Risk Management for Swedish Farmers - An empirical study on hedge ratios for Swedish wheat

Sammanfattning: The paper investigates data on purchasing price of wheat from Swedish grain buyer Lantmännen and MATIF future contracts on milling wheat in an attempt to replicate the conditions for a Swedish farmer trying to manage his risk on wheat by trading future contracts on the MATIF exchange. Two static linear regressions and four dynamic GARCH models are employed on a sample of 1679 daily returns and 339 weekly returns ranging from 2009-07-01 to 2016-01-11. All regressions are ran on both daily returns and weekly returns to investigate how the rebalancing frequency changes the outcome of the hedges. The correlation of spot and future price changes from 0.19 for daily returns to 0.49 for weekly returns and all weekly return hedges outperforms the daily hedges in variance reduction. It is however hard to find a general best model over both daily and weekly returns and for all samples. The simple OLS performs best in the daily sample with -3.19% in variance over the full sample compared to a no-hedge and in the weekly return the VECM-VECH reduces variance by -29% over the full sample compared to a no-hedge.