- Banded liquidation — reduce-only IOC limit orders at the mark, bounded to ±250 bps. Protects execution quality.
- Backstop liquidation — unbanded reduce-only market orders once a risk bucket falls below three quarters of its maintenance margin. Protects solvency by removing risk faster than it protects price.
- Auto-deleveraging — the terminal rung. When even the unbanded probe finds zero executable depth, the bankrupt position is force-closed against traders holding the opposite side.
What Monaco does differently
Some exchanges rank traders by profitability and leverage, then close the single highest-ranked account first. That concentrates the entire cost of an event on one trader — usually the one who was most right about the market. Monaco does not rank. The cost is spread pro-rata across the opposite side. Being the most profitable trader in a market makes you contribute proportionally more, not first and not entirely.When ADL triggers
A risk bucket must satisfy all of the following before ADL is even considered:- Auto-deleveraging is enabled for the venue.
- The bucket is already in backstop mode — tier 2 owns it, and tier 2 could not fill.
- The bucket is bankrupt: its equity is at or below zero. Above zero it still has equity to pay for aggressive fills, and the market slice keeps it.
The slow arm: a proven dead book
The bucket must have observed 12 consecutive unfillable-book laps — roughly 60 seconds of continuously unfillable book at the 5-second retry cadence. A lap counts as unfillable when the liquidation engine looked at the book and could not place a legal order against it. That covers an empty book and a book holding only dust: if the only resting size is below the market’s minimum order size, there is nothing to fill even though the ladder is not literally empty. “Consecutive” is literal. A lap deferred for any other reason — a stale oracle, a halted market, an errored step, a missing mark price — resets the count to zero. A stall that was never observed to be a dead book cannot accumulate toward a force-close. Sixty seconds is deliberately unhurried. A market maker that misses a single quoting cycle can leave the ladder empty for 15–30 seconds all by itself, and a threshold near that window would fire on an ordinary blink.The fast arm: the insurance fund is at risk
If the bucket’s projected shortfall exceeds 50% of the insurance fund’s remaining headroom, it becomes eligible after 3 unfillable-book laps — about 15 seconds — instead of the slow arm’s full 60. The insurance fund is a buffer to protect, not a balance to spend down. A hole big enough to consume half of what is left is not something to wait out — and it is precisely this fast arm that makes the slow arm affordable. Blinks get a patient 60 seconds because genuine fund risk does not have to wait nearly as long. The fast arm still waits those 15 seconds rather than firing on the first lap, and the reason is worth stating: as the fund approaches empty, “shortfall exceeds half the remaining headroom” becomes true for essentially any bankrupt bucket. Without a floor, the protection against ordinary blinks would vanish at exactly the moment ADL runs most often. Fifteen seconds covers one full market-maker quoting cycle, which is the shortest window that can tell a dead book from a restart. Note what this implies, because it is easy to be surprised by it later: the trigger is relative to the fund’s remaining headroom, not to an absolute size. After a drawdown has depleted the fund, a routine-sized hole can clear the 50% bar that the same hole would not have cleared a week earlier.There is no minimum event size
Monaco does not impose an absolute floor below which ADL cannot happen. A small position can be auto-deleveraged if the fund is depleted enough that it qualifies under the fast arm. This is a deliberate choice, not an oversight. A floor would mean that once the fund is exhausted, small holes accumulate as uncovered bad debt with no mechanism to clear them. Monaco would rather deleverage a small position than quietly socialize an unbounded number of small losses. The consequence is handled by telling you about it here, not by hiding it behind a threshold.Who is selected, and at what price
The pool is every account holding the opposite side of the bankrupt position in that market — both cross and isolated positions. The bankrupt account itself is excluded, as are accounts already mid-liquidation or in bad-debt resolution. Allocation happens in two tiers inside a single plan.Tier A — profitable holders, at the bankruptcy price
Holders with positive unrealized profit in that market are allocated pro-rata by absolute unrealized PnL, capped both at their position size and at their own unrealized profit. Absolute PnL, not a profit ratio: a ratio would penalize small accounts, could be diluted by splitting a position across accounts, and does not add up cleanly when it is combined. Absolute PnL is additive, so splitting a position across two accounts changes nothing about the total contribution. Tier A fills at the bankruptcy price — the price at which the bankrupt bucket’s equity is exactly zero. This is the industry-standard convention and, importantly, it costs the insurance fund nothing: the bankrupt bucket is made exactly whole, no more. It is not a clawback. Monaco does not reach backward into profit you have already realized.Tier B — the remaining opposite side, at the mark
Tier A alone cannot always absorb the whole position. A single very profitable holder whose contribution is capped at their own unrealized profit can be the entire opposite side, and the position would still not be closed. Whatever Tier A cannot absorb is therefore spread across the remaining opposite-side capacity — unprofitable holders’ full positions, plus profitable holders’ capacity above their Tier-A cap — pro-rata by remaining position size. Tier B fills at the mark, with no haircut. Because perpetual bookkeeping is zero-sum, the full opposite side is always large enough to absorb the bankrupt position. Every ADL episode closes the position completely, with no human in the loop.Tier B costs the insurance fund money
This is the part that is easy to leave out of a document like this, so it is stated directly: A Tier-B fill leaves the bankrupt bucket short of zero equity. Filling at the mark rather than at the bankruptcy price means the bankrupt bucket does not get made whole. That residual deficit is routed to Monaco’s existing bad-debt and insurance path — the same path any other uncovered liquidation shortfall takes. Zero fund draw is a property of Tier A, not of an ADL episode as a whole. An episode that needed Tier B drew on the fund, and Monaco meters exactly how much.Minimum slice size
Slices respect the market’s minimum order size, so ADL does not scatter unusable dust across hundreds of accounts. That floor is waivable by deterministic rule when the pool cannot otherwise absorb the position — a market fragmented into many small holdings must still be closeable. Remainders are allocated by largest remainder, with ties and leftover floor residue going to holders in descending order of unrealized profit. The residual advantage this rank ordering confers is bounded at roughly one lot per holder.Episodes, and why you are told afterward
An ADL event is an episode. The pool, the weights, the price, and every slice are fixed at the moment the episode opens. Execution then proceeds in chunks that carry out that fixed plan — they never recompute it. Episodes run one at a time per bucket; a bucket with an episode already open is never re-planned underneath it. If a selected counterparty’s position has shrunk below their planned slice by the time their chunk executes, their slice is clamped to whatever they actually still hold. Nobody is ever over-closed, and the shortfall is not redistributed onto the other counterparties in the plan — it rolls into a follow-on episode that re-plans against fresh state. Notifications fire when an episode closes, not when it opens. This is intentional: Monaco does not announce an episode in advance of executing it. Be clear about what that does and does not mean. It is not a guarantee that an in-flight episode is unobservable. A counterparty filled in an early chunk sees that fill in their own position and balance straight away, well before the episode closes. And ADL fills print as trades — they appear on the public trades endpoint and are folded into candles, volume and last-price like any other trade. Anyone watching the tape can see an episode in progress. Monaco also computes your standing ADL exposure internally from day one, so the risk is measured before it is ever surfaced.Known gaps
Three things are consciously not shipped in this first phase, recorded here rather than discovered later:- Statements do not yet label ADL fills. An ADL fill appears as an ordinary fill.
- Charts and share cards are not yet ADL-aware.
- Public trade data does not label ADL fills either. ADL trades print on the public trades endpoint and feed candles, volume and last-price, and nothing in that data marks them as ADL. To an outside observer they are indistinguishable from ordinary trades — visible, but not interpretable.
Summary
- ADL is tier 3, and only runs when the book cannot be filled at all and the bucket is bankrupt.
- Two trigger arms: 12 consecutive unfillable-book laps, or — after 3 such laps — a shortfall exceeding half the insurance fund’s remaining headroom.
- No minimum event size — a depleted fund makes routine-sized holes eligible.
- Cost is spread pro-rata by absolute unrealized PnL, never concentrated on one trader.
- Tier A fills at the bankruptcy price and costs the fund nothing. Tier B fills at the mark and leaves a residual deficit that the insurance path absorbs.
- Plans are fixed at episode open; nobody is over-closed; you are notified at close, though the fills themselves print publicly as they happen.