ECONOMETRICS: Weak Identification of Long Memory with implications for Inference; Professor Yu Jun (Singapore Management University)

Abstract

This paper explores weak identification issues arising in commonly used models of economic and financial time series. Two highly popular configurations are shown to be asymptotically observationally equivalent: one with long memory and weak autoregressive dynamics, the other with antipersistent shocks and a near-unit autoregressive root, often characterized as rough volatility in empirical work. This paper develops a data-driven semiparametric and identification-robust approach to inference that reveals these model ambiguities, investigates the implications of weak identification on forecasting, and documents the prevalence of weak identification in many realized volatility and trading volume series. Forecasting analyses at multiple horizons reveal advantages to long memory modeling; and the identification-robust empirical findings generally favor long memory dynamics in volatility and volume, a conclusion that is corroborated using social-media news flow data.

Click here to view paper.