Oil’s March-November gap matched random labels 81% of the time
SignalPilotLabs found that WTI’s March-to-November spread was no more unusual than random labels in 81% of 10,000 tests, undercutting calendar-based oil trades.
By Grace Kim
4 min read

What's News
- SignalPilotLabs measured 201 monthly WTI spot changes at Cushing from January 2010 through September 2026.
- March had the highest median monthly change, +4.6%, and rose in 10 of 17 years.
- November had the lowest median change, -2.9%, and rose in 7 of 16 years.
- A March-November gap of 7.5 percentage points matched random labels in 81% of 10,000 shuffles.
- November rose 3.0% in 2010-2017 but fell 6.1% in 2018-2026.
By Grace Kim
SignalPilotLabs found that the 7.5-percentage-point gap between WTI crude oil’s strongest and weakest calendar months matched random labels in 81% of 10,000 simulations.
How did SignalPilotLabs measure 201 months of WTI?
SignalPilotLabs examined every monthly WTI spot-price change at Cushing from January 2010 through September 2026, totaling 201 observations. The study used FRED’s DCOILWTICO series on TradingView.
The methodology compared each calendar month’s close with the previous month’s close. SignalPilotLabs noted that spot prices contain no futures roll within a month, so the series does not introduce a contract rollover into the measurement.
January through September each supplied 17 observations. October through December each supplied 16. For every month, SignalPilotLabs calculated the median change and counted the years in which WTI rose.
Plain takeaway: SignalPilotLabs ranked close-to-close spot returns without a futures roll inside the measurement.
Which months performed best and worst?
SignalPilotLabs’ ranking put March first and November last. Its full table shows each month’s median change and the number of years WTI rose:
- March: +4.6%, up in 10 of 17 years.
- December: +4.5%, up in 9 of 16 years.
- February: +3.9%, up in 11 of 17 years.
- April: +3.1%, up in 11 of 17 years.
- June: +2.6%, up in 10 of 17 years.
- July: +0.6%, up in 11 of 17 years.
- January: -0.4%, up in 8 of 17 years.
- August: -0.4%, up in 7 of 17 years.
- October: -1.0%, up in 7 of 16 years.
- September: -1.8%, up in 7 of 17 years.
- May: -2.1%, up in 7 of 17 years.
- November: -2.9%, up in 7 of 16 years.
The March-November spread was 7.5 percentage points. SignalPilotLabs’ shuffle test says that distance was ordinary rather than distinctive.
Is the 7.5-point spread real or luck?
SignalPilotLabs dealt the 201 monthly changes into random month labels 10,000 times. In each run, the author measured the best-to-worst gap among the 12 labels.
SignalPilotLabs noted that 12 labeled months will always produce a best and a worst. The test measured how wide a gap random labels could produce on their own.
A gap as wide as 7.5 points came up in 81% of the shuffles. In SignalPilotLabs’ assessment, the observed spread fell within the range random labeling alone could create.
What changes when the sample is split?
SignalPilotLabs divided the sample into two periods. February ranked best and September worst from 2010 through 2017; June ranked best and November worst from 2018 through 2026.
November’s result reversed across those periods: it rose 3.0% in the first half and fell 6.1% in the second. SignalPilotLabs used the reversal to show how unstable a calendar-month label can be across periods.
Plain takeaway: the winning and losing months changed when SignalPilotLabs changed the period.
SignalPilotLabs also checked TVC:USOIL, the futures CFD most charts show. The CFD produced the same March and November results, so SignalPilotLabs concluded that the roll was not responsible.
What does the chart show?
SignalPilotLabs’ chart uses monthly candles for the TVC:USOIL CFD from 2010 through 2026. The blue line traces a “typical year”: each month moves by its median change, then the line starts again from every December close.
In the chart, the shaded area shows the gap between that line and what oil actually did. SignalPilotLabs said most years leave the blue line far behind.
SignalPilotLabs drew the line after the first build in place of two strips of March and November cells, and said the figures remained unchanged.
Plain takeaway: SignalPilotLabs presents the chart as a comparison with a typical path, not a forecast.
Should traders use March or November as an oil signal?
SignalPilotLabs advised: “Don't plan an oil trade around the calendar month.”
SignalPilotLabs also found that gold’s worst month held up under the same test: only 2% of random shuffles matched its gap. Oil’s months did not.
SignalPilotLabs said the analysis cannot show which way oil will go in any month. Its limits are 17 years, one market and past data. The author closed: “Observations, not recommendations.”
The practical bar for any future oil-seasonality claim is the same random-label test; this sample does not clear it for March or November.
Source: tradingview.com
More from Grace Kim
Show full bio
Market editor covering people, launches and day-to-day practice in entrepreneurship, startups and venture funding at Business Bearings.
2,592 articles