It's not hard to build a machine learning model that can predict the price direction with slightly more than 50% accuracy. However, once you factor in the tradi

It's not hard to build a machine learning model that can predict the price direction with slightly more than 50% accuracy. However, once you factor in the trading fees, slippage and the spread, you will almost always lose money. I built a number systems that looked like easy money making schemes that would make small amounts on most trades, but once you factor in trading costs (which most people forget about initially) you will lose money most of the time. That applies to the stock market and also crypto currencies if you are trying to predict the price direction. I have heard also that predicting volatility in the equities market is easier and the better strategy. To make money off that you would need to use derivatives. It's less clear how to do this in crypto unless you are trading futures, and I think making money off the price volatility there requires a different strategy, making heavy use of limit orders and stop losses.