Where you are. You have a future: a linear payoff on a large exposure, with no money down beyond a deposit. This lesson is about that deposit. It is the entire cushion between a position and somebody else’s loss, and its size decides everything about how the position dies.
The number that decides how long you last
Ten thousand of your own money. A hundred thousand of exposure. Ten times leverage.
The position moves one percent your way and you have made a thousand: ten percent on your capital, from a one percent move. That multiplication is why anyone bothers.
Now run it the other way. One percent against you costs a thousand. Five percent costs five thousand, half the deposit. Ten percent costs ten thousand, and the deposit is gone entirely - on a move that a liquid market can produce in an afternoon and barely register as news.
The multiplier does not care about direction. At ten times leverage, a ten percent move is not a bad day. It is the end of the account.
The idea in one paragraph
Leverage multiplies percentage moves by a factor, and that factor sets the exact adverse move that erases your deposit: one divided by the leverage. At two times it takes fifty percent, at five times twenty, at ten times ten. Because the deposit is the only cushion, and because losses beyond it land on whoever lent you the exposure, no lender waits to find out. Instead they demand the cushion be topped up when it thins - the margin call you met in module 4 - and if it is not topped up they close the position themselves, which is liquidation. Liquidation is not a punishment and not a judgement about whether you were right. It is the lender’s protection, executed automatically, and it always happens before the deposit is fully gone rather than after.
The wipeout move, exactly
The exercise computes it for three leverage levels, and the arithmetic is worth doing once by hand because it is the whole lesson.
At leverage L, the position is L times the deposit. A move of x percent against you costs L times x percent of the deposit. Set that equal to the deposit and solve: x is one over L. Ten times leverage, ten percent. Twenty times, five percent. Fifty times - which retail crypto venues have offered - two percent, which is an ordinary Tuesday.
Why the position dies before the money does
If a lender waited until the deposit was exactly exhausted, it would be waiting until the moment it starts losing its own money - and in a fast market, by the time it acts the loss has already gone further. So the machinery triggers early.
The maintenance level is a threshold above zero: when equity falls to it, the lender calls for a top-up, and failing that, closes the position while there is still cushion left to sell into. That is why funds are liquidated with money still in the account, and why the complaint that “the position would have come back” is both often true and entirely beside the point. The lender is not managing your position. It is protecting its own.
The other half of the deposit’s job
The deposit is doing two jobs at once, and the next lessons separate them. It covers the loss that has already happened - the position is worth less than it was, and somebody must recognise that today rather than at expiry. And it covers the loss that has not happened yet, the plausible move between now and whenever the position could be closed.
Those are different quantities with different names, and lesson 12 builds both. For now the important thing is that a single number labelled “deposit” is doing both jobs, and that is exactly why it can look adequate right up until it is not.
Check yourself
1. At twenty times leverage, what move wipes out the deposit, and why is that number not reassuring?
Five percent, since the wipeout move is one over the leverage. It is not reassuring because five percent is an ordinary move in most liquid markets and a small one in volatile ones, so the account is exposed to being ended by a session that nobody would describe as dramatic.
2. Why does liquidation happen before the deposit is fully gone?
Because the lender is protecting itself, not managing your position. Waiting until the cushion is exactly exhausted means acting at the moment its own money starts to be lost, and in a fast market the price will have moved further before the close completes. Closing early, at a maintenance threshold, is what keeps the loss inside the deposit.
3. A trader complains that the position recovered the next day. Why is that not an argument against the liquidation?
Because the liquidation was never a prediction about the position. It was the enforcement of a threshold agreed in advance, protecting the lender from a loss beyond the posted cushion. Whether the position would have recovered is unknowable at the moment of the call and irrelevant to the party whose money was at risk.
4. Why is liquidation more dangerous to the market than to the individual account?
Because it is mechanical, simultaneous and one-directional. Participants with similar positions breach their thresholds on the same move, so forced selling arrives all at once into a market that is already falling, pushing prices further and triggering more. The individual outcome is a lost deposit; the aggregate outcome is a cascade.
Do this
Compute the fatal move at each leverage and watch the equity paths.
python3 code/leverage_paths.pyThe starter provides the equity path and leaves you wipeout_move: at leverage L, return the negative move that costs exactly the whole deposit. The assertions check it against minus one over L at each level and confirm that more leverage means a smaller move is fatal. Success prints the three paths and the line:
the deposit is the whole cushion: at ten times, a ten percent move against you is the end of itThe completed version is in solutions/leverage_paths.py.
What you can now do. You can compute the exact move that ends a levered position, explain why the lender closes it before that point, and say why the mechanism is more dangerous collectively than individually. The next lesson takes a contract with no expiry date at all and shows what has to replace the deadline.