摘要：This paper aims to enrich the understanding and modelling strategies for cryptocurrencymarkets by investigating major cryptocurrencies’ returns determinants and forecast their returns.To handle model uncertainty when modelling cryptocurrencies, we conduct model selection for anautoregressive distributed lag (ARDL) model using several popular penalized least squares estimatorsto explain the cryptocurrencies’ returns. We further introduce a novel model averaging approach orthe shrinkage Mallows model averaging (SMMA) estimator for forecasting. First, we find that thereturns for most cryptocurrencies are sensitive to volatilities from major financial markets. The returnsare also prone to the changes in gold prices and the Forex market’s current and lagged information.Then, when forecasting cryptocurrencies’ returns, we further find that an ARDL(p,q) model estimatedby the SMMA estimator outperforms the competing estimators and models out-of-sample.
关键词：cryptocurrencies; Mallows criterion; model averaging; model selection; shrinkage; tuning parameter choice
