Over 107 days of 20-level order-book snapshots, xyz:AMD quoted a median spread of 1.71 bps in US regular hours with $334k resting within 25 bps of mid. A $25k market order paid about 2.19 bps in regular hours and 3.48 bps on weekends, when depth within 25 bps fell to $116k. Taker fee is 0.90 bps with growth mode on; funding averaged +5.28% annualised over the last 30 days.
spread · rth
1.71 bps
median · p90 4.17 bps
depth ±25bps · rth
$334k
median, both sides
$25k slip · rth
2.19 bps
fills 100%
$25k slip · wknd
3.48 bps
fills 93%
taker fee
0.90 bps
fee scale 1.0 · growth mode on
funding apr · 30d
+5.28%
longs pay when positive · 7d +4.83%
vol 24h · 30d avg
$18.5M
notional
open interest
$15.4M
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
4,886
1.71
4.17
$6k
$327k
$334k
$334k
1.30
2.19
3.63
70%
3%
US extended 4–9:30, 16–20
7,218
1.65
4.23
$3k
$187k
$192k
$192k
1.52
2.86
4.91
24%
0%
Overnight 20–4
4,960
1.70
4.44
$3k
$171k
$176k
$176k
1.60
3.06
4.90
18%
0%
Weekend / holiday
9,128
1.68
4.04
$2k
$112k
$116k
$116k
1.77
3.48
5.27
3%
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 AMD, $50k or more→ sign infree · by email · when the data is rebuilt
taker buy · 24h
$10.3M
30d $269.8M
taker sell · 24h
$9.9M
30d $268.1M
imbalance · 24h
+1.8%
of $20.2M · 30d +0.3% of $537.8M
trades · 24h
20,709
38 large orders · 30d 491,703
takers · 24h
403
146 makers · 455 addresses
takers · 30d
4,612
4,796 addresses incl. makers
top-10 share · 30d
55%
of $537.8M by 4,612 takers · 8 make half
concentration · 30d
0.055
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: −$46k so far
by session · 30 d
session
taker buy
taker sell
net
imbalance
trades
share of notional
US regular 9:30–16:00
$161.2M
$168.4M
−$7.1M
-2.2%
189,792
61%
US extended 4–9:30, 16–20
$49.7M
$42.1M
+$7.5M
+8.2%
116,428
17%
Overnight 20–4
$25.8M
$25.7M
+$114k
+0.2%
76,827
10%
Weekend / holiday
$33.1M
$31.9M
+$1.2M
+1.8%
108,656
12%
sessions in New York time as in the liquidity table above · share = the session's taker notional over all four, $537.8M 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 $537.8M of 30-day taker notional (491,703 trades, 4,612 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/AMD.json · definitions in methodology
alert meAMD funding above or below a level I set→ sign infree · by email · when the data is rebuilt
last settlement
+5.48%
+0.0625 bps/h · 10-01 21:00 ET
next · estimate
+12.01%
+0.1371 bps/h · as of 10-01 22:15 ET
apr · 24h
+6.30%
21 of 24 hours observed
apr · 7d
+4.83%
165 of 168 hours observed
apr · 30d
+5.28%
549 of 720 hours · stdev 21.58%
paid by a long · 30d
33.08 bps
of notional, over 549 observed hours
hours above neutral · 30d
26%
at 63% · below 11% · negative 9%
range · 30d
+165%
highest hour · lowest -157%
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: +8.47% 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 = ±16.86% 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
+5.01%
97
14%
+5.81%
409
23%
US extended 4–9:30, 16–20
+6.94%
160
33%
+7.88%
678
33%
Overnight 20–4
+7.87%
97
38%
+9.16%
439
33%
Weekend / holiday
+2.77%
195
21%
+6.92%
839
30%
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
09-20 19:0009-20 23:00 UTC
weekend / holiday
+1.881
+165%
09-27 19:0009-27 23:00 UTC
weekend / holiday
-1.791
-157%
09-13 19:0009-13 23:00 UTC
weekend / holiday
-1.772
-155%
09-20 20:0009-21 00:00 UTC
weekend / holiday
+1.763
+154%
09-26 15:0009-26 19:00 UTC
weekend / holiday
-1.268
-111%
09-07 19:0009-07 23:00 UTC
weekend / holiday
-1.131
-99.10%
09-13 05:0009-13 09:00 UTC
weekend / holiday
-1.11
-97.22%
09-27 04:0009-27 08:00 UTC
weekend / holiday
+0.9578
+83.90%
09-13 03:0009-13 07:00 UTC
weekend / holiday
-0.9318
-81.62%
09-13 20:0009-14 00:00 UTC
weekend / holiday
-0.8817
-77.24%
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/AMD.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