AICryptoXpert

We tested 25 popular crypto trading rules. Most didn’t work.

AICryptoXpert research · 144 coins · September 2022 to October 2026 · CoinGecko daily data

Before launching, we tested 25 trading rules people commonly use, from “buy the dip” to “buy what’s pumping”, plus our own trend score. 21 looked positive in both halves of the data. None was positive in every single year, and none stayed clearly above zero in every year once each coin was counted once a month. One thing did hold up every time: how much a coin swings.

Check how risky your coins are · free Explore the results in the app

How we tested

The big claims at a glance

ClaimResult
Buy when a coin is “oversold”Didn’t hold up
Buy what’s pumpingStill testing
Buy coins that are beating BitcoinDidn’t hold up
Buy the dip in an uptrendDidn’t hold up
Wait for the signal to confirmDidn’t hold up
Calm coins go up moreDidn’t hold up
Our own scoreDidn’t hold up
Volatility tells you how big the swings will beHeld up

Buy when a coin is “oversold” · Didn’t hold up

The claim: Low RSI near support means a bounce is coming.

30 days later, vs an average day−1.5 pts
Went up within 30 days45% (any day: 45%)
Strict test (no overlapping trades)−1.4 ± 2.3 pts

No better than buying on a random day, slightly worse. Oversold coins often keep falling.

Buy what’s pumping · Still testing

The claim: Coins with strong recent momentum keep going up.

30 days later, vs an average day+6.3 pts
Went up within 30 days49% (any day: 45%)
Strict test (no overlapping trades)+7.2 ± 5.0 pts
2022 (bear market)−9.1 pts
Worked on91 of 141 coins

The strongest result we found, and it passed the strict test. But it lost 9 points in the 2022 bear market, and momentum tends to fail exactly when you most need it to work. We’re watching it in the live record before calling it.

Buy coins that are beating Bitcoin · Didn’t hold up

The claim: Coins outperforming Bitcoin have “relative strength” and keep leading.

30 days later, vs an average day+1.4 pts
Went up within 30 days40% (any day: 45%)
Worked on39 of 140 coins
90-day version+4.9 pts, but went up only 43% of the time

A few big winners lifted the average, but most of these coins went up less often than a random pick.

Buy the dip in an uptrend · Didn’t hold up

The claim: Pullbacks in a healthy trend are the best entries.

30 days later, vs an average day+1.9 pts
Went up within 30 days45% (any day: 45%)
Strict test (no overlapping trades)+0.9 ± 3.1 pts, can’t tell from zero

No measurable edge on daily crypto data. The dip often kept dipping.

Wait for the signal to confirm · Didn’t hold up

The claim: Waiting 3 days for a buy signal to hold filters out the fakes.

30 days later, vs an average day+3.3 pts
Went up within 30 days46% (any day: 45%)
Strict test (no overlapping trades)+1.0 ± 3.2 pts, can’t tell from zero

Waiting made signals later, not clearly better.

Calm coins go up more · Didn’t hold up

The claim: Low-volatility coins quietly beat the wild ones.

Typical (median) 30 days, calmest vs wildest group−0.3% vs −5.0%
Average 30 days, calmest vs wildest group+3.9% vs +10.5%
Ranked fairly, day by dayNo clear winner

This one fooled us at first. Calm coins had the better typical month, but the wild ones had the higher average because of a few huge winners. That’s a difference in how bumpy the ride is, not a reliable edge.

Our own score · Didn’t hold up

The claim: A high AICryptoXpert score means the coin will go up.

30 days later, vs an average day+3.0 pts
Strict test (no overlapping trades)+3.1 ± 3.4 pts, can’t tell from zero
Beat just holding Bitcoin39% of the time
Negative years2024 and 2025

Our score describes a coin’s trend well, and leaned positive on average, but not clearly enough to call it a forecast. We say so on every page and publish this result anyway.

Volatility tells you how big the swings will be · Held up

The claim: Wilder coins have bigger moves, in both directions.

Dipped 20%+ within 30 days15% calmest fifth → 51% wildest fifth
Rose 20%+ within 30 days25% calmest → 55% wildest
Average worst dip−10.4% → −22.3%
Held inEvery coin-size group

The one finding that held up. It doesn’t say which way a coin will go, only how rough the ride is likely to be. That’s what VoltX shows on every coin.

What we did with this

We built the app around the one finding that held up. VoltX rates every coin from Low Voltage to Overload by how hard it swings. It doesn’t predict direction; it tells you how rough the ride is likely to be. Our trend score is still in the app as a description of each coin’s trend, not a forecast, and every score is saved nightly so anyone can check it later on our Track Record page.

Run the free Risk Check

History, not a forecast. Past results don’t guarantee future ones. Educational research, not financial advice.