The arithmetic that makes 100x positions die in minutes, why expected value turns negative even for a winning strategy, and what to use it for instead.
At 100x with 0.5% maintenance margin, liquidation sits 0.5% from entry. Bitcoin's typical fifteen-minute range covers that distance several times a day. The question is not whether your analysis is right; it is whether the next fifteen minutes are calmer than average. Your edge, if you have one, needs hours to show up, and the position does not have hours.
Between the maintenance requirement and the liquidation fee, the exchange closes you before the market reaches the price your margin could theoretically cover, and keeps the remainder. The effective leverage you receive is lower than the number on the slider, while the probability of total loss is set by the number on the slider.
Suppose you win 60% of the time and your average win is equal to your average loss in price terms. A healthy system. At 100x, every loss is a liquidation, which is −100% of margin, while a win is capped by the first wick that scares you out, often +20% to +50% of margin. Sixty percent of +40% minus forty percent of −100% is negative. The strategy is good; the leverage made it a losing game.
It is a product that exists because it sells. If you want to understand it, open a 100x position on a simulator with live prices and watch how long it survives on an ordinary day. The answer is usually minutes. That experience, obtained at no cost, is the only thing 100x is good for.