I replayed 75 paid gold signals against real prices. Here's what happened

Signal groups sell a simple idea: someone better than you picks the trades, you copy them, and the profits roll in. I’ve paid for plenty over the years. This time, instead of trusting the screenshots, I tested one properly.

The group and the signals

It’s a Telegram channel that posts gold (XAUUSD) trades several times a day. I’m not naming it, because this test covers two and a half weeks, which isn’t enough to judge anyone for good. The point is the method, which you can use on any group.

A typical signal looks like this:

Buy (or sell) zone: a price range about $5 wide
Stop loss: about $11 away from the near edge of the zone
Take profits: TP1 about $3 away, TP2 further, TP3 about $10 away

I collected every trade signal the channel posted between 7 and 23 September 2026: 75 in total.

One thing I noticed straight away: 49 of the 79 trade posts were edited after they went out, mostly within a few minutes. I checked the posts that were never edited separately. They didn’t look any better, so the edits don’t change the result, but it’s worth knowing.

How I tested it

I replayed every signal against my own broker’s (Vantage) real one-minute gold prices, including the spread on every candle. The rules were strict, to keep it honest:

A buy limit only fills when the ask price actually reaches it, not just the chart price.
Market orders pay the spread plus a little slippage.
Stops slip a little too, and gaps fill at the next price, not your stop.
If a candle touches both the stop and the target, the stop wins, because you can’t know which came first.

All results are per 0.01 lot, where a $1 move in gold is worth $1.

Result 1: copy every signal straight away

This is what most people do: see the signal, buy or sell at the market price, with a $10 stop and a $10 target.


Trades	75
Win rate	49.3%
Average result per trade	−$0.21

A coin flip, minus the spread. The first half of the period and the second half both came out at roughly zero. Nothing here beats simply guessing.

Result 2: why the channel looked so good

The channel posted “TP1 achieved ✅” almost every day. So how does that fit with a coin flip?

The answer is in the entry. The signals use a limit order at the zone, so you only get in if price comes back to the zone. When I checked:

The limit order only filled about 1 time in 3.
In 49 of the 75 signals, TP1 was reached before price ever touched the entry zone.

In other words, most of the “wins” were trades nobody could have been in. Price moved in the right direction without coming back to fill the order first. The channel counts it as a win. Your account doesn’t see it at all.

What the win rate hides: layering into losers

The replay above follows each signal as it was first posted. But while I was in the group I noticed something the channel’s own numbers never show.

The group claims a 90–100% win rate. Here’s how. When a trade goes into drawdown, they tell members to add to it, opening another position (sometimes two or three) further against the move. That lowers the average entry. If price comes back even a little, everyone closes at break-even or a small profit, and it’s posted as a win.

The problem is when price doesn’t come back. I watched trades that had been layered two or three times carry on to the stop loss. Those were still reported as wins, or quietly never mentioned again.

This is what that does to an account:


Normal winning trade	one position, small profit, e.g. +$3
Losing trade after layering 3 times	three positions all hit the stop, e.g. 3 × −$11 = −$33

(Example sizes based on the group’s usual $3 first target and $11 stop, at 0.01 lot per position.)

One layered loss can wipe out ten or more wins. Then you spend the next two weeks trying to win it back. The win rate stays high because most trades do limp back to entry, but the losses are so much bigger than the wins that the account goes nowhere, or goes down.

If a group has a near-perfect win rate, ask two questions: do they add to losing trades, and do they post the losses?

Result 3: when the order did fill

When the limit did fill, TP1 was hit about 83% of the time. That sounds great until you look at the sizes: a $11 stop against a $3 first target. A high win rate with small wins and big losses can easily lose money.

The best version I found was following their full plan: take a third off at each of TP1, TP2 and TP3, and move the stop to break-even after TP1. That made +$1.96 per filled trade, but:

there were only 23 fills in the whole period
the statistical confidence was very low (a t-score of 0.79, where you’d want 2 or more)
in the second half of the period it made −$0.02 per trade

That isn’t evidence of an edge. It’s what luck looks like over 23 trades.

Result 4: my own idea, trading their stop level

I wondered if their stop-loss levels were worth trading the other way. If lots of people get stopped out there, maybe price bounces.

Price did reach their stop within 24 hours on 59 of the 75 signals. But after touching it, it bounced $5 first only 47% of the time (20 out of 43) and kept going $5 further the rest. Another coin flip. Every variation I tried lost money. Their stop isn’t a bounce level.

What I learned

“TP1 achieved” means nothing on its own. Always ask whether the entry actually filled first.
Look at the stop versus the target, not the win rate. An $11 stop against a $3 target needs a very high win rate just to break even.
Test before you pay. You can replay a group’s signals against real prices yourself. Mine took an evening.
Watch for edits. If posts change after they go out, screenshot them the moment they arrive.
Be wary of 90–100% win rates. They usually mean losing trades get layered until they limp back to entry, and the ones that don’t make it are big losses that aren’t counted.

The honest caveats

Two and a half weeks and 75 signals is a small sample. I can’t prove this group is bad over years. What I can say is that nothing in these 75 trades shows an edge once you use real prices, real fills and real costs.

That’s a big part of why I stopped following signals and started trading my own simple plan, logging every trade in public, wins and losses.