Killzone by volume burst instead of fixed clock

[LondonScalper]

I have been testing a different way to define my active window: wait for a volume burst instead of a fixed clock time.

The normal approach is a fixed killzone, say 07:00 to 10:00 London. My experiment keeps that outer window but adds a condition: I only look for continuation setups after tick volume on M1 exceeds roughly twice the average of the previous hour. Before the burst, I watch but do not trade.

Early results after about two months: fewer trades, slightly better average R, and fewer losses in the dead 07:00 to 07:45 stretch on quiet days. On busy days the burst comes almost immediately and nothing changes.

Some caveats. Tick volume on a retail feed is a rough proxy, not real volume. The threshold is arbitrary and I have not optimised it, because I do not want a curve-fitted number. And the sample is still small.

The fixed clock still runs everything else: news blackouts, my flatten time and the ticket limit. The burst only decides whether I look for continuation inside a window that was already allowed. I would not use it to extend trading outside my normal hours.

Has anyone tried something similar with real futures volume?

[Fairman]

Has anyone tried something similar with real futures volume?

The main thing futures volume gives you is size rather than update count, so a burst can't come from a feed simply sending more quotes. 6E or 6B volume during London is reasonable, though CME volume before about 07:00 London is thin enough that the previous-hour average itself is tiny, so a 2x threshold gets crossed easily early on. Comparing against the same time of day over the last 20 sessions instead of the previous hour would fix that. On the arbitrary threshold, you could check sensitivity without optimising: run 1.5x, 2x and 3x and see whether the results change direction or just degrade gradually. If small changes flip the result, the edge probably isn't in the burst.

[Shadow Trader]

LondonScalper, waiting for a volume burst inside the outer killzone is something I tested for about six weeks last winter. Fewer tickets, slightly better average R, and far fewer losses in the dead stretch before 07:45 on quiet mornings.

Fairman's note about a same clock baseline across prior sessions is the fix I needed. Using only the previous hour made early London look busy simply because the base was tiny, so a 2x threshold fired almost every day by 07:10.

I now use a 20 session same clock average for the threshold and keep the burst as a gate only, never as a reason to trade past my flatten time. The fixed clock still owns news blackouts and the ticket limit.

Sensitivity across 1.5x, 2x and 3x did not flip my results. That gave me more confidence that the edge was not hiding in the exact multiplier.

I have been on the same clock baseline for two months now. Have you tried it yet, or are you still on the previous hour average?