Skip to content
Market Update

Weekly Funding Rate Recap — Mar 23–29, 2026

Zirodelta Research
Quantitative Research5 min read

48,069 settlements. Three exchanges. One token — NOM — dominated the extremes like nothing we've tracked in weeks.

Exchange Summary: Mar 23–29

ExchangeSettlementsAvg Rate (%)Positive %Max Rate (%)Min Rate (%)
KuCoin20,860-0.009046.4%+1.2887-2.0000
Binance14,804-0.007368.1%+0.4644-2.0000
Bybit12,405-0.010874.6%+0.2172-2.2484

All three exchanges closed the week negative. Bybit ran the deepest average at -0.0108%, despite having the highest positive-settlement ratio at 74.6%. That combination — mostly positive but deeply negative when it swings negative — tells you the extremes are doing the heavy lifting on the mean.

KuCoin sits on the opposite end. Only 46.4% of settlements were positive, meaning the majority of KuCoin resolutions printed negative. Its max rate of +1.2887% came from a single IRUSDT spike. Its floor was -2.0%, the hard cap.

Day-by-Day Shape of the Week

DateSettlementsAvg Rate (%)Positive %
Mar 2313,536-0.009357.2%
Mar 2413,284-0.008862.7%
Mar 2513,222-0.004465.9%
Mar 269,392-0.014154.9%
Mar 274,689-0.009259.4%
Mar 283,098-0.011952.2%
Mar 292,668-0.011651.3%

The week started recovering. March 25 had the least negative average (-0.0044%) and the highest positive share (65.9%). Then March 26 hit and the trend reversed.

The settlement count drop from 13,000+ to under 5,000 mid-week isn't a data gap. It reflects which symbols actively settled on those days — Thursday and Friday had fewer qualifying hourly and 4-hourly settlements across the tracked pairs.

The NOM Story

This is the signal of the week. NOMUSDT — the Network of Machines token — entered a funding rate spiral that ran for roughly 30 hours straight.

Date/Hour (UTC)Binance Avg Rate (%)NOM Mark Price
Mar 27, 06:00+0.0050$0.001856
Mar 28, 11:00-0.1498$0.002079
Mar 28, 13:00-0.7615$0.002443
Mar 28, 17:00-1.0359$0.002765
Mar 29, 00:00-1.2790$0.002555
Mar 29, 08:00-0.5187$0.002654
Mar 30, 01:00-1.0470$0.003268

At its worst, NOM funding hit -1.28% per period at 00:00 UTC on March 29. That's annualized at roughly -11,199% if sustained. It wasn't sustained — rates mean-revert — but the direction was consistent and extreme for long enough to inflict real damage on anyone short.

The price move confirms the signal: NOM went from $0.001797 on March 27 to a 48-hour high of $0.004188. That's a +133% move, cross-validated across Binance, Bybit, Gate, and KuCoin simultaneously.

The crowd data on NOM shows long accounts at approximately 38% — well below the 50% neutral line. Shorts were the majority. Negative funding at -1%/period is the market paying those shorts to hold on, because the long demand couldn't absorb the short pressure. Eventually it could. The squeeze followed.

This is the Settled use case in real time. Anyone tracking NOM's funding rate market before March 28 had the signal. The rates were pricing in exactly what happened next.

SIREN: The Second Act

SIRENUSDT continued its own stretch of elevated negative funding — for a different reason.

Symbol48h Price RangePeak MoveAvg Funding Rate (%)
SIRENUSDT (Binance)$1.33 → $1.92+44.7%-0.093
NOMUSDT (Binance)$0.0018 → $0.0042+133.0%-0.512
PTBUSDT (Bybit)$0.00096 → $0.00261+170.5%-0.216
GFUSDT (KuCoin)$0.00107 → $0.00315+195.4%+0.048

SIREN's funding averaged -0.093% per period during the move, a holdover from the pump covered in last week's crowd trap analysis. The token moved another 44% on top of an already-elevated baseline. Shorts kept paying. Longs kept collecting.

GF is the outlier: positive funding through a 195% move. That means longs paid shorts the entire time. Structural buyers, no squeeze — just genuine demand.

Crowd Signal Watch: STRK

One setup standing out at week's end: STRKUSDT on Bybit.

Long accounts at 81.9% of all tracked positions. Funding rate sitting at approximately -0.008% per period — the market paying longs marginally, not enough to signal a squeeze, but the crowd skew is extreme.

The last time a symbol hit 80%+ long concentration on Bybit, the position reversion came within 24–48 hours in 7 of the past 10 instances in our data. That's not a guarantee. It's a setup to watch.

See STRK's live funding data and open Settled markets.

What the Week Tells You

Negative bias dominated. The positive-settlement ratio on KuCoin was below 50% — meaning more settlements printed negative than positive. The extreme outliers on the downside (NOM, SIREN, PTB) weren't noise: they were the week.

The Ornstein-Uhlenbeck mean reversion model flagged NOM's extreme divergence as a reversion candidate before the move. The half-life on a -1.28% rate is measurable in hours, not days.

Three exchanges reporting negative average rates in the same week — with Bybit's positive-settlement ratio at 74.6% while still printing a negative mean — reflects a market structure where the long tail of alts carries most of the downside weight while blue-chips stay near neutral.

That divergence is exactly where edge lives.


Explore the data:

Zirodelta Research

The research arm of Zirodelta. Data-driven analysis of crypto sentiment markets, model development, and market microstructure research. Data-driven. Real-time. Across 6 exchanges and 3,700+ perpetual futures.

Learn more about Zirodelta
HomeMarketsRatesBoardPortfolio