The functional central limit theorem and weak convergence to stochastic integrals. I: Weakly dependent processes (Q2716481)

The paper gives new conditions for the functional central limit theorem (FCLT) and of weak convergence of stochastic integrals. The authors work with the concept of near-epoch-dependence (NED) of a mixing process. Let \( X_nt\) denote a triangular array of random variables. Then \( X_nt\) is called \( L_2\)-NED on random variables \(V_nt\) if for \(m\geq 0\) NEWLINE\[NEWLINE{\|X_nt - E(X_nt\mid{\mathcal F}_{n,t-m}^{t+m}) \|_2}\leq d_{nt}{\nu}(m),NEWLINE\]NEWLINE where \({\mathcal F}_{nt}^t={\sigma}(V_ns,\dots, V_nt)\), \(t\geq s\), \(d_nt\) is an array of positive constants, and \({\nu}(m) {\rightarrow}0\) as \(m{\rightarrow}\infty\). The authors' assumptions are robust in cases of partially specified models, in which aspects of the short-run data generation process are unknown. The main FCLT dominates the existing ones in the economic literature that use comparable assumptions. The results for stochastic integrals convergence impose virtually the same conditions as the FCLT. Some explanations are presented how to apply the FCLT in the theory of unit root testing and for cointegrating regressions. Some useful properties of mixingales are established as auxiliary results.