RMB misalignment: What does a meta‐analysis tell us?

Using a dataset of 3108 dichotomous renminbi (RMB) misalignment estimates from 95 studies, we conduct a meta‐analysis based on Bayesian model averaging (BMA) to assess the effects of study characteristics on these RMB misalignment estimates. The findings demonstrate that selected study characteristic types systematically influence RMB misalignment estimates, especially time series versus non‐time series approach and co‐integration versus non‐cointegration. Though lean toward RMB undervaluation inference, results from the BMA meta‐analysis and the corresponding median probability model and synthetic practice evaluation do not offer a conclusive statistical inference on RMB misalignment. Specifically, the evidence varies with the configuration of model specification and does not corroborate the RMB undervaluation claim for the periods of “2009–2010” and “2011–2014.”