There is a particular kind of loss that stings more than the others: the one where you were right, eventually, but the exchange closed your position before "eventually" arrived. That is liquidation. It is not a punishment, not a conspiracy, and not bad luck. It is an automated, entirely mechanical process with rules you can read in advance — and most traders only read them after the fact. This article explains what liquidation is, the arithmetic that triggers it, and the structural habits that make it a remote event rather than a recurring one.
- Liquidation happens when your account's equity falls below the exchange's maintenance margin — the minimum collateral required to keep a leveraged position open. It is automatic, not discretionary.
- Leverage does not change how much you can lose on a move; it changes how little a move needs to be before you lose everything committed to that position.
- The variable you actually control is position size relative to total capital — not leverage, not the stop level, and certainly not the market.
- Avoiding forced exits is an architecture problem: sizing, margin mode, collateral buffer and a predefined invalidation point, decided before entry.
What liquidation actually is
When you trade with leverage on a derivatives venue like Bybit or Binance, you are not buying an asset outright. You are posting collateral — margin — against a position whose notional value is larger than that collateral. The exchange is effectively extending you exposure, and it protects itself with a simple rule: your position must always be backed by a minimum amount of equity, called the maintenance margin.
As the market moves against you, your unrealised loss erodes that equity. When equity drops to the maintenance margin threshold, the venue's risk engine takes over. It does not call you. It does not wait for your view to be vindicated. It closes the position at whatever price the order book offers, and the remaining collateral — often very little — is returned to you. In many cases a liquidation fee is applied on top.
The term "margin call" comes from traditional brokerage, where a human or an automated system asks you to top up collateral before closing anything. Crypto venues compress this: some offer a brief warning state, but the practical reality is that the warning and the forced exit can occur within the same candle. Treat the margin call as a courtesy, not a safety net.
Initial margin vs maintenance margin
Two numbers matter. Initial margin is what you must post to open the position. Maintenance margin is the lower floor you must stay above to keep it. The gap between them is your entire tolerance for being wrong before the decision is taken out of your hands.
Maintenance margin is usually expressed as a percentage of notional value, and — crucially — it is tiered. The larger your position, the higher the percentage the venue demands. A position that was comfortably within a 0.5% maintenance tier can, once scaled up, sit in a 1% or 2% tier. Traders who scale into winners sometimes discover this the hard way: adding size moves them into a stricter tier and pulls the liquidation price closer, even though the trade is working.
The arithmetic, without the mystique
Leverage is often described as "borrowing power". A more useful framing: leverage is a divisor on your margin of error.

Ignore fees and funding for a moment and take the simplified case of an isolated long position. The adverse move required to wipe out your posted margin is roughly 100% divided by the leverage multiple. At 2x, that is a 50% move. At 10x, 10%. At 25x, 4%. At 50x, 2%. Maintenance margin means liquidation actually arrives slightly before those numbers — you are closed out a little above zero equity, not at it.
These are illustrative figures, not predictions, and real venue formulas include funding payments, fees and tier adjustments. But the shape of the relationship is the point. A 4% move in a major crypto asset is not an extreme event; it is a Tuesday. Choosing 25x leverage is therefore not choosing "more upside" — it is choosing to be removed from the market by ordinary volatility.
Leverage does not increase how much you can lose on a given move. It decides how small a move needs to be before you lose it all.
Why your real risk is size, not leverage
Here is the distinction that reframes everything. Suppose you hold 100 units of an asset. Whether you posted 100% of the value as collateral or 10% with 10x leverage, a 5% adverse move costs you the same amount in currency terms: five units. The loss is identical.
What differs is survival. In the unleveraged case, you still hold the position and can decide what to do. In the 10x case, you are halfway to a forced exit. Leverage does not amplify loss per move; it amplifies the probability that a temporary drawdown becomes a permanent, realised one.
The operational consequence: size your positions by the loss you are prepared to absorb in currency terms, then let leverage be whatever number falls out of that calculation. Most traders do the reverse — pick a leverage setting they find emotionally acceptable, then size up to fill the margin available. That ordering is how accounts get emptied.
Cross margin versus isolated margin
Venues offer two collateral modes, and the choice changes the entire failure profile of your account.
In isolated margin, each position is backed only by the collateral explicitly assigned to it. If the trade is liquidated, you lose that assigned amount and nothing more. The rest of the account is untouched. The loss is capped and knowable in advance.
In cross margin, the entire account balance backs every open position. This makes liquidation far less likely for any single trade — there is much more equity to absorb an adverse move — but if it does occur, the consequences are account-wide. One position can consume the collateral supporting all the others.
- Isolated margin: higher probability of individual liquidation, strictly bounded loss. Good for defined, speculative expressions.
- Cross margin: lower probability of liquidation, unbounded consequences. Only defensible with disciplined total exposure and genuine collateral surplus.
- Neither mode is safer in the abstract. Cross margin with small total exposure is conservative; cross margin with maximum exposure is the most fragile configuration available.
The mechanisms that make liquidations worse than expected
Slippage and thin order books
Liquidation is a market order placed at the worst possible moment. When many positions are liquidated at once, the risk engine dumps size into an order book that other participants have already stepped away from. The fill you receive can be meaningfully worse than the theoretical liquidation price displayed on your screen.
Liquidation cascades
Because leveraged positions cluster around popular entry levels, their liquidation prices cluster too. A move that breaches one cluster triggers forced selling, which pushes price into the next cluster, which triggers more. This is why crypto wicks are so violent and so brief: a short burst of mechanical, price-insensitive selling, followed by a rapid retrace once it exhausts. Many liquidated traders watch price return to their entry within minutes. The position does not come back with it.
Funding costs on perpetuals
Perpetual futures use a funding rate to tether the contract to spot. If you are positioned with the crowd, you pay funding periodically, and those payments come out of your margin. A position held through a sustained period of elevated funding can drift toward its liquidation price even in a flat market. Slow erosion, same outcome.
Auto-deleveraging
In extreme conditions, if the venue's insurance fund cannot absorb liquidation losses, profitable counterparties on the other side may have positions closed automatically. It is rare, and it is a reminder that leveraged derivatives carry venue-level risk beyond your own decisions.
**The honest risk.** Any leveraged position can be liquidated — including one sized conservatively — because markets can gap, order books can thin out, and no formula anticipates every regime. Liquidation does not merely cost you the position; it converts a reversible drawdown into a permanent realised loss and removes you from the asset at its worst price. Capital protection measures reduce frequency and severity. They do not eliminate the possibility. Trading leveraged instruments can result in the loss of all capital committed, and past behaviour of any strategy is no indication of future results.
How to avoid being forced out
Avoiding liquidation is not a matter of cleverness in the moment. It is a matter of decisions made before the position exists.
- Define the invalidation point first. Ask where price would have to go for your reasoning to be wrong, and place your stop there — at a structural level, not at a round number of currency you find tolerable. If the structural stop is too far away for comfortable sizing, the trade is too large, not the stop.
- Size from the loss, not from available margin. Decide the maximum percentage of total capital you accept losing on a single idea. Work backwards: that figure, divided by the distance to invalidation, gives position size. Leverage is simply the arithmetic residue.
- Keep a collateral buffer. Margin that is deployed cannot defend anything. A meaningful share of account equity held unused is not idle — it is the distance between a drawdown and a forced exit.
- Let the stop-loss do the work, never the liquidation price. If your stop is placed correctly, liquidation should be a theoretical construct you never approach. If your liquidation price sits close to your stop, your sizing is wrong by a wide margin.
- Respect tier effects when scaling. Adding to a position changes the maintenance requirement. Recalculate the liquidation price after every adjustment rather than assuming it moved in your favour.
- Account for funding and fees in long holds. On multi-week positions, carry costs are not a rounding error.
The structural view
A single liquidation is painful; a pattern of liquidations is diagnostic. It usually indicates that position sizing is driven by conviction rather than by capital preservation rules. Conviction is not a risk parameter. It feels strongest precisely when positioning is most crowded and the liquidation clusters are deepest.
This is why disciplined operators separate the decision of *what* to trade from the decision of *how much*. The first is analysis. The second is arithmetic, decided in advance, and applied identically whether the setup looks ordinary or irresistible.
Where this fits with copy-trading
If you follow a strategy rather than trading yourself, these mechanics still apply to your account — because the positions are opened on your account. That makes a few things worth verifying before you commit capital to any copy-trading arrangement.

- Where does the capital sit? With URIEL, your funds remain on your own Bybit or Binance account. We never custody them.
- What can the connection actually do? A trade-only API key permits opening and closing positions. It carries no withdrawal rights, and you can revoke access in one click.
- How is exposure managed? URIEL operates discretionary copy-trading: decisions are taken by the strategy, within a defined framework for sizing and exposure — not by an automated rule executing blindly.
- What is the intended horizon? Twelve months minimum. Leverage-driven approaches evaluated over weeks tell you almost nothing except how volatile the recent regime was.
The principle is the same one that governs sound position sizing: the structure should mean that no single event — a bad trade, a venue outage, a disagreement with your provider — can remove your ability to act. You stay in control.
Can I lose more than my deposit through liquidation?
On major venues, standard retail accounts use negative-balance protection and an insurance fund, so losses are generally capped at the collateral backing the position — the assigned margin in isolated mode, or the full account balance in cross margin. Policies differ by venue and product, so check the documentation of the platform you actually use.
Does a stop-loss protect me from liquidation?
It should, if sized correctly: a stop placed well before the liquidation price closes the position on your terms. But a stop is a market instruction, not a guarantee. In a fast gap or a thin order book, it can fill materially worse than intended. Position sizing is the primary defence; the stop is the execution of that decision.
Is low leverage always safer than high leverage?
Not automatically. Leverage measured against an oversized notional position is still dangerous. What matters is total exposure relative to total capital, and how far price must travel to reach your liquidation price. A 3x position using your entire balance can be riskier than a 10x position using a small fraction of it.
Why did price reverse immediately after I was liquidated?
Because liquidations are a cause of the move, not only a consequence of it. Forced closes inject price-insensitive orders into the book, overshooting fair value. Once the cluster is cleared, price often retraces. This is a structural feature of leveraged markets, not a sign that anyone targeted you individually.
Should I add margin to a losing position to push the liquidation price away?
It buys room, but it also increases the capital committed to an idea the market is currently disputing. Occasionally it is a considered adjustment within a plan. More often it is an unplanned escalation that converts a bounded loss into a larger one. If the decision was not made before entry, treat it with suspicion.
Does URIEL use leverage?
URIEL operates a discretionary strategy within a defined risk framework, and leverage may be used as part of it. What does not change is the structure: your capital remains on your own exchange account, the API key has no withdrawal rights, and you can cut access at any moment.
Closing thought
Liquidation is often discussed as a dramatic event. It is better understood as the predictable terminal point of a sizing decision made earlier — usually calmly, usually without arithmetic. The traders who rarely encounter it are not the ones with better forecasts. They are the ones who decided, before entering, exactly how much of their capital a single disagreement with the market was allowed to cost.
That is the whole discipline: make the boring decision first, so the dramatic one never has to be made for you.
