We ran the popular strategies for real
Eight widely used strategies, each run on six coins under the same conditions. 10 of 48 combinations made money.
Most of what is on this page loses money, and that is the point. This is not a list to copy; it is what the rules people recommend most actually did across seven years of real candles. The numbers come from the same engine that runs in your browser, and every row opens in the backtester so you can rerun it yourself.
Look first at how many of the six coins were profitable. A strategy that only worked on one coin usually found that coin's luck. The median is a more honest number than the best.
| Strategy | TF | Profitable | Median | Best |
|---|---|---|---|---|
| RSI oversold bounce | 1h | 0/6 | -89.9% | -77.9% |
| RSI in at 30, out at 70 | 1h | 0/6 | -94.7% | -46.3% |
| RSI with a trend filter | 1h | 0/6 | -7.6% | -0.9% |
| Golden cross trend following | 4h | 3/6 | +3.2% | +190.7% |
| Bollinger mean reversion | 1h | 0/6 | -100.0% | -98.2% |
| Supertrend momentum (both ways) | 4h | 5/6 | +224.0% | +336.6% |
| MACD cross with ADX filter | 1h | 0/6 | -97.1% | -86.4% |
| Donchian channel breakout | 4h | 2/6 | -58.7% | +161.7% |
Data through 2026-10-16 · 0.06% fee and 0.02% slippage per side · 90% position size · 1x leverage
Past performance does not guarantee future results. This table is a record of applying rules to past data, not a recommendation to trade them.
How these numbers were produced
Every figure in the table above came out of the same engine that runs in your browser when you press Run. There is no separate offline pipeline with different assumptions, so anything here can be reopened, changed and rerun. Each strategy was tested on six coins at its own timeframe, which is 48 runs and 43,946 trades in total, recomputed every week as new candles arrive.
- Costs are on. 0.06% fee and 0.02% slippage per side, charged on every fill.
- No look-ahead.A signal on one candle fills at the next candle's open, and a candle that touches both the take profit and the stop is resolved as the stop.
- 1x leverage and 90% position size, so these are the results before anyone amplifies them.
- Cross-checked. The engine is verified trade for trade against an independently written Python implementation, which is what the methodology page describes.
What all 48 runs add up to
Taking every strategy and coin together rather than one row at a time:
| Across all 48 runs | Value |
|---|---|
| Runs that made money | 10 of 48 |
| Median return | -74.7% |
| Best single run | +336.6% |
| Worst single run | -100.0% |
| Median max drawdown | 88.8% |
| Runs with a drawdown above 80% | 30 of 48 |
10 of these 48 runs made money, and the median one returned -74.7%. That is the honest headline of this page. These are not obscure rules; they are the strategies most commonly taught, run on the most liquid crypto markets over seven years.
The drawdown line deserves as much attention as the return line. The median run fell 88.8% from its peak, and 30 of the 48 fell more than 80%. Even among the 10 runs that finished profitable, the median maximum drawdown was 72.6%, and 6 of them exceeded 50%. A profitable backtest here is not the same thing as a holdable one.
Trade count and timeframe
Here are the eight strategies sorted by how much they trade across the six coins: Golden cross trend following 248 trades, RSI with a trend filter 318 trades, Supertrend momentum (both ways) 2,110 trades, Donchian channel breakout 2,773 trades, RSI in at 30, out at 70 2,996 trades, RSI oversold bounce 4,504 trades, MACD cross with ADX filter 12,772 trades, Bollinger mean reversion 18,225 trades.
The 4 that trade least account for 10 of the 10 profitable runs, and the 4 that trade most account for 0. The 3 strategies on 4-hour candles hold 10 of them; the 5 on hourly candles hold 0.
How to read the columns
- Profitable counts how many of the six coins finished above zero. A 5/6 is a far stronger claim than a single good result, because the rules were not adjusted per coin.
- Median is the middle of the six, not the average. It is harder for one spectacular coin to flatter it.
- Best is the single luckiest market. Read it as the ceiling, and read the median as the expectation.
Opening a strategy shows its per-coin table with drawdown, win rate, profit factor and Sharpe, and a link that loads the whole configuration into the backtester so you can change one thing and see what moves.
What this gallery cannot tell you
Each strategy is shown at one parameter setting on one timeframe. Our indicator guides sweep those settings, and on the MACD rule alone the result ranged from a near-total loss to a large gain depending on the periods chosen. Treat a row here as one point in a large space rather than a verdict on the indicator.
Nor is any of this a forecast. Seven years of history includes two bull markets and one deep bear market, and a strategy that suited those may not suit what comes next. The point of the gallery is narrower and more useful: it shows what the popular answers actually did, with costs charged, before you spend months rediscovering it with your own money.