Methodology — SPECTRE Markets

01 — WHAT WE PUBLISHThe board

Each session, before the open, we publish the day's full structure for NQ, ES, BTC and ETH, plus a short written read. It's the whole desk's worth of levels — the same families every day, each computed its own way:

Gold (GC) ships every family below except dealer gamma. Its options complex isn't part of our chain pipeline, and we'd rather publish nothing than publish a gamma level we can't stand behind — so GC gets pivots, value areas, VWAP, prior periods, opening range and SMC, and no gamma at all. The board says so on the page.

Dealer gammaThe gamma flip (HVL), the call & put walls, and the GEX strike ladder — where dealer hedging defends, pins, and accelerates.
PivotsFive systems — Classic, Camarilla, Fibonacci, Woodie and DeMark — at daily scope; Classic also weekly and monthly.
Value areasVAH, VAL and POC — where volume actually traded — for the session and the overnight, plus naked POCs.
VWAPVolume-weighted fair value and its standard-deviation bands.
Prior sessionYesterday's high, low and close — the references every desk trades off.
Smart moneyOrder blocks and imbalances (FVGs) — the institutional footprint in the tape.
Expected moveOn the BTC and ETH boards; the NQ and ES boards do not carry it. The one-standard-deviation move the options price for one session: the implied variance to the nearest expiry, spread over its trading sessions. From late October 2026; earlier days used the rule-of-16 scaling, about 20% wider mid-week, and are graded as published. On BTC and ETH the expected-move band is graded on that same one-session basis: the nearest-expiry straddle move, often one to two weeks long, is scaled by the square root of one over the sessions to that expiry before it is checked against the next day's range. Since late October 2026 the whole crypto record is graded this way.
GoldenConfluence — where three or more independent families stack in a tight band. Rare, high-conviction, re-derived every session.

02 — WHERE THE NUMBERS COME FROMThe dealer book

Options dealers who sell contracts hedge their risk in the underlying. The size and sign of that hedging — their gamma exposure (GEX) — changes with price, and it concentrates at specific strikes. Near heavy positive gamma, dealers buy dips and sell rips, so price tends to pin; past the flip into negative gamma, they hedge with the move, so price tends to accelerate. Those concentration points are the levels we publish.

We compute them from publicly available options data — the listed chains for each instrument — back-solving implied volatility and summing dealer gamma per strike and expiry. The inputs are public; the read is ours. We don't fabricate, hand-pick, or back-date a single level.

On the read and the ideas

The daily desk note is commentary — it describes what the structure suggests and where reactions are likely. Every path now also carries a free trade idea — a defined entry, an invalidation and a target — published as General research, not personal advice. We publish a measured 1:1 reaction rate for each level type; it is not a trading win rate or a P&L claim. An idea is not an instruction to buy or sell. Reads are AI-generated and labeled as such. Read the risk & disclaimers.

03 — HOW WE GRADE OURSELVESThe public record

Every level is published in advance and timestamped. After each session, we score the levels against what the tape actually did and post the result to a public track record anyone can read. The record is honest by construction: because levels are set before the move and never edited, a good run can't be faked after the fact and a bad day can't be hidden. We report a reaction rate, not a profit-and-loss claim — and the ideas we add are research, not a track record, so we claim no points total. We publish a measured 1:1 reaction rate for each level type; it is not a trading win rate or a P&L claim.

03b — THE GRADING RULEWhat “reacted” means

Each level is scored on intraday bars across the published history — not a single end-of-day close. The exact span and the exact number of scored touches are stated live in the coverage note beneath the proof panel on the home page, rendered from the same payload as the rates themselves, so this page quotes no figure that could drift away from it. A level reacted if, after price first tagged it, the market moved ≥0.25% in the level's favour (up off a support, down off a resistance) at some point in the session. The reaction rate is simply reacted touches ÷ total touches, computed the same way for every family — so the numbers are directly comparable. A touch whose own bar also reaches the 0.25% counts as neither: one bar can't show which came first, so it is left out of both sides of the rate (the same bar rule as our trade-idea grading).

And we go a level deeper

Reaction rate is honest but blunt — within an instrument the families look similar. So we also grade every individual line with a symmetric 1:1 (±0.25%) test and rank it best-to-worst per instrument. That's where the real signal is: each line gets its own 1:1 win rate (did price reach +0.25% in the level's favour before −0.25% against; 50% = a coin flip), so you can see which hold and which don't. The full per-line ranking — which levels to trust, which to be wary of — lives in the terminal. A reaction rate, not a P&L claim: published sessions plus structure levels recomputed for earlier sessions with no look-ahead. We show the laggards alongside the winners.

04 — WHO RUNS ITA methodology, not a personality

There's no guru, no track-record screenshot, no “95% win rate.” What stands behind SPECTRE Markets is this page and the public record — the data source is transparent, the computation is consistent, and the results are posted whether they're good or bad.

See today's free sample View the track record