The 76,322 ETH Withdrawal Spike: Compounding Validators Wake Up
Something unusual happened on February 21, 2026. While scanning Ethereum's withdrawal data, a pattern emerged that didn't fit the normal rhythm of the network. 52 validators withdrew over 1,000 ETH each in a single hour — a volume we haven't seen since the Pectra upgrade activated on January 25.
The Discovery
Most Ethereum withdrawals are predictable. Validators with 0x01 credentials receive their consensus rewards — typically 0.01-0.02 ETH every few days. It's a steady heartbeat, background noise in the machine.
But the Pectra upgrade introduced something new: 0x02 compounding withdrawal credentials. These validators don't receive automatic payouts. Instead, rewards compound in-place until the balance exceeds the 2048 ETH maximum effective balance, at which point the excess is withdrawn.
This creates a different pattern. Rather than frequent small withdrawals, compounding validators accumulate rewards until they hit the threshold, then release a flood.
The Data
Querying canonical_beacon_block_withdrawal from the Xatu dataset:
SELECT
toDate(slot_start_date_time) as day,
countIf(withdrawal_amount >= 1e12) as large_withdrawals,
sumIf(withdrawal_amount, withdrawal_amount >= 1e12) / 1e9 as large_eth
FROM canonical_beacon_block_withdrawal
WHERE slot_start_date_time >= now() - INTERVAL 14 DAY
GROUP BY day
ORDER BY day
The results show a clear escalation:
| Date | Large Withdrawals (≥1,000 ETH) | Total ETH |
|---|---|---|
| Feb 10 | 0 | 0 |
| Feb 13 | 31 | 57,993 |
| Feb 14 | 22 | 44,165 |
| Feb 20 | 18 | 36,771 |
| Feb 21 | 52 | 76,322 |
| Feb 22 | 0 | 0 |
February 21 stands out. The 52 large withdrawals that day represent 76,322 ETH — more than double the daily average of the preceding week.
The Hourly Breakdown
Zooming into February 21 by hour reveals the concentration:
SELECT
toHour(slot_start_date_time) as hour,
countIf(withdrawal_amount >= 1e12) as large_withdrawals,
sumIf(withdrawal_amount, withdrawal_amount >= 1e12) / 1e9 as large_eth
FROM canonical_beacon_block_withdrawal
WHERE slot_start_date_time >= '2026-02-21 00:00:00'
AND slot_start_date_time < '2026-02-22 00:00:00'
GROUP BY hour
ORDER BY hour
- 04:00-06:00 UTC: 5 validators, 10,080 ETH
- 12:00 UTC: 42 validators, 56,827 ETH — the spike
- 14:00-19:00 UTC: 4 validators, 8,018 ETH
The 12:00 UTC cluster is remarkable. 42 validators, all withdrawing 1,000+ ETH within the same hour. These aren't random — they're a coordinated cohort.
Who Are These Validators?
Looking at the validator indices from the February 21 12:00 UTC batch:
SELECT
withdrawal_validator_index,
withdrawal_amount / 1e9 as eth_amount,
withdrawal_address
FROM canonical_beacon_block_withdrawal
WHERE slot_start_date_time >= '2026-02-21 12:00:00'
AND slot_start_date_time < '2026-02-21 13:00:00'
AND withdrawal_amount >= 1e12
ORDER BY eth_amount DESC
LIMIT 5
| Validator Index | ETH Withdrawn | Withdrawal Address |
|---|---|---|
| 2109254 | 1,818.03 | 0x096BC969... |
| 2109323 | 1,817.98 | 0xb1241d13... |
| 2110655 | 1,816.95 | 0x2d82F61B... |
| 2110999 | 1,816.74 | 0x16C33c16... |
| 2111060 | 1,816.56 | 0x4698a71c... |
All indices are in the 2.1M range — recently deposited validators. The withdrawal amounts (~1,816 ETH each) suggest these validators were deposited with 0x02 credentials from inception, accumulated rewards for roughly 4 weeks, and hit their first automatic withdrawal when exceeding the 2048 ETH cap.
The Bigger Picture
As of February 24, only 1.16% of validators have upgraded to 0x02 credentials:
SELECT
substring(withdrawal_credentials, 1, 4) as cred_type,
count() as validator_count,
round(100.0 * validator_count / sum(count()) OVER (), 2) as pct
FROM canonical_beacon_validators_withdrawal_credentials
WHERE epoch_start_date_time >= now() - INTERVAL 1 DAY
GROUP BY cred_type
| Credential Type | Count | Percentage |
|---|---|---|
| 0x01 | 4,076,196 | 70.1% |
| 0x00 | 1,671,434 | 28.74% |
| 0x02 | 67,539 | 1.16% |
The vast majority of validators still use 0x01 credentials with automatic reward distribution. But the 0x02 cohort is growing, and their withdrawal pattern is fundamentally different.
Why This Matters
Compounding validators create lumpier withdrawal flows. Instead of a steady stream of small payouts, we get periodic bursts when multiple validators hit the 2048 ETH threshold simultaneously.
February 21 may be a preview. As more validators upgrade to 0x02 credentials — especially existing validators migrating from 0x01 — we could see larger, less predictable withdrawal events.
For staking operators, this changes liquidity planning. For the network, it means withdrawal volume will become more bursty. For analysts, it's a new pattern to track.
The 76,322 ETH withdrawn on February 21 didn't stress the network. But it was a signal. The compounding validators have arrived.
Data from Xatu (Ethereum consensus layer telemetry) covering February 10-24, 2026. Queries executed against the canonical_beacon_block_withdrawal table.