Exploit momentum in Cryptocurrency Market

Researchers put efforts into explanations of the momentum phenomenon and improvements strategy since emergence in 1993. Interested anomalies appearing as exhibited traditional asset markets, adequate studies are launched on nascent emergers last decade, cryptocurrency market. Recent have shown that there is hardly any cross-sectional To explore anomaly additionally market, this paper implemented a time-series winners for improvement. Previous introduced detecting turning point between long-term slow factor short-term fast contributes to predicting trend well. Furthermore, threshold decided by certain machine learning model suggests better performance. In paper. A multilayer perceptron (MLP) utilized learn weights factors. The combination shows advantages MLP learned weighted preferable.