When a venue's fair price drifts away from where trades print

The idea: the last trade ran away from the exchange's fair price, and the exchange's own machinery will drag it back. In a calm market it gives you nothing: almost all of those gaps are either untradeable or close in the opposite direction from the one you bet on.

Why it sounded right

There is a real, documented mechanism here, and it is not about guessing. A perpetual contract has two prices: the price of the last trade, and the fair price the exchange computes from other venues. Funding is charged off the difference between them, and liquidations and forced closes are priced off the fair one. So the exchange itself has built-in levers pulling the traded price toward the fair price. This is not arbitrage between exchanges, it is the internal machinery of a single venue.

Then comes the beautiful extension, the part that really hooks you. Liquidation fires off the fair price, but you enter at the traded price, which is already far in your favour relative to fair. So your room before liquidation is the usual room plus the gap itself. You get a position that supposedly cannot be liquidated — take maximum leverage and wait calmly. And funding usually drips your way while you do.

The third layer is the public channels that post these signals. There, almost every gap closes, and closes within minutes. It looks like ready-made statistics. But the channels publish the ones that closed. The ones that did not never make it into the feed.

How we tested it

We stopped looking at other people's feeds and started recording everything ourselves: every gap our own polling of the venues could see, and its honest outcome — closed, widened, or still open when the window ran out.

Then three controls. First: which price actually moved. A gap can close two ways — the traded price came back to fair, or fair caught up to the traded price. In the first case you made money; in the second, liquidation is walking toward you. Those are different outcomes and they have to be counted separately, otherwise "the gap closed" and "the trade made money" get confused. Second: is there anything to trade. We checked live bids and offers at that moment, not the last print. Third: how the venue computes its own fair price — from outside venues, or by smoothing its own trades.

How it ended

Most of the signals are hollow. On some venues almost every gap we found was just a stale print: there had been no trade for a long time, the orders were already sitting at the fair price, and there was nothing to trade. After that cleanup only a small slice of the whole pile is left.

Of what remains, the gap more often gets closed by the fair price than by the traded price. Which means a gap usually says that the traded price is right and the fair price is lagging behind a real move — so betting against it is a mistake. Our own records and the public feeds agree on this. Most cases do not close at all before the window ends: you sit in a position, carrying risk, and exit around flat, and the rare but large losses come from exactly there.

And then the contradiction that closes the idea for calm markets. The room before liquidation is only real where the fair price is built from outside venues. But those venues are efficient — gaps there barely happen. And where gaps happen often, the fair price is glued to the venue's own trades, and then there is no room at all: liquidation sits almost at your entry, and leverage works against you. On thin coins, on top of that, the distance between the bid and the offer is as big as the gap itself.

Lists of "coins that gap often" do not help either — they do not carry forward. The gaps are event-driven: every crash and every vertical spike brings new coins, they gap for a day or two and disappear. In flat markets all that is left is dust.

What this teaches

• "The gap closed" is not the same as "the trade made money." Always look at which of the two prices moved: the other leg closes the gap too, and then you are down with a green checkmark next to it.
• A public signal feed shows you the cases that survived. You cannot count the share of winners from it — you need your own collection, with the ones that never closed written down too.
• Check that you can actually fill against live orders, not against the last traded price. A stale print produces the prettiest gaps and the emptiest ones.
• Ask how the venue computes its fair price. If it smooths its own trades, then the "gap" is its own lag, and the promised room before liquidation is imaginary.
• Where the mechanism is clean there is no flow; where there is flow the mechanism is dirty. That mismatch is worth checking first — it closes an idea faster than any profit calculation.
• Only the event-driven part is still alive: crashes and violent spikes, when the gaps are big and real. In a calm market there is nothing here to measure.