Why most pre-market stops are false

US stocks don't trade only in the regular session. Before it there is pre-market, after it after-hours. In US Eastern time:

• pre-market: 4:00 – 9:30;
• regular session: 9:30 – 16:00;
• after-hours: 16:00 – 20:00.

Many brokers let stop orders work in these extended hours too. The problem is that prices in these hours can't always be trusted.

What we measured

Our scanner tracks TP and SL levels for every signal, including outside regular hours. In early September 2026 we examined these exits closely: after a stop triggered in pre/post, did that day's regular session actually take the price down to the stop?

As of 2 September 2026 we re-ran 482 signals in our database from scratch:

• with no filtering, 106 stops triggered outside regular hours;
• 56 of them were false — the lowest price of the following regular session never touched the stop.

So more than half of the extended-hours stops were caused not by a real drop but by a single isolated bad print. By strategy, the share ranged from 60% to 72%.

What a bogus print looks like

A few real examples:

• RCEL. Entry 4.65, first target (TP1) 4.84. In pre-market a price of 4.89 printed twice, at exactly the same price. 45 minutes later the market opened at 4.64 and never went above 4.64 all day. The report said "TP1 hit, +4%".
• INNV. Within a single 5-minute bar the price "jumped" between 12.09 and 10.00, then the session opened at 11.58. A −4% loss should actually have been a +2% gain.
• DLB. A lone 59.50 print in pre-market while surrounding prices were 61.19–61.21. The regular session never went below 60.02.

Why this happens

1. Very thin liquidity. Few buyers and sellers in pre-market. One small order can "fling" the price far away for an instant.
2. Volume is invisible. Many public data feeds don't provide volume at all for extended hours. When we measured, 100% of pre-market bars had zero volume — so it's hard to tell a real trade from a stale quote.
3. Data gets corrected later. Some bogus bars are later fixed at the source — but the stop has already fired.

A bogus print is isolated: nothing near it before or after. A real drop persists — the price keeps trading around that level for several bars.

The fix: confirm, don't ignore

The easiest option is to ignore extended-hours prices entirely. We tested that too, and it turned out to be one of the worst options. Why: when news breaks (a bad earnings report, say), the pre-market drop is real — and that's exactly when a stop should fire.

So we check a stop not against a single print but with confirmation:

• bars with impossibly long wicks are dropped entirely;
• the stop level must be confirmed by nearby closing prices (tolerance tied to the stop distance);
• in pre/post a level only counts as broken if two consecutive closes are below it.

On those same 482 signals, extended-hours stops fell from 106 to 43, and false ones from 56 to 10, while the real stops were kept. The problem isn't gone entirely, but most of it is closed.

Practical takeaways for traders

• Check which hours your broker's stop orders are active in (the "outside regular trading hours" setting).
• If you do use a stop outside regular hours, use a stop-limit, not a stop-market — otherwise one isolated print can sell your whole position at the worst price.
• When you see a sharp pre-market move, check whether there is news before acting. If there isn't, it's most likely thin-market noise.
• You can see our signals' results openly: track record.

This article is for information only, not investment advice. Figures are our internal measurement as of September 2026.