A Comparison of Seasonality Using Different Methods

Heating oil and soybeans represent two fundamentally sound seasonal markets. Even before we analyze the patterns, we can expect heating oil to post higher prices during the winter months and soybeans to be high during the summer. Heating oil is accumulated beginning in midsummer in anticipation of the winter season, while soybeans are subject to speculative volatility during planting and growing seasons.
For the purposes of comparison, we will use four basic methods of calculating seasonality: average price, median price, percentage change from the previous month, and the moving average deseasonalizing. For our purposes, more complicated methods are unnecessary.
According to these techniques, it is the anticipation of winter that drives prices up, while February and March, typically colder, show lower average prices as most consumers use up the oil they had committed to buy during the previous September and October. Three of the methods for calculation gave similar results, while % Change seems to lead by one month.
When applied to soybeans, all four basic seasonal calculations produce patterns that show the expected winter and summer cycles, with most lines rising and falling smoothly throughout the year. Again, it should be noted that the seasonal pattern of the percentage price changes leads the normal and detrended methods because it acts in the same way as momentum; a decline in the percentage change does not mean that current prices are lower than the previous month, but that prices did not increase by as much as the previous month.