Gold Market Hours And Best Times To Trade

For the first four months of 2024, I traded gold exclusively in the Asia session — 11pm to 7am my time — because it fit my day job. I told myself the market was the market. The results told a different story: 63 trades, −$4,280, a win rate of 38%, and a journal full of entries like "stopped by $4 wick, then went my way." I was not a bad trader in those months. I was a decent trader fishing in a pond with no fish. When I finally plotted my P&L by session, the picture was brutal and obvious: my London-session trades were profitable (+$3,100 over 41 trades) and my New York trades were flat, while Asia had eaten everything. I had spent four months blaming my strategy for what was actually a scheduling problem. Gold market hours are not a footnote to your trading plan. For me, they were the entire difference between red and black.

Gold trades nearly 24 hours a day, but it does not trade the same at every hour. Volatility, spread, and follow-through concentrate in specific windows, and the best times to trade gold are narrower than most traders think. This is the full breakdown from my logs: when gold actually moves, how to score each session mathematically, the overlay tool I use to see session structure on the chart, and the mistakes that cost me $4,280 before I learned to check the clock first.

Scoring Every Session: The Numbers Behind "Best Times"

"Trade London and New York" is advice, not analysis. I wanted a formula — a way to score any hour of the day for gold tradeability and compare sessions objectively. So I built one from a year of my H1 data: average true range per hour (movement), average spread per hour (cost), and follow-through rate (what percentage of $10+ hourly moves extended another $10 in the same direction within three hours). The session score combines them:

London open scores (11.40 × 0.58) ÷ 0.32 = 20.7. Run the same math on the Asia dead zone (2am–5am GMT): ($4.10 × 0.31) ÷ $0.55 = 2.3. London outscores Asia nine to one — not because Asia is "bad," but because the formula exposes what Asia actually offers: small ranges, poor follow-through, and wider spreads, the exact cocktail that manufactured my 38% win rate. The CME Group's trading-hours data confirms the volume side of this: gold futures volume concentrates overwhelmingly in the London-New York overlap, and volume is what turns price movement into tradeable movement. Score your own broker's spreads by hour for a week — the numbers will tell you when to show up.

Setting Up the Market Structure Overlay for Session Trading

Knowing the best hours is useless if you cannot see them on the chart. The Market Structure Overlay draws session boundaries, key institutional levels, and structure markers directly on my gold charts, so the "when" is always visible next to the "where." Here is my exact configuration:

The kill-zone-only structure filter is the setting that changed my trading most. Before it, I took structure breaks at 3am GMT with the same confidence as 9am breaks — and paid for it. Now the overlay only flags breaks inside my two kill zones, and everything outside is visual context, not signal. My Asia session analysis goes deeper on why those hours underperform, but the short version is in the formula above: low score, low priority. The overlay also plots the prior day's high and low, which double as the stop-hunt magnets I wrote about in my carry-trade notes — session structure and institutional behavior are the same picture viewed from two angles.

Executing the Two Windows That Actually Pay

My entire gold week now fits into two windows: 7–10am GMT (London open drive) and 1–4pm GMT (New York morning / London-New York overlap). Everything else is either management of existing positions or no-trade time. The London window is the trend-setter: gold picks a direction for the day in the first two hours of London more often than not, and my log shows 63% of my winning day trades entered between 7 and 10am GMT. The playbook there is simple — wait for the first 30-minute range to form, then trade the break with the day's bias, stop on the other side of the range.

The New York window is the volatility window: US data, futures volume, and the overlap's liquidity combine to produce the day's largest hourly ranges — averaging $13.80 in my data versus $11.40 for London open. But it is also the choppiest: false breaks are 40% more common in the first 20 minutes after major US releases. My rule is no fresh entries for 15 minutes after high-impact US news, then trade the post-news trend. Outside those two windows, my rules are strict: Asia session — no new day trades, manage swings only. Late New York (after 4pm GMT) — no new entries; liquidity drains and spreads widen into the rollover. Friday after 3pm GMT — close day trades, never open new ones; weekend gap risk is not compensated by anything on a Friday afternoon.

Position sizing flexes with the window too: full size in the two kill zones, half size in the shoulder hours (6–7am and 12–1pm GMT), zero new risk elsewhere. This is not superstition — it is the session score converted into a risk budget. When I backtested the sizing overlay against flat sizing on the same signals, the window-weighted version added 0.31R per trade on average. Time of day is a position-sizing input, not just a scheduling preference.

Why Gold Behaves Like Three Different Markets in 24 Hours

The mechanics behind the session differences are structural, not mysterious. Asia hours (11pm–7am GMT) are dominated by regional hedging flows and thin order books — real money moving slowly, algos mostly idle. The result is tight $4–$8 ranges that look tradeable on M15 but rarely follow through; breakouts fail 69% of the time in my data. London (7am–12pm GMT) brings European institutional flow, the LBMA auctions, and genuine two-way price discovery — this is when gold establishes its daily direction. New York (12pm–5pm GMT) adds COMEX futures volume, US macro data, and the deepest liquidity of the day — the largest ranges, but also the most violent reversals as data reprices expectations.

The rollover row is the one beginners ignore and veterans respect: for about an hour around 5pm New York time, liquidity vanishes, spreads triple, and price can spike $5 on nothing. I have a standing rule — no entries, no stop adjustments, no "quick scalps" during rollover. The BabyPips market-hours guide covers the session overlaps well if you want the general forex framing; gold follows the same rhythm with roughly triple the amplitude. And note the spread column: Asia's $0.55 versus New York's $0.30 means every Asia trade starts nearly twice as deep in the hole. Session selection is cost selection.

Testing the Kill Zones on a Live NFP Week

First week of October 2025, NFP Friday ahead. I ran a deliberate test: trade only inside the two kill zones all week, using the Market Structure Overlay's session boxes as my on-chart referee, and log every setup the overlay flagged. Five trading days, same 0.5-lot size throughout, 1:2 fixed targets:

Five setups, four trades, three winners, +$2,020 on the week. But the row I am proudest of is Thursday: zero trades. The old me — the Asia-session me — would have forced two or three entries into that chop out of boredom and given back the week's gains. The overlay's kill-zone filter gave me permission to do nothing, which is the hardest permission in trading. Friday's NFP trade deserves a note too: I waited the full 15 minutes after the release, let the initial spike-and-reverse play out, then entered the post-news trend at 1:50pm. The overlay's session box showed me exactly where the NY structure had formed. Discipline about when to trade beat cleverness about entries, all week long.

The Session Mistakes That Cost Me $4,280

Mistake one was the meta-mistake: trading Asia for four months because it fit my schedule, never once checking whether my strategy fit the session. The −$4,280 was not a strategy failure; it was a calendar failure. If your life only allows Asia hours, trade a strategy built for Asia — range fades, not breakout momentum — or do not trade gold intraday at all. Mistake two: Friday afternoon entries. I lost $1,100 across three Friday-afternoon trades in 2024, all for the same reason — thin liquidity, unpredictable flows, weekend gap risk with zero compensation. Fridays after 3pm GMT are for closing, not opening. I have not broken this rule in over a year.

Mistake three: trading through the rollover. Twice in 2024 I left tight stops near the 5pm New York hour and got wicked out by $4–$6 spikes on nothing — spreads at $0.90+, price jumping like a startled cat. Now I widen stops or close positions before 4:45pm New York time. Mistake four: treating all of New York as one session. The 12–2pm GMT stretch (pre-US-data positioning) behaves nothing like the 2–4pm stretch (post-data trends) — my win rate is 61% in the latter and 44% in the former. Lumping them together cost me months of misattributed results. And mistake five, the subtlest: ignoring daylight-saving shifts. London and New York change clocks on different weekends, and for two to three weeks each year the overlap shifts by an hour. I once traded the "London open" an hour early for a full week before noticing. Check the clock, then check it again in March and October.

Gold Trading Hours: The Questions Worth Asking

What are the best hours to trade gold?
7–10am GMT (London open) and 1–4pm GMT (New York morning/overlap), by every metric I track: largest hourly ranges ($11.40 and $13.80), tightest spreads ($0.30–$0.32), and highest follow-through rates (55–58%). My session-score formula rates London open at 20.7 versus 2.3 for the Asia dead zone. If you can only trade one window, make it the London-New York overlap — it concentrates the day's volume and direction into four tradeable hours.

Can you trade gold 24 hours a day?
Technically yes — spot gold trades roughly 23 hours on weekdays with a one-hour halt around 5pm New York time. Practically, no: the Asia session offers $4 ranges with $0.55 spreads and 31% follow-through, which is a statistical desert for momentum strategies. I traded Asia exclusively for four months in 2024 and lost $4,280 learning that lesson. Trade when the market trades, not when your calendar is free.

Is gold more volatile in the London or New York session?
New York edges it on raw range — $13.80 average hourly versus $11.40 for London in my year of data — thanks to US macro releases and COMEX futures volume. But London is cleaner: fewer violent reversals, better follow-through on breaks. I take trend-continuation trades in London and treat New York as the breakout-and-reversal session. If you hate whipsaws, favor London; if you hunt big moves and manage reversals well, New York pays better.

Should I hold gold positions over the weekend?
Only swing positions with wide stops and a deliberate reason — never day trades, never tight stops. Weekend gaps of $10–$20 on the Sunday open are routine, driven by futures repricing on weekend news. My rule: day trades close by Friday 3pm GMT, swings keep stops at least $25 from Friday's close or get closed too. The gap is not your enemy if you sized for it; it is a margin call if you did not.

What happens to gold during the daily rollover?
Between roughly 5pm and 6pm New York time, liquidity collapses, spreads triple to $0.90+, and price can spike $5 on no news. Nothing tradeable happens in that hour — it is pure plumbing noise. I close or widen stops before 4:45pm New York and never open new positions until spreads normalize. Two of my 2024 stop-outs came from ignoring this hour; zero have come since I started respecting it.

How do daylight-saving changes affect gold trading hours?
London and New York shift clocks on different weekends each spring and autumn, so the overlap moves by an hour for two to three weeks twice a year. If your kill zones are clock-based (mine are: 7–10am and 1–4pm GMT), verify them against actual session opens in March and October. I traded the wrong "London open" for a full week once before catching it. Put a recurring calendar reminder on both clock-change weekends — it is the cheapest edge in this entire article.