This is not about an error in the forecast. It is about what happens when the forecast was right — and the position did not survive to the target. At 25× leverage that happens to one in four a correct forecast. Here is the anatomy of one trade.
Analysis on the Deribit price archive (BTC/ETH, 2024–2026). The option premium is a Black-Scholes model at market volatility. Every number is a measurement. A journal of decisions, not signals.
Dossier · evidence
The same signal to buy, the same $1,000 of risk capital, the same market move. The only difference is the instrument.
−$1 000
On the way to the target, price dropped for an hour by −4%. That is enough for the exchange to take the entire margin. Liquidation irreversible: price then went exactly where the trader thought, but they were no longer in the position.
+$3 605
The same bet on a rise. It survived the −4% wick without consequence — liquidating a bought option impossible by construction. It reached the +5% target and booked the profit.
The same bet. The same move. One result is zero, the other is $3,605. The difference is not in the forecast. The difference is that one instrument has a built-in forced-closure mechanism and the other does not.
A systemic error
It seems that “−4% against the position” is a rarity. We measured across the entire BTC and ETH archive how often such a wick happens within a day. It turned out to be an ordinary week.
| Leverage | Liquidates on a move against of | How often such a wick happens per day |
|---|---|---|
| 10× | −10% | ≈ every 125th day |
| 25× | −4% | every 8th day |
| 50× | −2% | every 3rd day |
| 100× | −1% | more than half of all days |
| 125× | −0.8% | 60% of days |
Measured across ~43,400 daily windows, BTC+ETH, 2024–2026. The threshold is ≈ 1/leverage; the real liquidation arrives a little earlier still (maintenance margin).
This is where a trader’s worst feeling comes from: the forecast was right, and the capital vanished. At 25× leverage, of the bets that eventually touched a modest +2% target, one in four died on the way to it. For a +5% target it is already 62%. The signal’s stop-loss usually sits at −3…−5%, but the exchange closes the position at −4% before that stop can fire. You do not even manage to exit as planned — micro-noise carries you out.
Why this applies to any signal channel rather than one in particular — the breakdown of the leveraged-signal model →
You did not lose to the market. You lost to the instrument you used to bet on it.
Asymmetry
Usually “safer” means “less profitable”. Not here. On the horizon that fast signals work with, the same bet through an option gives more, no less:
The signal worked exactly as promised — the move came quickly. The 25× future gave $1,250. The same move through an option — $3 605. Three times more, and impossible to liquidate.
And if the signal had lied and the move had not come? The future would have been liquidated on the first wick. The option would have cost you only the premium — not the deposit. And you would have waited for the next entry.
Why this horizon? How much did the premium cost? Which structure is used here and why? If these questions have occurred to you, you are already ahead of most people trading blind. The answers are the course.
Honestly: the same structure behaves differently over a longer horizon, and there are moves where futures would have earned more. Where exactly that line runs — we work through it on real cases inside.
And one more thing futures cannot do
Ahead of an earnings report, a Fed decision or a large expiry, one thing is often clear: there will be a strong move. The direction is not. Futures are helpless here: they forces you to pick a side, and half the time you will choose wrong.
In options you do not have to make that choice. There is a structure that earns on the move itself — in either direction, and costs you, as always, only the premium. How it is built and when it makes sense — a separate conversation, and it is in the course.
Stop guessing the direction. You bet on how aggressively the market will move.
The course is not about “signals that work”. It is about an architecture in which a correct forecast survives to the result, and a mistake costs exactly what you decided in advance. The first intake is limited — to keep the quality of the analysis, not to gather a crowd.
Join the first intake → First — the desk’s live data