wald-type Break-Point Statistics in Time Series Regression Models

“…The use of partial-sum processes for model residuals when testing the null hypothesis of no structural break using residual-based statistics appears in various applications in the literature (Katsouris, 2021). Firstly, Brown et al (1975) proposed the OLS-CUSUM test constructed based on cumulated sums of recursive residuals for testing for the presence of a single structural break in coefficients of the linear regression model (see, also Krämer et al (1988) and Ploberger and Krämer (1992)).…”

“…Severe time series fluctuations manifesting as market exuberance are considered by econometricians as early warning signs of upcoming economic recessions which are not usually explained by common boom and bust cycles (see, Greenwood et al (2020), Baron et al (2021) and Katsouris (2021)). Furthermore, the economic aspects of prolonged economic policy uncertainty as well as as more recently pandemics can impact the robustness of parameter estimates due to increased model uncertainty.…”