Everybody loves a good strategy comparison. I see people arguing online every day about which indicator or method is “the best” for trading gold. Rather than join the arguments, I decided to actually test three of the most popular strategies on 5 years of XAUUSD daily data and see what the numbers say.
I tested three strategies that people actually use: EMA crossover (the classic), RSI divergence trading (the contrarian favorite), and breakout-pullback (my personal approach). All were tested on XAUUSD daily charts from August 2021 to August 2026. Risk per trade was set at 1% of equity. No commissions, no slippage (since we’re on daily charts with gold’s liquidity, this is reasonable).
Let’s get into the results.
Strategy 1: 50/200 EMA Crossover
The granddaddy of trend-following strategies. Buy when the 50-period EMA crosses above the 200-period EMA (golden cross), sell when it crosses below (death cross). Simple, mechanical, and everyone knows it.
I tested it with 2ATR stop loss and a 3ATR take profit. Here’s what I got:
| Metric | EMA Crossover |
|---|---|
| Total Return | +34.7% |
| Number of Trades | 23 |
| Win Rate | 39.1% |
| Profit Factor | 1.6 |
| Max Drawdown | -18.3% |
| Sharpe Ratio | 0.58 |
OK, so it makes money — barely. A 34.7% return over 5 years is about 6% per year compounded. That’s worse than the S&P 500 with way more drawdown. The 18.3% max drawdown is the real killer here. You’d be sitting on an 18% loss for months wondering if the strategy was broken.
Why does it perform so poorly? Because EMA crossovers are incredibly slow. By the time the golden cross happens, most of the move is already over. You buy late, you sell late, and you give back a lot of profit in the transitions.
Verdict: Technically profitable, but not worth the emotional ride.
Strategy 2: RSI Divergence Trading
RSI divergence is the contrarian trader’s bread and butter. The idea: when price makes a higher high but RSI makes a lower high, that’s bearish divergence — go short. When price makes a lower low but RSI makes a higher low, that’s bullish divergence — go long.
I tested it on the daily chart with the standard 14-period RSI. Entries on the candle after divergence confirmation. 2ATR stop, 2ATR target.
| Metric | RSI Divergence |
|---|---|
| Total Return | -22.4% |
| Number of Trades | 67 |
| Win Rate | 44.8% |
| Profit Factor | 0.8 |
| Max Drawdown | -31.6% |
| Sharpe Ratio | -0.34 |
Oof. Negative returns, a 31% drawdown, and a profit factor below 1. This strategy loses money.
Why? Because divergence is everywhere if you look hard enough. The strategy generated 67 trades in 5 years — way too many for a daily chart strategy. Most of those “divergences” were just noise. RSI divergence works great in hindsight when you’re pointing at a chart, but as a mechanical trading rule, it’s terrible.
It gets worse in strong trends. In a powerful uptrend, you get bearish divergence after bearish divergence, and every single one fails. That’s how you get a 31% drawdown — fighting a trend with a counter-trend strategy.
Verdict: Don’t trade divergence mechanically. If you use it, use it as a confirmation tool within a broader trend-following framework, not as a standalone entry signal.
Strategy 3: Breakout-Pullback Entry
This is the strategy I actually trade, so I was curious to see how the numbers stacked up. The rules: only trade in the direction of the 200-day EMA (long only above it, short only below it), enter on a pullback to the 20-period EMA after a recent 50-day high/low breakout, with 1.5ATR stop and 3ATR target.
| Metric | Breakout-Pullback |
|---|---|
| Total Return | +89.2% |
| Number of Trades | 41 |
| Win Rate | 51.2% |
| Profit Factor | 2.1 |
| Max Drawdown | -9.7% |
| Sharpe Ratio | 1.24 |
There we go. 89.2% over 5 years is about 13.6% annualized. A 2.1 profit factor is solid. A max drawdown under 10% is excellent for a trend-following strategy. The Sharpe ratio of 1.24 means decent risk-adjusted returns.
What makes this strategy work better? Three things:
- Trend filter. The 200-day EMA ensures you’re only trading in the direction of the major trend. You don’t short a raging bull market just because RSI looks “overbought.”
- Pullback entry. Buying after a pullback is way better than buying the breakout itself. You get a better entry price, a tighter stop, and a higher reward-to-risk ratio.
- Fewer trades, higher quality. Only 41 trades in 5 years — about 8 per year. Less screen time, lower transaction costs, and only the best setups.
Side-by-Side Comparison
| Metric | EMA Crossover | RSI Divergence | Breakout-Pullback |
|---|---|---|---|
| Total Return (5yr) | +34.7% | -22.4% | +89.2% |
| Win Rate | 39.1% | 44.8% | 51.2% |
| Profit Factor | 1.6 | 0.8 | 2.1 |
| Max Drawdown | -18.3% | -31.6% | -9.7% |
| Trades/Year | ~4.6 | ~13.4 | ~8.2 |
| Sharpe Ratio | 0.58 | -0.34 | 1.24 |
The breakout-pullback strategy wins on every single metric. It makes more money, has a higher win rate, a better profit factor, lower drawdown, and a better Sharpe ratio.
Important Caveats
Before you run off and start trading this exact setup, let me be clear about the limitations:
- This is daily chart data only. Results would be different on 4-hour or 1-hour timeframes.
- No commissions or slippage. For daily chart trades on liquid gold, this is minor, but it would reduce returns by maybe 5-10% total.
- Past performance ≠ future results. Just because this worked for the last 5 years doesn’t mean it’ll work for the next 5. Market regimes change.
- These are mechanical rules. Real trading involves judgment. I skip setups that don’t “look right” even if they meet the mechanical criteria.
The Real Takeaway
The biggest lesson from this backtest isn’t which strategy “won.” It’s that the style of trading matters more than the specific indicators. Trend-following with proper risk management (breakout-pullback) beats both lagging trend indicators (EMA crossover) and counter-trend strategies (RSI divergence) by a wide margin.
If you’re new to backtesting, start simple. Test the strategies you actually use, not the ones that look good on YouTube. Use realistic assumptions. And remember — a strategy that makes 10% a year with 10% drawdown is way better than one that makes 30% a year with 50% drawdown.
The numbers don’t lie, but only if you’re honest about how you run the test.
Want more strategy analysis and backtest results? Follow me on Telegram @DongyiTrade or shoot me an email at contact@dongyitrade.com.
