Funding on Hyperliquid io markets: who pays, and when
What it costs to hold a position on each EntropyIO (io) HIP-3 market. Every hour each market settles a funding rate; positive means longs pay shorts. Part of that rate is the deployer's built-in carry, the neutral rate, and the rest is what the crowd pays on top. Below: how many markets pay above neutral now, when in the US week longs pay most, every market's 24-hour, 7-day and 30-day rate, the largest settlements, and the same names on other deployers.
last settlement 2026-10-10 00:00 UTC · 2026-10-09 20:00 ETnext-rate estimate 2026-10-10 00:46 UTCsince 2026-08-19generated 2026-10-10 00:46 UTCdataindex.json
longs pay above neutral
8 of 9
last settlement above the market's neutral rate
at neutral
1 of 9
settled exactly at the neutral rate
shorts pay
0 of 9
last settlement below zero
median apr · 30d
+24.81%
across 9 markets · over observed hours
neutral rate
1.37% APR
0.01% per 8 h × multiplier 0.125 = 0.01562 bps/h · 7 switch to 0.5 on weekdays
hours above neutral · 30d
68%
median market · the rest at or below
largest settlement · 30d
+1,343%
EWY · +15.33 bps in one hour · 09-18 11:00 ET
hours observed · 30d
695 of 720
median market · collector gaps are not filled
Read this first. Positive funding means longs pay shorts, negative means shorts pay longs. The neutral rate is what a market settles at when the perp trades on its oracle: Hyperliquid's base interest of 0.01% per 8 hours times the deployer's funding multiplier, 0.125 on io, so 1.37% APR. It is the deployer's carry, not crowd positioning; only the part above it says longs are paying up. Seven markets (DRAM, EWY, GPRO, IONQ, NBIS, SNDK and TCNT) ran under more than one funding setting in the last 30 days (neutral 1.37–5.48% APR), and every hour is compared with the neutral rate in force when it settled. io switches their multiplier between 0.5 on weekdays and 0.125 at weekends: 51 of the 53 changes in 30 days follow that pattern. Every mean is over the settlements we observed, shown with its hours: our collector has gaps and we never fill them. This page is not live: it is rebuilt nightly and at 09:05 ET, and the next-rate estimate is the running estimate at the as-of time, not a settled rate.
When longs payUS-equity markets · New York time
hour of week · mean funding apr
mean of hourly rates over every US-equity market and every settlement in the archive, annualised · an hour is the hour the settlement covers · 168 of 168 hours with data · hover a cell for its mean and n
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Mon
Tue
Wed
Thu
Fri
Sat
Sun
shorts paylongs pay · full shade = ±83.75% APR or more · empty = no settlements
by session · 30 d
session
median apr
markets above neutral
hours observed
US regular 9:30–16:00
+26.15%
100%
983
US extended 4–9:30, 16–20
+24.53%
100%
1,634
Overnight 20–4
+24.51%
89%
1,045
Weekend / holiday
+14.65%
89%
1,653
median apr = median over markets of each market's mean rate in that session over the last 30 days
markets above neutral = share of markets whose session mean exceeded their own neutral rate
hours = market-hours observed in the session, all markets summed
sessions as in the liquidity data
All marketsAPR = hourly rate × 24 × 365 × 100, in percent · positive = longs pay
last = the last settled hourly rate, annualised
next est = the running estimate of the next rate at the as-of time, "—" when it is older than 2 hours
24 h, 7 d, 30 d = mean of the hourly rates observed in the 24, 168 or 720 hours ending at the last settlement; "n of N h" below a value = observed of expected hours
max, min = highest and lowest single hour in 30 days
above neutral = share of observed hours settled above the market's neutral rate in force at that hour
sum = what a long paid over the observed hours, in bps of notional
other venues = the same symbol on another HIP-3 deployer: 30 d APR and its observed hours, listed when it has at least 7
"neutral varies" = the market's neutral rate changed within the 30 days, with its range; each hour is compared with the neutral rate in force when it settled
definitions in methodology
Largest settlements · 30 dby size of the hourly rate, either sign
settled = settlement time, New York (ET) with UTC below; the rate covers the hour before it
session = the US session of that hour
rate = the hourly rate in bps of notional
APR = that one hour annualised, which is why a single hour can read in the thousands of percent
Same name, another deployer30 d APR on io versus xyz
each venue's 30 d APR is over its own observed hours (shown)
neutral = that deployer's carry: base interest 0.01% per 8 h × its funding multiplier, so the same price premium settles at different rates on different deployers; compare the gap above each venue's neutral, not the raw APRs
io − xyz = difference in percentage points, positive when longs pay more on io
listed when the other venue has at least 7 observed hours in 30 days
sorted by the largest difference
Dailycross-market median and share above neutral, per UTC day
median apr
median over markets with at least 12 observed hours that day of each market's daily mean · dashed = neutral · hover a day
share above neutral
share of those markets whose daily mean exceeded their neutral rate · hover a day
UTC days from the first settlement in the archive; today (2026-10-10) is still open and not drawn
a day with too few observed hours has no point and the line bridges it with a dashed segment
raw: /api/io/funding/index.json