Identifiability and estimation of structural vector autoregressive models for subsampled and mixed-frequency time series.

By Alex Tank, Emily B. Fox, Ali Shojaie. 2017

28 citations recorded by the source corpus.

Causal inference in multivariate time series is challenging because the sampling rate may not be as fast as the time scale of the causal interactions, so the observed series is a subsampled version of the desired series. Furthermore, series may be observed at different sampling rates, yielding mixed-frequency series. To determine instantaneous and lagged effects between series at the causal scale, we take a model-based approach that relies on structural vector autoregressive models. We present a unifying framework for parameter identifiability and estimation under subsampling and mixed frequencies when the noise, or shocks, is non-Gaussian. By studying the structural case, we develop identifiability and estimation methods for the causal structure of lagged and instantaneous effects at the desired time scale. We further derive an exact expectation-maximization algorithm for inference in both subsampled and mixed-frequency settings. We validate our approach in simulated scenarios and on a climate and an econometric dataset.

Subfields: Control and Systems Engineering, Electrochemistry, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Control and Systems Engineering

In Electrochemistry

In Signal Processing

Citations in the corpus

Causal Discovery from Temporally Aggregated Time Series (2017, methodology): “This work was recently extended to mixed frequency data by structural VAR modeling (Tank et al., 2017) .”
Invariant Causal Prediction for Sequential Data (2017, background): “Another line of work starts from high-frequency data sampled from a Granger model without instantaneous effects and tries to infer the instantaneous effects appearing in low-frequency sub-sampled data (e.g., Gong et al. 2015; Tank, Fox, and Shojaie 2017).”
Granger Causality: A Review and Recent Advances (2021, methodology): “11(c), not only leads to the kind of mistaken inferences discussed above, but also to further mistakes unique to the mixed - frequency case; see Tank et al. (2019) for examples.”
DYNOTEARS: Structure Learning from Time-Series Data (2020, background): “There is also an extensive literature on learning SVAR models in the econometrics and statistics literature (Demiralp and Hoover, 2003; Lanne et al., 2017; Reale and Wilson, 2001, 2002; Swanson and Granger, 1997; Tank et al., 2019).”