CMStatistics 2016: Start Registration
View Submission - CFE
A0579
Title: Returns signal momentum Authors:  Jiadong Liu - Queens University Belfast (United Kingdom) [presenting]
Fotis Papailias - Queens University Belfast (United Kingdom)
Dimitrios Thomakos - University of Peloponnese (Greece)
Abstract: A new type of momentum based on the probability of past returns signs is introduced. Position signals are generated when the equally weighted average of past returns signs exceed a certain threshold. We consider various fixed and time-varying values for this threshold. We form a portfolio which consists of 55 of the world's most liquid commodity and financial futures. Investment strategies using returns signal momentum result in higher returns, Sharpe ratio and lower maximum drawdown when compared to the $1/N$, moving average strategies and the time series momentum. Additional statistical tests illustrate the robustness of the method. Returns signal momentum can, hence, be considered as an effective strategy for speculation and hedging by market participants.