Data for: 3487540

Published: 15 November 2019| Version 1 | DOI: 10.17632/w4x4d9fyzp.1
Contributors:
Alexander Fish,
,

Description

A portfolio of commodity index-ETF pairs was constructed and its performance was compared against the Bloomberg commodity index. The portfolio was trained on data from 2011 to 2014 to determine the optimal parameters for trade-execution thresholds. Using these parameters and the optimal exposures found from optimizing the risk-adjusted expected returns, the trading strategy was tested from 2015 to 2018 with the portfolio rebalanced at the beginning of each year. Comparing the returns of the portfolio against the Bloomberg Barclays Commodities Index as a benchmark, it was found that the portfolio outperformed the benchmark in total returns and Sharpe ratio for 2015, 2017, and 2018, but it underperformed for both metrics in 2016. Performing a paired t-test on the portfolio and benchmark returns proved that the portfolio does not significantly outperform the benchmark in any given year.

Files

Categories

Finance, Algorithms, Convex Optimization, Computational Finance

Licence