Price is an output, not a cause
"Why did it move?" is the first question after every candle, and the honest answer is rarely the headline. Prices in crypto are the visible output of a stack of forces — liquidity, positioning, flows, narrative and structure — that operate at different speeds and interact with each other. Understanding that stack is what separates forecasting from reacting, and this guide walks it from the slowest force to the fastest.
Layer 1 — Global liquidity (months to years)
Crypto is a long-duration, risk-first asset class, and its macro beta is violent. When global money conditions ease — rates cutting, money supply expanding, risk appetite returning — capital reaches for the smallest, most elastic boat in the ocean. When conditions tighten, that boat empties first. The major crypto cycles of the last decade have tracked the liquidity cycle with an almost mechanical quality: expansions lift everything, contractions discount everything, regardless of individual coin news flow.
The practical monitoring set: policy rate direction, global money supply trend, and — crypto's own liquidity gauge — stablecoin net issuance. When stablecoins are being minted, fresh buying power is entering the system; when they're being redeemed en masse, capital is leaving. This is the closest thing the asset class has to a money-flow indicator, and it leads price more often than it follows.
Layer 2 — Flows and structural demand (weeks to months)
Inside the liquidity regime, flows decide who's actually buying. Spot ETF share creations, exchange reserve drawdowns (coins leaving venues into custody), and treasury-style corporate accumulation are the durable versions — money that buys and doesn't flip. Their signature is supply absorption: price grinds up on thinning sell-side liquidity.
Their opposite — coins flowing onto exchanges in size — is distribution pressure that precedes most major drawdowns. Neither is a timing signal alone, but both change the odds of the next large move's direction, which is precisely what a forecast engine wants to know.
Layer 3 — Positioning and leverage (days to weeks)
Derivatives are the market's gearbox, and leverage is where smooth trends become violent dislocations. The key readings:
- Funding rates — the rent longs pay shorts (or vice versa) on perpetuals. Deeply positive funding at extremes means the crowd is max-long and paying to stay there; the squeeze fuel for a violent flush is fully stored.
- Open interest — total leveraged positions outstanding. Price breaking out while open interest explodes is a crowd piling in; a breakout on falling open interest is shorts covering — much weaker continuation fuel.
- Liquidation maps — where forced sellers cluster. Cascades trigger where the leverage stacked, and they explain why crypto's worst hours are so concentrated.
This layer is why technical structure alone is incomplete: a perfect-looking breakout into max-crowded positioning has an empirically different outcome distribution than the same pattern at neutral positioning. Fundamental data, not chart noise.
Layer 4 — Sentiment regimes (days)
Positioning indices (Fear & Greed style aggregates) are weak signals in the middle of their range and powerful at the rails, where they flip contrarian. Peak euphoria — euphoric readings, trending tickers, friends asking which coin to buy — means the marginal buyer is already in and exit liquidity is being manufactured. Peak panic means the marginal seller is already out and bounces arrive on no news. The emotion isn't the signal; the exhaustion of who's left to convert is.
Layer 5 — Reflexivity: the crypto-specific accelerant
Soros's observation — market perception changes the fundamentals that perception is pricing — runs hotter in crypto than in any other major market, because crypto's "fundamentals" are partly made of attention and adoption. Price up → headlines → new users → new demand → price up, until the loop's fuel exhausts. It runs in reverse just as cleanly. This loop is why crypto trends overshoot every valuation model, and why regime detection (which loop state are we in?) matters more than static fair-value estimates.
Layer 6 — Idiosyncratic events (instant, unpredictable in kind)
Token unlocks releasing months of supply into the market, exchange incidents, regulatory headlines, network outages — the event layer is unforecastable in kind but predictable in effect shape: liquidity gaps and volatility spikes. The forecast engine's job here is not clairvoyance; it is building bands wide enough that the event layer doesn't invalidate an otherwise sound distribution, and shrinking exposure to names with known event calendars (overhangs) approaching.
Putting the stack together
Read top-down: liquidity sets the allowable direction, flows confirm who's buying, positioning measures how fragile it all is, sentiment marks the rails, and reflexivity amplifies whatever layer is currently dominant. That is exactly the architecture of the FARAZ fusion engine — 24 technical models for structure and timing, a fundamental layer for liquidity, flows, funding and sentiment, and a blending layer that lets recent reality re-weight the stack. The fusion guide covers the mechanics, and every coin prediction page shows the methodology in action.