Over 107 days of 20-level order-book snapshots, xyz:GME quoted a median spread of 6.07 bps in US regular hours with $164k resting within 25 bps of mid. A $25k market order paid about 5.29 bps in regular hours and 33.12 bps on weekends, when depth within 25 bps fell to $11k. Taker fee is 0.90 bps with growth mode on; funding averaged +5.64% annualised over the last 30 days.
spread · rth
6.07 bps
median · p90 16.82 bps
depth ±25bps · rth
$164k
median, both sides
$25k slip · rth
5.29 bps
fills 99%
$25k slip · wknd
33.12 bps
fills 27%
taker fee
0.90 bps
fee scale 1.0 · growth mode on
funding apr · 30d
+5.64%
longs pay when positive · 7d +1.90%
vol 24h · 30d avg
$1.4M
notional
open interest
$2.6M
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
4,888
6.07
16.82
$6k
$76k
$164k
$310k
4.35
5.29
14.32
67%
19%
US extended 4–9:30, 16–20
7,195
8.32
18.92
$2k
$14k
$55k
$126k
7.83
15.43
33.92
22%
0%
Overnight 20–4
4,937
7.84
17.52
$2k
$13k
$45k
$92k
7.94
17.06
37.06
13%
0%
Weekend / holiday
9,155
15.06
33.37
$986
$1k
$11k
$25k
19.19
33.12
41.36
2%
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 GME, $50k or more→ sign infree · by email · when the data is rebuilt
taker buy · 24h
$1.7M
30d $18.6M
taker sell · 24h
$1.7M
30d $18.5M
imbalance · 24h
-1.7%
of $3.4M · 30d +0.2% of $37.1M
trades · 24h
4,616
8 large orders · 30d 75,637
takers · 24h
135
60 makers · 153 addresses
takers · 30d
1,248
1,313 addresses incl. makers
top-10 share · 30d
46%
of $37.1M by 1,248 takers · 13 make half
concentration · 30d
0.038
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: +$3k so far
by session · 30 d
session
taker buy
taker sell
net
imbalance
trades
share of notional
US regular 9:30–16:00
$9.1M
$8.8M
+$379k
+2.1%
27,782
48%
US extended 4–9:30, 16–20
$4.9M
$5.2M
−$274k
-2.7%
23,271
27%
Overnight 20–4
$1.8M
$1.7M
+$76k
+2.1%
11,470
10%
Weekend / holiday
$2.8M
$2.9M
−$113k
-2.0%
13,114
15%
sessions in New York time as in the liquidity table above · share = the session's taker notional over all four, $37.1M 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 $37.1M of 30-day taker notional (75,637 trades, 1,248 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/GME.json · definitions in methodology
alert meGME funding above or below a level I set→ sign infree · by email · when the data is rebuilt
last settlement
+17.30%
+0.1975 bps/h · 10-01 21:00 ET
next · estimate
+119%
+1.353 bps/h · as of 10-01 22:15 ET
apr · 24h
+5.90%
24 of 24 hours observed
apr · 7d
+1.90%
168 of 168 hours observed
apr · 30d
+5.64%
608 of 720 hours · stdev 29.78%
paid by a long · 30d
39.17 bps
of notional, over 608 observed hours
hours above neutral · 30d
33%
at 44% · below 24% · negative 20%
range · 30d
+308%
highest hour · lowest -192%
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: +13.34% 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 = ±31.83% 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
+7.88%
100
37%
+13.76%
424
56%
US extended 4–9:30, 16–20
+10.02%
181
39%
+15.79%
722
48%
Overnight 20–4
+6.77%
118
31%
+12.99%
466
42%
Weekend / holiday
+0.15%
209
26%
+9.51%
878
40%
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
09-21 00:0009-21 04:00 UTC
weekend / holiday
+3.517
+308%
09-20 21:0009-21 01:00 UTC
weekend / holiday
+2.453
+215%
09-28 04:0009-28 08:00 UTC
weekend / holiday
-2.197
-192%
09-22 00:0009-22 04:00 UTC
overnight
+2.157
+189%
09-27 21:0009-28 01:00 UTC
weekend / holiday
-1.398
-122%
09-13 21:0009-14 01:00 UTC
weekend / holiday
+1.302
+114%
09-06 12:0009-06 16:00 UTC
weekend / holiday
-1.272
-111%
09-26 12:0009-26 16:00 UTC
weekend / holiday
+1.157
+101%
09-20 22:0009-21 02:00 UTC
weekend / holiday
+1.127
+98.75%
09-05 07:0009-05 11:00 UTC
weekend / holiday
-1.075
-94.19%
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/GME.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