Bitcoin Price Volatility Forecasting Shows Potential for Risk Management Improvement.
The study looked at how the price of Bitcoin changes over time. They used different models to predict how much the price might go up or down. The best model they found was SGARCHGED (1,1) with a generalised error distribution. This model was good at predicting volatility, but not the best at forecasting future prices. They also tested the model to see how much money might be lost on average, finding that a 1.2% loss could happen 1% of the time. Overall, the volatility of Bitcoin prices was high at the start of 2018 but then started to decrease.