Buy the dip means nothing until you say which dip

What an extreme is, and why it matters

There’s no magic in it, and no certainty either. It’s statistics. In those conditions the market has historically tended to recover, which means that if you repeat the trade many times you end up in profit. It tells you nothing about the next single trade.

And extremes are rare. In our archive, over 9 years, bitcoin closed down eight percent or worse 57 times out of 3,284 days, or about 6 times a year. If you buy every five percent drop, you’re mostly buying the noise between the extremes.

How you recognize one, in practice

The second is speed: how far the price has fallen over a short stretch of time. That’s different from the size of the fall. A three percent drop in an hour says something different from three percent over a week, even though the number is the same and the chart, seen from a distance, looks identical.

When the two measures hit extreme values together, the odds shift. When only one of them does, it doesn’t mean anything yet: that’s just the market going down, which by definition it does almost half the time.

The strength of the recent move on a 0 to 100 scale: below 30 is the standard oversold threshold, but in the crypto market the extreme starts below 20 and becomes rare below 10. The speed of the fall over short intervals: what counts is how fast the price fell, more than how far. The two measures only count when read together.

Why fast crashes are different

It’s a domino effect that feeds itself until the positions left to close run out. At that point the pressure behind the fall disappears, simply because it has run out of fuel, and what’s left is a lower price with nobody forced to sell any more.

That’s why a ten percent drop in a few hours is a different thing from a ten percent fall spread over three weeks. In the first case the derivatives market has cleaned itself out. In the second someone is getting out calmly, and people who get out calmly usually know why.

The extreme can get more extreme

The practical consequence is simple: buying during an extreme means starting to build a position, not buying everything at once. You know you might have to buy more further down, and the second part of the plan exists for exactly that.

Then there are three cases where the drop shouldn’t be bought at all. When it’s slow and steady, two or three percent a day for weeks, because that’s a decline rather than a dip. When the measures aren’t really extreme. And when there’s a real reason behind it, like a new rule, an attack on a protocol or an exchange going under: then what happened matters more than any indicator.

The drop shouldn’t be bought when it’s slow and gradual (two or three percent a day for weeks, which is a decline rather than a dip), when the indicators aren’t at truly extreme values, or when there’s a real fundamental cause such as hostile regulation, an attack on a protocol or an exchange failing.

How I use it

The edge in working this way comes from staying out when everyone is buying and buying when everyone is scared. Guessing the bottom has little to do with it. It’s the same thing said the other way round, but with a rule in front of it instead of a gut feeling.

And it’s a statistical edge: it works over many repetitions and promises nothing about the single trade. Anyone who uses it expecting the next trade to go well has understood the opposite of what’s written here.