How Long Does a Trade Last? 6 Signal and Timeframe Facts

A PatternForge signal usually holds until the next candle closes on your timeframe, so about 15 minutes on M15 and an hour on H1. Your trade is a different story, because it lasts until price touches your stop or your target, and that could be 20 minutes or three days.

Last Tuesday I took an H1 SELL, watched the clock pass the hour mark, and closed my position because I figured the signal had expired. Price dropped another 40 pips while I stared at my cold coffee. If you've ever mixed up how often a signal updates with how long a trade should run, this post saves you from that exact mistake.

Key Takeaways

• The card refreshes about every 5 seconds, but a new BUY or SELL direction usually appears only after a candle closes.
• There is no daily signal quota. On one symbol and one timeframe, expect 0 to a few clear direction changes per session.
• Signals follow your chart's candle closes and whether the tab is open, not a fixed time of day.
• PatternForge reads your chart's timeframe, or the one you pin. If it can't read one, it falls back to H1.
• SL and TP are prices, not timers. An H1 SELL does not end after one hour.

1. Refresh Rate vs. Direction Change vs. Trade Duration

Three different clocks run at the same time, and mixing them up is the most common mistake I see.

The first clock is the refresh. The second is the direction change. The third is how long your trade actually stays open.

Here's the same idea as a quick picture:

Poll every ~5s  -->  Often the same / HOLD  -->  New side after a close
                H1 does not mean the trade lasts 1 hour
                SL/TP are prices, not timers

Why does this trip people up? Because a card that updates every few seconds feels like a card that has new opinions every few seconds. It doesn't.

Polling is like checking your oven through the glass. You look often, but the cake only changes when it changes.

2. How Many BUY and SELL Signals Per Day?

Ever wonder how many signals you'll get in a day? The honest answer is that nobody can promise you a number, and PatternForge doesn't try.

There's no fixed daily quota. It's a live chart research overlay, not a scheduled signal newsletter that sends you three picks before lunch.

That last row confuses people, so here are the caps:

Those numbers describe how many signals you can keep saved. They say nothing about how many get generated per day.

So what's realistic? On one symbol and one timeframe, you may see 0 to a few clear BUY/SELL changes per session, not dozens of independent setups. Open more charts and timeframes and you'll see more updates, but they're still tied to bar closes and data quality.

I almost skipped the next point, but I think it's the most useful one. A long stretch of HOLD isn't the tool being broken. Sometimes the best trade is no trade, and I wrote a whole post about when no-trade is the edge.

And if history or the quote fetch fails, you'll see NO DATA. That's not a trade signal. It means the tool couldn't get what it needed.

3. What Time of Day Do Signals Appear?

You know that itch to wait for "the right hour" to trade? PatternForge doesn't work on that kind of schedule.

Signals show up whenever three things are true:

1. A supported chart page is open in Chrome.
2. Symbol and timeframe are detected, or pinned by you.
3. Market data for that symbol is available.

There's no built-in "signals only at 9:30" timetable. Your candle closes set the rhythm.

Notice that the session strip is a warning light, not a starting gun. I treated it like a gun for about two weeks. It cost me.

News is the same story. A warning near a big release is a nudge to think about your exposure, and that's why I keep a hard daily loss limit in place on volatile days.

4. How Long Does Each Signal Take to Calculate?

My dad used to time everything. Boiling water, red lights, the guy ahead of him at the bank.

So I'm the kid who timed the extension, too. Here's the thing: there are two kinds of "how long," and they have nothing to do with each other.

A) Time to compute one update. This is the technical part, and it's quick. With history cached, a typical update takes about 1 to 3 seconds.

A cold start or a slow data source can take longer. And polling every few seconds does not mean you get a brand-new edge every poll.

B) Time until the direction can change. This is the one that matters for trading. It usually lines up with the next bar close on that timeframe.

Until the recent window of bars moves, HOLD can last across many polls. That's expected, not a freeze.

Wait, let me rephrase that, because "expected" sounds dismissive. What I mean is: if the card says HOLD for forty minutes on H1, nothing is wrong. The market just hasn't given it a reason to change its mind.

5. Which Timeframes Does PatternForge Support?

Here's what nobody tells you about timeframes: the same symbol can show different results on M15 and H1, and both can be right.

That's because the bars are different, so the matches are different. The engine isn't locked to one timeframe. It follows your chart's interval when it can read it, or the one you pin.

Manual pin options commonly include M5, M15, M30, H1, H4, D1, and W1. If the page doesn't expose a timeframe, PatternForge defaults to H1 until you pin one.

This is the rambling paragraph, so bear with me. I started on M5 because it felt exciting, and I think I confused "lots of action" with "lots of opportunity," which is a bit like thinking a busy highway is a good place for a picnic. Then I moved to H1 and got bored for a week, then I tried D1 and got really bored, and somewhere in there I realized the timeframe wasn't the problem. My patience was. Anyway, I landed on M15 and H1 for most things.

Not sure which patterns you're even looking at on those charts? I put together 11 chart patterns every trader should know, and it pairs well with this.

6. Why an H1 SELL Doesn't End After One Hour

I made this mistake in public, so let me save you the embarrassment.

An H1 label tells you which candles the signal was built from. It doesn't tell you how long to hold. Your stop loss and take profit are prices, not timers.

Here's a made-up example to show what I mean:

The ratio is simple:

$$R = \frac{\text{Reward}}{\text{Risk}} = \frac{60}{30} = 2$$

So that's a 2R setup. The trade ends when price hits 1.0880 or 1.0790. If that takes 25 minutes, fine. If it takes two days, also fine. The clock on the wall has no vote.

What happens when you close early because "an hour passed"? You cut winners short and turn a decent 2R plan into a 0.3R scratch. Do that enough times and your win rate, risk-reward, and expectancy quietly fall apart.

It also sets you up for the next mistake. You close early, price runs, you jump back in angry, and now you're in revenge trading territory. Ask me how I know.

Here's the routine I use now. Try it on your next chart:

1. Open your chart and check the timeframe. Make sure it's the one you meant, or pin it manually.
2. Read the direction. BUY, SELL, HOLD, or NO DATA. Treat NO DATA as "no information," nothing more.
3. Write down your SL and TP as prices. Do this before you click anything.
4. Check the session strip and news warning. They're cautions, not triggers.
5. Walk away from the refresh. Look again at the next candle close, not the next five seconds.

I was eating a slightly bruised pear while I rewrote that list, which has nothing to do with trading. It was a pretty good pear, though.

If you want to see how the card, the historical matches, and the trade journal fit together on a live chart, the PatternForge AI Chrome extension is the place to start. And when you want the longer walkthrough, my PatternForge AI chart pattern signals guide covers the rest.

Frequently Asked Questions

Is there a daily limit on BUY/SELL signals? No. There's no fixed daily quota. Stored signals have tier caps (about 20 on Basic, 200 on Pro, 500 on Pro+), but that's storage, not output.

Why does the card refresh every few seconds if the signal rarely changes? Polling checks for updates. By default it runs about every 5 seconds, and you can set it anywhere from 2 to 30 seconds. A new direction usually needs a candle close.

What time of day do signals appear? Whenever a supported chart is open, the symbol and timeframe are detected or pinned, and data is available. There's no fixed schedule.

How long does one update take? Usually about 1 to 3 seconds with cached history. A cold start or slow data source can take longer.

Which timeframes can I use? M1 through MN are supported in the engine. Common manual pins are M5, M15, M30, H1, H4, D1, and W1. If the timeframe can't be read, it defaults to H1.

Does an H1 SELL mean I should exit after one hour? No. SL and TP are prices. The trade lasts until price reaches one of them.

Why am I seeing HOLD for so long? That's normal. HOLD can persist across many polls until the recent 20 to 30 bar window changes enough.

What does NO DATA mean? History or the quote fetch failed. It's not a trade signal.

One Last Thought

Signals update on candle closes. Trades end at prices. Once I stopped treating those as the same thing, my screen got a lot quieter and my results got a lot calmer.

So, quick question before you go: which timeframe are you watching right now, and what's your stop price? If you can't answer both in five seconds, that's your next task.