Ethereum Up/Down order book microstructure on Polymarket

• 715 ETH Up/Down markets in the archive, led by 491 five-minute markets
• Average market spread of 0.0139 (median 0.0100), second only to Bitcoin
• Average top-of-book liquidity of roughly $27k per market over three days
• 2.94% of the last 24 hours of ETH book snapshots were crossed - the highest rate tracked
• All 715 markets settle on chainlink_twap reference prices

As of 2026-10-17 the archive holds 715 Ethereum Up/Down markets: 491 on the 5-minute horizon, 209 at 15 minutes and 15 at 4 hours. That is the second-largest crypto series after Bitcoin, and the 69% share of 5-minute markets keeps the ETH sample dense.

327 markets have resolved with a recorded winner, split 172 up to 155 down (52.6% up). That is mildly up-biased over the sample, roughly the same lean Bitcoin shows, which is consistent with both assets drifting during a partially up-trending measurement window rather than any grading asymmetry.

Ethereum's average spread is 0.0139, essentially a one-tick book, but it is looser than Bitcoin's 0.0115 in the same period - the two assets do not quote identically even though both are liquid. The median tells the sharper story: at 0.0100 both assets sit one tick wide most of the time, and the mean gap comes from the occasional wide-moment tail, which is slightly heavier for ETH.

Depth is roughly half of Bitcoin. The trailing three-day average top-of-book liquidity is about $26,844 per ETH market against $57,147 for BTC. For a 5-minute contract that is still ample for retail-to-mid-size orders, but a size that would be a rounding error on BTC can start moving the ETH touch.

Compared with the rest of the basket ETH is comfortably in the top tier: 0.0139 versus XRP and Solana near 0.026, Dogecoin 0.036, BNB 0.042 and HYPE 0.10.

import requests

BASE = "https://api.polyorderbooks.com/v1"
HEADERS = {"X-API-Key": "your-key"}

slug = "eth-updown-5m-1787486400"
r = requests.get(
    f"{BASE}/markets/{slug}/metrics",
    headers=HEADERS,
    params={"start_ts": 1787486400, "end_ts": 1787486700, "resolution": "1s"},
)
r.raise_for_status()
for pt in r.json()["data"][:5]:
    print(pt["t"], "spread=", pt["spread"], "liquidity=", pt["liquidity"])

ETH has the highest crossed-book rate in the archive: 20,705 of 703,643 snapshots in the last 24 hours, or 2.94%. Ethereum re-prices at least as aggressively as Bitcoin even though it is a little less deep, which is exactly the combination that produces crosses - thin relative to how fast the mid moves.

That combination is worth flagging for execution modeling. The ETH touch is tight, but it is also a moving target, so a naive fill that trusts the last snapshot price will be optimistic more often on ETH than on a similar BTC trade.

The archive keeps the raw crossed flag on every snapshot, so a study can measure the cross frequency for any individual market rather than relying on an average. Over a single ETH market the rate varies with the volatility of that 5-minute window.

Every ETH market resolves on a chainlink_twap reference price just like Bitcoin. The reference feed recorded 67,109 ETH observations on 2026-10-17 from 07:38 UTC onward, covering both the spot print and the native 60-second TWAP used for grading.

That coverage lets you reconstruct the settlement decision after the fact. Join the reference rows at the market's close time to the final book snapshot and you can verify the graded direction against the TWAP the exchange consumed, not just the last trade price.

The Starter plan is free, includes order books, prices, and metrics at 250ms resolution, with 3 days of history, 60 requests/min, 1,000/day, 1 free AI backtest, and 3 strategy backtests. No credit card required.

Paid data windows from $19/mo extend history to 30–120 days at higher throughput. Backtest AI add-on is +$19/mo or standalone at $29/mo.

• Polymarket ETH data
• Crypto microstructure data
• Order book API guide