Advertisement IC Markets

Market closed errors at open: clock and session mismatches

Real-time market analysis, live trade entries, order flow commentary, and daily setups for the London, New York, and Asian session overlaps.
Post Reply
LondonScalper
Posts: 965
Joined: Sat Sep 05, 2026 7:54 am

Market closed errors at open: clock and session mismatches

Post by LondonScalper »

If you see "market closed" at the open, check your clock before you blame the broker.

Every time this has happened to me, the cause was on my side. Once it was the VPS clock drifting four minutes after a Windows update disabled time sync. Once it was a session template in my EA that still used summer time in November. Once it was me trying to trade a symbol whose trading session on that broker started an hour later than I assumed.

My checklist when it happens:
- Compare the VPS clock with an internet time server.
- Open the symbol specification and read the actual trading sessions for that day.
- Check if a holiday schedule applies; brokers publish these and they change.
- If an EA is involved, print the time it thinks it is to the log.

Only after that do I message support, and when I do, I include the server time from the platform, not my own.

The DST weeks in March and October are the worst, because the UK, US and broker server all change on different dates. I mark those weeks on my calendar and double-check every time-based rule before the first session.
Recommended broker for automated trading & scalping IC Markets
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Market closed errors at open: clock and session mismatches

Post by Shadow Trader »

One more cause for your list: the first quote of the session arriving later than the session start on the spec sheet. My broker lists the Sunday open at the server's midnight, but on a quiet week the first EURUSD tick has landed two or three minutes after that, and every order sent before the tick came back as market closed.

So my EA no longer trusts the clock alone. It waits for a fresh tick with a timestamp inside the session, then adds a further two minutes before it is allowed to send anything.

The other one that looks identical is a symbol switched to close only. After an account migration I had a gold symbol that still showed prices but rejected new orders, and the journal message looked so similar that I lost twenty minutes chasing the clock.

On the DST point, servers that run on New York close time move with the US date rather than the UK one. The two or three week gap in March and the one week gap in autumn are when a session template written in London time is most likely to be an hour out.

Do you log the server time of the first tick each Sunday? It is a cheap way to spot drift before the Monday open.
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Market closed errors at open: clock and session mismatches

Post by Shadow Trader »

LondonScalper, one small disagreement on the first item of your checklist. On MT4 and MT5 the session check happens on the broker server, so a drifting VPS clock only bites when the EA reads local time instead of server time.

In my own code that meant swapping every TimeLocal call for TimeCurrent. The drift problem went away even before I fixed time sync on the machine.

The other window that throws market closed for me is the daily break around rollover. On my broker the indices pause from 23:59 to 00:05 server time, and an EA trying to trail a stop during those six minutes gets the same rejection a new order would.

Retry logic is worth a look while you are in there. My old EA resent every second, which meant around 360 rejected requests across one of those breaks, and the broker emailed to ask what was hammering the server.

Now it backs off to one attempt every 30 seconds after the second rejection, and writes the error code plus server time to the log. That single line has told me more than any support ticket.

Does your EA treat a rejected stop modification differently from a rejected entry? I think the stop case deserves a loud alert, since the position is still open.
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Market closed errors at open: clock and session mismatches

Post by Shadow Trader »

Twice a year, the United States and Europe change clocks on different dates. American daylight saving begins in the second week of March and ends in the first week of November, while Europe moves its clocks at the end of March and the end of October. For a few weeks, the usual time gap between London and New York is one hour shorter or longer than normal.

The effect is subtle but real. During those weeks, the London-New York overlap shrinks to three hours or stretches to five, and fixed clock-time rules shift relative to the actual market. Data releases scheduled in New York time land at a different point in the European session than participants have grown used to.

Australia and New Zealand add another layer, since their clocks change in the opposite season. The Sydney open therefore moves relative to Tokyo and London twice a year, altering when Asian liquidity thins or builds.

Strategies defined by a local clock, such as a rule to trade only the first ninety minutes after the London open, can drift without anyone noticing. Anchoring session rules to exchange hours rather than to a broker's server clock, and rechecking them during the odd weeks, reduces the chance of an unplanned mismatch.
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Market closed errors at open: clock and session mismatches

Post by Shadow Trader »

A losing day creates pressure to repair something, and the nearest thing to modify is the system. Indicator settings get tweaked, targets get shrunk, and a sound method is rewritten on the strength of a handful of trades. A structured review can make the urge to tinker more productive.

One approach sorts every loss from the day into three buckets. The first holds execution errors, where a rule was broken, such as an early entry or a moved stop. The second holds valid trades that simply lost, where every criterion was met and price went the other way. The third holds environment mismatches, such as running a range strategy into a trending release.

Each bucket points to a different response. Execution errors call for behavioural work, not strategy changes. Valid losers call for nothing, since they are the cost of an edge that is probabilistic. Environment mismatches may justify a new filter, but only after the same mismatch appears repeatedly across many sessions.

The system itself should be altered only on a schedule and with a sample large enough to mean something. Until then, the red day is filed, labelled, and left alone, and the next session begins from the same rules.
Post Reply