Year: 2019
Type: dissertation
Abstract: Empirical Time Series is too linear. After the 2008 great depression, the board members \nof the central banks realize that they were unable to foresee the fi nancial meltdown until \nit was too late, due to the linear structure of the models used for the forecasts, claiming \nthe need for non-linear models. The fi eld of non-linear time series model is too vast, and \nsometimes these models are too complex to use them for forecasting. Furthermore, most of \nthe economic variables are persistent, viewed as unit roots, adding an extra level of di culty \nin the study of non-linear time series models. The challenge is to develop non-linear models \nwith persistent variables. \nThreshold models are a class of non-linear model characterized by different regimes, determined \nby a threshold variable. These regimes can represent the different stages of the economic \ncycles, for example, economic expansions and recessions, periods with high volatility \nand low volatility in the stock market, among many other examples. Many of the advantages \nof the threshold models are the simplicity of estimation using least square estimation, \ninterpretation of the non-linear structure, and testing. \nIn this dissertation, we study threshold models with unit roots from two different perspectives. \nIn one had we introduce a univariate analysis and on the other hand, a multivariate \nanalysis. \nIn the fi rst chapter, titled "Threshold Stochastic Unit Roots Models" co-authored with \nJesús Gonzalo and Raquel Montesinos, we present the univariate analysis by introducing a \nnew class of stochastic unit-root (STUR) processes. This new model, namely the threshold \nautoregressive stochastic unit root (TARSUR) process, is strictly stationary, but if we do \nnot consider the threshold effect, it can mislead to conclude that the process has a unit \nroot. The TARSUR models are not only an alternative to xed unit root models but present \ninterpretation, estimation, and testing advantages to the existent STUR models. \nThis study analyzes the properties of the TARSUR models and proposes two simple tests \nto identify this type of processes. The fi rst test will allow us to detect the presence of unit \nroots, which can be fi xed or stochastic, and the asymptotic distribution (AD) of this test \npresents a distribution discontinuity depending if the unit root is fi xed or stochastic. The \nsecond test we propose is a simple t-statistic (or the supremum of a sequence of t-statistics) for \ntesting the null hypothesis of a fi xed unit root versus a stochastic unit root hypothesis. It is \nshown that its asymptotic distribution (AD) depends if the threshold value is identi ed under \nthe null hypothesis or not. When the threshold parameter is known, the AD is a standard \nnormal distribution, while in the case of an unknown threshold value, the AD is a functional \nof Brownian Bridge. A Monte Carlo simulation shows that the proposed tests behave very \nwell in a nite sample, and the Dickey-Fuller test cannot easily distinguish between exact \nunit roots and threshold stochastic unit roots. The study concludes with applications to U.S. \nstock prices, U.S. house prices, U.S. interest rates, and USD/Pound exchange rates. \nThe second chapter, we present the multivariate analysis with "Multiple Long Run Equi- \nlibria Through Cointegration Eyes". In this chapter, we introduce threshold effects in the \ncointegration relation. Cointegration has succeeded in capturing the unique long-run linear \nequilibrium. Speci c non-linearities have been incorporated into cointegrated models but \nalways assuming the existence of a single equilibrium. In this study, we explore the possibility \nof different long-run equilibria depending on the state of the world (i.e., good and bad \ntimes, optimism and pessimism, frictional coordination) in a threshold framework. Starting \nfrom the present-value model (PVM) with different discount factors and depending on the \nstate of the economy, we show that this type of PVM implies threshold cointegrated with \ndifferent long-run equilibria. We present the estimation and inference theory, and the study \n nishes with two empirical applications where the variables are not linearly cointegrated but \nthreshold cointegrated. \nThe third chapter, we continue in the multivariate framework and introduce the paper \ntitled "Quasi-Error Correction Model". Cointegration captures single long-run equilibrium \nrelationships between economic variables and the error correction model (ECM) is the mechanism \nin which the equilibrium is maintained. In this study, we introduce the quasi-error \ncorrection model (QECM), derived from the cointegration relation with threshold effects, \nwhere each regime represents a different equilibrium relation between the variables. In contrast \nto the linear ECM, the QECM has a regressor which captures the switching between \nequilibria, capturing the dynamic structure of the equilibrium change. This regressor will \npose a problem similar to the non-linear error correction models, where the model cannot be \nbalanced using the traditional de nitions of integration. We present the estimation and the \ninference theory and nish with an empirical application for U.S. interest rate of instruments \nwith different maturities.
Author: Peng Zhou, Jun Yi
Language: English
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7065868470
