JultikaUniversity of Oulu repository

Abstract

I create a pairs trade on the commodity futures curve, which captures the roll returns of commodity futures and minimizes the standard deviation of the returns. The end results is a strategy that has an annualized arithmetic return of 6,04% and an annualized standard deviation of 2,01%. Transaction costs and liquidity are also accounted for.

The goal was to create and backtest a trading strategy that tries to capture the roll return component of commodity futures returns. In order to reduce the very high spot price volatility of commodity returns a market neutral systematic arbitrage was introduced through a pairs trade. The pairs trade involves taking a counter position relative to the position that is designed to capture the roll return, with as small of a negative expected return as possible.

In practice capturing the roll return component means taking a long position into the largest dollar difference of a backwarded futures curve. And the pairs trade component is then a short position into the same curve, but with the smallest dollar difference. If the commodity futures curve was in contango, the procedure was reverts.

It can be concluded, that both of the targets of this research were reach; capturing the roll returns of the commodity futures and minimizing volatility through a statistical arbitrage pairs trade.

The trades designed to capture the roll returns of commodity futures returned 5,55% annually (which was 91,9% of the portfolios total return), compared to the annualized return of the benchmark index of 0,5%.

The pairs trade designed to minimize the standard deviation of these returns, lowered the annualized standard deviation of the portfolio’s returns from 6,37% to 2,01%, making it almost fully market neutral.