Time Series Analysis for Futures Investment

The project aims to develop a trading strategy that maximizes Sharpe ratio when investing in the 88 futures securities on Quantics platform. Sharpe ratio measures risk-adjusted return. It is determined by the ratio of average return to the average volatility of the returns.

Majority Voting

python majority_voting.py

It uses 3 technical indicators (Moving Average Convergence Divergence,Relative Strength Index and BollingerBand) to measure the market conditions and make an investment decision based on the majority sentiment reflected by the indicators.

ARIMA&SARIMA

It uses time series models ARIMA/SARIMA with various modifications to predict the close price. modifications including:

• pure ARIMA/SARIMA model

python ARIMA_pure.py

• ARIMA/SARIMA model with model evaluation

python ARIMA_with_eval.py

• ARIMA/SARIMA model with momentum indicator

python ARIMA_with_indicator.py

Holt-Winters

python ExpoSmoothing_clustering.py

It uses the Holt-Winters model combined with clustering when deciding the portfolio allocation ratios.