AbstractThe behavior of a commodity's price‐return variance over time is critical to both the theory and practice of commodity option valuation. In this paper three models are used to forecast soybean price variance for the period during which a seasonal increase in variance has been found in previous studies. A time‐series model outperforms the ordinary least squares and naive models. The significance of the forecast error levels is then examined in terms of expected deviations above and below a price target for a put hedge. The resulting trade‐off between risk and return is shown by strike price and variance expectation.