Applying an Active Trading Strategy to Financial Bubbles - Forward Looking Risk Management

This paper examines the performance of the The Log-Periodic Power-Law (LPPL) model and the Generalized Hurst Exponent (GHE) in forecasting a bubble’s crash date. Furthermore, it is investigated whether applying a market-timing strategy to financial bubbles results in obtaining higher returns compared to a passive trading strategy. The LPPL model and the GHE approach are fitted to historical price time series of several indexes. Based on the information gathered from this calibration, an active trading strategy is applied. The aim of the active trading strategy is to obtain significant higher returns, whilst reducing the risk, compared to a passive trading strategy. Furthermore, the performance of the LPPL and the GHE model on the Bitcoin is examined. Both the LPPL model and the GHE approach are accurate when fitting historical bubbles. When applying the LPPL model and GHE approach to a trading strategy, the models overall show to be beneficial. The LPPL model leaving the market at the OLS estimated critical time provides the best trading strategy. For the Bitcoin, the models do not provide a higher return.