Arizona time, daily. Approximately halfway through the UTC daily candle.
NOVEX INTELLIGENCE / OPERATING STANDARD
The receipt matters as much as the read.
Every Daily Market Regime begins with public evidence, keeps interpretation labeled, and ends with a permanent record. This is the exact system behind the archive.
PUBLIC
INPUTS EXPLICIT
FORMULAS HUMAN
APPROVAL PERMANENT
ARCHIVE
INPUTS EXPLICIT
FORMULAS HUMAN
APPROVAL PERMANENT
ARCHIVE
4GATES
BTC, ETH, and SOL rolling statistics from one named venue: Coinbase.
Observed facts, labeled desk read, and the unknowns the evidence cannot resolve.
One canonical Clutch post. One evidence reply. One Novex repost. One archived edition.
01 / SOURCE HIERARCHY
Public evidence before interpretation.
Novex uses credential-free public endpoints for the daily report. Each source has a narrow job, and every edition from 036 forward publishes the exact hash-verified public snapshot behind the read. No provider is treated as proprietary Novex data.
PUBLICSENTIMENT
CoinMarketCap
Fear & Greed value, classification, latest public update, and exact-date daily history comparisons.
Open public endpoint ↗PUBLICGLOBAL ESTIMATE
CoinGecko
Total crypto market capitalization, aggregate 24-hour volume, their reported 24-hour changes, and broad dominance estimates.
Open source ↗PUBLICSTABLECOINS
DefiLlama
Circulating supply for assets classified as pegged to the US dollar.
Open source ↗PUBLICVENUE TAPE
Coinbase Exchange
Rolling BTC/USD, ETH/USD, and SOL/USD product statistics. One venue is not the market.
Open source ↗PUBLICPERPS DESK
Orderly via Novex Scanner v3
Execution-eligible perpetual-market breadth, catalog coverage, estimated funding, marks, and open interest for BTC, ETH, and SOL. Venue-specific context, not the whole market.
Open public Scanner feed ↗Source marks identify each public-data role. They do not imply sponsorship, endorsement, or that a provider operates Novex.
02 / METRIC CONTRACT
Every displayed number has a definition.
The report does not mix labels, calculations, and interpretation into one claim. Each measure keeps its source and scope attached.
M01
Fear & Greed
The current CoinMarketCap public value and classification. Daily and seven-day changes are exact-date point changes, not percentages.
M02
Total crypto market cap
The CoinGecko global estimate at capture time. Its 24-hour change is the provider's reported estimate and can change after publication.
M03
Aggregate crypto volume
The CoinGecko global 24-hour volume estimate and its reported 24-hour percentage change. This is aggregator-wide turnover context, not exchange-specific executed volume.
M04
USD stablecoin dominance
DefiLlama peggedUSD supply ÷ CoinGecko total crypto market cap × 100
This is a calculated share of market capitalization. It is not “cash waiting to enter,” future buying power, or a directional signal.
M05
Rolling Coinbase change
(last price − rolling open) ÷ rolling open × 100
“Moved most” means the largest absolute rolling percentage change among BTC, ETH, and SOL on Coinbase.
M06
Rolling range
(rolling high − rolling low) ÷ rolling open × 100
“Traded widest” identifies the largest calculated range among the three named Coinbase markets.
M07
Execution-eligible perps breadth
Advancers, decliners, and flat markets are counted only across complete Novex Scanner v3 markets eligible for the breadth denominator. Incomplete markets remain visible in coverage and are excluded from the count.
M08
Estimated funding
The Orderly estimated funding rate is converted to basis points with rate × 10,000. It is an estimate at capture time, not a guaranteed settlement rate.
M09
Open interest comparison
(current OI ÷ prior verified OI − 1) × 100
The comparison uses the prior standard fixed-time brief only when its capture age is inside the accepted window. Otherwise the comparison is labeled unavailable. Edition 036 is the first evidence-backed perps snapshot, so no prior Orderly open-interest baseline existed.
03 / CAPTURE WINDOW
A snapshot is not a closing price.
The daily workflow runs at 07:00 America/Phoenix. Arizona remains on UTC−7 throughout the year. Other audiences may move their clocks, but the Arizona publishing time stays fixed.
The report describes the market at capture time, approximately halfway through the UTC daily candle. Coinbase statistics use each product's rolling 24-hour window. Orderly-backed perps values use the Scanner capture timestamp. Those are different clocks and remain labeled.
TIME DISCLOSURE
CAPTURE ≠ CLOSE
Later market movement does not retroactively make the original snapshot inaccurate.
04 / EDITORIAL LAYERS
Facts, judgment, and uncertainty stay separate.
What the sources establish.
Values, changes, classifications, venue scope, capture times, and calculated comparisons.
How Clutch interprets the state.
A concise read grounded in the observed layer. It remains labeled as interpretation.
What the evidence cannot prove.
Causation, global confirmation, future direction, and where stablecoin supply moves next.
HARD BOUNDARY
No targets, stops, position sizing, personalized recommendations, guaranteed outcomes, or manufactured urgency.
05 / PUBLICATION GATES
One verified chain from report to archive.
- 01
Capture and validate
Public responses, timestamps, formulas, ranges, and source freshness are checked. Every output receives a SHA-256 receipt.
- 02
Build the publication package
The report, X copy, animated video, captions, evidence card, ALT text, reply copy, accounts, and repost sequence are bound to one approval code.
- 03
Human approval
Clutch reviews the exact Telegram bundle. Nothing publishes before the matching, unexpired approval code is received.
- 04
Publish and read back
The Clutch video post and evidence-card reply are read back and verified. Novex then reposts the canonical Clutch post.
- 05
Archive the completed receipt
Only the completed publication receipt can generate the permanent edition page. The approved film, captions, transcript, evidence card, exact public JSON snapshot and SHA-256, original X receipt, searchable ledger, RSS item, and sitemap route move together.
06 / FAILURE POLICY
Missing evidence produces a stop, not filler.
Hold the package. Do not substitute a remembered or estimated value.
Fail validation. Preserve the evidence and issue no approval request.
Stop publication when a visible provider label, transcript claim, manifest scope, or public evidence hash disagrees.
No fallback image-only post. The full video package must pass.
Keep verified X posts intact, alert operations, and resume the archive update from the completed receipt.
07 / CORRECTIONS
The record can change, but never silently.
A material error in a metric, formula, date, source, venue label, scope statement, or interpretation receives a timestamped correction on the affected edition. The visible modified date, structured data, RSS entry, and archive record are updated together.
Typography that does not alter meaning can be corrected without a formal note. A later market move is not a correction to an accurately timestamped snapshot.
SEPTEMBER 1, 2026
Editions 036 to 044 received a visible source-label correction: CoinMarketCap replaces the incorrect Alternative.me attribution, and the Orderly-backed Novex Scanner v3 feed is now named. Published metrics, transcripts, timestamps, signed snapshot hashes, and X receipts did not change.
CORRECTION ENTRY
WHAT CHANGED
The exact claim or value.
WHEN
UTC timestamp of the amendment.
WHY
The source or process error.
08 / ANALYTICS AND PRIVACY
Measure whether the work gets used.
The publishing layer can record report views, qualified reads, product clicks, X clicks, Telegram clicks, RSS clicks, and public-source opens through named events. The event payload is limited to the event name and page path.
The Intelligence layer does not read wallet addresses, private account data, credentials, positions, or trading activity.
Disclaimer: Market information, not personalized financial advice. Aggregator estimates and venue-specific prices can change quickly.