Receipt page · Premium Add-on
How the Premium Add-on suite measures
This is a receipt page. Dossier §8.2 fixes what one contains, and the five sections below are that anatomy in order: the versioned methodology, the pre-registrations, the proof corpus, the changelog, and the invitation to recompute any published number without asking me for anything. Nothing has been measured yet. Three of the five sections are therefore empty, and they are empty in a way you can inspect — the structure is printed, the slots are named, and each one says NOT YET PUBLISHED rather than showing you something plausible. Looking for what this family will contain instead? That is the Premium Add-on suite hub.
Receipt anatomy · dossier §8.2Premium Add-on suiteReceipt version v1.0
- Instruments in scope
- 5
- Metrics defined
- 21
- Pre-registrations posted
- NOT YET PUBLISHED
- Artifacts published
- NOT YET PUBLISHED
- Measurements published
- NOT YET PUBLISHED
- 01Methodology, versionedProse and formal definitions. Every metric carries its name, its definition, the estimator that computes it, its known failure modes, and what it does not establish. The page carries a version, and the version is printed on it.DEFINED · v1.0
- 02Pre-registrationsDeclarations made and dated before any measurement runs: decision thresholds, exclusion rules, completeness bars. Immutable once posted — an amendment appends beneath the original with its reason, and never overwrites it.NOT YET PUBLISHED
- 03Proof corpusThe published artifacts themselves: dataset descriptors, content hashes, download links, and the version tag of the code that generated each one. A result whose inputs are not downloadable is not a receipt.NOT YET PUBLISHED
- 04ChangelogAppend-only, dated, written by a person. What changed on this page, when, and why. A correction is a new entry naming the entry it corrects — entries are never edited away.1 ENTRY
- 05Independent recomputationThe exact numbered steps a stranger follows to re-derive a published number without asking me for anything. Where the recomputation disagrees, the disagreement is the finding.PROCEDURE STATED
Methodology, versioned
version v1.0last changed 2026-08-0121 metrics · 5 instruments
The Premium Add-on suite measures market terrain: dealer-positioning structure read from listed FX options, volatility and liquidity regimes identified under a never-repaint constraint, dependence-aware breaks in reported positioning, the propagation structure of forced-flow cascades, and joint change points across a multivariate panel. Every output is described state with uncertainty attached — never a forecast.
Terrain is the category where narration is cheapest and measurement rarest. Every instrument in this suite draws the same line on every surface: it describes the structure you are standing on, publishes its uncertainty and its sample sizes, and leaves the directional judgement — and the risk — with the user.
Every metric below carries five fields, because four of them are the fields a methodology usually leaves out. Definition states what the quantity is, closely enough that a stranger could implement it. Estimator states how it is computed and which parameters must be fixed before the data is read. Known failure modes states where the estimator breaks — written now, while nothing depends on it, rather than conceded later under questioning.Does not establish states the claim the metric will not support, however natural the reading. The last field is the one that costs something to publish, which is why it is published.
A structural map of forced-flow cascades — how liquidation chains have propagated through leveraged positioning.
P14.1Cascade anatomy
- Definition
- The propagation structure of a liquidation chain across price levels and instruments, separated from ordinary discretionary flow.
- Estimator
- Sequenced flow reconstruction across levels and instruments under a declared separation rule. Every mapped cascade carries the window and the rule that produced it.
- Known failure modes
- Separating forced from discretionary flow is inferential and it is the hard part. Where the separation cannot be made with the data available, the episode is excluded under a declared rule rather than mapped ambiguously.
- Does not establish
- A mapped cascade does not establish that participants were liquidated — only that flow propagated in the recorded pattern.
P14.2Resolution-versus-fizzle rate
- Definition
- The proportion of triggered chains that completed against those that did not, published with the sample size behind it.
- Estimator
- Outcome classified under a completion rule fixed in advance and counted over every qualifying episode; raw counts publish with the ratio.
- Known failure modes
- Completion is a defined threshold, not a natural fact. A threshold set after inspecting outcomes produces whatever rate the author wanted.
- Does not establish
- A completion rate does not establish that the next chain completes.
P14.3Chain preconditions
- Definition
- The observed positioning states in which cascades have historically initiated — crowding, one-sided leverage, thin absorbing liquidity — described as measured structure.
- Estimator
- Precondition states measured at the initiating bar of every mapped episode and reported as a distribution across episodes, never as a single profile.
- Known failure modes
- Preconditions are measured on episodes that occurred. The states that looked identical and produced nothing are the harder and more important sample, and their count publishes alongside.
- Does not establish
- A precondition is not a trigger. Its presence does not establish that a cascade will follow.
Standing limit. A map, not a signal. It records how cascades have propagated under which preconditions; any use for timing is the user’s judgement and risk.
Full instrument page →
Dealer-positioning terrain read from listed FX options. The engine is built; the listed-options feed it consumes is not under licence, so no map, sample output or figure publishes from it.
P7.1Strike-level open-interest congregation
- Definition
- Open interest aggregated by strike and expiry across the listed chain, rendered as terrain rather than asserted in a morning note.
- Estimator
- Direct aggregation of a chain snapshot by strike and expiry. Every map carries the snapshot it was computed from and the conventions used to derive each quantity.
- Known failure modes
- A snapshot taken at a different time of day yields different terrain. A chain with stale or missing strikes produces holes, and those holes must be drawn as holes rather than smoothed across.
- Does not establish
- Open interest is the market’s aggregate. It does not establish any dealer’s inventory or any participant’s position.
P7.2Risk reversals and skew migration
- Definition
- The standing skew across the chain and its movement day over day, tracked as a dated series rather than quoted as a mood.
- Estimator
- Risk reversal computed at declared deltas from the chain’s own quotes, differenced across snapshots taken at a fixed time of day.
- Known failure modes
- Delta and quoting conventions differ by venue. A convention change mid-series produces a migration that is an artifact of the convention rather than a movement in the market.
- Does not establish
- A skew shift does not establish direction. It states where the market is paying up for protection.
P7.3Expiry gravity
- Definition
- Notional expiring at each strike across the calendar, mapped forward so an expiry concentration is a dated object with a size rather than a rumour.
- Estimator
- Notional aggregated by strike and expiry date from the chain snapshot and projected forward on the exchange’s own expiry calendar.
- Known failure modes
- Listed notional excludes over-the-counter optionality, which in FX is substantial. The mapped gravity is a real but partial view, and it is labelled partial on the map itself.
- Does not establish
- An expiry concentration does not establish that price will be drawn toward it.
P7.4Snapshot provenance
- Definition
- The chain snapshot, its as-of timestamp, and the derivation conventions, attached to every map — because a positioning map without its as-of timestamp is decoration.
- Estimator
- The snapshot is content-hashed at ingestion and the digest travels with the map. A map without its snapshot is not published.
- Known failure modes
- The current failure is upstream of the estimator: the listed-options feed is not licensed, so no snapshot exists, so no map publishes. The instrument states that rather than simulating one in the meantime.
- Does not establish
- Provenance does not establish that the chain is complete — only that the map came from the snapshot it names.
Standing limit. Terrain, not direction. It describes where positioning congregates; it does not forecast what price does about it.
Full instrument page →
Multivariate change-point detection across a matrix of related series, with per-channel attribution: which series moved, decomposed per changepoint.
P17.1Joint break
- Definition
- A departure of a multivariate panel’s joint behaviour from its estimated regime — the move no single-series monitor sees, because no single series moved far enough alone.
- Estimator
- Hotelling T-squared over a Ledoit-Wolf shrunken covariance estimate, with detection thresholds calibrated against surrogate data rather than asymptotic formulas.
- Known failure modes
- When channels are many and history is short the covariance estimate is unstable. Shrinkage stabilises it at the cost of sensitivity to genuinely small joint moves, and the trade-off is a registered parameter rather than a silent default.
- Does not establish
- A joint break is a structural statement about the panel. It is not a trade signal and it does not establish a cause.
P17.2Sequential drift
- Definition
- The accumulation of small, persistent departures a snapshot test would dismiss — the slow bleed into a new regime rather than the clean fracture.
- Estimator
- A cumulative-sum statistic over the panel with reference and decision parameters fixed before the run.
- Known failure modes
- CUSUM trades detection delay against false alarms through exactly those parameters. A setting tuned after the fact is not a detection; it is a description of what already happened.
- Does not establish
- Detected drift does not establish where the drift ends.
P17.3Retrospective segmentation
- Definition
- A partition of the full history into internally consistent segments, so a detected break arrives with the regime it ended and the regime that followed.
- Estimator
- Pruned exact linear-time segmentation under a penalty fixed in advance and published with the segmentation.
- Known failure modes
- The penalty sets the number of segments. A penalty chosen to produce a pleasing partition is a curve-fitting exercise wearing a detector’s clothes.
- Does not establish
- A retrospective segment boundary is not a live detection and must not be read as one.
P17.4Per-channel attribution
- Definition
- The decomposition of a detected changepoint into which series moved and by how much — the difference between "something changed" and "rates and the dollar changed; equities did not".
- Estimator
- Each channel’s contribution to the detection statistic at the changepoint, reported per channel rather than as an anonymous multivariate alarm.
- Known failure modes
- Attribution across strongly correlated channels is not unique. Where two channels are near-collinear the split between them is reported as indeterminate rather than arbitrarily assigned to one.
- Does not establish
- Attribution does not establish causation between channels.
P17.5Surrogate calibration
- Definition
- Detection thresholds set against surrogate data generated to match the panel’s null structure, so false-alarm behaviour is measured rather than assumed.
- Estimator
- The threshold is chosen at a declared false-alarm rate over a surrogate ensemble whose generation procedure and seed publish with the threshold.
- Known failure modes
- Surrogates encode an assumption about the null. A surrogate that is too well-behaved understates the false-alarm rate the detector will actually produce on real data.
- Does not establish
- A calibrated threshold does not establish that detections are real. It states the false-alarm rate the calibration targeted.
Standing limit. A detected changepoint is a structural statement about the data, not a trade signal. Where the surrogates say the data cannot support a break call, the honest null publishes instead.
Full instrument page →
Dependence-aware positioning analytics on Commitments-of-Traders data — the version with the arithmetic still attached.
P12.1Positioning z-score
- Definition
- A category’s current net position expressed in standard deviations from its own history, with the lookback window printed beside the number.
- Estimator
- Standardisation over a lookback fixed before the data is read. The window is part of the number, and is never selected after seeing which window looks dramatic.
- Known failure modes
- A z-score is a claim about a reference window. Non-stationary open interest makes long windows misleading and short windows noisy, and neither problem is visible in the score itself.
- Does not establish
- An extreme reading does not establish that a reversal is due. It states where the reading sits in its own history.
P12.2Flip detection
- Definition
- A crossing of a category from net long to net short, or back, recorded with its date and its magnitude.
- Estimator
- Sign change on the net series between consecutive reports, with magnitude taken as the change in net position across the crossing.
- Known failure modes
- A category oscillating around zero produces many crossings of no consequence. A minimum-magnitude bar is declared in advance rather than chosen to make the series look clean.
- Does not establish
- A flip is an event in a series. It does not establish a turn in price.
P12.3Dependence-aware joint change point
- Definition
- A structural break in the positioning panel treated as one multivariate object, because the categories are components of the same open interest rather than independent series.
- Estimator
- Hotelling T-squared with covariance shrinkage for joint mean shifts, CUSUM for sequential drift, and PELT for retrospective segmentation, with per-channel attribution at each detected break.
- Known failure modes
- Treating dependent categories as independent double-counts every move; conversely, shrinkage on a short panel can suppress a real break. Thresholds are surrogate-calibrated rather than taken from asymptotic formulas.
- Does not establish
- A detected break is a statement about the panel. It does not establish a cause and it is not a signal.
P12.4Resolution-versus-fizzle split
- Definition
- How often a positioning extreme has been followed by a move and how often it has not, published only with the denominator behind it.
- Estimator
- Outcome classified under a rule fixed in advance — horizon, threshold, and what counts as a resolution — and counted over every qualifying occurrence in the sample. Raw counts publish beside the ratio.
- Known failure modes
- The classification rule determines the answer. A horizon chosen after seeing outcomes manufactures whatever split is wanted, which is why the rule is pre-registered and the counts are published with it.
- Does not establish
- A base rate does not establish what happens next. It is a count over a past sample under a stated rule.
P12.5Report-date to publication-date lag
- Definition
- The interval between the date a positioning report describes and the date it became public, carried on every reading.
- Estimator
- Both dates are recorded at ingestion and displayed together; the lag is their difference and is never smoothed away.
- Known failure modes
- Vendors sometimes restate a report. A restatement changes the pairing and is recorded as a new event rather than overwriting the original reading.
- Does not establish
- Displaying the lag does not remove it. A weekly, lagged, self-reported dataset stays weekly, lagged and self-reported.
Standing limit. Every resolution or fizzle publishes with its sample size visible — a rate without a denominator is marketing, not measurement.
Full instrument page →
State-space identification of volatility and liquidity regimes, built around one constraint: a printed label never repaints.
P11.1Regime label
- Definition
- The identified volatility or liquidity state for a bar — trending, mean-reverting, transition, stress — printed once and never revised.
- Estimator
- State-space identification with the model form and state definitions pre-registered before the feed goes live. A bar’s label is computed from information available at that bar only, which is what makes the printed history equal to the live one.
- Known failure modes
- Prefix stability costs accuracy at boundaries: the label printed at a turn is the label with the least information behind it, and it stands. The feed says so on every surface rather than quietly sharpening the history later.
- Does not establish
- A label does not establish what the next state will be, nor that the state definitions are the right partition of market behaviour.
P11.2Per-label confidence
- Definition
- The model’s confidence in each printed label, with boundary bars marked as boundary bars at the moment of printing.
- Estimator
- Posterior state probability at the bar, published with the label rather than after it. Bars below the registered confidence bar are marked as boundary bars.
- Known failure modes
- Confidence is model-conditional. A model wrong about the number of states is confidently wrong, which is why the state count is part of what is pre-registered rather than tuned.
- Does not establish
- High confidence does not establish that a label is correct.
P11.3Transition receipt
- Definition
- The record of the inputs behind every state change, kept so a transition can be recomputed instead of trusted.
- Estimator
- Inputs at the transition bar are recorded and content-hashed at print time; a recomputation from the receipt must reproduce the transition exactly.
- Known failure modes
- A receipt records the inputs, not their quality. Upstream data revised after the print would break recomputation, which is why the inputs are hashed at print rather than referenced by location.
- Does not establish
- A reproducible transition does not establish that the transition was correct.
P11.4Repaint audit
- Definition
- The check that a printed label has not changed — the single property the whole instrument exists to hold.
- Estimator
- Every published history is compared against the independently archived print sequence. Any divergence is a defect and publishes as one.
- Known failure modes
- The audit only works if the print sequence is archived independently of the published history. Archiving both from the same store would make the audit circular, and it is deliberately designed not to be.
- Does not establish
- An audit showing no repaint does not establish that the labels are accurate — only that they are the labels that were printed.
Standing limit. Prefix stability costs accuracy at regime boundaries, and the feed says so on every surface. It identifies the regime you are in, with stated lag — it does not predict the next one.
Full instrument page →
Versioning rule: this page is v1.0. A change to any definition, estimator, failure mode or limit above increments the version and appends an entry to the changelog in §04 naming what changed. Definitions are never edited silently, because a definition that can move after a result is published is not a definition — it is a degree of freedom.
Pre-registrations
A pre-registration is a declaration made and dated before the measurement runs: the thresholds that will decide, the rules that will exclude, and the sample bar below which the honest null publishes instead of a number. Its entire value comes from its ordering. Posted before the answer is known it is a constraint; posted afterwards it is a description of a result, which is a different and much cheaper object wearing the same clothes.
Pre-registration record · Premium Add-onNOT YET PUBLISHED
No pre-registration has been posted for the Premium Add-on suite. Not one that is pending review, not one that is drafted and unhashed — none. This block is the structure a registration will occupy, printed empty on purpose, because the alternative is a page that describes a discipline while quietly implying it has already been exercised.
A pre-registration is worth exactly the provability of its ordering. It has to be posted, dated and content-hashed while the answer is still unknown; posted afterwards it is a description of a result, which is a different and much cheaper object. So the first registration cannot be backdated into this slot, and the slot stays visibly empty until one is posted in the only way that counts.
The field schema of a pre-registration record for the Premium Add-on suite: each field, what it will hold, and its current value. Every value reads NOT YET PUBLISHED because no registration exists.| Field | What it will hold | Value |
|---|
registration_ref | The permanent identifier this registration is cited by. | NOT YET PUBLISHED |
|---|
scope | The instruments and the measurement window the declaration binds. | NOT YET PUBLISHED |
|---|
declared_utc | When the declaration was posted — necessarily before any data was touched. | NOT YET PUBLISHED |
|---|
first_data_utc | When collection began. This must fall after the line above, and the ordering is the evidence. | NOT YET PUBLISHED |
|---|
thresholds | Every decision threshold, fixed while the answer was still unknown. | NOT YET PUBLISHED |
|---|
exclusion_rules | What will be dropped from the sample, and on what stated grounds. | NOT YET PUBLISHED |
|---|
completeness_bar | The minimum sample below which the honest null publishes instead of a number. | NOT YET PUBLISHED |
|---|
document_sha256 | The content hash of the registered document itself. Any later edit changes it, visibly. | NOT YET PUBLISHED |
|---|
amendments | Appended corrections, each with its own date and reason. The original text stays. | NOT YET PUBLISHED |
|---|
What a Premium Add-on registration must fix in advance. The lists below are classes of declaration, not declarations. They name the decisions that have to be made before the data is touched, because each one is a decision that could otherwise be made afterwards, in the direction that flatters the result. No value below has been registered.
Thresholds
- The deltas at which risk reversals are computed, and the quoting convention they are computed under.
- The state count in the regime model, and the confidence below which a bar is marked a boundary bar.
- The false-alarm rate a detection threshold is calibrated to, and the surrogate procedure that calibrates it.
Exclusion rules
- Which strikes and expiries enter an aggregation, and how an incomplete chain is drawn as incomplete.
- The lookback window for every normalisation, fixed before the series is read.
- What separates forced flow from discretionary flow, and what happens to an episode where the separation cannot be made.
Completeness bars
- The minimum chain coverage below which a map is not published at all.
- The minimum panel history required before a joint break can be called.
- The minimum episode count below which a resolution-versus-fizzle split publishes as a null rather than a rate.
Immutability, stated before it is tested. Once a registration is posted it is not edited. If it is wrong, an amendment is appended beneath it carrying its own date and the reason for the change, and the original text stays where it is, readable, above the correction. A registration that quietly improved after the data arrived would be indistinguishable from one that was right all along — which is precisely why the append rule is written here, now, while there is nothing yet to be tempted by.
Amendment rule, stated in advance: a posted registration is never edited. An amendment is appended beneath the original carrying its own date and its reason, and the original text stays above it, readable. This page will show both.
Proof corpus
The corpus is the set of artifacts a published measurement ships with — not a description of them, the artifacts themselves, downloadable, each with the digest that proves you received the bytes I measured and the code tag that produced them. A result whose inputs cannot be downloaded is not a receipt; it is an assertion with better typography.
The corpus for this suite is empty. Every row below is a slot, and every slot is NOT YET PUBLISHED. The table is printed anyway, because a reader should be able to see the exact shape of what will arrive — and because a page that described a corpus without showing how empty it currently is would be making the claim it exists to refuse.
Five columns: artifact, contents, content hash, code tag, download. Scroll sideways if they do not all fit.
Proof corpus for the Premium Add-on suite: the seven artifact classes a published receipt carries, what each will contain, and its current state. Every content hash, code tag and download reads NOT YET PUBLISHED, because no artifact from this suite's proof corpus has been published yet.| Artifact | What it will contain | Content hash | Code tag | Download |
|---|
| Registered methodology document | The versioned document these definitions are taken from, in the exact form it was registered — estimators, parameters, and the limits stated above. | SHA-256NOT YET PUBLISHED | NOT YET PUBLISHED | NOT YET PUBLISHED |
|---|
| Pre-registration record | The dated declaration: thresholds, exclusion rules and completeness bars, plus any amendments appended beneath the original with their reasons. | SHA-256NOT YET PUBLISHED | NOT YET PUBLISHED | NOT YET PUBLISHED |
|---|
| Source snapshot set | The option-chain snapshot, positioning panel or multivariate matrix the instrument consumed, each with its as-of timestamp and its own digest. | SHA-256NOT YET PUBLISHED | NOT YET PUBLISHED | NOT YET PUBLISHED |
|---|
| Result set | Per-metric results with intervals and effective sample sizes, and the honest nulls wherever a sample could not support a metric. | SHA-256NOT YET PUBLISHED | NOT YET PUBLISHED | NOT YET PUBLISHED |
|---|
| Can-fail proof transcript | For every test in the battery: the planted defect, the refusal that was expected, and the outcome that was observed. | SHA-256NOT YET PUBLISHED | NOT YET PUBLISHED | NOT YET PUBLISHED |
|---|
| Kill-ledger extract | Hypotheses registered against this suite and killed by the data, each with the run that killed it. Published with the same visibility as a registration. | SHA-256NOT YET PUBLISHED | NOT YET PUBLISHED | NOT YET PUBLISHED |
|---|
| Generation code | The tagged commit that produced the result set, with its build receipt. Named here because a result whose code version is unstated cannot be re-run. | SHA-256NOT YET PUBLISHED | NOT YET PUBLISHED | NOT YET PUBLISHED |
|---|
Corpus state, per instrument
Per instrument, so that the emptiness cannot hide behind a suite-level summary.
Changelog
Append-only, dated, written by a person. It records changes to this page — it is not a measurement log, and it will not become one. A correction is a new entry that names the entry it corrects; nothing here is ever edited away, because a changelog you can rewrite is a marketing surface with a monospace font.
2026-08-01 · v1.0
Receipt page established for the Premium Add-on suite, carrying all five parts of the §8.2 anatomy: the versioned methodology with a formal definition, estimator, failure modes and non-claim for each of the 21 metrics its 5 instruments measure; the pre-registration structure with no registration in it; the proof-corpus table with no artifact in it; this changelog; and the independent-recomputation procedure. Supersedes the earlier per-suite methodology summary at this URL, which carried the battery outline without the receipt anatomy. No measurement, pre-registration, artifact or hash accompanies this version — every receipt slot below is empty as a matter of fact, not of omission.
Independent recomputation
Independent recomputation is what makes the rest of the page checkable rather than merely well-written. It is the exact sequence a stranger follows to re-derive a published number from this suite using only artifacts I published — no account, no request, no conversation with me at any point.
Today the procedure terminates at step 1, because no measurement from the Premium Add-on suite has been published and there is no receipt to open. The steps are written now, in the specific form they will take for this suite, precisely so that they exist before the first result does and cannot afterwards be shaped to fit one.
Open the receipt and take its four identifiers.
Every published measurement links a receipt carrying four: the registration reference, the methodology-document digest, the input-manifest digest, and the generation-code tag. If any one is missing, stop — the result is not recomputable and should not be treated as though it were, including by me.
Verify the methodology document against its digest.
Download it, hash it, compare. A mismatch means the method you are about to apply is not the method that was registered, and everything after this step would be measuring a different thing.
Check the ordering before you check anything else.
The registration timestamp must precede the first-data timestamp on the manifest. If it does not, the registration is a description of a result rather than a constraint on one, and no statistic downstream can repair that.
Download the source snapshot named in the manifest and verify its digest.
The option-chain snapshot, the positioning panel, or the multivariate matrix. Every snapshot carries its as-of timestamp; a map or a break without one is not recomputable and should not be treated as though it were.
Apply the registered exclusion rules yourself.
The convention set, the lookback window, the family declaration, and the minimum-sample bar below which the honest null publishes instead of a number.
Check out the instrument at the code tag on the receipt and run it.
Where thresholds are surrogate-calibrated, the receipt names both the surrogate-generation procedure and its seed. Both are required to reproduce a threshold; neither is optional.
Recompute the published quantity and compare.
Strike aggregations and expiry notionals to the published precision. Regime labels bar for bar — a single relabelled bar is a repaint and therefore a defect. Detected changepoints as an exact set, with their per-channel attribution.
Run the can-fail proof.
On a surrogate series containing no break, confirm the detector’s false-alarm rate matches the rate its calibration targeted. A detector that never fires on noise has been tuned, not calibrated.
If your number differs, the difference is the finding.
Send it with your inputs and the version you ran. A confirmed discrepancy publishes as a correction appended beside the original — and the original stays exactly where it is, unedited, because the error is the part of the record that proves the discipline is real.
The point of publishing this before there is anything to check: a recomputation procedure written after a result is a procedure written by someone who already knows which steps would be inconvenient.