Seasonal filters for corn-wheat spread breakouts

Seasonal review as a permission filter

The case study treats seasonal review as a time-framed look at past spread behavior. Recurring first-half versus second-half patterns are identified, then used as a permission filter rather than a standalone forecast.

The traded object is an intermarket-spread: a paired long-and-short position in two related futures, here July or December corn versus the matching wheat contract, so the book follows the price difference rather than either outright.

Two calendar rules define the bias

Two calendar rules define the seasonal-bias. Corn tends to gain on soft red winter wheat in the first half of the year, and wheat tends to gain on corn in the second half, tied to each crop's planting and early-maturation window.

That seasonal-bias is used only as a permission filter and not as a forecast of the next print. The seasonal-window then restricts the book to one side of the spread in each half-year: long corn versus short wheat only in the first half, and short corn versus long wheat only in the second half.

A 10-session extreme admits or ends the trade

Trend following is implemented as a 10-session extreme breakout on the CBOT July or December corn/wheat spread. A close through the prior 10-day high or low admits a position. The opposite extreme exits it.

That rule is a trend-following-filter. It does not predict the next move. It waits until price action of the spread has already changed direction, then stays with that direction until an opposing extreme appears. Breakout-confirmation is the close through the prior 10-session extreme that admits the trade. The same extreme logic, reversed, ends the trade.

In the first six months, the procedure only buys July corn and sells July wheat after the July spread closes at or above its prior 10-session high. It exits that long-corn/short-wheat book when the same spread closes at or below its prior 10-session low.

In the second six months, the procedure only sells December corn and buys December wheat after the December spread closes at or above its prior 10-session low. It exits that short-corn/long-wheat book when the same spread closes at or above its prior 10-session high.

How the rules were exercised

Rules were exercised in a spreadsheet on 1991-1996 data, with December 1 through the last July-contract session used for the first rule set and a mid-June start used for the second rule set.

On the tabulated second-half sample, the December seasonal-breakout book recorded 19 trades, 14 of them winning. Reported results were stated in cents per 5,000-bushel spread and without commissions or slippage.

The author presents the seasonal-window as a way to cut counter-seasonal breakouts rather than as a multi-parameter optimizer. The procedure uses one 10-day lookback and two calendar halves, with the claim that the same filter idea can be applied beyond this grain pair.