forex-scalping.com

Some prop firms ban holding over the weekend, some ban it only on certain account types, some allow it. As a scalper you might think it doesn't concern you. It can, in a couple of ways.

• Friday afternoon trades. If the firm closes positions automatically at a fixed time (say 21:45 server time), a late trade could be closed in thin liquidity with a wide spread.
• Forgotten pending orders. A limit order left overnight on Friday may count as "holding" depending on the rules.
• Server time confusion. Cut-off times are in server time, which might differ from yours by one or two hours, and shifts with daylight saving.
• Violation versus closing. At some firms, holding is a hard breach, at others they just close the trade.

What I do:

1. No new trades after 19:00 London time on Fridays.
2. All pending orders cancelled on Friday at the end of my session.
3. A calendar reminder 30 minutes before the firm's cut-off time, in my local time.

Has anyone had a challenge failed for a weekend violation they didn't see coming? Curious about the details.

Look at returns across different time scales and the character of price changes. At the level of individual ticks, moves tend to reverse a little, thanks to bounce and market-maker behaviour. At a few minutes, there's a mild tendency for moves to continue, especially around news or heavy flow. At several hours, behaviour depends heavily on the day's events.

For a scalper, that matters because the intuition built on one scale doesn't carry over to another. A trader who learns that "price snaps back" on a ten-second chart may apply the same belief on a five-minute chart and get run over by momentum.

A few practical consequences follow.

A fade strategy tends to work better at very short holding times, where bounce dominates, and gets worse as the holding period lengthens. A breakout or trend-following approach starts to work as the horizon stretches into minutes, but trading costs eat into it at the shortest scales.

The crossover point varies by pair, by session and by volatility regime. On a quiet Asian EURUSD, the zone where reversion gives way to continuation might sit at thirty seconds. On a busy London GBPUSD, it might be fifteen.

Measuring it is simple. Compute the autocorrelation of returns at different horizons from tick or one-second data, and plot it. The shape usually shows a negative value at the shortest lags and a slight positive hump later.

That curve says more about where a strategy belongs than most indicators do.

At what holding time does behaviour switch on the pairs being traded?