Read this before you pay anyone
Before you pay for stock signals, ask six questions
You are about to pay a monthly fee for stock calls, and nothing on the seller's page tells you whether the record behind those calls is real. Six questions settle it — they work on any seller, and they work on us. Each one below is followed by how we answer it, with the number we are publishing today rather than the number that would look best.
1. Can you see the losing calls?
Why it matters: A seller who shows only wins has shown you nothing you can price. The loss column is the only part of a record that costs money to publish.
Ask the seller: “Show me your last twenty closed calls, wins and losses, with dates and prices.”
How we answer it: Every call ever logged is in one table — 395 of them (333 Buy, 59 Sell, 3 Watch) — with the losing ones in it. See the public scorecard.
2. Is the record dated, and can you see what was known when?
Why it matters: A record without dates cannot be audited: a call can always be right in hindsight.
Ask the seller: “What date was each call made, and at what price?”
How we answer it: Each call carries its session date and the level it was called at; the newest session in the record is 2026-10-21. See the live forward-test record.
3. Was the record measured before or after the calls were sold?
Why it matters: A backtest and a live record are different animals, and a good backtest is easy to produce after the fact.
Ask the seller: “How did the calls do in public, in the order they were published?”
How we answer it: Separately from the two-year backtest, a live forward test has been running 51 of 91 days: 228 resolved, 54 won and 174 lost (24%), on a virtual bankroll that is worth $6,965.43 of its $10,000 start. It is currently behind its own backtest, and that is published as it stands. See the paper portfolio.
4. What happens on a day when nothing fires?
Why it matters: A daily minimum in a sales page is a promise the market does not make. Services that fill quiet days are inventing calls to look busy.
Ask the seller: “How many days a month do you publish nothing, and will you say so on the day?”
How we answer it: Silent days are published as silent days. The daily feed carries only calls that actually fired in the newest session, and a page with no pick says so instead of manufacturing one. See today's free pick.
5. Does every call carry a stop-loss and a sell price?
Why it matters: Without a stop, a loss has no defined size; without a target, you cannot tell a win from a lucky exit.
Ask the seller: “What is the stop and the target on each call — and who set them?”
How we answer it: Every Buy call carries a suggested sell price, an estimated time frame and a stop-loss derived from historical volatility and the length of the recent swing — estimates, labelled as estimates, never guarantees. See how the next sell price is worked out.
6. Who is the seller, and what happens when you want out?
Why it matters: A refund policy and a named company are the two things a seller who intends to stay cannot fake.
Ask the seller: “Which company takes the card, and what is the refund rule in writing?”
How we answer it: The company, its registration number, the refund rule and the support address are printed on every page — they are in the footer you are reading now, and the policy is written out in full rather than summarised. See the refund policy.
Our own numbers, in the same format we are asking for
Read from the public record through the 2026-10-21 session when this page was served — not typed in, and not rounded up.
- Calls logged: 395 (333 Buy · 59 Sell · 3 Watch)
- Algorithms tested in the two-year backtest: 50 · cut in the open for losing money: 12 · still allowed to signal: 38
- Live forward test: 51 of 91 days · 228 resolved · 54 won · 174 lost (24%)
- Virtual bankroll: $6,965.43 of $10,000 · -30.35% · $1,000 notional per call
That is a losing record so far, and it is the first honest answer to question three. A service that hides a number like that is the one you should walk away from — including this one, if a record in the red is not something you want to base decisions on.
What this page does not claim
It does not claim these six questions make money for anyone, and it does not claim a good record predicts the next call — past performance never guarantees future results. Nothing here is financial advice: every signal on this site is an educational estimate, we do not personalise advice, and we do not hold, execute or manage anyone's money.
Questions, answered plainly
The ones people ask before they trust a signal service with a card.
How can I tell whether a signal service's record is real?
Ask for the losing calls, dated, and check that the same call appears with the same numbers on every page. A real record is timestamped, reproducible and includes the losses; a highlight reel shows winning trades without a date, an entry price or a sample size. If a service cannot show you a loss, you are not looking at a record — you are looking at advertising.
What is a good win rate for a stock signal service?
A win rate on its own means nothing. What matters is the sample size, the average win against the average loss, and whether the number was measured before or after the calls were published. A 40% win rate with wins twice the size of losses can beat a 70% win rate with the opposite split. Anyone quoting a win rate without the average win, the average loss and the number of trades is asking you to trust them instead of checking.
Why would a signal service have silent days?
Because patterns do not align every day, and a service that fires regardless is padding its feed to look busy. On this site, silent days are published as silent days: the feed carries only calls that actually fired in the newest session, and the page says so rather than inventing a pick. A daily minimum is a marketing promise, not a market fact.
What should each signal include besides the ticker?
The reason it fired (which pattern, and that pattern's own record), a graded strength, a suggested sell price, an estimated time frame and a stop-loss — plus the date and the price it was called at, so the outcome can be scored by someone who is not the seller. Without a stop-loss, a loss has no defined size, and without a date, a call cannot be audited at all.
How do I check a signal service for free, without signing up?
Use the seller's own published record before you create an account. On this site the scorecard, the live forward-test record and the paper portfolio are all readable without a login, and the engine's cut algorithms are published with the numbers that ended them, so you can see what was removed as well as what survived.
Should I subscribe to The Stock Proof?
That is your decision, and this page exists so you can make it on evidence rather than on our copy. Read the loss column first: we publish every call, wins and losses, and the live forward test has run behind its own two-year backtest so far. If a record like that is not what you want to base decisions on, do not subscribe — the free pages stay free either way. Signals here are educational estimates, never financial advice.
More detail: the full public scorecard · the live forward-test record · the paper portfolio · pricing.