Over 107 days of 20-level order-book snapshots, xyz:SMH quoted a median spread of 9.74 bps in US regular hours with $164k resting within 25 bps of mid. A $25k market order paid about 10.66 bps in regular hours and 42.64 bps on weekends, when depth within 25 bps fell to $12k. Taker fee is 0.90 bps with growth mode on; funding averaged +18.75% annualised over the last 30 days.
spread · rth
9.74 bps
median · p90 18.69 bps
depth ±25bps · rth
$164k
median, both sides
$25k slip · rth
10.66 bps
fills 92%
$25k slip · wknd
42.64 bps
fills 54%
taker fee
0.90 bps
fee scale 1.0 · growth mode on
funding apr · 30d
+18.75%
longs pay when positive · 7d +14.95%
vol 24h · 30d avg
$736k
notional
open interest
$2.2M
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
4,896
9.74
18.69
$4k
$35k
$164k
$280k
6.85
10.66
18.43
66%
25%
US extended 4–9:30, 16–20
7,187
9.96
24.38
$3k
$12k
$61k
$135k
8.67
17.02
34.66
38%
1%
Overnight 20–4
4,921
9.97
25.93
$3k
$14k
$60k
$135k
8.28
16.47
34.42
37%
0%
Weekend / holiday
9,143
14.53
35.75
$742
$2k
$12k
$31k
18.10
42.64
80.08
13%
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 SMH, $50k or more→ sign infree · by email · when the data is rebuilt
taker buy · 24h
$624k
30d $9.7M
taker sell · 24h
$607k
30d $11.3M
imbalance · 24h
+1.4%
of $1.2M · 30d -7.5% of $21.0M
trades · 24h
3,555
1 large orders · 30d 67,911
takers · 24h
137
51 makers · 157 addresses
takers · 30d
638
688 addresses incl. makers
top-10 share · 30d
48%
of $21.0M by 638 takers · 12 make half
concentration · 30d
0.039
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: +$1k so far
by session · 30 d
session
taker buy
taker sell
net
imbalance
trades
share of notional
US regular 9:30–16:00
$4.5M
$6.3M
−$1.8M
-16.6%
25,210
51%
US extended 4–9:30, 16–20
$2.9M
$2.8M
+$61k
+1.1%
20,420
27%
Overnight 20–4
$761k
$944k
−$183k
-10.8%
8,699
8%
Weekend / holiday
$1.6M
$1.2M
+$334k
+12.0%
13,582
13%
sessions in New York time as in the liquidity table above · share = the session's taker notional over all four, $21.0M 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 $21.0M of 30-day taker notional (67,911 trades, 638 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/SMH.json · definitions in methodology
alert meSMH funding above or below a level I set→ sign infree · by email · when the data is rebuilt
last settlement
+11.66%
+0.1331 bps/h · 10-01 21:00 ET
next · estimate
+9.86%
+0.1125 bps/h · as of 10-01 22:15 ET
apr · 24h
-1.91%
24 of 24 hours observed
apr · 7d
+14.95%
167 of 168 hours observed
apr · 30d
+18.75%
618 of 720 hours · stdev 35.24%
paid by a long · 30d
132.26 bps
of notional, over 618 observed hours
hours above neutral · 30d
61%
at 17% · below 22% · negative 20%
range · 30d
+180%
highest hour · lowest -172%
daily funding apr
mean of the observed hourly rates per UTC day, annualised · up = longs paid, down = shorts paid · dashed = neutral 5.48% · 109 complete days · hover a day · today (2026-10-02) is still open and not drawn: +8.57% 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 = ±39.72% 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
+25.12%
115
79%
+12.55%
433
54%
US extended 4–9:30, 16–20
+26.44%
184
68%
+20.79%
717
50%
Overnight 20–4
+23.89%
118
64%
+19.13%
459
46%
Weekend / holiday
+5.05%
201
42%
+6.72%
852
43%
mean of the hourly rates settled in each session, annualised · 30d = the 720 hours ending at the last settlement, all = since 2026-06-15 · above neutral = share of those hours settled above 5.48% APR
largest settlements · 30 d
settled · ET
session
rate bps/h
APR
09-20 20:0009-21 00:00 UTC
weekend / holiday
+2.056
+180%
09-27 19:0009-27 23:00 UTC
weekend / holiday
-1.963
-172%
09-13 19:0009-13 23:00 UTC
weekend / holiday
-1.176
-103%
09-13 05:0009-13 09:00 UTC
weekend / holiday
-1.164
-102%
09-21 05:0009-21 09:00 UTC
US extended
+1.15
+101%
09-20 00:0009-20 04:00 UTC
weekend / holiday
-1.119
-97.98%
09-27 05:0009-27 09:00 UTC
weekend / holiday
+1.106
+96.88%
09-27 22:0009-28 02:00 UTC
weekend / holiday
+1.014
+88.81%
09-28 00:0009-28 04:00 UTC
weekend / holiday
+1.01
+88.50%
09-27 06:0009-27 10:00 UTC
weekend / holiday
+0.9914
+86.85%
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/SMH.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