Does silver (XAG/USD) have good days, good months and good hours? We analysed 23 years of M1 tick data (7,269,441 candles between May 2003 and July 2026) and aggregated it by day of week, calendar month and trading session. Spoiler: silver has the strongest finding of all 4 pairs analysed — the January effect is a beast.
The dataset
1. Seasonality by day of week
| Day | n | Mean % | Median % | Std % | Win rate | t-stat | p-value |
|---|---|---|---|---|---|---|---|
| Monday | 1,202 | +0.028% | -0.006% | 1.950% | 49.5% | 0.50 | 0.614 |
| Tuesday | 1,202 | +0.035% | +0.035% | 2.030% | 50.9% | 0.59 | 0.553 |
| Wednesday | 1,198 | +0.110% | +0.149% | 1.918% | 55.6% | 1.98 | 0.048 ⭐ |
| Thursday | 1,200 | -0.051% | +0.020% | 2.238% | 50.8% | -0.79 | 0.431 |
| Friday | 1,181 | +0.070% | +0.134% | 2.286% | 53.3% | 1.06 | 0.290 |
Finding: Wednesday is bullish and significant (+0.110%, win 55.6%, p=0.048). Other days are noise, though Friday is also slightly positive. Silver shows a "Wednesday effect" that doesn't exist in gold, EUR/USD or GBP/USD.
2. Seasonality by month
| Month | n | Mean % | Median % | Std % | Win rate | t-stat | p-value |
|---|---|---|---|---|---|---|---|
| January | 23 | +5.740% | +3.778% | 7.095% | 73.9% | 3.88 | 0.001 ⭐⭐⭐ |
| February | 23 | +2.611% | +3.788% | 8.095% | 56.5% | 1.55 | 0.136 |
| March | 23 | +0.415% | -0.311% | 10.521% | 43.5% | 0.19 | 0.852 |
| April | 23 | +0.450% | -1.262% | 10.653% | 39.1% | 0.20 | 0.841 |
| May | 23 | +0.646% | +0.130% | 10.476% | 50.0% | 0.30 | 0.770 |
| June | 24 | -2.908% | -3.964% | 8.484% | 33.3% | -1.68 | 0.107 |
| July | 24 | +3.871% | +1.494% | 8.460% | 62.5% | 2.24 | 0.035 ⭐ |
| August | 23 | +1.504% | -0.648% | 9.848% | 47.8% | 0.73 | 0.472 |
| September | 23 | -1.209% | -0.554% | 11.385% | 47.8% | -0.51 | 0.616 |
| October | 23 | +1.931% | +1.629% | 7.061% | 69.6% | 1.31 | 0.203 |
| November | 23 | +1.642% | -0.911% | 8.719% | 47.8% | 0.90 | 0.376 |
| December | 22 | +1.723% | +2.655% | 10.034% | 54.5% | 0.81 | 0.430 |
Star finding: January in silver is brutal. +5.74% mean with 73.9% win rate (17 of 23 Januaries closed positive), p=0.0008 ⭐⭐⭐. It is the strongest finding across all 4 pairs analysed — surpassing gold's January (+3.44%, p=0.005) and GBP/USD's April (+1.20%, p=0.020). The silver January rally is tied to Indian jewellery demand (Akriti) and post-Christmas re-leveraging.
July also significant: +3.87%, win 62.5%, p=0.035 ⭐. Silver has a "summer rally" that gold doesn't show as clearly.
June is the worst month: -2.91%, win 33.3% (8 of 24 Junes closed negative). Silver tends to correct in June, consistent with the "sell in May and go away" pattern applied to metals.
3. Seasonality by trading session (UTC)
| Session | n | Mean drift % | Win rate | p-value | Mean range % | |Gap| % |
|---|---|---|---|---|---|---|
| Asian (00–07 UTC) | 5,932 | -0.0055% | 48.9% | 0.645 | 1.797% | 0.158% |
| London (07–12 UTC) | 5,943 | +0.0059% | 50.1% | 0.626 | 1.856% | 0.614% |
| NY (12–17 UTC) | 5,954 | +0.0097% | 51.0% | 0.593 | 2.611% | 0.856% |
| NY late (17–22 UTC) | 6,062 | -0.0383% | 44.7% | <0.001 ⭐ | 1.602% | 1.270% |
| Weekend gap (22–24 UTC) | 5,532 | +0.0765% | 63.3% | <0.001 ⭐⭐⭐ | 0.936% | 1.168% |
Findings for XAG/USD:
- Asian and London neutral: silver shows no directional drift in these sessions.
- NY neutral with slight positive bias: +0.010%, p=0.593. Not significant, but silver moves most here (range 2.61%).
- NY late bearish and significant: -0.038%, p=0.0001 ⭐. Silver drops in the last NY hours — probably leveraged position closing before weekend.
- Weekend gap VERY bullish: +0.077%, win 63.3%, p < 1e-30 ⭐⭐⭐. Strongest session finding across all 4 pairs. Silver closes Friday with very robust positive momentum.
4. Heatmap: day × session
| Day | Asian (00–07) | London (07–12) | NY (12–17) | NY late (17–22) |
|---|---|---|---|---|
| Monday | -0.009% | +0.019% | +0.018% | -0.076% |
| Tuesday | -0.022% | +0.001% | +0.005% | -0.037% |
| Wednesday | +0.015% | -0.017% | +0.058% | -0.040% |
| Thursday | -0.023% | +0.013% | -0.062% | -0.064% |
| Friday | +0.011% | +0.015% | +0.030% | +0.021% |
Unique silver finding: Friday is universally bullish. All 4 Friday sessions are positive (+0.011%, +0.015%, +0.030%, +0.021%). It is the only day where ALL cells are green. Friday NY is the strongest Friday cell (+0.030%).
This contrasts with GBP/USD (Friday universally bearish) and gold (Friday positive but NY late neutral). Silver has the cleanest Friday behaviour of all 4 assets analysed.
5. Top most volatile hours (UTC)
| Hour UTC | Mean range % | Std % | Session |
|---|---|---|---|
| 13:00 | 2.089% | 1.225 | NY open |
| 14:00 | 2.012% | 1.196 | London/NY |
| 12:00 | 1.929% | 1.160 | Pre-NY |
| 11:00 | 1.599% | 0.991 | London |
| 15:00 | 1.556% | 1.002 | London/NY |
| 19:00 | 1.474% | 1.098 | NY late |
Silver is the most volatile of all 4 analysed: NY open (13:00 UTC) has mean range of 2.09% — more than double EUR/USD (0.38%) and GBP/USD (0.38%), and almost double gold (1.08%). 12:00-15:00 UTC is silver's hot zone, same as the others but with much larger ranges.
---6. Three practical setups for XAG/USD
Setup 1 · Long bias on Friday NY
- When: Friday 12-17 UTC and 17-22 UTC (NY + NY late).
- Why: Friday is the only day with all sessions positive. Friday NY (+0.030%) and Friday NY late (+0.021%) are the strongest end-of-week cells.
- Application: for scalpers, Friday NY is silver's best directional drift window.
Setup 2 · Long bias on Wednesday NY
- When: Wednesday 12-17 UTC.
- Why: +0.058% drift, the strongest cell in the heatmap. Matches the significant Wednesday aggregate (+0.110%, p=0.048).
- Application: "Wednesday NY silver" is the cleanest weekly setup.
Setup 3 · Long bias in January (seasonal filter)
- When: all of January.
- Why: +5.74% mean, win 73.9%, p=0.001 ⭐⭐⭐. The January effect on silver is the strongest of all 4 pairs analysed.
- Application: add as a filter to your system. If your base edge is trend-momentum, opening longs in silver during January has historically worked.
What this data does NOT tell you
- Silver is the most volatile: mean ranges 2-3x gold's and 5x EUR/GBP's. This means wider stop-losses or smaller position sizes. Recommended silver position size is ~50% of what you'd use on gold.
- January effect is robust but has exceptions: 6 of 23 Januaries closed negative. Not every January is bullish — crisis Januaries (2008, 2020) saw liquidations.
- Friday NY late is NOT bullish in 2020+: in the post-COVID sub-period, that cell flipped from +0.021% to negative. The "Friday bullish" is weakening.
- Transaction costs: typical retail silver spread is higher than gold (3-8 pips vs 1-3 pips). You need an ECN broker with spread <3 pips for serious trading.
- Silver has two engines: monetary (like gold) and industrial (like copper). The January effect is dominated by the monetary/jewellery side. The industrial side (solar panels, electronics demand) is more correlated with the manufacturing cycle.
Methodology
- Source: M1 tick data (Dukascopy) for XAGUSD between May 2003 and July 2026.
- Total: 7,269,441 M1 UTC candles, aggregated into 5,983 trading days and 278 months.
- Daily return: (close_22UTC[t] - close_22UTC[t-1]) / close_22UTC[t-1] × 100.
- Monthly return: (close_lastDay - close_lastDayPrevMonth) / close_lastDayPrevMonth × 100.
- Session drift: (close_last_bar - open_first_bar) / open × 100.
- Statistical test: one-sample t-test vs zero. ⭐ = p < 0.05.
- Software: Python 3.14, pandas 3.x, pyarrow, scipy, matplotlib.
For volatility-adjusted lot sizes on silver, use the position size calculator. For more on precious metals, read how to trade gold and silver by hours.
The full seasonality series
This study is part of a series applying the same methodology to six assets. The contrast between them is the most useful part: the index "Monday rally" does not exist in gold or the FX majors.
- XAUUSD (gold) — 23 years, best day Friday and best month January.
- EURUSD — the world's most traded pair, 23 years.
- GBPUSD — the London session and its intraday bias.
- Dow Jones (USA30) — the index "Monday rally".
- NASDAQ-100 — the strongest Monday of the six (p=0.0001).