Liquidation Cascade

What is a liquidation cascade, and how does it differ from a single liquidation event?

A liquidation cascade refers to a sequence of liquidation events that trigger one another — the first liquidation's collateral gets sold, this sell pressure pushes the underlying asset's price further down, the price drop then causes another batch of positions with thinner safety margins to hit their liquidation threshold too, and once that batch's collateral gets sold, price gets pushed down again, repeating in a cycle that forms a self-accelerating downward spiral. The key characteristic of the whole process is a 'causal chain': each subsequent liquidation is, to some extent, directly caused by the price impact from the previous liquidation — not each independently happening to coincide in timing.

The key difference from a single liquidation event lies in 'whether the scale self-amplifies': when a single position gets liquidated, the collateral sold is usually small relative to the pool's overall liquidity, with limited price impact, not triggering a cascade effect; a liquidation cascade, by contrast, involves a scale large enough, or enough positions clustered near similar liquidation prices, that the first batch's price impact is sufficient to push price into the next batch's liquidation range, letting the sell pressure's scale keep compounding and amplifying through the cycle, with the final decline far exceeding what any single liquidation event alone could cause.

Why does a liquidation cascade happen, and what structural factors underlie it?

A few common contributing factors: positions clustering around similar liquidation prices — if a large number of traders in the market use similar leverage multiples and open positions within a similar price range, these positions' liquidation thresholds cluster within a narrow price band, and once price falls below this band, it triggers a large batch of positions being liquidated simultaneously or in quick succession — this clustering is a prerequisite for a cascade to be possible; instant insufficient market liquidity — the sell pressure liquidation generates needs enough buy-side demand to absorb it, and if market depth is currently insufficient (say, panic sentiment causing buy orders to collectively pull away), even a relatively moderate liquidation sell pressure could cause a price impact far exceeding the normal proportion, accelerating the trigger of the next round of liquidation; cross-protocol, cross-asset transmission effects — if multiple protocols rely on the same batch of highly correlated collateral assets, or share similar oracle price sources, a liquidation cascade occurring within one protocol could transmit through price linkage to other seemingly independent protocols, turning a problem originally confined to a single venue into a cross-protocol systemic event.

These factors usually need to coexist for a cascade's probability and scale to significantly increase — a single factor's presence doesn't necessarily trigger a cascade effect.

How does a liquidation cascade actually unfold, and what does one complete cycle look like?

A typical liquidation cascade's unfolding process involves several stages:

1. Initial trigger: some external factor (such as macro market news, a large holder selling off, or the flash crash covered in an earlier article) causes the underlying asset's price to see an initial decline — this decline magnitude might still be within normal volatility range on its own
2. First batch of liquidations executes: the price drop causes a batch of positions with the thinnest safety margins to hit their liquidation threshold, the liquidation mechanism activates, and this batch's collateral gets sold
3. Price impact amplifies: this batch's liquidation sell pressure pushes price further down, and if market liquidity currently isn't sufficient to fully absorb this sell pressure, the price decline magnitude might exceed what this batch's liquidation sell pressure would theoretically cause
4. Second batch of liquidations triggers: the new, lower price causes the next batch of positions — with a slightly higher safety margin, but still insufficient to withstand this wave of decline — to also hit their liquidation threshold, entering the next round of the cycle
5. The cycle continues or terminates: this cycle keeps going until any of the following holds — the market sees sufficient buy-side force emerge (possibly dip buyers, or the arbitrageurs and liquidators covered in an earlier article themselves) to absorb the sell pressure, price hits a range where liquidation thresholds happen to be relatively sparse, or the protocol itself triggers some emergency mechanism (such as pausing liquidation or deploying its insurance fund)

The entire cascade might run through one complete round within minutes, or persist for hours under continuously insufficient liquidity — the specific duration depends on the market structure's actual condition at that moment.

What's the practical impact of a liquidation cascade on everyday users, and how can they lower their probability of getting caught up in one?

For a user holding a leveraged position, a liquidation cascade's most direct risk is 'your position's safety margin looked sufficient, yet still got liquidated within the cascade' — because the decline a cascade causes can far exceed the volatility range that fundamentals or normal market conditions would suggest, a position that wouldn't be liquidated under normal circumstances might still hit its liquidation threshold within a cascade's extreme decline. This means when assessing your own leveraged position's safety margin, you can't estimate purely using historical normal volatility range — you need to reserve extra buffer space to handle this kind of abnormal decline that only appears under such extreme conditions.

For a depositor (even without using leverage), if a liquidation cascade ultimately causes large-scale bad debt (a concept covered in an earlier article), this gap might ultimately need to be jointly borne by the entire pool, even if your own deposit operation was entirely normal. Concrete ways to lower your probability of getting caught up include: reserving a more generous safety margin than typically recommended when using leverage, rather than using leverage close to the platform's allowed limit; avoiding concentrating a large amount of capital in a single protocol or single collateral asset — diversified allocation lowers a single cascade event's impact proportion on your overall assets; paying attention to whether the pool you're in has a high concentration of liquidation thresholds (checkable via on-chain analysis tools showing position size at similar price levels) — the higher the concentration, the larger the scale a cascade could reach once triggered.