Reliable Trading Signals: The 7-Point Checklist to Run Through

A reliable trading signal is recognisable by four things: a described calculation logic, validation on the candle close, a stop and targets announced before entry, and a complete track record you can inspect trade by trade. If any one of the four is missing, you are not following a signal: you are following an opinion. The seven points below turn that intuition into a control procedure.

Point 1 — Insist that the calculation logic be described

You don't need the source code. You need to know what the signal is built on: a moving average crossover, a trend-following indicator, a break of structure, a statistical model. That information changes everything, because it tells you which market regime the method will struggle in.

A trend-following system loses regularly in a tight range. That's not a hidden defect, it's its nature. If the provider owns it and explains it, you'll know that three losses in a row in a flat market are normal. If they refuse to say anything at all — "proprietary algorithm", "artificial intelligence" with no further detail — you'll quit at the first bad run, with no framework to interpret it.

The question to ask: "Under what conditions does your method fail?" A serious provider answers in three sentences. A salesman changes the subject.

Point 2 — Check that the signal is validated on the close, not mid-candle

This is the most technical point and the one most often dressed up. Many indicators compute their state on the current candle. As long as that candle isn't finished, its state can flip several times. A buy arrow appears, price pulls back, the arrow disappears. On the historical chart, it never existed.

The consequence: the track record you see contains only the signals that survived. The others were erased by the calculation itself. This is called repainting, and it's enough to make any method look superb on screen and mediocre live.

The test is simple. Log every signal the moment it appears, with the time and the price. A week later, compare your notebook to the chart. If signals have vanished, you know where you stand.

That is exactly why the PIPSTER PRO engine only validates a Supertrend signal on Heikin-Ashi at the candle close: a published signal stays in the track record, whether it ends up a winner or a loser.

Point 3 — Refuse any signal without a stop announced in advance

A signal that says "buy XAUUSD now" with no invalidation level is unusable. You can't work out your position sizing, you can't measure your risk, and you can't judge performance afterwards. The stop isn't an accessory: it's the unit of measurement for everything else.

Also check how the stop is built. Two families dominate:

• Volatility stop (ATR): a distance proportional to recent movement. Consistent, mechanical, but sometimes parked in the middle of a zone where price comes to hunt liquidity.
• Structural stop: placed beyond the last significant swing low or high, with a volatility-based buffer. Wider, but anchored to a level the market genuinely has to break to invalidate the idea.

A provider who offers both and explains when to pick one over the other shows they've thought about execution, not just entry.

Point 4 — Ask for targets expressed in R, not in pips

A "+300 pips" target means nothing until you know the stop distance. 300 pips of gain for 400 pips of risk is a bad deal. 300 pips for 100 pips of risk is a good one. Only the ratio matters, and that ratio is written in R: 1 R is the entry-to-stop distance.

Here's a full illustration. A $10,000 account, risk fixed at 1% per position, i.e. $100.

• Buy signal on gold, entry at $2,412.00.
• Structural stop below the last swing low, at $2,398.00 — that's $14.00 of risk, so 1 R = $14 per ounce.
• Target 1 at 1 R: $2,426.00. Target 2 at 2 R: $2,440.00. Target 3 at 3 R: $2,454.00.
• Size: $100 ÷ $14 = 7.14 ounces, i.e. 0.07 lot rounded down.

If the stop is hit, the loss is roughly $98, i.e. −1 R. If target 2 is reached, the gain is roughly $196, i.e. +2 R. The maths is identical on a $500 account or a $100,000 one: only the number of lots changes. That's why levels in R are universal, and why the position size calculator is the first reflex before every order.

Point 5 — Demand a verifiable track record, trade by trade

A screenshot of profits is not a track record. A track record is the complete sequence: every signal, its entry, its stop, its outcome, in chronological order, with no gaps. Losing months included.

What you're looking for in it:

1. The total number of signals. Below thirty, no conclusion is possible.
2. The worst run of consecutive losses. That's what decides whether you'll hold on psychologically.
3. The average win/loss ratio in R, cross-referenced with the win rate.
4. The distribution of the gains. If three trades explain the whole result, the method depends on calendar luck.
5. Consistency between calm periods and volatile ones.

One point often forgotten: the order of testing within a single candle. If one candle touches both the stop and the target, which outcome does the provider record? The honest answer is the stop. Any backtest that assumes the target first mechanically overstates the result. The PIPSTER PRO backtest is replayed candle by candle with that pessimistic order openly applied.

Point 6 — Check latency, delivery channel and reproducibility

A perfect signal received twenty minutes too late is no longer the same signal. Check three concrete things.

Latency. Between the candle close and the alert landing, how many seconds? On a 5-minute timeframe, a minute of delay can eat a significant share of your 1 R.

The channel. A sound and browser alert at the moment of the signal beats a message buried in a chat group. You need to know when to look, not to monitor permanently.

Reproducibility. Looking at the chart, can you understand why the signal appeared there? If yes, you'll be able to decline it when the context doesn't suit you — fifteen minutes before a macro release, for instance. If not, you're dependent.

Point 7 — Measure what the signal doesn't cover: gaps, correlation, news

No provider controls execution. Three blind spots remain your responsibility.

Gaps. At the Sunday evening open, or after a surprise announcement, price can jump beyond your stop. The order is then filled at the first available price and the loss exceeds 1 R. That can't be fixed technically; it's managed upstream, by cutting size ahead of a tense weekend.

Correlation. Three simultaneous buy signals on gold, silver and EUR/USD are not three risks of 1%. They're often a single risk of 2.5%. A correlation check before stacking positions avoids nasty surprises.

News. A technical signal has no view on a central bank meeting due in ten minutes. Filtering is on you.

The control table to keep to hand

How to apply the checklist in two weeks

1. Pick a single instrument and a single timeframe. Don't test everything at once.
2. Open a notebook and log every signal received: time, entry price, stop, targets, in R.
3. Execute on a demo account, without altering the stated levels. You're testing the method, not your creativity.
4. After two weeks, compare your notebook to the provider's public track record. Look for missing signals.
5. Work out the total in R, the number of trades, the worst losing run.
6. Decide: carry on for another thirty signals, or stop. There is no third option.

Counting in R rather than in currency makes this evaluation independent of your capital. To then translate the result into real amounts, the profit and loss calculator does the job in seconds.

What you gain by imposing this checklist

These seven points don't improve a signal. They weed out the providers who aren't really providing one — those whose track record is reconstructed, whose arrows move, whose stop only exists after the fact. It's a defensive filter, and it's the most profitable kind of filter there is.

It also makes you capable of judging your own system with the same severity. A trader who demands confirmed closes, announced stops and a complete track record from someone else ends up demanding them of themselves.

If you'd like to see this checklist applied on a terminal — signals validated on the close, ATR or structural stop, targets in R, pessimistic backtest replayed candle by candle — the Discovery pack from PIPSTER PRO is free and enough to start taking your first readings.