A 2% BTC Move at 10x Already Hits My Calculator's ADL Red Zone

My own ADL Risk Score Calculator has a red-zone line built into it: any ranking score above 50 turns the row red. I built that threshold in months ago and never sat down to ask how easy it actually is to cross. So I ran the exact formula — leveraged ROE% ÷ maintenance margin rate — across a handful of leverage and asset-tier combinations. A 2% favorable BTC move at 10x leverage lands on exactly 50.00. Not close to it. Exactly on the line.

The formula, and how fast it climbs with leverage

Auto-deleveraging exists because every centralized perp exchange guarantees a liquidated trader never loses more than their margin — the exchange's insurance fund eats the rest. When a liquidation is too big for the fund to absorb, the exchange force-closes part of the most profitable positions on the opposite side instead, at the bankruptcy price, no stop order, no warning. Who gets picked first is decided by a ranking score: leveraged ROE% divided by the position's maintenance margin rate. Higher score, higher up the queue.

Take a $60,000 BTC long on the 0.4% maintenance-margin tier (the blue-chip rate most majors sit on) and move the price up 2%. At 10x leverage that's a 20% leveraged ROE. Divide by 0.4 and the score is 50.00 — my calculator's own threshold for "high ranking score, watch it." Push the leverage further on the same 2% move and the score keeps climbing in a straight line with it, because leverage sits directly in the ROE numerator:

None of that requires a violent move. A 2% BTC candle happens most days. It's the leverage multiplying a routine move into a queue-topping score, not the size of the move itself.

The part that surprised me: "safe" assets rank higher, not lower

I expected the riskier, higher-maintenance-margin tiers — the small-cap and meme-coin listings — to dominate the ADL queue, since that's where exchanges price in more risk. It's backwards. MMR sits in the denominator of the score, so a smaller maintenance margin rate produces a larger score for the same profit. I held a $60,000 position at 20x through the same 10% favorable move and only changed which asset tier's MMR applied:

Same $60,000 position, same 20x, same 10% move, same $12,000 profit in every row. Only the score changes — by a factor of 12.5x between the top and bottom rows — because BTC's thin 0.4% maintenance requirement gives the exchange far less room before it needs to trim someone's position to rebuild its insurance buffer, and profitable BTC longs are the deepest, most liquid pool of "someone" to trim from. The instrument that feels safest to hold at high leverage is exactly the one whose winners sit closest to the front of the deleveraging line.

Losing positions never show up here at all

For contrast, I ran the same $60,000 BTC position at 20x through a 10% move against the trade instead of in its favor. ROE: -200%. Score: -0.80. Nowhere near the queue, and it never will be — the formula flips sign on losing ROE specifically so a losing position collapses toward its own liquidation price instead of climbing anyone's deleveraging ranking. ADL is a tax on being right and leveraged, not a tax on being wrong.

None of this means a score of 500 gets you closed today. It's a relative ranking against every other trader holding the same side of the same contract, and it only activates at all when a liquidation on the opposite side blows past what the insurance fund can absorb — a rare event, not a daily one. But rare doesn't mean irrelevant if you're holding the highest-leverage, lowest-MMR side of the book when it happens. Run your own entry, mark price and leverage through the ADL risk score calculator — it's the same formula, live, with the same red line at 50.