You could be forgiven for thinking that the price of Bitcoin is hard to predict. We should know. Longtime readers may recall our January 2018 cover story, headlined “How High Will Bitcoin Go?” which Robert Hackett and I were writing as the Bitcoin price touched $20,000 for the first time. As we went to press less than a week later, Bitcoin was already crashing. Such is the life of a magazine journalist trying to keep up with digital technology.

Today Bitcoin is hovering around $58,000, after breaking $60,000 for the first time over the weekend. But for all the seemingly capricious moves that Bitcoin makes—sometimes it goes up when stocks do but sometimes it goes down; sometimes people trade it the way they trade Tesla but sometimes they trade it the way they trade GameStop—there is one reliable signal for whether it’s going up or down.

It’s a pattern as old as time: the Fibonacci sequence—as central to mathematics as it is to nature, dictating the way seashells spiral and the way petals grow on a flower—can also help predict the price of Bitcoin.

Fibonacci works like this: Each new number in the sequence is the sum of the two preceding figures. Starting with 0, it goes 1, 1, 2, 3, 5, 8, 13, and so on. (The sequence derives from a thought experiment about rabbit breeding from 12th century Italian mathematician Leonardo Pisano, a.k.a. Fibonacci.)

JC Parets, an expert in technical analysis of markets who for the last decade has also run a research service called All Star Charts, has found a way to apply this pattern to Bitcoin.

Back in 2017, Parets first blew my mind with this revelation when he correctly predicted that Bitcoin would surpass $6,500, then go on to rise above $7,400 (a bullish trend that kept on going at the time).

Parets figured this out by breaking down Bitcoin’s price movements into segments. After a crash, if the price recovered nearly 62%, then went on to regain 100% of its old high, Bitcoin was likely to bounce further, 162% off the low. After that, if Bitcoin didn’t fall back to the previous level, it was likely to rise to nearly 262% from the last bottom (because 262 is the sum of the previous two percentage figures: 100+162=262).

This led Parets to predict in mid-December, when Bitcoin was trading just below $20,000, that Bitcoin $30,000 was nigh; it hit $30,000 in the early days of 2021. His next prediction was about $45,000, which Bitcoin surpassed on Feb. 8.

Now, because Bitcoin is holding above that level, Parets is targeting $70,000 next.

If the math is hard to follow, Parets’s chart sketches it out:

It's not exactly clear why this pattern works so well. At its root, technical analysis is intended to measure and predict shorter-term market sentiment, in contrast to fundamental analysis focused on long-term underlying value. And Parets' Fibonacci pattern certainly echoes the boom-and-bust cycle of the crypto market.

Bitcoin also bounces around between its big moves, and may hover above or below the key thresholds in the meantime. (“Think of it like a mattress,” Parets says. “Supporting and resistance levels have some give to [them].”) But as long as Bitcoin stays above $45,000, he thinks there’s reason to be bullish.

Watch out, though, if things turn the other way: “A break that holds below $45,000 could send it back towards $30,000 or lower,” Parets adds.

For those considering whether it’s a good time to buy Bitcoin, the Fibonacci sequence can be useful; Parets says of his clients alone, several hundred firms include the technique in their investment processes—though they likely combine Fibonacci with other indicators and strategies “to come up with a favorable risk vs. reward proposition,” he explains.

With an asset as unique and volatile as Bitcoin, any guidance helps.

Says Parets, “I don't know how anyone looks at these assets without those tools.”