Explaining salmon futures returns with factor models

Explaining returns on salmon futures contracts via risk factors is relevant to financial investors, risk managers, and researchers. A prior study applied the Capital Asset Pricing Model (CAPM) to weekly futures payoffs and found it adequate. We identify several econometric and subject-matter limitations in that approach and propose a refined analysis with partly different findings. Our test regression assumes a linear relationship between stock market excess return and futures relative returns (rather than payoffs) and adopts monthly sampling to reflect the futures market’s low liquidity. Under this specification, the CAPM is rejected for six contract maturities between one and twelve months. The estimated systematic risk is negligible in the full sample, though noticeable in earlier subsamples. We then extend the test regression to include autoregressive terms and monthly dummy variables. While systematic risk remains small, returns are clearly autocorrelated and exhibit nonzero alphas in some months of the year, violating the CAPM. Liquidity permitting, the associated abnormal returns may have been exploitable in real time. These results are broadly consistent across full-sample and rolling-window analyses and align with those from the Fama-French three-factor model.