A0404
Title: Application of machine learning models and interpretability techniques to identify the determinants of bitcoin price
Authors: Jose Manuel Carbo Martinez - Bank of Spain (Spain) [presenting]
Sergio Gorjon Rivas - Bank of Spain (Spain)
Abstract: Historically, the price of bitcoin has been subject to large and abrupt fluctuations, as demonstrated once again by its sudden drop following the all-time high of $68,000$ in November 2021 and, more recently, on the occasion of the crypto-asset market turmoil sparked by the likes of the Terra/Luna crash or the Celsius Networks collapse. Thus, a legit question arises as to which are the determinants that influence bitcoin the most. We attempt to answer that question, using a flexible machine learning model, specifically a Long Short Term Memory (LSTM) neural network, to establish the price of bitcoin as a function of a number of economic, technological and investor attention variables. Our LSTM model replicates reasonably well the behaviour of the price of bitcoin through different periods of time. We then use an interpretability technique called SHAP to understand with are the most important features of the LSTM outcome. We conclude that the importance of the different variables in the formation of the price of bitcoin changes substantially throughout the analysed period. What's more, we also find that not only does its influence vary, but that new explanatory factors seem to appear often over time that, at least, for the most part, remain initially unknown.
