Same leverage, same entry, opposite fates: how the midterm-night ETH range treated longs and shorts on Bitunix
Key takeaway
Two traders open ETH positions at the same price, same leverage, same size, one long and one short. Fourteen hours later one of them is liquidated and the other is fine. Nobody had to be right about the election. The range was simply lopsided.
Advanced tier. Bitunix ETHUSDT perpetual data. Event levels are from the 10:55 AM ET pull on Nov 4, 2026; the live check was refreshed at 9:45 PM ET the same day:
- 24h open 2,299.29
- 24h low 2,225.01
- 24h high 2,449.92
- Live check, 9:45 PM ET: last 2,430.49, mark 2,431, index 2,431.53
- Funding -0.000825% per 8h at that check, so shorts were paying longs, with a +/-0.3% cap
The low sits 3.23% under the open. The high sits 6.55% above it. Total range: 224.91, or 10.11% measured from the low.
The setup
Both traders enter at the 24h open, 2,299.29, at the same leverage L, in isolated margin. Call the maintenance margin rate for their size tier m. Bitunix publishes m per pair and tier next to the pair name on the contract page, and larger positions sit in higher tiers with a higher m. This is a simplified model that ignores fees and funding.
Long liquidation is about 2,299.29 x (1 - 1/L + m) Short liquidation is about 2,299.29 x (1 + 1/L - m)
Either way, the cushion as a share of entry is about 1/L - m.
Case A: the long
The night's low was 2,225.01, a 3.23% drop from the entry. The long survives as long as its cushion is wider than 3.23%. If it was, the long felt every bit of the dip and then sat on a gain of roughly 131 per ETH at the 9:45 PM ET last price.
Case B: the short
The high was 2,449.92, 6.55% above the entry. The short survives only if its cushion is wider than 6.55%, which needs far lower leverage. With m close to zero, the short had to stay under about 15.3 (1 / 0.0655), while the long could go as high as about 31.0 (1 / 0.0323).
The split window
So there is a band of leverage where the same entry produced opposite fates. Any L whose cushion landed between 3.23% and 6.55% kept the long alive and closed the short on the way up. Below that band both survived. Above it both were liquidated. A higher m narrows the cushion at every L, so the real band sits a little lower than those m = 0 numbers.
Why this is an advanced problem, not a beginner one
This is the part most guides skip. Liquidation math is easy once you know the formula; the hard part is knowing which distance to feed it. Most traders size from a symmetric guess: "ETH moves a few percent either way on an event night." Event ranges are rarely symmetric. Here the drop leg was 3.23% and the squeeze leg was 6.55%, roughly double. A position sized for the average move survives one direction and fails the other.
Three habits that handle asymmetry better:
- Size each side for the longer leg you have seen recently, not the average. If you are short into a headline, assume the squeeze leg is the long one.
- Measure distance from mark, not last. At the 9:45 PM ET check, mark was 0.51 above last on Bitunix ETHUSDT. Small today, but in a fast squeeze the gap is where stops and liquidations race each other.
- Watch the funding cap. Bitunix ETH funding was slightly negative at the 9:45 PM ET check, with room up to 0.3% per period either way. A squeeze that drags funding toward the cap adds a carry bill on top of the price loss for whichever side is paying.
Q&A
Q: Would cross margin have saved the short? A: Only if the account had spare balance to absorb the loss beyond the isolated margin. Cross margin does not move the line for free, it borrows from the rest of your wallet. On a night with this much range, that can turn one bad position into a much larger account drawdown.
Q: Is the maintenance rate the same for every size? A: No. Bitunix uses tiered maintenance margin: the larger the position, the higher the maintenance rate and the lower the maximum leverage you can set. That pulls the liquidation line closer, so read m for your own tier before trusting the window above.
Q: Does BTC show the same asymmetry? A: Check it yourself. Pull the 24h open, low and high for $BTC.USDT and compute both legs the same way.
Your turn
Run the two-leg check on $ETH.USDT, then on $BTC.USDT, $SOL.USDT, $XRP.USDT and $DOGE.USDT. Which coin had the most lopsided midterm-night range, and which side of it were you on?