We have been trading this seasonal spread annually every year since 1994, except for 1995. The spread has a good track record, but suffered its worst loss to date last year. All of the pertinent statistics are included below.
In its current form, we buy RBOB Gasoline futures and sell Heating Oil futures, with both contracts expiring in February of the following year. The spread is entered in late November and is normally exited in late December or early January. We do not use options in the spread. We have looked many times at trying to incorporate options, but the time value premium of February options bought in November is just too damaging to the results of the spread. If one were to use deeply in-the-money options in order to simulate the results with futures (and to minimize time value premium expense), he would actually be harming the rate of return – because the futures have a favorable spread margin.
From a fundamental point of view, the spread seems counterintuitive. Why would one want to buy Gasoline and sell Heating Oil heading into the winter. Shouldn’t demand for Heating Oil be increasing while demand for Gasoline is decreasing at that time? Yes, it probably should, but markets tend to discount such things in advance. I can’t say why they discount it so early in the season, but the track record of the spreads says that they do.
The graph above shows the composite spread (each year’s daily prices averaged over all the years going back to 1992)...