Free research · Golden cross (50/200 moving average)

The Golden Cross on Gold: 9 Years of XAUUSD, Every Timeframe

Short answer: trading every 50/200 cross made less than simply holding gold, on every chart that produced a cross. The selling side lost money on all of them.

Tested onGold (XAUUSD)
DataJanuary 2017 to September 2026 (9 years 8 months)

Short answer: trading every 50/200 cross made less than simply holding gold, on every chart that produced a cross. The selling side lost money on all of them.

Read this first, what this test can and cannot tell you

What we tested

The 50-period average crossing the 200-period average on a closed candle (simple and exponential averages). Golden cross → buy at the next open; death cross → sell. Two ways: (1) hold for 5, 20 or 60 candles; (2) always in the market, buy at a golden cross, reverse at the death cross. Charts M1 to W1.

Holding 20 candles after each cross (simple averages)

Chart Signal Crosses Trades taken Win rate Average per trade (bp, after spread) Verdict
M1 golden → buy 11088 10921 43.0% -0.81 no edge (not significant)
M1 death → sell 11089 10909 41.6% -1.04 no edge (not significant)
M5 golden → buy 2129 2101 49.2% -0.16 no edge (not significant)
M5 death → sell 2129 2108 44.4% -1.70 no edge (not significant)
M15 golden → buy 683 677 46.2% +1.70 no edge (not significant)
M15 death → sell 683 677 42.4% -3.09 no edge (not significant)
M30 golden → buy 345 344 50.3% +4.57 no edge (not significant)
M30 death → sell 345 341 39.9% -5.72 no edge (not significant)
H1 golden → buy 176 168 48.8% +3.45 too few trades to judge
H1 death → sell 177 170 41.8% -7.23 too few trades to judge
H4 golden → buy 49 48 56.2% +31.58 too few trades to judge
H4 death → sell 50 49 42.9% -18.21 too few trades to judge
D1 golden → buy 6 6 66.7% +53.63 too few trades to judge
D1 death → sell 7 7 85.7% +267.24 too few trades to judge
W1 golden → buy 0 0 , , no crosses
W1 death → sell 0 0 , , no crosses

Always in the market vs just holding gold (bp, after spread)

Chart Average Trades System total (bp) of which buying of which selling Just holding gold (bp)
M1 SMA 22176 -16322 -1753 -14569 +12847
M1 EMA 21351 -16051 -1653 -14398 +12847
M5 SMA 4257 +6501 +9654 -3153 +12818
M5 EMA 4005 +5399 +9088 -3689 +12818
M15 SMA 1365 +6938 +9838 -2900 +12746
M15 EMA 1409 +4219 +8489 -4270 +12746
M30 SMA 689 +4445 +8709 -4264 +12888
M30 EMA 663 +6746 +9870 -3124 +12888
H1 SMA 352 +948 +6909 -5961 +12822
H1 EMA 334 +1964 +7382 -5418 +12822
H4 SMA 98 +2728 +7740 -5012 +12709
H4 EMA 84 +6242 +9469 -3226 +12709
D1 SMA 12 +11269 +11614 -345 +12105
D1 EMA 16 +8038 +9989 -1951 +12105

Every daily golden and death cross, one by one (simple averages)

Trade Signal Opened Closed Result (bp, after spread)
1 death → sell 2018-06-15 2019-01-02 +162
2 golden → buy 2019-01-02 2021-01-07 +4064
3 death → sell 2021-01-07 2021-06-09 +159
4 golden → buy 2021-06-09 2021-08-04 -445
5 death → sell 2021-08-04 2021-12-28 -8
6 golden → buy 2021-12-28 2022-01-13 +70
7 death → sell 2022-01-13 2022-01-21 -75
8 golden → buy 2022-01-21 2022-06-20 -4
9 death → sell 2022-06-20 2022-12-29 +178
10 golden → buy 2022-12-29 2023-08-22 +480
11 death → sell 2023-08-22 2023-11-30 -762
12 golden → buy 2023-11-30 2026-06-14 +7448

How we tested

Think we got it wrong? Show us. Send your rules, your data source and your result (form fields: your definition · timeframe · data source and dates · costs you charged · your result and sample size · a link or file). If your test holds up, we publish the correction on this page, dated, with your name if you want it.

Request a backtest

Want a concept, a variation or another market tested the same way? Tell us what to test and where to send the result. It is free.

We only use your contact to send you this result.

Disagree? Bring results.

We publish numbers, not opinions, and we answer with numbers. If you believe a concept works and our test missed it, send us your exact rules, the market and timeframe, your data source and dates, the costs you charged, and your result with its sample size. We will run your version and reply with the actual statistics: every trade, the win rate, the result after costs, year by year. If your test holds up, we correct this page, dated, with credit to you if you want it.

Claims without results will not change a page. A screenshot of a winning trade is one trade; the tables here are thousands.

Send it to support@fxsynapseai.com with the subject “Research correction”, or use the request form above.