Scarb Glossary
live
← All sections
Tested — does not work

Buying the exit from a range to the upside

Half of the range-breakout rule — buying the exit upward — turned out to be indistinguishable from simply buying bitcoin at the same moment. The whole difference was sitting in the sells.

Why it sounded right

Buying the exit upward is the more intuitive half of the move. Price sat still for a long time, the sellers above have been cleared out, and above the range there is nothing: nobody expected price up there, so there are few orders and the area is cheaper to travel through. On top of that, the protective buys of everyone who sold inside the range sit above it, and their filling adds to the move up. Selling downward has the same mechanism, but buying upward feels cleaner: the market goes up more often than not, and going the same way as the market is easier to live with.

The numbers back the buys at first, too. They make up about half the sample, they are positive on average, a high share of them finish up, and the trade is short-lived — in, take a small target, out. In a live run the buys actually looked like the best part of the rule: they were making money while the sells were losing it. It is easy to conclude "trade the upside only, leave the downside alone."

There is a second, subtler argument. Buys open at the same time are barely connected to one another: coins rise scattered, each for its own reason. Simultaneous sells are the opposite — a crash takes everything down at once, and a hundred sells in one day are effectively one observation. Hence the logical thought: buys are genuinely independent trades, so they are the ones to lean on, while the sells are one big bet on a crash.

How we tested it

The control here is different from the one for the whole rule. Somebody else's coin is no threat to the buys: if simultaneous buys are not tied to each other, then by construction you cannot reproduce them on another coin. So we took the market itself as the bar. For every trade we computed what a plain entry into the biggest coin would have given over the same window, in the same direction, with the same fees and the same funding. The question was put directly: does the rule pick the coin and the time better than "buy bitcoin for that many hours"?

Separately we checked whether the difference was hiding in the exit. We compared a fixed target, a trailing stop armed at that target, exiting on a return into the range, no stop at all, and a time limit. And we counted both halves of the window separately, so we would not present something that only held in the early years as a law.

How it ended

The sells beat the bar, and beat it clearly: while the market on average went against them, they made money. The buys do not. The difference between buying by the rule and simply buying the biggest coin for the same amount of time was small enough to be chance. So for the buys, "you might as well just buy bitcoin" is literally true: the range, its boundary and the margin past it add nothing to what the market's rise gives you anyway.

The independence argument does not convert into money either. Yes, the buys are unconnected, and under a bigger position size their drawdown holds up better than the sells' does. But the payoff barely grows with size, because the capital just sits idle: buys too rarely overlap. Spreading risk does not pay when there is nothing to spread. And both sides together still beat either one alone. The benefit of the buys being independent turned out to be in the speed of testing, not in the portfolio: you can learn the truth about them several times faster than about the sells.

Where this could be wrong. The live run showed the opposite sign: there the buys were up and the sells were down, and the buys made money while the market went against them. There were few trades, so there is no conclusion in them. And the whole measured window was mostly a rising market, which flatters the "just buy and hold" bar rather than the rule.

What this teaches

More in this section

A long shelf, then a break upward
The idea: price sits in a well-worn range for months, and leaving it means a run to the next such range. When we tested it, breaking out of the shelf paid exactly as much as a random day on the same coin.
Big orders in the book as a reference point
The idea: a big order in the book is a level price will bounce off, or a level it will aim for. When we tested it, price reached such a wall no more often than it reached a randomly picked level at the same distance.
Buying into a vertical spike
The idea: the coin has gone vertical, so a move has started and you should get on board while it runs. When we tested it, buying the spike was not a zero but a loss: far fewer trades finished up than with a random entry.