Bitcoin is one of the most exciting and dramatic stories at the intersection of technology and finance in the past decade. Bitcoin was created by a pseudonymous author in 2008, and began operations in January of 2009 with the goal of creating the first purely digital payment system without reliance on any central authority.
The exciting aspect of Bitcoin for our research is the creation ex nihilo of an entirely new class of tradable asset that has value somewhat independently from the traditional physical economy. This offers a natural experiment which allows us to examine the real-time evolution of an important financial market from its beginning. Our research focuses on the evolving efficiency of the Bitcoin market since its inception.
Over time, financial markets have become more and more efficient, making it harder to predict asset returns and extract profit from active trading strategies. Once a market is competitive enough and previously effective return predictors well known enough, the actions of market participants to take advantage of these strategies eliminate their effectiveness.
Methods and Expected Results
Bitcoin markets in the early days were characterized by low volumes and no professional/institutional investor participation. This situation changed dramatically from the earliest trades in 2010 to the present. We hypothesize that along the way some "traditional" technical trading strategies (strategies relying purely on past price/volume data, such as moving averages, on balance volume, bollinger bands, etc.) may have been profitable, and then gradually decayed in profitability as the amount of professional and institutional investors in the market increased and these "obvious" strategies got priced out of the market. Technical strategies have been in use for over a century as attempted means to determine patterns in changes of supply and demand for securities from analysis of historical price and volume data. While many studies have found they fail to produce excess returns after trading costs, others have found them to have predictive power over some time periods. By examining the applicability of technical trading strategies and the change thereof over time in the Bitcoin market, using historical Bitcoin market data (sourced from bitcoincharts.com), we can shed new light on both how profitable these strategies may be in less efficient markets, as well as attempt to distinguish between competing theories for why they become less so over time.