Price discovery, technical analysis and weak-form efficiency: evidence from cryptocurrencies

Abstract

The continued prevalence of Bitcoin and other cryptocurrencies in the media, in online forums, and even casual conversation continue to add to the buzz of this controversial asset class. What adds to the controversy around cryptocurrencies is that they do not have fundamental value, so we can only observe the price of the cryptocurrency. The question thus arises as to whether we can discover the price of a cryptocurrency, and whether this means that the market is predictable. The present study explores the price discovery process in the cryptocurrency market. This is achieved by using state space modelling, via the Kalman filter, and technical analysis. The sample period covers 2010 to 2020 and consists of the top 30 cryptocurrencies based on market capitalisation. The Kalman filter is passed through all 30 cryptocurrencies and used to generate the efficient equilibrium price level which is further utilised to yield buy and sell opportunities. Additionally, the ability of a family of five technical trading rules (Moving Average (MA), trading range breakout, Moving Average Convergence, Divergence (MACD), Relative Strength Index (RSI), and Bollinger Bands (BB)) to generate returns in excess of the passive buy and hold benchmark is explored. The results show superior performance of the technical trading rules and the Kalman filter across the full sample as well as in the sub samples even after including transaction costs. Overall the results provide strong evidence in favour of the ability of the Kalman filter and technical analysis to “time” the market