Predicting salmon futures returns : an empirical study from 2007 to 2019

Abstract

In order to investigate the factors that influence the return of salmon futures, we construct a fully hedged, passive, front month rolling portfolio of long positions in these contracts. We show that the excess return on such a portfolio is affected by momentum, spot volatility, term structure and seasonality, but not by systematic risk or basis. When including transaction costs, the return on this portfolio is less than that of the market, but with a much higher volatility. However, when predicting subsequent monthly excess returns using a simple regression model based on the factors we have identified, we are able to construct a portfolio that significantly outperforms the market with no systematic risk. Keywords: Commodity futures, salmon, multi-factor model, asset pricing