How This Works

What Unlock Market Intelligence measures

Unlock Market Intelligence (UMI) is a live quantitative and editorial intelligence layer built for professional readers — institutional investors, traders, regulators, and fintech executives operating in digital asset markets. It does not predict prices. It characterises the current market regime and identifies which signals are driving it.

Two composite scores are computed for each asset on a continuous basis: a Direction Score measuring bullish vs. bearish signal alignment, and a Stability Score measuring structural quality and consistency of the market environment. Together they produce a Regime Label — the system's read of the current market state.

All scoring resets at 09:00 GST (Gulf Standard Time, UTC+4) each day. This defines the active cycle — the window over which all intra-day moves, regime shifts, and signal changes are evaluated.

How the scores are built

The Direction Score is a weighted composite of independent market signal components. Each component produces a score from 0 (strongly defensive) to 100 (strongly constructive), with 50 representing a neutral, unconfirmed state. Components cover price action, cross-asset dynamics, venue breadth, on-chain and network structure, and institutional positioning.

The Direction Score is the weighted composite of these signals. The Stability Score is derived separately — it measures the structural quality of the current regime: how consistent the signals are, how broad the venue coverage is, and how much price volatility is present. Stability is directionally neutral: a strongly bearish market can score high on stability if the signals are coherent and the move is confirmed across sources.

Components are weighted to reflect their relative information value for directional quality in the digital asset context. Price action carries the highest weight; macro and sentiment signals carry the lowest.

Reading the Direction and Stability scores

Both scores run from 0 to 100. The midpoint (50) represents a genuinely neutral state — not a failure to signal, but the absence of a confirmed directional or structural read.

A high Direction Score with a low Stability Score indicates a directional move that is not yet confirmed by structure — it may be early, thin, or fragile. A high Stability Score with a neutral Direction Score indicates a market holding form but awaiting a catalyst. Both scores together produce the Regime Label.

What each regime means

The system maps Direction and Stability scores to one of five regime labels. These are not predictions — they are characterisations of the current moment, updated on each data tick.

Where the data comes from

The composite price and scoring inputs are assembled from multiple independent sources on every collection cycle. Sources that deviate significantly from the cross-venue consensus are excluded automatically to protect against outlier distortion.

The system draws from four categories of data: live exchange prices from both centralised and decentralised venues; on-chain and network signals for each asset (gas activity, hash rate, staking, stablecoin flows); broad market context including global market cap, dominance, sentiment indicators, and altcoin rotation; and institutional positioning signals from US spot ETF activity. Data sources include CoinGecko, CoinMarketCap, Etherscan, and direct blockchain network APIs.

All data inputs are refreshed on a continuous rolling basis with per-source intervals calibrated to data volatility. Price data updates every 30 seconds; on-chain and macro signals update on longer intervals appropriate to each source.

The 09:00 GST cycle and why it matters

All performance metrics on this platform are measured against the cycle open — the first confirmed price snapshot after 09:00 GST (UTC+4) each day. This boundary was chosen because it aligns with the start of the Gulf business day, which represents a meaningful market open for MENA institutional participants and overlaps with early European session activity.

The "cycle change %" shown on asset pages reflects the move from that morning's cycle open, not the standard 24-hour exchange window. This is intentional: the standard 24h metric is timezone-agnostic and often obscures the actual intra-session structure that professional readers need to interpret.

All scores, regime labels, and cross-asset metrics reset at cycle open. The first 15–30 minutes of each cycle are a "Building" period as the system accumulates sufficient ticks for confident scoring.

BTC / ETH correlation, beta, and lead-lag

The cross-asset page computes three metrics across three time windows (active cycle, last 60 minutes, last 15 minutes) from synchronised live price ticks.

• Correlation — Pearson correlation of per-minute returns. Values near 1.0 indicate tight co-movement; near 0 indicates decoupling.
• ETH Beta — ETH's sensitivity to BTC moves. A beta of 1.2 means ETH tends to move 1.2% for every 1% BTC moves. Higher beta means ETH amplifies BTC. Lower means ETH is driven more by its own factors.
• Lead-Lag — Identifies which asset leads price discovery by scanning lag offsets of ±5 one-minute buckets. A clear lead requires a correlation of ≥ 0.35 at the optimal offset. "Simultaneous" means no detectable directional lead.

Relative strength (ETH return minus BTC return in the window) feeds directly into the BTC/ETH Transmission component of the scoring matrix — ETH outperformance adds positive weight to ETH's direction score and mildly negative weight to BTC's.

How editorial judgement interacts with the model

The system operates in two layers: a quantitative model that updates automatically on every collection cycle, and an editorial layer where Unlock Blockchain analysts can apply informed context and calibration.

Analysts can activate macro and geopolitical pressure signals to reflect conditions not fully captured by quantitative data — regulatory shifts, monetary policy decisions, geopolitical risk events, or structural market changes. These inputs influence the scoring in a bounded way: they inform the read without overriding the underlying market data.

Daily editorial reads — when published — take precedence over the AI-generated signal on all public-facing pages. The AI brief is suppressed entirely when an editorial read is present. This hierarchy is intentional: AI signals provide analytical continuity between editorial publications; they do not replace analyst judgement.

How the AI brief is generated

When no editorial read has been published for the current cycle, the platform generates an AI intelligence brief for each asset. The brief consists of three structured sentences: a framing of the current regime, identification of the dominant driver or constraint, and a specific threshold to watch for regime confirmation or invalidation.

The brief is generated once per cycle and cached — not regenerated on every page load. It draws from live scoring data: direction and stability scores, the dominant signal component, institutional flow tone, market sentiment, and any active macro context. The brief is analytical in nature, not predictive, and does not constitute financial advice.