I did something almost nobody does.
Not "I watched 3 YouTube videos and formed an opinion." Not "I bought a course and opened a demo account."
I tested 200 trading strategies on 6 years of real Bitcoin data. Systematically. Every single one. With walk-forward analysis, parameter sweeps, and out-of-sample tests.
The result? 195 were garbage.
This article is the summary of everything I learned along the way. Every strategy category. Every result. Every lesson. If you only read ONE article on this website, make it this one.
The Method: How I Tested
Before you see the results, you need to understand the method. Because a backtest is only as good as its rules.
My rules:
- 6 years of real data (2019-2025). Bull market AND bear market. No cherry-picking.
- Walk-forward analysis. The strategy is trained on data it has never seen. No overfitting.
- Parameter robustness. Not testing ONE setting. Hundreds. If only one works, it's luck.
- Costs included. Fees, slippage, spread. In reality, every trade costs money.
- Compared against HODL. Every strategy must beat simple holding. Otherwise it's useless.
If you want to know more about the method: I wrote an entire article about it.
-> Backtesting: How to Test Strategies the RIGHT Way
Category 1: Technical Indicators
These are the strategies YouTube gurus love to sell. RSI, Bollinger Bands, MACD, Stochastic. Colorful lines on the chart. Looks professional.
Doesn't work.
| Strategy | Win Rate | Return (6 yrs) | Vs. HODL |
|---|---|---|---|
| RSI (buy below 30, sell above 70) | 28% | -43% | ❌ Much worse |
| Bollinger Bands (classic) | 35% | -38% | ❌ Much worse |
| MACD Crossover | 31% | -27% | ❌ Worse |
| Stochastic Oscillator | 33% | -19% | ❌ Worse |
| RSI + Bollinger Combo | 38% | -12% | ❌ Worse |
Why they all fail: Technical indicators measure the past. They tell you where Bitcoin WAS, not where it's GOING. In a trending market like Bitcoin, "oversold" is often just the beginning of a bigger drop.
Remember: If an indicator is taught in every beginner course, it lost its edge long ago. Otherwise all trading teachers would be rich.
-> RSI: Why the Most Popular Indicator Fails -> Bollinger Bands: The 497-Euro Course That Delivers Nothing
Category 2: Sentiment Indicators
"Be greedy when others are fearful." Buffett's most famous quote. Thousands follow it in the crypto market. And lose.
| Strategy | Win Rate | Return (6 yrs) | Vs. HODL |
|---|---|---|---|
| Fear & Greed (buy at "Extreme Fear") | 41% | +12% | ❌ Worse |
| Search trend data as contrarian indicator | 39% | +8% | ❌ Worse |
| Social media sentiment | 36% | -5% | ❌ Worse |
Why they fail: Sentiment indicators measure the mood of the crowd. But the crowd reacts to price, not the other way around. When Bitcoin drops, sentiment turns bad. When Bitcoin rises, it turns good. The indicator FOLLOWS price. It doesn't lead it.
-> Fear & Greed: Buffett's Advice Burns Your Money
Category 3: Mean Reversion ("Buy the Dip")
The idea: Bitcoin has fallen, so it MUST go back up. Buy the dip!
| Strategy | Win Rate | Return (6 yrs) | Vs. HODL |
|---|---|---|---|
| Buy at -10% dip | 42% | -51% | ❌ Much worse |
| Buy at -20% dip | 38% | -47% | ❌ Much worse |
| 3 red days, then buy | 31% | -4% | ❌ Worse |
| Smart dip (with trend filter) | 52% | +50% | ❌ Half as good as HODL |
Why they fail: Bitcoin is not yogurt on sale. A falling price doesn't mean "cheap." It means: the market is currently valuing Bitcoin lower. And it can go EVEN lower.
The best "dip" strategy managed +50%. HODL managed +100%. You would have earned twice as much by doing NOTHING.
-> Buy the Dip: The Most Expensive Lie -> 3 Red Days: Now It HAS to Go Up. Right?
Category 4: Trend-Following
And here's where it gets interesting. Because ONE category actually worked.
| Strategy | Win Rate | Return (6 yrs) | Vs. HODL |
|---|---|---|---|
| Simple trend filter (1 signal) | 58% | +320% | ❌ Worse |
| Trend filter + momentum | 64% | +890% | ✅ Better |
| Multi-signal trend filter (3 filters) | 71% | +2,942% | ✅ 3x better |
Why it works: Trend-following doesn't fight AGAINST the market. It follows it. When Bitcoin rises, you're invested. When it falls, you're in cash. You don't try to predict. You react.
That sounds simple. It is. But the execution requires discipline no human can sustain. That's why I built a bot.
The multi-signal trend filter with 3 independent filters beat HODL by a factor of 3. With LESS risk (drawdown -35% vs. -77%).
An honest word about the +2,942%: This number is historically correct, but the bulk of it comes from massive Bitcoin growth during those 6 years. Bitcoin went from ~$3,500 to over $80,000. That won't repeat at this scale because the market cap is much larger today. Realistic expectation: Bitcoin grows slower (logarithmic growth curve). But the bot can likely keep beating HODL because it avoids the big crashes. The 3x factor matters more than the absolute number.
-> The Most Boring Strategy That Works -> The Watchdog: My Bot That Did Nothing for 184 Days
Category 5: Breakout Strategies
Breakout means: Bitcoin breaks through a high and you jump on.
| Strategy | Win Rate | Return (6 yrs) | Vs. HODL |
|---|---|---|---|
| Simple breakout (new 50-day high) | 55% | +85% | ❌ Worse |
| Breakout + momentum confirmation | 68% | +195% | ❌ Slightly worse |
| Multi-signal breakout (3 filters) | 71% | +197% | ❌ Slightly worse |
The surprise: Breakout strategies are BETTER than indicators. But they still don't reliably beat HODL. The best ones barely manage it, most don't.
However: They have extremely low drawdown (under 3%) and are in cash 95% of the time. As a COMPLEMENT to a trend filter, they're valuable.
-> The Scout: Short-Term Opportunities
Category 6: Daytrading / High-Frequency
The worst category. By far.
| Strategy | Win Rate | Return (6 yrs) | Vs. HODL |
|---|---|---|---|
| Regime-adaptive (complex) | 33% | -15.5% | ❌ Much worse |
| RSI Extreme + Trend | 28% | -15.6% | ❌ Much worse |
| Funding Streak | 39% | -7.1% | ❌ Worse |
| Funding Rate Flip | 49% | -0.3% | ❌ Worse |
Why it fails: Too many trades = too many fees. 150 trades per year at 0.2% per trade = roughly 26% of your capital gone (compound effect: 0.998^150). Just for fees. Add to that: slippage, spread, emotional mistakes, sleep deprivation.
97% of daytraders lose money long-term. That's not an opinion. That's a study with 20,000 traders.
-> The Sad Daytrader Statistics -> Daytrading: 99.4% Lose Money
The Summary in One Table
| Category | Tested | Profitable | Beat HODL | Best Return |
|---|---|---|---|---|
| Technical Indicators | ~60 | 5% | 0% | -12% |
| Sentiment Indicators | ~20 | 20% | 0% | +12% |
| Mean Reversion (Dips) | ~30 | 15% | 0% | +50% |
| Trend-Following | ~40 | 60% | 25% | +2,942% |
| Breakout | ~30 | 45% | 5% | +197% |
| Daytrading/HF | ~20 | 10% | 0% | -0.3% |
Result: Out of 200 strategies, 5 beat HODL. All 5 were trend-following.
The following chart makes the difference between categories visible at a glance. Hover over the bars for details on each category:
The 5 Lessons from 200 Tests
1. Less is more. The best strategies have 1-3 rules. Not 7. Every additional condition makes it worse, not better.
2. Follow the trend, don't fight it. 97% of the strategies I tested tried to fight the trend. "Bitcoin dropped, so it MUST go up." No. It doesn't have to.
3. Parameter robustness beats returns. A strategy that works with MANY settings is better than one that makes 10% more with ONE perfect setting. Because the "perfect" setting is almost always luck.
4. Fees are the silent killer. The more often you trade, the more certain you are to lose. My bot makes 4 trades per year. Not 4 per day.
5. HODL is the benchmark, not the enemy. Every strategy that doesn't beat HODL is useless. Because HODL costs zero effort. If you put in MORE effort and earn LESS, you're doing something wrong.
Remember: The best strategy is usually the most boring one. And the second best is no strategy at all. Just hold.
-> 200 Strategies Tested, 7 Mistakes Made -- My personal lessons
-> Backtesting: How to Test Strategies the RIGHT Way -- The method behind the 200 tests
-> HODL Beats 27 out of 30 Strategies -- Why doing nothing beats almost everything
-> Bot Trading Explained -- Why I automated the 5 survivors
-> DCA Savings Plan Calculator -- What a savings plan would have returned
-> Is Your Trading Worth It? -- Calculate your expected value
-> Our 7 Bots -- The systems behind the survivors
-> What Does the Bot Say Right Now? -- Current signal
Your Dominic, the guy who tested 200 strategies so you don't have to.
Disclaimer: This is not financial advice. All backtest results are based on historical data and do not guarantee future results. Only invest what you can afford to lose.
