Bitcoin fell to about $75,600 in the 10 p.m. ET hour on Tuesday, November 3, 2026, as U.S. midterm results came in, then reached about $81,130 twelve hours later, a 7.3% low-to-high swing on Coinbase hourly data. CoinGlass data cited by CryptoSlate showed about $378 million of bearish positions liquidated within a single hour of the rebound. For anyone running leveraged automated strategies, the useful lesson is not which side was right. It is that a move of this size can close a position on either side before the configured exit runs. This crypto margin checklist turns one election night into six checks you can walk through, step by step, before the next scheduled event.
This article uses public market reporting as context. It is not a trading signal, a price forecast, or a recommendation to use leverage. Venues are named only because their data or reporting is cited; nothing here implies that Gimmer supports or integrates with them.
What happened on election night
Unchained reported that bitcoin dropped to about $75,600 on Coinbase in the 10 p.m. ET hour on Tuesday and reached $81,130 in the 10 a.m. ET hour on Wednesday. Coinbase’s public hourly candles confirm both extremes: a low of $75,600.64 in the candle that opened at 03:00 UTC on November 4, and a high of $81,130.79 in the candle that opened at 15:00 UTC. CoinDesk’s live coverage tracked the same recovery through the U.S. session, with bitcoin pulling back to just above $80,000 by the afternoon.
The liquidation numbers depend on the window. CryptoSlate cited CoinGlass for about $378 million of bearish positions closed within one hour, with shorts making up about 81% of liquidations; it did not state the hour’s exact start and end. Unchained’s CoinGlass snapshot on Wednesday showed $950 million over 24 hours, made up of $560 million in shorts and $391 million in longs. Bitcoin Magazine reported the same 24-hour total and said $403 million of it was in bitcoin positions, 76% of them shorts.
Those reports describe three different records:
- Price record: what one venue printed, hour by hour. Other venues and data sources can show a different low for the same night.
- Liquidation record: an aggregate estimate of positions closed by venues under their own margin rules, and it changes with the window you choose.
- Account record: what happened to your own position, your collateral, and your orders. No headline contains this one.
The 24-hour split is the detail worth keeping. Shorts took the larger share, but $391 million of longs were closed in the same window. A sharp two-way move does not need to pick a side to hurt leveraged positions.
Why a two-way move catches leveraged strategies
Scheduled events tend to expose four common mistakes in margin automation:
- Right thesis, wrong survival math: a position can be correct about the eventual direction and still be closed during the first leg against it.
- Configured exit treated as the first exit: Gimmer’s Margin guide warns that liquidation can occur before a configured exit, because of exchange rules, rapid moves, gaps, interest, lack of liquidity, or failed orders.
- Leverage read as a setting, not a multiplier: a 7.3% move on 10x notional is about 73% of posted margin before fees and interest. Near 13.7x, that move equals the whole margin, and maintenance requirements would normally end the position earlier.
- Closed candles read as live coverage: strategies that evaluate closed candles see a move after the interval ends. In the Coinbase data, the 15:00 UTC hour alone ranged from $79,190.02 to $81,130.79, about 2.5%, so a one-hour strategy could only respond once that candle had closed.
The multiplication above is arithmetic on one reported swing, not a suggested leverage level. Real outcomes also depend on entry price, venue rules, and fees.
A six-step crypto margin checklist
Write one row per strategy and market. Run it before a scheduled event rather than during one.
| Step | Question to answer | Evidence to keep |
|---|---|---|
| 1. Side | Can this strategy be long, short, or both, and which direction is open now? | Allowed directions, current position, and the rule that closes each side. |
| 2. Leverage | Do the venue and this market support the leverage you set? | Configured leverage, account mode, and the venue’s current market terms. |
| 3. Notional | What is the combined exposure if every market opens at once? | Sizing mode, per-market value, simultaneous positions, and total notional. |
| 4. Buffer | Which comes first on a fast move: your exit or the venue’s liquidation? | Stop distance, maintenance margin and liquidation terms, and reserve collateral. |
| 5. Open drawdown | What did equity do while positions were still open? | Open PnL, drawdown, and worst simultaneous exposure from a backtest window with a two-way move. |
| 6. Finality | Did the position actually close, and at what price? | Final order state, fill price, remaining collateral, and the venue’s account view. |
The step that usually gets skipped is the fourth. A stop distance means little until it has been compared with where the venue would act first.
Use Gimmer’s Margin execution to run the checklist
Gimmer’s Margin execution guide opens with a direct instruction: “Use Margin execution only after modeling leveraged exposure.” The feature is labeled high risk, and it is an execution mode rather than a separate decision engine. The steps below follow the guide in order.
Step 1: Choose the engine, then select Margin
Start from a current decision engine (AI Decision, Indicators, Code Editor, Swing, or Portfolio) and select Margin as execution. Confirm which of the four position directions your rules can produce: Long, Short, Close long, and Close short. A strategy that can open a short needs an equally clear rule for closing it. Connect an exchange account that supports Margin only when you need exchange-connected simulation or live execution, and confirm the account mode and trading permission. Never enable withdrawals on the key.
Step 2: Set leverage and work out the notional
The guide asks you to set leverage that both the venue and the market support, and to understand the resulting notional exposure. Then set percentage or fixed sizing and calculate combined exposure if several markets can open at once. The earlier margin risk-control checklist covers that review before any backtest runs.
Step 3: Configure exits, and compare them with liquidation
Configure the stop loss, trailing behavior, position limits, and engine-specific rules. The risk controls guide explains that a stop requests an exit and that the fill can differ because of gaps, slippage, and liquidity. Our stop-loss checklist covers that side in detail. For Margin, add one more comparison: check the venue’s maintenance margin and liquidation terms, and record whether your exit sits safely before them.
Step 4: Backtest a two-way window, then simulate
Run a backtest over a date range that includes a sharp reversal, and review long and short trades separately. The Margin guide says drawdown should reflect current equity, not only realized closed trades, so read Open PnL next to Realized PnL and look for worst simultaneous exposure. Compare the window with calmer periods using the market-regime condition map. Backtests are models: they cannot reproduce every liquidation, partial fill, or outage, so follow with simulation and watch Runtime activity, positions, and orders.
Step 5: Work through the pre-live list
Live use is a separate, deliberate choice. Before it, the guide asks you to confirm that every market supports the requested side and leverage, check borrowed-asset availability, interest, maintenance margin, and liquidation terms, risk only an amount you can afford to lose, keep reserve collateral, monitor the first open and close through final order states, and keep an independent exchange login available for emergency intervention.
A short example: one strategy, two legs
Consider a synthetic Indicators strategy running in simulation. It can go long or short on one market, uses 10x leverage, and has fixed sizing. All numbers are invented.
- The operator notes a scheduled event and fills in the six-step table. Posted margin is 100 units, so notional is 1,000 units.
- They model a move of 7% against the position. That is about 70 units of the 100 posted, before fees and interest.
- They compare it with the stop. The stop sits 4% from entry, which looks fine on paper, but the assumed venue terms would close the position at a similar distance once maintenance margin is included. The buffer is too thin to rely on.
- They replay a two-way backtest window. On the long side the stop triggers on the first leg, and the modeled fill is worse than the stop level. The later rebound does not help, because the position is already closed.
- They reduce modeled leverage in a copy of the strategy, rerun the same window, and record Open PnL, drawdown, and worst exposure for both versions.
- They keep both versions in simulation and write down a finality rule: an exit counts only after the final order state and the remaining collateral are confirmed.
The result is a documented comparison of two configurations, not evidence that either one will work. The example is not a recommendation to trade, to use leverage, or to use any specific setting.
What this margin checklist cannot do
- It cannot predict election reactions, the bitcoin price, or the next liquidation wave.
- It cannot stop a venue from liquidating a position under its own rules, and it cannot guarantee that a stop or trailing exit will execute.
- It cannot make aggregated liquidation figures exact. Windows differ, and the totals are estimates built from what venues report.
- It cannot make a backtest or simulation a forecast; both describe how configured rules behaved on the data provided.
- It cannot replace reading your venue’s current margin terms, which can change.
Trading digital assets involves substantial risk. Leverage amplifies gains, losses, fees, and liquidation risk. Backtests and simulations do not guarantee future results, and risk controls can fail to execute.
Frequently asked questions
Does a stop loss protect a margin position from liquidation?
Not reliably. Gimmer’s Margin guide says liquidation can occur before a configured exit, for reasons that include rapid moves, gaps, and failed orders. Compare your exit with the venue’s terms, and do not treat either as a guarantee.
Why did both longs and shorts get liquidated?
The market moved sharply in both directions within one day. In the CoinGlass snapshot Unchained reported, shorts were the larger share of the 24-hour total, but longs were closed too. Positions on either side can be hit if the move against them is large enough for their leverage.
Should I reduce leverage before every scheduled event?
That is your decision, and this article cannot make it for you. A practical step is to measure first: backtest a window with a two-way move, read open drawdown, and check how much buffer the current setting leaves.
Measure the buffer before the next event
Election night showed how quickly one market can punish both directions. The practical response is not to guess the next reaction. It is to know, for each leveraged strategy, which sides it can hold, what its total notional is, whether its exit comes before the venue’s liquidation, what open drawdown looks like on a two-way window, and how you will confirm that a position really closed.
Next step: open Gimmer’s Margin execution guide, fill in the six-step table for each leveraged strategy, and replay a backtest window with a sharp reversal before you change any live setting.
— The Gimmer Team