DST shift checklist so London open is still London open

[LondonScalper]

Clocks go back in the UK on the last Sunday of October, and in the US a week later. That one-week gap, plus a similar one in March, is when my process is most likely to break. This year I've written a checklist, and I'm sharing it because I'm sure I'm not the only one who's been caught.

Before the UK change:

Check the broker's server time. Many brokers keep their server on a time zone that moves with New York, so my chart's candle times relative to London shift by an hour at different points. Note the new offset.

Check every EA or indicator that uses hard-coded hours. My session-range indicator uses server time inputs. I keep a spreadsheet with the correct inputs for each period.

Review the calendar for the gap week. US releases at 13:30 London will come at 12:30. The US cash open moves from 14:30 to 13:30.

Check alarms on my phone that reference market events rather than clock times.

After the UK change:

Confirm that the Asian range indicator is measuring the right hours. Look at the chart and check that the range boxes line up with what you expect.

Make sure the daily rollover break is where you think it is.

After the US change:

Repeat all of the above, put the inputs back, and check again on Monday morning before 07:00.

It's tedious. It's also about twenty minutes, compared with a morning spent trading the wrong range.

[Fairman]

Clocks go back in the UK on the last Sunday of October, and in the US a week later.

Timely one, with the UK change on 25 October this year and the US on 1 November. One more item for the list: some markets in the routine don't change at all, or change in the other direction. Japan has no daylight saving, so anything keyed to Tokyo, like the 09:55 fix, moves by an hour against London on the UK change and again versus a New York based server on the US change. Australia moved its clocks forward in early October, so Sydney is already on a different offset from a month ago, and AUD data times relative to London shift twice in a few weeks. Checking the Asian range box on the Monday after each change, as you say, catches most of it.

[Shadow Trader]

Your checklist is basically what I run, and Fairman's point on Japan and Australia is the bit that used to catch me.

I keep a one page sheet with four columns: UK winter, UK summer, US winter, US summer. For each I write the broker server offset, the Asian range start and end in server time, and the local clock time of 13:30 US releases. Updating that sheet on the Sunday of each change is the whole job.

The EA hard-coded hours are the dangerous ones. I had a session filter that still thought London open was an hour late for a full week after a US change. It filtered out the real open and let me trade the dead zone instead.

I also move phone alarms to event names where I can, like "US cash open", not "14:30". The ones that stay as clock times get a recurring calendar reminder two days before each DST change.

Twenty minutes of checks is cheap against one morning of trading the wrong Asia range.

Do you change anything on the prop platform side, or only on the personal terminals?

[Shadow Trader]

One more item for the Monday after each change is the calendar app on my phone. Some apps shift event times on their own, while others keep last week's local times until you refresh them.

I once sat waiting for a 13:30 print that had already gone off at 12:30 because the app was stale. Now I find the next big US release after each change and check its local time against a second source before the week starts.

I also look at the rollover on the chart late on the Sunday. If the daily candle break is not where my sheet says it should be, the server offset note is wrong, and I fix it before Asia builds a range I will misread.

The gap week between 25 October and 1 November is still the messiest for me. I run half size on any trade that depends on a timed session filter that week, because that is where a wrong input hurts most.

LondonScalper, is your broker server on New York time, or on a fixed offset that never changes? I've found a fixed offset easier, since only the event times move and the server clock stays put.