Method & limits

How the signals are produced, and what this service deliberately is not.

The cadence

The market is reviewed three times every trading day, at fixed times. Nothing is published in between. The schedule exists to stop the service from reacting to every twitch in the tape, and to make the output comparable from one day to the next.

The watchlist

Fifteen instruments, all American: three stock indices and twelve companies. Narrow coverage means every instrument gets real attention, and the liquidity of these names keeps the gap between a simulated result and an achievable one as small as we know how to make it.

The third column is measured, not estimated: it is the correlation of weekly returns over the last twelve months against the closest of the three indices, calculated on 17 September 2026. It exists for one purpose, explained below.

Four things worth knowing about this list before you use it

At most two open ideas per group, and no more than six in total. The three indices move together between 56% and 92% of the time: they are not three bets, they are one. The same holds, to a lesser degree, within each group of shares. That is why the second idea opened in the same group comes out with the suggested risk halved and with a line saying which idea it is linked to: whoever takes it knows they are adding to the same bet, not diversifying. A third is not published. An index idea and a technology idea in the same direction count as the same group, because NVDA moves with the indices up to 72% of the time.

The defensives and energy are there precisely because they do not follow the market. WMT at 13%, KO at 16%, LLY at 13%, XOM at −9%: when the market falls as a block, these do not always fall with it. They are the only thing stopping this service from being one idea repeated six ways.

On a single share, the stop is not guaranteed. A stock can open far from your level on an earnings release, an analyst revision, a trial result or sector news: the price jumps past the stop without trading through it, and the loss is then larger than the one planned. An index, being the average of hundreds of companies, does this far more rarely. That is why on shares we prefer ideas that close within the day, and any idea on a share held longer says so explicitly.

Earnings are treated as a wall, not as a risk to weigh up. We open no idea on a share if its results are due within three days, and no multi-day idea that straddles that date. This is not a case-by-case judgement: the calendar is loaded into the system and publication is refused automatically, because it is the one risk on our list against which a stop level does nothing at all.

The reviews are not interchangeable

The US market opens at 15:30 Rome time, and that decides what each review can do.

• 08:30 opens no new ideas. The open is seven hours away and the pre-market is so thin that any level calculated there is unreliable. That review does a different job: it revisits open positions, checks whether overnight news has invalidated a thesis, writes up outstanding post-mortems and prepares the list to watch in the afternoon.
• 13:00 can work on indices only, estimating the price from the future. Not on single shares: a stock's pre-market is thin and a price estimated there is not reliable.
• 18:30 is the real review. Indices and shares have been trading for three hours, the price is directly verifiable and the spread is at its tightest. It is the only review in which we publish on single shares, and it is the one most ideas should come from.

What it costs to execute

The published record is before whatever your broker charges you, so it falls to us to give you the sum to do. The number that matters is not the commission itself but how much of your risk it eats: total cost as a percentage of the position, divided by the width of the stop.

A concrete example, because the order of magnitude surprises people. On a share with an average daily range of 2.5%, our typical stop is worth about 1.25%. If your broker charges you nothing on US shares — and many now do not — the cost is negligible. If it charges 0.25% per side, the round trip is 0.50%, which is 40% of what you were risking, before the market has moved at all. The same service, with the same idea, is sensible in the first case and pointless in the second.

We do not know which is your case and we will not pretend to: that number is one you have to look up yourself, once, at your own broker. It is the most profitable sum you will do on this site.

What we left out, and why

• Commodities, gold included — out by choice. An editorial decision taken on 17 September 2026: this is a service on indices and shares, and a commodity is neither. Gold in particular is quoted XAU/USD and appears on every currency platform, so it read as a currency pair even though it is technically a metal. There is a cost and we will name it: it was the least correlated instrument we had. That role now belongs to the defensive shares and energy.
• Currencies — not covered. EUR/USD and GBP/USD move together 81% of the time and USD/JPY is the same risk with the sign flipped: holding them would be holding the same position three times.
• Non-US indices — out by choice. An editorial decision, not an analytical one, and it has a measured cost: with an all-American list a single US data release touches every instrument at once. The macro filter reasons about real exposure instead of blocking a whole day, and the defensives sit outside the market's move, but neither of those removes that concentration.
• Crypto — not covered. Bearish ideas are not genuinely actionable on most European retail accounts, and where they are, the round-trip cost can reach 79% of the risk before the market moves.

What each review does

1. Context. Establish the higher-timeframe trend, the current volatility regime, and what the macro calendar holds in the next 24 hours.
2. Levels. Mark the structural levels that matter: prior highs and lows, the previous day's range, round numbers, the levels where the market has already reacted.
3. Setups. Look for a location where price meeting a level would give a specific, falsifiable reason to be long or short.
4. Checklist. Test each candidate against the rules below.
5. Publish. Anything that passes is published in full to every registered user at the same moment. Anything that does not is discarded, and the review is still written up as a market note.

The checklist

An idea is only published if it clears every one of these:

• A defined invalidation level. There must be a specific price at which the idea is simply wrong. No invalidation, no trade.
• Reward-to-risk of at least 1.5 to the first target. Measured from entry to stop against entry to target 1.
• The stop sits behind structure, not at an arbitrary distance chosen to make the ratio look good.
• No first-tier data release inside the horizon of an intraday idea on the affected instrument.
• Not a duplicate. No second open idea expressing the same directional bet on a correlated instrument beyond the second, and no more than six open ideas in total.
• No earnings in sight, if the idea is on a single share: neither within three days of entry, nor inside the idea’s expected life.
• Liquid hours only. No entries in thin sessions where the spread distorts the result.

Most reviews produce nothing. A day with no signal is a normal outcome, not a failure — it is published as a market note so you can see what was considered.

Risk, expressed as a percentage

Every idea risks a percentage of a notional account between entry and stop — typically 1%, never more than 2%. We never publish a position size in lots, contracts or currency, for two reasons: we do not know your account, and a percentage keeps the record readable at any capital. That is why the track record lets you view the same history against $100, $1,000 or any other figure — the percentages do not change, only the currency amounts do.

Results are recorded in R, multiples of the risk taken. An idea that made twice what it risked is +2R; one that hit its stop is −1R. R is what makes a 30-pip idea and a 300-point idea comparable.

How an idea is closed

Once published, an idea is followed until one of these happens, and the outcome is then permanently recorded:

• Target reached — closed at target 1 or target 2.
• Stop reached — closed at the stop, recorded as −1R.
• Expiry — the horizon runs out with neither level touched; closed at the market price, for whatever it gives.
• Cancellation — the entry was never reached and the setup is no longer valid; recorded at 0R and still shown, so the count of published ideas always reconciles.

Records are append-only. A published idea is never edited, re-priced or removed, and each one carries a visible event log.

How the method learns

A signal service that never revisits its own output is just a random idea generator with a website. Every closed idea therefore goes through a post-mortem before the next review is allowed to use the slot, and the conclusions are published on the what we learned page.

Outcome and process are graded separately

This is the part most track records get wrong. A single trade tells you almost nothing about whether the decision behind it was good: an idea can be built correctly, sized correctly, and still lose, because the outcome of any one trade is dominated by chance. Another can win after a real error — entering late, ignoring a data release — and reward exactly the behaviour that should be corrected.

So each review records two independent things: the result (in R, unchangeable) and a verdict on the process — sound, flawed, or inconclusive. Losses with a sound process are left alone. Wins with a flawed process are flagged just as loudly as losses. Learning from profit and loss alone is how a method teaches itself the noise.

From a mistake to a rule

When a review finds a real process error it is filed against a fixed vocabulary — stop too tight, entry too late, data release overlooked, hidden duplicate, stale data, and so on. A closed vocabulary is what makes the errors countable: the same mistake made three times is visible as a pattern instead of three unrelated anecdotes.

A recurring pattern becomes a rule, written as a single actionable line, and every subsequent review receives the full list of active rules before it is allowed to publish anything. That is the mechanism by which the service actually changes: not by intending to do better, but by carrying a written constraint forward. A rule keeps a counter of how many times its error has occurred, so a mistake seen five times outweighs one seen once.

Learning from the numbers, not just the narrative

Alongside the written reviews, results are broken down by instrument, asset class, direction, review slot, horizon, conviction score and exit type. If conviction-5 ideas quietly underperform conviction-3 ideas, the numbers say so even when every individual write-up sounded convincing.

The honest limits of this

• Small samples lie. Below roughly twenty closed ideas, almost any pattern you can see is noise. The learning page labels the sample size rather than hiding it, and rules are meant to rest on at least a handful of occurrences.
• Adapting can make things worse. A method that rewrites itself after every loss ends up fitted to the recent past instead of the market. Rules are added deliberately and can be archived when the evidence stops supporting them.
• Learning is not a trend. Nothing here guarantees that results improve over time. The process gets more disciplined; the market does not become more predictable. Anyone promising that an AI "gets better every week" at making money is selling something.

Where the AI sits in this

An AI model performs the review and drafts the rationale, working from public market data and the rules above. The checklist is applied mechanically; the model is not free to publish an idea that fails it. What the model does bring is the reading of context — and that reading can be wrong. AI systems misinterpret data, work from stale prices, miss what has not yet shown up in the tape, and can be confidently wrong. The checklist reduces that risk; it does not remove it.

Every idea shows the model that produced it. That is a transparency choice, and it also means you can judge the record against the tool that generated it.

What this method deliberately is not

• Not a backtest. The record starts from the day the service went live and grows forward. There is no historical simulation and no curve fitting, because there is no history to fit to.
• Not a strategy you can outsource. The ideas are a starting point for your own analysis, not instructions.
• Not personalised. Everyone sees the same idea, whatever their account size, experience or situation. See the Recommendation Disclosure.
• Not a promise. The published record is hypothetical, and past results say nothing reliable about future ones.