Oil (WTI) Seasonality: Average Monthly Returns Since 2000
How oil has performed in each calendar month since 2000: the average change and how often the month finished higher. Seasonal tendencies are weak and vary a lot from year to year. This is a statistical reference, not a trading signal.
Average change by month
| Month | Average change | Average | Years up |
|---|---|---|---|
| January | +2.0% | 54% | |
| February | +3.5% | 73% | |
| March | +2.6% | 58% | |
| April | +2.3% | 65% | |
| May | +3.0% | 46% | |
| June | +2.1% | 65% | |
| July | +1.5% | 69% | |
| August | −0.7% | 50% | |
| September | −0.9% | 41% | |
| October | −1.3% | 48% | |
| November | −2.7% | 42% | |
| December | +0.3% | 58% |
Month-end close versus the previous month-end close, from daily closes of front-month futures, 2000 to the last completed month.
What the pattern says about oil
February has had the highest average monthly change at +3.5%, and it has also finished higher most often, in 73% of years. November has been the weakest on average at −2.7%, and September has been up least often at 41% of years.
Averages hide wide differences. In any month the result can be far above or below the average, and most calendar months show only a small edge over a coin flip. Interest rates, currencies, supply and news usually matter much more than the calendar, so use this as background and not as a reason to buy or sell.
See also the oil price statistics, the live oil price and the economic calendar.
Frequently asked questions
What is the best month for oil?
Since 2000, February has had the highest average monthly change (+3.5%) and February has been up most often (73% of years). Averages hide large differences between individual years, so this is not a reliable timing signal.
What is the worst month for oil?
November has had the lowest average change (−2.7%) and September has been up least often (41% of years).
Is oil seasonal?
Some months show consistent tendencies, often linked to jewellery and festival demand, driving season, heating demand or planting cycles depending on the market. The pattern is weak compared with the effect of interest rates, currencies, supply and news.
How is this calculated?
For each calendar month we compare the month-end close with the previous month-end close, from 2000 to the last completed month, using daily closes of front-month futures, then average the results.
Information only, not investment advice. Past performance does not guarantee future results. Methodology.