Over 107 days of 20-level order-book snapshots, xyz:ARM quoted a median spread of 3.56 bps in US regular hours with $348k resting within 25 bps of mid. A $25k market order paid about 5.33 bps in regular hours and 7.83 bps on weekends, when depth within 25 bps fell to $107k. Taker fee is 0.90 bps with growth mode on; funding averaged +8.30% annualised over the last 30 days.
spread · rth
3.56 bps
median · p90 8.37 bps
depth ±25bps · rth
$348k
median, both sides
$25k slip · rth
5.33 bps
fills 100%
$25k slip · wknd
7.83 bps
fills 79%
taker fee
0.90 bps
fee scale 1.0 · growth mode on
funding apr · 30d
+8.30%
longs pay when positive · 7d +13.29%
vol 24h · 30d avg
$2.9M
notional
open interest
$6.3M
both sides · 2026-10-02
by sessionNew York time · medians over 107 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
4,898
3.56
8.37
$2k
$135k
$348k
$363k
3.36
5.33
8.81
69%
8%
US extended 4–9:30, 16–20
7,228
2.93
6.22
$2k
$62k
$161k
$173k
3.20
6.08
11.73
21%
4%
Overnight 20–4
4,971
3.17
7.37
$1k
$52k
$154k
$166k
3.73
6.97
13.85
19%
2%
Weekend / holiday
9,148
3.03
7.16
$1k
$38k
$107k
$122k
3.65
7.83
17.96
6%
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 ARM, $50k or more→ sign infree · by email · when the data is rebuilt
taker buy · 24h
$1.5M
30d $40.8M
taker sell · 24h
$1.3M
30d $44.6M
imbalance · 24h
+4.2%
of $2.8M · 30d -4.4% of $85.4M
trades · 24h
3,284
4 large orders · 30d 142,231
takers · 24h
135
74 makers · 169 addresses
takers · 30d
1,313
1,395 addresses incl. makers
top-10 share · 30d
44%
of $85.4M by 1,313 takers · 14 make half
concentration · 30d
0.037
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: +$23k so far
by session · 30 d
session
taker buy
taker sell
net
imbalance
trades
share of notional
US regular 9:30–16:00
$16.1M
$18.2M
−$2.1M
-6.2%
46,789
40%
US extended 4–9:30, 16–20
$13.1M
$13.7M
−$522k
-1.9%
47,312
31%
Overnight 20–4
$4.8M
$5.1M
−$271k
-2.7%
18,412
12%
Weekend / holiday
$6.8M
$7.6M
−$814k
-5.7%
29,718
17%
sessions in New York time as in the liquidity table above · share = the session's taker notional over all four, $85.4M 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 $85.4M of 30-day taker notional (142,231 trades, 1,313 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/ARM.json · definitions in methodology
alert meARM funding above or below a level I set→ sign infree · by email · when the data is rebuilt
last settlement
+20.37%
+0.2325 bps/h · 10-01 21:00 ET
next · estimate
+33.26%
+0.3797 bps/h · as of 10-01 22:15 ET
apr · 24h
+14.17%
18 of 24 hours observed
apr · 7d
+13.29%
162 of 168 hours observed
apr · 30d
+8.30%
529 of 720 hours · stdev 24.78%
paid by a long · 30d
50.13 bps
of notional, over 529 observed hours
hours above neutral · 30d
52%
at 38% · below 10% · negative 8%
range · 30d
+136%
highest hour · lowest -232%
daily funding apr
mean of the observed hourly rates per UTC day, annualised · up = longs paid, down = shorts paid · dashed = neutral 5.48% · 110 complete days · hover a day · today (2026-10-02) is still open and not drawn: +16.54% 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 = ±19.43% 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
+10.59%
88
53%
+10.62%
396
51%
US extended 4–9:30, 16–20
+13.14%
157
62%
+12.65%
673
53%
Overnight 20–4
+13.83%
96
73%
+10.06%
437
50%
Weekend / holiday
+0.37%
188
33%
+5.52%
830
34%
mean of the hourly rates settled in each session, annualised · 30d = the 720 hours ending at the last settlement, all = since 2026-06-14 · above neutral = share of those hours settled above 5.48% APR
largest settlements · 30 d
settled · ET
session
rate bps/h
APR
09-13 20:0009-14 00:00 UTC
weekend / holiday
-2.654
-232%
09-13 19:0009-13 23:00 UTC
weekend / holiday
-2.281
-200%
09-13 04:0009-13 08:00 UTC
weekend / holiday
-2.211
-194%
09-13 05:0009-13 09:00 UTC
weekend / holiday
-1.864
-163%
09-20 20:0009-21 00:00 UTC
weekend / holiday
+1.554
+136%
09-20 00:0009-20 04:00 UTC
weekend / holiday
-1.091
-95.58%
09-27 19:0009-27 23:00 UTC
weekend / holiday
-1.004
-87.99%
09-06 11:0009-06 15:00 UTC
weekend / holiday
-0.9564
-83.78%
09-07 12:0009-07 16:00 UTC
weekend / holiday
-0.9113
-79.83%
09-07 14:0009-07 18:00 UTC
weekend / holiday
+0.9063
+79.39%
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 · rebuilt nightly and at 09:05 ET, not live · raw: /api/xyz/funding/ARM.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 107 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