BITCOIN · ETHER · SOLANA · FEES IN
Five strategies every trading video names. Here each one runs by code on years of real prices — fees and slippage in, the last 30 % of the history kept apart as a check — next to simply buying and holding.
2 of the five beat simply holding Bitcoin over the whole 9 years, after fees; 3 did on the last 30 %, kept apart. The best that traded here: 20-bar breakout (+2,928 %).
| Strategy | After fees | Kept apart |
|---|---|---|
| Golden cross | +439 %behind holding | −29 %behind holding |
| RSI 30 / 70 | +141 %behind holding | +27 %beat holding |
| MACD cross | +1,958 %beat holding | +32 %beat holding |
| Bollinger bounce | −33 %behind holding | −3.7 %behind holding |
| 20-bar breakout | +2,928 %beat holding | +32 %beat holding |
| Buy and hold | +1,274 % | +25 % |
“Kept apart”: the last 30 % of the history, from 2024-02-10, run on its own with a fresh account — the fairer test of the two. Every trade pays 0.1 % a fill and 0.05 % slippage.
What one dollar became in each strategy, and in simply holding — log scale. The part kept apart starts 2024-02-10.
Range: 2017-08-18 → 2026-11-20 · 243 points · Source: Binance spot BTCUSDT 1d bars; fees 0.1 % a fill + 0.05 % slippage
How often each beat simply holding the same coin, after fees — over the whole history, and on the part kept apart.
| Strategy | Beat holding | Kept apart |
|---|---|---|
| Golden cross | 0 of 3 | 1 of 3 |
| RSI 30 / 70 | 0 of 3 | 2 of 3 |
| MACD cross | 3 of 3 | 2 of 3 |
| Bollinger bounce | 0 of 3 | 1 of 3 |
| 20-bar breakout | 3 of 3 | 2 of 3 |
Long only, the whole account in each trade, no stop and no target — the textbook versions. A signal on a day's close fills at the next open.
Golden cross — Ride the long trends, sit out the long falls.
Buy when the 50-day average crosses above the 200-day average; sell when it crosses back below (the death cross).
On Bitcoin: 9 trades — not enough to trust; loses money out of sample (profit factor 0.31); max drawdown -67 % — most people would not sit through that.
Try your own version in the Strategy test →RSI 30 / 70 — Buy what fell too far, sell what rose too far.
Buy when the 14-day RSI climbs back above 30 (out of oversold); sell when it climbs above 70 (overbought).
On Bitcoin: 14 trades — not enough to trust; holds up out of sample (profit factor 11.20, win rate 75 %); max drawdown -52 % — most people would not sit through that.
Try your own version in the Strategy test →MACD cross — Catch the turn in momentum early.
Buy when the MACD line (12, 26) crosses above its 9-day signal line; sell when it crosses back below.
On Bitcoin: 120 trades — a usable sample; holds up out of sample (profit factor 1.25, win rate 36 %); max drawdown -54 % — most people would not sit through that.
These are past prices of three coins that went up a great deal — holding them is a hard bar to clear, and a strategy that sits out half the time gives up much of the rise. What the classics often do better is the ride: a smaller deepest fall.
The same rules on other coins, other bars or other years give other numbers; the shorter 4-hour history covers about five years. A test on the past is not a forecast — research and education, not investment advice.
Source: Binance spot BTCUSDT 1d bars, run by the site's backtester (the Strategy test's) · read 2026-11-21 11:04 UTC
Bollinger bounce — A stretch below the band snaps back to the average.
Buy when a day closes below the lower Bollinger band (20 days, 2 deviations); sell when a day closes back above the middle (the 20-day average).
On Bitcoin: 53 trades — a thin sample, treat the numbers as rough; loses money out of sample (profit factor 0.95); max drawdown -71 % — most people would not sit through that.
Try your own version in the Strategy test →20-bar breakout — New highs lead to more new highs.
Buy when a day closes above the highest high of the 20 days before it; sell when one closes below the lowest low of the 10 before it (the turtles' rule).
On Bitcoin: 44 trades — a thin sample, treat the numbers as rough; holds up out of sample (profit factor 1.37, win rate 38 %); max drawdown -54 % — most people would not sit through that.
Try your own version in the Strategy test →