Guides · FARO team · July 15, 2026 · 6-minute read

How to tell if a trader is reliable: 7 checks before you follow anyone

Every day, thousands of people start following a trader because of a screenshot: a winning trade, a green percentage, an “I told you so”. And every day, some of those people lose money following someone whose real track record they never saw — because nobody showed it to them in full.

This guide is a list of concrete checks to evaluate any analyst or signal channel before you listen to them. You don’t need to know about trading to apply it: you need to know what to ask.

1. Ask for the complete list, not the highlights

The simplest check, and the one almost nobody does. A reliable trader can show you all their trades from the last few months: the winners and the losers, with dates. One who isn’t will show you a selection.

The usual trap is called selection bias (cherry-picking): if someone publishes 30 ideas a month and only shows you the 10 that went well, their “track record” is spotless and their reality can be ruinous. The question that exposes it: “Can I also see the ones that went wrong?”. The reaction to that question tells you more than any percentage.

And it is not a theoretical precaution. We went to count how many show it in full: we analyzed 25 Spanish-language trading signal channels and none publishes the result of all its trades, not even as an aggregate summary. What stands out in the finding is that they hardly promise returns either: they don’t lie with numbers, they simply don’t publish the ones that would let you judge them.

2. Check whether they delete their failures

On Telegram, X or YouTube, scroll back two or three months in their channel. Look for specific old calls (“buy X”, “support is at Z”) and check what happened next.

• If the failed calls are still visible, that is a good sign: they own their mistakes.
• If the old history is suspiciously clean, or only the wins survive, you know how they work.

A channel can delete any message without a trace. That is why a track record on an editable platform, however nice it looks, is a statement of intent, not proof.

3. Be wary of percentages without a sample size

“90% win rate” means nothing on its own. Over how many trades? In what period? Counting the ones that were left open?

• 90% over 10 trades can be luck (and probably is).
• 60% over 200 trades, with the losses included, is a serious figure.
• Professional traders rarely sustain win rates above 55-65% over the long term. Whoever promises more either has a small sample or doesn’t count everything.

Rule of thumb: a percentage without its “n” (number of trades) next to it is marketing, not statistics. That is how FARO’s methodology applies it: every win rate comes with its sample, and with few trades it is marked as provisional.

4. The win rate is not enough: ask about returns — and about open positions

You can be right 70% of the time and lose money: all it takes is winning little on the wins and losing a lot on the losses. That is why the figure that matters is not only how often they are right, but how much is left at the end.

And a nuance almost nobody looks at: do the returns they show include open positions valued at today’s price, or only the closed ones? The classic trick is to close the winners quickly (and show them off) while the losers stay “open” indefinitely — so they never count as a loss. If the published return is only for closed trades, ask how many are still open and in what state.

5. Check whether their signals were really executable

There is a subtle way to inflate a track record: publishing entries at prices the market barely touched. If the signal said “buy at 100” and the price never went below 102, nobody could make that trade — but in the channel’s history it counts as a win if the price rose afterwards.

The metric that measures it is called fill rate: of all the published signals, in how many did the price really touch the entry? A low fill rate means the track record is theoretical, not tradable. Few platforms publish it; ask for it.

6. Look for the conflict-of-interest disclosure

An uncomfortable but necessary question: do they hold a position in what they recommend? Holding one is not bad — not saying so is. The European rules on investment recommendations (the Market Abuse Regulation) require anyone who disseminates recommendations to disclose their conflicts of interest. A serious analyst does it without being asked: “I recommend X and I have held a position since Z”.

And the serious version: be especially wary of recommendations on small, illiquid securities, where the recommender themselves can profit from their followers buying after them.

7. Follow the money: how do they make a living?

• If they live off a paid VIP channel, their incentive is to sell you the subscription — and the winning screenshots are their advertising.
• If they live off broker commissions for every account you open with their link, their incentive is for you to trade a lot, not to trade well (and they should disclose that affiliation).
• If their complete track record is public and verifiable, and they ask nothing of you to see it, their incentives and yours are a lot more alike.

None of these models is illegitimate in itself. But knowing the incentive tells you what they are optimizing for.

If an analyst passes all seven, it does not mean they will be right next time — past performance never guarantees future results. It means something more valuable: that you can trust that what they show you is true.

Why we wrote this

FARO exists because these checks should be automatic, not detective work. On getfaro.org, every signal an analyst publishes is sealed within a few minutes —it cannot be edited or deleted from the app, and any intervention is recorded— and it closes by itself at the real market price. The complete track record, with wins and losses, sample size, fill rate and declared conflicts, is public for anyone. Free.

With one exception, which is stated on their profile: whoever publishes in deferred mode does not show the result of their live positions, because publishing it would reveal their entry and their stop. For them, FARO says how many they have at stake, counts them in their levels as the maximum loss their stop allows (−1R) and reveals them in full when they close, with their result and their conflict-of-interest disclosure, win or lose.

We don’t give signals or predict anything. We just make the list above check itself.

And if what you are evaluating is an automated system, there is one more check the list above does not cover: that its parameters have not been touched along the way. How to commit to them before you start, in a way that can be verified from outside, is in sealing parameters in advance: how to make your forward test verifiable.