Documentation
What you're looking at
An options map looks like an aircraft dashboard the first time. This page explains, without formulas, what each thing in STRAVEX means and why we look at it.
The problem
None of what this page explains is secret. Gamma exposure, dark pool prints, aggressive tape, implied volatility — it is all published, and a professional desk reads all of it before the open.
The problem is that it lives in six different places, arrives in different formats and goes stale at different speeds. Assembling it by hand takes the first hour of the session, which is exactly the hour where the day is decided. So most people end up doing one of two things: trading on a fraction of the picture, or trading on how it feels.
What the system does
STRAVEX puts the whole picture on one screen and asks the same six questions every day, about the same universe, in the same order. Not because six is a magic number, but because a question you ask only when you remember it is not a method — it is a mood.
- One screen
- Institutional flow, gamma walls, dark pool levels, defenses, aggressive tape and market tide converge in a single view instead of six terminals.
- One arithmetic
- The risk plan is computed once, on the server. The screen, the Excel export and the assistant read the same numbers — they cannot disagree.
- One entry time
- The plan is sealed at the method's entry time. No hunting for a better print later; the record is the one from the moment the decision was made.
- A concrete answer
- Reference price, stop, target and risk-reward. Not a bullish arrow: a number you can size a position against.
What that buys you is not certainty. It is the first hour of the session back, and a decision made on the whole picture instead of the part you had time to look at.
What makes it different
Any platform can show metrics. Two things are harder, and they are the two this one is built around:
- It says when it does not know
- A source that failed produces an explicit gap, not a plausible number. Coercing missing data into something believable is the most expensive way to be wrong, because nobody checks it again.
- It grades itself
- Every plan is stored with a timestamp and cannot be edited afterward. At the close it is graded against what price actually did — including the ones that went wrong.
The track record separates what was sealed before the fact from what was reconstructed after it, and reports them apart. Reconstructed results are not evidence of an edge and are not presented as one — mixing them would produce a prettier number that measures nothing.
And the verdict is read by expectancy, not by hit rate. A nearer target gets touched more often for mechanical reasons; a high hit rate on plans that pay little is not performance. What matters is what the whole set pays.
GEX — gamma exposure
When you buy an option, someone sold it to you. That someone is almost always a market maker, and they have no interest in betting on direction: they make money on the spread. So they hedge — buying or selling shares until they are flat.
The problem is that being flat breaks on its own as the price moves. Gamma measures how fast it breaks. And every time it breaks, the market maker has to go into the market and rebalance — not because they want to, but because their book demands it.
That turns the dealers’ aggregate position into a mechanical force on price:
Positive gamma
Dealers sell into strength and buy into weakness. It dampens. Price tends to stay pinned in a range and breakouts cost more than usual.
Negative gamma
Dealers buy into strength and sell into weakness. It chases. Moves get amplified and a small push travels a long way.
That is why the same good print moves a stock 0.4% on Tuesday and 3% on another day: the news did not change, the side the dealer book was sitting on did.
Walls
That pressure is not spread evenly: it concentrates at the strikes where a lot of options are open. We call those concentrations walls.
A wall is not a prediction. It is a place where, if price gets there, it will run into people who are forced to trade against the move. Sometimes it holds. Sometimes it gives way, and then the move accelerates — because whoever was slowing it down now has to chase it.
Dark pool
A fund that wants 400,000 shares does not buy them on the screen. If it did, price would run away while it bought: its own order would ruin its entry.
So it uses off-exchange venues — dark pools — where the trade crosses without being posted first. It gets reported afterward, and that is where we see it.
A single block says nothing — it could be a reallocation, a hedge, an index rebalance. What says something is repetition at the same price.
Institutional defense
When the same price accumulates several large blocks, someone with size is taking a position there over and over. That is a defense: not a line somebody drew, but a price where money has already committed.
Every defense carries a freshness, and that is what decides how much it weighs:
- ACTIVE
- Being defended today. The only one a stop can lean on with confidence.
- RECENT
- Defended not long ago, but not today. Useful as reference, not as support.
- COLD
- Historical. It explains the terrain; it holds nothing.
And when a defense lands on the same price as a gamma wall, we mark it A+. Those are two different actors — an institution and a dealer — standing on the same number for reasons that have nothing to do with each other. It is the strongest coincidence the system produces.
Options flow
Every option that trades carries directional exposure. Adding it up through the day tells you which way options money is leaning — not how much traded, but which side.
This is different from volume. Volume counts contracts; flow counts intent. A record-volume day with flat flow means a lot changed hands and nothing was decided.
Iceberg
The gap between what was visible on the screen and what actually traded.
Big money hides. It slices the order, spreads it out, executes where nothing prints at the moment. When a bar’s flow is enormous and almost none of it showed up on the visible tape, somebody built a position without wanting it noticed — and that usually says more than size alone.
The aggressive tape
In every trade there is someone who waits and someone who does not. The one who waits posts a price and sits. The one who does not crosses the spread: pays the offer just to be in now.
That urgency is the signal. We separate what executed by lifting the offer from what executed by hitting the bid, and compare the two sides. A market where 90% of the aggressive money is buying is saying something total volume does not.
Market tide
The same idea, but across the whole market at once: how much net premium is going into calls versus puts, minute by minute.
It is the wind. It does not decide a trade, but it moves the bar: if the tide is running against the side you want, everything else has to be better to make up for it.
Multi-day accumulation
A single day is noise. Twenty-one sessions of accumulated flow start to be a position.
The dollar figure on its own is misleading: three million is enormous in a small name and invisible in NVDA. So we compare it against that ticker’s own history— if today’s total sits in the 90th percentile of what that name usually moves, that is a flag even when the number looks modest.
Vanna and charm
Two second-order forces that create hedging flow without anyone making a decision:
- Vanna
- How the dealer's hedge changes when volatility changes. If volatility drops, dealers buy shares even though price has not moved.
- Charm
- How the hedge changes from the passage of time alone. It matters near expiration, and it explains Friday moves that answer to no headline.
They are the reason the market sometimes moves and nothing explains it. Nothing explains it because it was not a decision: it was hedging arithmetic.
Open interest
How many contracts are still open at each strike. It is the balance, not the activity: volume says what traded today, open interest says what stayed alive.
Where it piles up, that price matters to a lot of people — and when price gets near it, there are positions that will have to do something.
IV rank and term structure
Implied volatility is what the market is charging for uncertainty. On its own it says little; compared, it says a lot:
- IV rank
- Where today's volatility sits against the last year of that same ticker. 30% can be very expensive in a quiet name and cheap in a jumpy one.
- Term structure
- Whether the near expiration costs more than the far one. When it inverts, the market is saying it expects something soon — an earnings report, a decision, a print.
The six layers
None of the above decides a trade on its own. What STRAVEX does is ask the same six questions, every day, about the same universe — and count how many line up:
- Multi-day
- Is the last few weeks' accumulation on this side?
- Intraday flow
- Are the most recent bars going this way?
- Defense
- Is there institutional money defending the side the stop would sit on?
- Tape
- Does today's aggressive money favor this side?
- Tide
- Is the broad market with us, or do we have to row against it?
- Risk plan
- Does the geometry of price justify the shot from here?
The risk plan
The last layer has no opinion about direction: it does arithmetic. It takes the most solid level on the risk side, the first one price would meet on the reward side, and compares how much is risked against how much can be made.
Out of that come three words — qualifies, marginal, discard — and there is one confusion worth heading off from the start:
Why everything is recorded
Every day, at the method’s entry time, the full plan for each ticker is saved: reference price, stop, target, risk-reward and verdict. Timestamped, and it cannot be edited afterward.
At the close it gets graded against what price actually did. Not to decorate a statistic, but because it is the only way to know whether the method works. A platform that only shows metrics never has to answer that question.
The series are not mixed: the one sealed in the morning measures an intraday strategy, the one sealed at the close measures a swing strategy, and the one reconstructed backward is not forward-tested and is reported separately. Averaging them would produce a well-formed number that measures nothing.
Need a hand?
Every question goes to one place, and a person reads it. Include your ticker and the time of the cut you were looking at — with those two, the answer comes back with your own numbers instead of a general explanation.
Open a support ticketWeekdays, within one market session. You get a reference to quote. Nothing here is investment advice.
What changed
Cancelling during the trial now keeps your remaining days. You are not charged on day 4, and the days you were promised stay yours.
A failed payment no longer locks you out on the spot. Your card bouncing starts a grace window instead of an immediate block, so a rotated card does not cost you the session.
Plan upgrades now charge the prorated difference right away instead of deferring it to the next invoice.
Reactivating after a cancellation works again, and it never charges twice. If you already used your free trial, you can subscribe with billing starting today instead of being turned away.
The risk plan now shows `alternativePlan` when the noise filter discards an institutional defense that sat too close to price — so you can see what the discard cost instead of redoing the arithmetic.
Verdicts that had no plan to grade are labelled NO_PLAN, separate from NO_DATA. A day the system stood aside is not a gap in the record.
MCP server live at app.stravex.io/mcp — the same analytics from your own assistant, with every rule's exact threshold in the tool descriptions.
Passkey sign-in. No password, no second factor to type.
This page was last revised on .