Over 44 days of 20-level order-book snapshots, xyz:NET quoted a median spread of 10.92 bps in US regular hours with $18k resting within 25 bps of mid. A $25k market order paid about 30.42 bps in regular hours and 103.17 bps on weekends, when depth within 25 bps fell to $2k. Taker fee is 0.90 bps with growth mode on; funding averaged +12.08% annualised over the last 30 days.
spread · rth
10.92 bps
median · p90 53.54 bps
depth ±25bps · rth
$18k
median, both sides
$25k slip · rth
30.42 bps
fills 50%
$25k slip · wknd
103.17 bps
fills 39%
taker fee
0.90 bps
fee scale 1.0 · growth mode on
funding apr · 30d
+12.08%
longs pay when positive · 7d -4.57%
vol 24h · 30d avg
$181k
notional
open interest
$556k
both sides · 2026-10-02
by sessionNew York time · medians over 44 days
session
snaps
spread
p90
top
depth ±10
depth ±25
depth ±50
$5k
$25k
$100k
$100k fills
$250k fills
US regular 9:30–16:00
1,747
10.92
53.54
$230
$598
$18k
$98k
16.31
30.42
38.48
20%
0%
US extended 4–9:30, 16–20
2,628
21.64
59.83
$462
$0
$6k
$19k
30.32
57.37
65.94
11%
0%
Overnight 20–4
1,810
25.30
66.23
$500
$0
$3k
$14k
35.98
74.60
102.09
9%
0%
Weekend / holiday
3,041
28.72
53.73
$494
$0
$2k
$11k
39.83
103.17
300.39
16%
0%
bps unless stated · RTH 9:30–16:00 · extended 4:00–9:30 and 16:00–20:00 · overnight 20:00–4:00 on weekdays · weekend runs Friday 20:00 to Monday 4:00 and includes NYSE holidays · slip = mean of the buy and sell side of a market order walking the visible 20 levels · fills = share of snapshots the size fits inside those levels · "—" = not fillable in the median snapshot
trade flowtaker side · 24 h to 2026-10-02 00:41 UTC · 30 d = last 30 complete UTC days · all markets
alert melarge orders in NET, $50k or more→ sign infree · by email · when the data is rebuilt
taker buy · 24h
$84k
30d $2.9M
taker sell · 24h
$32k
30d $2.6M
imbalance · 24h
+44.8%
of $117k · 30d +5.1% of $5.4M
trades · 24h
971
0 large orders · 30d 39,388
takers · 24h
46
38 makers · 68 addresses
takers · 30d
560
602 addresses incl. makers
top-10 share · 30d
50%
of $5.4M by 560 takers · 10 make half
concentration · 30d
0.040
Herfindahl over takers · 1 = one taker
daily net taker flow
USD per UTC day · taker buy − taker sell · up = net buying, down = net selling · 108 complete days · hover a day for its value · today (2026-10-02) is still open and not drawn: +$99 so far
by session · 30 d
session
taker buy
taker sell
net
imbalance
trades
share of notional
US regular 9:30–16:00
$1.9M
$1.3M
+$554k
+17.1%
24,601
60%
US extended 4–9:30, 16–20
$495k
$679k
−$183k
-15.6%
7,710
22%
Overnight 20–4
$210k
$172k
+$39k
+10.1%
3,375
7%
Weekend / holiday
$259k
$388k
−$129k
-20.0%
3,702
12%
sessions in New York time as in the liquidity table above · share = the session's taker notional over all four, $5.4M in 30 days
hour of week · taker notional
New York time · 30 days summed per hour · brighter = more volume · log scale between the 5th and 95th percentile · 168 of 168 hours with trades · hover a cell for USD and trades
large orders: fills grouped by transaction hash, taker and side, $50k or more per order; fills without a hash are not grouped · share = the address's taker notional over all $5.4M of 30-day taker notional (39,388 trades, 560 takers) · addresses are public on-chain data and link to the Hyperliquid explorer; an address can be a subaccount, vault, market maker or bot and is not labelled · taker notional is volume, not a change in open interest · rebuilt nightly and at 09:05 ET, not a live tape · raw: /api/xyz/flow/NET.json · definitions in methodology
alert meNET funding above or below a level I set→ sign infree · by email · when the data is rebuilt
last settlement
-3.19%
-0.03637 bps/h · 10-01 21:00 ET
next · estimate
-160%
-1.822 bps/h · as of 10-01 22:15 ET
apr · 24h
+41.56%
11 of 24 hours observed
apr · 7d
-4.57%
145 of 168 hours observed
apr · 30d
+12.08%
636 of 720 hours · stdev 62.03%
paid by a long · 30d
87.71 bps
of notional, over 636 observed hours
hours above neutral · 30d
46%
at 21% · below 33% · negative 31%
range · 30d
+230%
highest hour · lowest -357%
daily funding apr
mean of the observed hourly rates per UTC day, annualised · up = longs paid, down = shorts paid · dashed = neutral 5.48% · 45 complete days · hover a day · today (2026-10-02) is still open and not drawn: -30.63% over 2 h so far
hour of week · mean funding apr
New York time · whole archive · the hour a 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 = ±73.89% APR or more · empty = no settlements
funding by session
session
apr · 30d
hours · 30d
above neutral · 30d
apr · all
hours · all
above neutral · all
US regular 9:30–16:00
+15.96%
107
42%
+19.78%
152
49%
US extended 4–9:30, 16–20
+13.88%
183
42%
+36.64%
259
51%
Overnight 20–4
+12.98%
123
51%
+27.63%
176
55%
Weekend / holiday
+8.25%
223
48%
+17.53%
314
54%
mean of the hourly rates settled in each session, annualised · 30d = the 720 hours ending at the last settlement, all = since 2026-08-18 · above neutral = share of those hours settled above 5.48% APR
largest settlements · 30 d
settled · ET
session
rate bps/h
APR
09-27 19:0009-27 23:00 UTC
weekend / holiday
-4.075
-357%
09-13 11:0009-13 15:00 UTC
weekend / holiday
+2.626
+230%
09-18 07:0009-18 11:00 UTC
US extended
+2.417
+212%
09-08 02:0009-08 06:00 UTC
overnight
+2.358
+207%
09-10 02:0009-10 06:00 UTC
overnight
-2.296
-201%
09-23 21:0009-24 01:00 UTC
overnight
-2.201
-193%
09-13 06:0009-13 10:00 UTC
weekend / holiday
-2.193
-192%
09-13 07:0009-13 11:00 UTC
weekend / holiday
-2.128
-186%
09-20 20:0009-21 00:00 UTC
weekend / holiday
+2.048
+179%
09-30 05:0009-30 09:00 UTC
US extended
+2.012
+176%
same name on parapara:NET · neutral 6.57% APR
para apr · 7d
-28.74%
168 hours observed · xyz -4.57%
para apr · 30d
-11.28%
566 hours observed · xyz +12.08%
para last
+55.80%
+0.637 bps/h
neutral rates
6.57% vs 5.48%
para vs xyz · multipliers differ
daily apr · xyz:NET vs para:NET
last 30 UTC days · each day's mean of its observed hours · a missing day breaks the line · hover a day
xyz:NETpara:NET
positive = longs pay shorts · the neutral rate is the deployer's carry (base interest 0.01% per 8 h × funding multiplier 0.5), not crowd positioning · every mean is over observed hours; collector gaps are not filled · other deployers use other multipliers, so their APRs are not like for like · rebuilt nightly and at 09:05 ET, not live · raw: /api/xyz/funding/NET.json · definitions in methodology
slippage curve
median bps over mid vs order size · a line ends where the size stops filling in the median snapshot
hour of week
New York time · median per cell over 44 days · brighter = better · the breathing cell is now
weekly · spread
median bps, all sessions
weekly · depth ±25 bps
median USD both sides, all sessions
cost calculator
per round trip = 2 × median slippage + 2 × taker fee · funding uses the 30-day mean hourly rate, annualised, charged for a position held all day · slippage between tested sizes is interpolated on a log scale · ignores maker rebates, impact that persists after your fill, and fee tiers
history4 market changes to xyz:NET observed in the changelog · newest first
2026-08-18 20:46listedannounced→tradingtrading enabled, announced 2026-08-14 · OI cap $25M · 10× max · growth mode on · marginTableId 10 · szDecimals 2 · fundingMultiplier 0.5
2026-08-17 15:46growth modedisabled→enabledgrowthMode observed changing from disabled to enabled
2026-08-17 13:46OI cap—→$25.0MoiCap observed changing from unknown to $25M
2026-08-14 14:56announced—→announcedmarket announced, trading not enabled
observed = first collector snapshot showing the new value, UTC; the change happened between the previous snapshot and that one · announced = the market appeared with trading disabled; listed = trading enabled; trading halt = off and back within 48 h, before → after = mark either side · full changelog · atom feed