Everything worth asking before you sign up.

It is the published research of a working trading desk. Each session the desk reads the market, grades every mover against its actual news and writes the trade plans. That finished session is published here and rendered as-is — nothing is recomputed or re-derived along the way, so you are reading what the desk read.

No. There is no alert to follow, no entry to copy and no performance claim. What you get is the evidence a desk uses before the open: what moved, why it moved, what the plan around it looks like and how confident that plan is in its own parts. Deciding whether to trade any of it is entirely yours.

The dashboard is designed to be read top to bottom in one pass: where the market stands, what moved and why, the plans for the names that earned one, then what is coming. The calendars and the barometer are there when you want them rather than something to check daily. The point is not that it is fast for its own sake — it is that the reading you would not have had time to do has already been done.

Because a hundred and fifty overnight movers is a time problem, not a discipline one. Read properly, each name needs the release, the guide against consensus, and how the stock behaved after its last four reports — so in practice you skim, and the one that mattered was the company that beat on both lines and guided the year down in the fourth paragraph. The AI asks the same questions of every name, at the fortieth as much as the first, and shows its reasoning so you can overrule it where you disagree.

Someone who already trades and wants a faster, better-organised pre-market read than they can assemble themselves — particularly around catalysts, earnings behaviour and macro prints. If you are looking for something to follow without forming your own view, this is the wrong product.

Eight things: the session's market update, graded movers with their catalysts, full trade plans, the option-flow board with whale positioning, earnings history with the desk's own annotated charts, an earnings calendar, an economic calendar with a read on every print, political catalysts, and a macro barometer scored against 1929, 2000, 2008 and 2020.

Three jobs. It reads the news behind a move and grades it 1-10, with sentiment and forward guidance judged separately. It writes the trade plans and scores each scenario into six named components — catalyst, setup, direction, levels, geometry and data health — each showing the reason it scored what it did. And it reads each economic print for what it meant for the index, then summarises the month's prints into what the set adds up to.

It is writing an argument, not placing an order. A plan on this site has never been executed by the thing that wrote it. The confidence score is deliberately broken into parts so you can disagree with one of them — a scenario can score well on catalyst and badly on geometry, and that is visible rather than blended into a single number.

The clearest test is where the read disagrees with the surface. A payroll number that beat its forecast by a wide margin is graded mixed, not bullish, because it also removed rate cuts from the curve. A stock that beat on both lines is graded bearish when the guide came in light. Those are on the site now, and beat/miss is kept in its own column precisely so the two can visibly disagree.

Regularly, and in two different ways: the reasoning can be poor, and the data underneath can be stale or missing. The site is built to make the second kind visible — every panel is stamped with when it was captured, a board that has gone stale says so, and a value the desk did not produce prints as a dash rather than a zero. The first kind is why the reasoning is shown at all.

The session content — market update, graded movers, trade plans, news — is published once per build, each trading day. The option-flow board is designed to refresh through the session, the economic and earnings calendars monthly, and the macro barometer when its scan runs. Every page prints its own last-refresh time, so you never have to guess which of those actually happened today.

No, and the site never implies otherwise. Everything is a published snapshot with a capture time attached. If a board has not been refreshed within its own cadence it prints a warning saying how long it has actually been, rather than looking current.

Because the desk did not produce that value. A missing figure is never filled with a zero or an estimate. The clearest case is overnight pricing: index levels are only shown converted from futures when a validated basis exists, so a level either prints in both frames or prints in neither.

Because they genuinely are, and hiding that would be worse. Whale positioning is collected per name and some names have not been polled recently, so each reading carries its own date and anything over a week old is marked. The macro barometer moves slowly by design — most of its inputs update monthly or annually.

Not by name. What travels with each figure is the KIND of source behind it — an official statistic, an international series, a market market measurement — plus the series code where one exists. That distinction matters more than any brand name: the same dealer book scores very differently depending on how it is measured, so a reading means little without knowing which kind produced it.

All of it. During the beta there is no split between accounts — every approved member reads every page: the market update, every graded mover on both days, the full trade plans with their confidence breakdowns, the whole option-flow board with whale positioning, the complete earnings record with the desk's annotated charts, both calendars, the political calls and all fourteen barometer indicators.

Two, and only one of them is about permissions rather than content: an admin can review and remove people, and a member reads everything else. There is no tier that sees less of the research than another.

No. There is no payment form anywhere on the site and no card is ever asked for. The beta is free for everyone approved into it, and if that ever changes you will be told before it does, not after.

Ask on the request page. It is invite-only during the beta, so each request is read and approved by a person rather than dropping into an automatic queue — usually the same day. You will get an email when you are in, and from then on signing in is a six-digit code to your address.

Because the beta is small on purpose. Every account is someone whose feedback can be acted on, and keeping the number low means a question gets an answer rather than a ticket number.

You will be told before anything changes, not after. Nothing on the site can take a payment today — there is no payment form anywhere in it — so there is no way to be moved onto a paid plan without an explicit, separate step that you take yourself.

Sign-in is a six-digit code emailed to your address. There is no password to choose, reuse, forget or leak. The code works once, expires in ten minutes, and dies after five wrong tries so it cannot be guessed at.

Your email address, your name and whatever you wrote when you asked for access, your plan, and when you last signed in. That is the whole list. Sign-in codes and session cookies are stored hashed, never in the clear, so a database leak would not hand anyone a working session.

No. Everything here is research and education. Nothing on the site is a recommendation to buy or sell any security, and no one behind it is acting as your adviser. Position sizing and risk are entirely your own decisions.