Every ARB volume spike this site detects is followed up: the price 5 minutes, 15 minutes and 1 hour later is compared with the price at the spike. The tables count how often price kept going in the spike's direction (the hit rate) and how far, by exchange and by buy or sell side, with every figure's sample size. Past outcomes, not a forecast.
Across 84 buy- and sell-side ARB volume spikes with a 1-hour outcome (27 Sep 2026), price an hour later had moved in the spike's direction 48.8% of the time; the median follow-through was -0.03% (middle half -0.82% to +0.83%). After 15 minutes the hit rate was 57.3% over 89 spikes (median +0.10%). Buy-side spikes: 43.8% hit after 1 hour (n = 32); sell-side: 51.9% (n = 52).
Last 7 days119 ARB volume spikes27 Sep 2026Descriptive stats, not a signalAs of 21:57 UTC
Hit rate +5m
45.2%
median -0.04%n = 93
Hit rate +15m
57.3%
median +0.10%n = 89
Hit rate +1h
48.8%
median -0.03%n = 84
ARB volume spikes 7d
119
23 mixed (no side)27 Sep 2026
All ARB spikes · Last 7 days
119 volume spikes, 27 Sep 2026 (UTC)
After
Spikes measured
Hit rate
Lower quartile
Median
Upper quartile
Mean
Date range (UTC)
5 min
93
45.2%
-0.18%
-0.04%
+0.33%
+0.02%
27 Sep 2026
15 min
89
57.3%
-0.33%
+0.10%
+0.55%
+0.11%
27 Sep 2026
1 hour
84
48.8%
-0.82%
-0.03%
+0.83%
-0.04%
27 Sep 2026
ARB by exchange
27 Sep 2026 (UTC)
+5m
+15m
+1h
Hit rate
Median
Hit rate
Median
Hit rate
Median
Binance
44.9%n = 49
-0.05%-0.14% … +0.33%
57.4%n = 47
+0.09%-0.34% … +0.62%
44.2%n = 43
-0.23%-0.83% … +0.75%
Bybit
51.6%n = 31
+0.02%-0.11% … +0.34%
56.7%n = 30
+0.09%-0.32% … +0.45%
48.3%n = 29
-0.06%-0.98% … +0.52%
Hyperliquid
30.8%n = 13
-0.14%-0.43% … +0.20%
58.3%n = 12
+0.33%-0.46% … +0.59%
66.7%n = 12
+0.69%-0.31% … +1.28%
ARB by direction
27 Sep 2026 (UTC)
+5m
+15m
+1h
Hit rate
Median
Hit rate
Median
Hit rate
Median
Buy-side spikesa hit is price up
54.5%n = 33
+0.09%-0.12% … +0.42%
63.6%n = 33
+0.26%-0.50% … +0.75%
43.8%n = 32
-0.29%-1.36% … +1.01%
Sell-side spikesa hit is price down
40.0%n = 60
-0.08%-0.21% … +0.27%
53.6%n = 56
+0.04%-0.32% … +0.47%
51.9%n = 52
+0.02%-0.68% … +0.77%
Last 20 ARB volume spikes
All history, newest first · price change after each spike
+5m / +15m / +1h: the raw price change after each spike (not signed by side), — until that outcome is recorded. Each day links to that day's spike archive page. Live ARB price, funding and open interest: ARB overview.
Method
Counting. The same volume spikes as the live spike scanner, for markets whose base asset is ARB. For each buy- or sell-side spike, the follow-through at +5m, +15m and +1h is the change from the spike minute's close to the close that many minutes later, with the sign flipped for sell-side spikes; a hit is a follow-through above zero. Mixed spikes (23 in this window) are left out of hit rates, medians and quartiles. HIP-3 markets (Hyperliquid builder perps such as stock tickers) are shown apart from Hyperliquid's own perps, since a HIP-3 ticker can share a name with an unrelated coin.
Reading it. Every figure shows its sample size (n) and each table its date range; cells with fewer than 30 spikes are faded. This page is only indexed once ARB has 50 or more spikes with a 1-hour outcome in the last 30 days. These are descriptive statistics of past spikes, before fees, slippage and funding, from a price you could not have traded at. They are not a trading signal and make no claim of an edge. Not financial advice. Compare with spike outcomes for all coins.
ARB spike outcome FAQ
How is the hit rate calculated?
For each buy- or sell-side ARB volume spike we compare the close of the spike minute with the close 5, 15 and 60 minutes later. A buy-side spike is a hit when price went up, a sell-side spike when it went down (exactly unchanged is a miss). Mixed spikes, with no clear taker side, are left out of hit rates, medians and quartiles. The method is the same as on spike outcomes for all coins.
Can I trade on these numbers?
No claim is made that you can. These describe what happened after past spikes, before fees, slippage and funding, measured from a price you could not have traded at. Small samples swing a lot. Nothing here is financial advice.