Gold seasonality, checked
Since November 2021, gold's average month was best in January (+4.6%, up in 3 of 5 years) and worst in June (-3.0%, up in 1 of 5). Five years of data is a small sample: the gaps between months are within the noise, so treat this as history, not a rule.
Gold month by month
| Month | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 | Average | Median | Up |
|---|---|---|---|---|---|---|---|---|---|
| January | - | -1.7% | +5.7% | -1.2% | +6.7% | +13.3% | +4.6% | +5.7% | 3 of 5 |
| February | - | +6.2% | -5.3% | +0.2% | +2.1% | +7.9% | +2.2% | +2.1% | 4 of 5 |
| March | - | +1.5% | +7.8% | +9.2% | +9.3% | -11.6% | +3.3% | +7.8% | 4 of 5 |
| April | - | -2.1% | +1.0% | +2.4% | +5.3% | -1.0% | +1.1% | +1.0% | 3 of 5 |
| May | - | -3.1% | -1.4% | +1.8% | +0.0% | -1.7% | -0.9% | -1.4% | 2 of 5 |
| June | - | -1.6% | -2.2% | -0.0% | +0.4% | -11.7% | -3.0% | -1.6% | 1 of 5 |
| July | - | -2.3% | +2.4% | +5.2% | -0.4% | +0.9% | +1.2% | +0.9% | 3 of 5 |
| August | - | -3.1% | -1.3% | +2.3% | +4.8% | +10.0% | +2.5% | +2.3% | 3 of 5 |
| September | - | -2.9% | -4.7% | +5.3% | +11.9% | -6.6% | +0.6% | -2.9% | 2 of 5 |
| October | - | -1.6% | +7.3% | +4.1% | +3.7% | - | +3.4% | +3.9% | 3 of 4 |
| November | -0.5% | +8.2% | +2.6% | -3.2% | +5.4% | - | +2.5% | +2.6% | 3 of 5 |
| December | +3.0% | +3.1% | +1.3% | -1.2% | +2.4% | - | +1.7% | +2.4% | 4 of 5 |
Change in gold from the last trading day of the previous month to the last trading day of the month (bid, minute data). Green - gold rose, orange - fell; deeper colour - bigger move. Months in progress are not shown.
By weekday
| Weekday | Trading days | Average change | Gold rose on |
|---|---|---|---|
| Monday | 257 | -0.027% | 49% of days |
| Tuesday | 260 | +0.074% | 55% of days |
| Wednesday | 258 | +0.089% | 54% of days |
| Thursday | 258 | +0.098% | 51% of days |
| Friday | 254 | +0.116% | 56% of days |
Close-to-close change of each trading day (day starts 6 pm New York time the evening before). With a typical daily range near 2%, an average of a tenth of a percent is far inside the daily noise.
How much of this is a pattern?
Seasonality in gold is real in the long run - jewellery demand around Indian festivals and weddings, central-bank buying and the January fund flows are often cited - but it is weak next to what drives gold day to day: real interest rates, the dollar and risk sentiment. In our five years, each calendar month has only 5 observations, and one strong year can move a month's average by several percent. A month that was "up in 4 of 5 years" can be a coincidence: with five years, four ups out of five happens by pure chance about one time in five for any month.
That is why we show every year behind the average, not only the average. Use this table to see what happened, not to time a trade; when more years accumulate, the same table will say more.
How we measure
- Daily closes of gold (bid) from the broker feed we record, from 27 October 2021; trading day from 6 pm New York time. Only complete trading days are used.
- The monthly change is the close of the last trading day of the month against the close of the previous month's last trading day, in percent.
- No adjustment for the price level: percent changes are compared across years.
Questions
Which month is best for gold?
January has the highest average change since November 2021 (+4.6%, up in 3 of 5 years), but with five years of data one strong year can decide it.
Which month is worst for gold?
June: an average of -3.0%, gold up in 1 of 5 years.
Is there a best day of the week to buy gold?
The differences between weekdays are a tenth of a percent or less, far inside gold's typical daily range of about 2%; in our data no weekday is distinguishable from noise.
Is gold seasonality reliable?
In the long run seasonal demand (festivals, weddings, central-bank buying) matters, but over five years each month has only five observations and the pattern is weak next to interest rates, the dollar and risk sentiment.
Page data last updated: · Measured and edited by Adrian Troy · How we measure