Why Do Perps Markets Need Their Own Regimes?

Spot price tells you where an asset trades. It says nothing about how the leverage underneath it is positioned — and on perps venues like Hyperliquid, that positioning is the trade. A coin can be flat on the day while funding pins at +0.1% per 8h, open interest hits a record, and one long-liquidation cluster is all that stands between calm and cascade.

Those signals don't exist in spot markets. They are derivatives-native: funding-rate extremes, OI imbalance, liquidation clusters, and the perp-vs-spot basis. Reading them needs different endpoints than a candle feed.

This post is the hub for two of them — Regime #4 (Derivatives-Native) and Regime #8 (Carry Trade / Basis) — out of the 14 in the regime taxonomy. We cover what each measures, the exact endpoints, and when to reach for which. For the deep dives, we cross-link the detailed posts rather than repeat them.

Regime #4: The Four Derivatives-Native Setups

Regime #4 is the directional layer. It is where leverage positioning turns into a tradable signal on a timescale of minutes to days. Four setups dominate, each backed by a live endpoint.

For the cascade mechanics in full, see the liquidation data API, and for positioning bias the long/short ratio as a contrarian indicator.

Pulling Funding, OI, and Liquidations

All three sit under /api/v1/market-intelligence/. They default to the Hyperliquid perps coin-set; pass exchange=all, binance_spot, or asterdex to widen or narrow it, and symbol=BTC for a single coin.

curl -H "X-API-Key: cdk_live_yourkey" \
  "https://cryptodataapi.com/api/v1/market-intelligence/funding-rates?symbol=BTC"

Each funding entry carries symbol plus two arrays — stablecoin_margin_list and token_margin_list — with per-exchange funding_rate, interval, and next_funding_time:

{
  "symbol": "BTC",
  "stablecoin_margin_list": [
    {"exchange": "Binance", "funding_rate": 0.0102,
     "interval": 8, "next_funding_time": 1749657600000},
    {"exchange": "Hyperliquid", "funding_rate": 0.0125,
     "interval": 1, "next_funding_time": 1749639600000}
  ],
  "token_margin_list": []
}

Liquidations return liquidation_usd_4h, long_liquidation_usd_4h, and short_liquidation_usd_4h (plus 1h/12h/24h windows) per coin. Open interest returns the top ~25 perps, each with an exchange_list of per-venue open_interest_usd and open_interest_change_percent_4h.

The funding field exposes a key nuance: with exchange=hyperliquid you get HL coins, but the margin lists still show every venue's funding. See the cross-exchange funding rates guide for that distinction.

Reading OI Divergence as a Fragility Gauge

Funding tells you who is paying; OI divergence tells you whether the positioning is fragile. The /liquidity/oi-divergence endpoint ranks the tracked universe by 4h divergence — open-interest change minus price change.

import httpx

HDR = {"X-API-Key": "cdk_live_yourkey"}
BASE = "https://cryptodataapi.com/api/v1"

# Coins where OI is rising faster than price = fragile longs
rows = httpx.get(
    f"{BASE}/liquidity/oi-divergence", headers=HDR,
).json()["rows"]

for r in rows[:5]:
    print(r["coin"], r["divergence_4h"],
          "oi", r["oi_change_4h_pct"],
          "px", r["price_change_4h_pct"])

Each row carries coin, price, open_interest_usd, price_change_4h_pct, oi_change_4h_pct, and divergence_4h (also 1h and 24h windows). A positive divergence_4h means OI is climbing faster than price — new leverage stacking into a stalling move. That is the textbook setup before a forced unwind.

Coins without enough history for a window report null for it, so guard for missing values before you size anything.

Regime #8: When Basis Becomes Its Own Regime

Regime #4 is directional and fast. Regime #8 — Carry Trade / Basis — is slower and structural, measured in days to weeks. The futures basis (perp vs. spot spread, expressed as annualised APR) is not just a funding signal; it is a regime-health indicator in its own right.

The reason it earns a separate regime: basis describes the term structure of leverage, not a single trade. The bands map to distinct market states:

You reconstruct basis from the same primitives — persistent funding via /market-intelligence/funding-rates joined to OI from /market-intelligence/open-interest. The signal is the persistence, not any single 8h print.

Regime #4 vs Regime #8: Same Inputs, Different Questions

Both regimes are built from funding, OI, and basis. The difference is the question they answer and the timescale they answer it on. Read them together — directional setups inside a fragile basis regime are how the sharpest reversals form.

Regime #4 — Derivatives-NativeRegime #8 — Carry / Basis
What it measuresDirectional positioning extremesTerm structure of leverage (perp−spot APR)
Core signalFunding >0.1%/8h, OI imbalance, cascade clusterBasis 8%+ = crowded; <0% = collapse
TimescaleMinutes – daysDays – weeks
BiasBoth (fade or ride the squeeze)Regime health indicator
Primary endpoint/market-intelligence/funding-rates,
/liquidity/oi-divergence		/market-intelligence/funding-rates + /open-interest

Put simply: #4 tells you a trade is overheated right now; #8 tells you whether the whole leverage structure is fragile enough for that trade to cascade.

When to Reach for the Derivatives & Carry Regimes

These two regimes earn their place in three concrete jobs:

What these regimes aren't: standalone alpha. Funding extremes can stay extreme for days, and basis can grind higher long after it looks crowded. Pair them with the rest of the 14-regime framework so you trade the context, not just the number. These are regimes #4 and #8 of 14 — perps-native by design, and most powerful when read against the other twelve.

Forward-looking companion: the new quant engine turns these into probabilities — its funding and open-interest prediction heads forecast where carry and leverage are headed, so you can position ahead of the crowding instead of reacting to it.