Ideas with the homework attached.
Anyone can shout a ticker. A Flux idea arrives with the thesis in plain English, the levels it lives and dies on, a confidence score you can interrogate, and a backtest that shows how the setup behaved the last thousand times it appeared.
Illustrative sample feed · not advice
This is what an idea looks like.
Five from a recent session, expanded in full. Nothing here is a recommendation — it's the shape of the output: what Flux saw, where the setup breaks, and how confident the ensemble is that the pattern is real.
Sample ideas and levels shown are illustrative and generated for demonstration only. They are not recommendations, not personalised, and not a solicitation to buy or sell any security. Confidence scores describe model agreement — not the probability of profit.
Five parts. None of them optional.
If any one of these is missing, the idea doesn't ship. That rule is what stops the feed turning into a wall of tickers with arrows next to them.
One paragraph, plain English, naming the specific thing that happened — the sweep, the filing, the level reclaimed. If it can't be written in a sentence, it isn't understood.
How strongly the engines agree, after AlphaRank re-scores the signal against five years of outcomes. It measures conviction in the pattern, never the odds of making money.
Entry zone, invalidation stop, target and the resulting reward-to-risk. The stop is the honest part: it's the price at which the thesis is simply wrong.
The same setup replayed across historical sessions, with the sample size shown so you can tell a pattern from a coincidence.
Position size against your book, correlation with what you already hold, and total exposure. An idea that breaks your rules never reaches the feed.
Illustrative example · hypothetical figures
From raw signal to a number you can argue with.
Confidence isn't a vibe. It's the output of four steps, and you can open any of them to see why a score is 88 and not 62.
Signal
One of the six engines flags something concrete — a sweep, a breakout, a filing, a level reclaimed. Raw, unranked, and far too noisy to trade on its own.
Cross-check
The other engines are asked whether they see it too. A breakout with flow behind it and a catalyst underneath scores very differently from one standing alone.
Backtest
The setup is replayed across historical sessions, walk-forward, with costs. If the edge disappears out of sample, the idea is dropped here and you never see it.
Score
AlphaRank re-scores the survivor against five years of outcomes and assigns the confidence figure — then the Risk engine decides whether it fits your book at all.
A high confidence score means the engines agree and the historical pattern was consistent. It is not a probability of profit, not a guarantee, and not a substitute for your own judgement.
The last engine to see an idea can kill it.
A good setup in the wrong size, on top of three correlated positions, is a bad trade. Every idea passes the Risk engine before it reaches your feed — and every live order passes it again on the way out.
Position sizing
Size is derived from the distance to the stop and your own risk-per-trade rule — not from how exciting the idea sounds.
The stop
Every idea carries an invalidation level before it carries a target. No stop, no idea.
Correlation
Checked against what you already hold. Four semis in a row is one position wearing four tickers.
Exposure
Gross, net and sector exposure against your limits. Ideas that would breach them are held back and labelled, not quietly sent anyway.
Figures are illustrative and describe typical product behaviour and simulated backtests. They are not performance results and do not predict future outcomes.
The fair questions.
Do you place trades for me?
Not unless you explicitly set that up. By default Flux researches and surfaces ideas, and you decide what happens next — paper first, if you want. If you do enable automation, every order still passes the Risk engine's guardrails, and you can stop it at any time.
What does the confidence score actually mean?
It measures how strongly the six engines agree on a setup, after AlphaRank re-scores it against historical outcomes. High confidence means the pattern is consistent and corroborated. It is explicitly not a probability of profit and not a measure of how much you might make.
Are these signals guaranteed?
No. Nothing in markets is, and any product that says otherwise is lying to you. Ideas can be wrong, data can lag, and historical patterns break. That's why every idea carries a stop, a sample size and a backtest you can inspect — so you can size the risk you're actually taking.
Is this financial advice?
No. Flux is research and automation software. It is not a broker-dealer, not an investment adviser, and nothing it produces is personalised advice or a recommendation. Every decision, and its outcome, is yours.
Where do the entry, stop and target come from?
From the structure of the setup: the level that triggered the signal, the price that would invalidate the thesis, and the next level the tape has historically respected. They're a starting frame to adjust to your own plan — not instructions.
Can I test an idea before risking anything?
Yes, and we'd rather you did. Every idea has a one-tap path to paper trading and to the backtest behind it, on every plan including the free one.
Read the homework. Then decide.
Open the feed, pick an idea apart, and paper trade it before a single dollar is at risk.