Tracked Signals

Strategy tests

TradingLab '86% win rate' MACD strategy: review and backtest

No edge YouTube strategy · tested 10 Oct 2026 · figures re-checked 11 Oct 2026

What does TradingLab's '86% win rate' MACD strategy claim?

The video "BEST MACD Trading Strategy [86% Win Rate]" (2021) presents a MACD and 200 moving average setup with an 86% win rate in its title. Neither its description nor two independent summaries of the video give a trade count, market, timeframe or test method for the 86%.

How we tested it

Rules as tested: buy when the MACD line (12, 26, 9) crosses above its signal line with both lines below zero and the price closes above the 200-period moving average (EMA); sell on the mirror image. Entry at the next bar's open, stop just beyond the 200 EMA (0.1 x ATR past it), target 1.5 times the risk, one trade at a time. The video's support and resistance step is a judgement call and was not tested; a mechanical stand-in was run separately. All of these choices were written down before the first test ran. Main test: gold on 1-hour bars, January 2020 to July 2026. Checks: 15-minute and 4-hour gold, two variants of the rule, and the same rule on 8 markets at 30 minutes and 12 markets at 4-hour and daily bars: 14 versions in all.

42.1% winsthe '86% win rate', measured on gold 1-hour174 of 413 trades reached the 1.5R target; 95% range 37.5-46.9%
40% / 44.1%win rate needed to break evenbefore / after costs, with a 1.5R target
-0.039Raverage per trade, after costs+0.065R before costs; 413 gold 1-hour trades
$8,403$10,000 risking $100 a tradeafter costs, 2020 to 2026, no compounding; $12,665 before costs
-0.219Raverage per trade on 15-minute gold, after costs1,509 trades
8 of 8markets negative after costs on 30-minute barsgold, silver, 3 currency pairs, S&P 500, Bitcoin, Ether
Before trading costs unless stated. R = the amount risked on a trade.

Costs are spread, slippage and overnight financing, about 0.10R per trade on gold 1-hour. They turned +0.065R per trade into -0.039R. The worst losing run on the 1-hour test was 37.8R, $3,782 on the $10,000 account.

Does TradingLab's '86% win rate' MACD strategy work?

Not in our test: it did no better than chance. No test is shown for the 86%, and nothing in our test supports it. With a target 1.5 times the risk, a strategy needs 40% wins to break even before costs and, here, 44.1% after them. On gold 1-hour it won 42.1%: just above the first line and below the second. Its entries did a little better than random entries in the same direction, by about 0.08R a trade, but 11% of 200 random-entry runs did as well or better, so that is not evidence of an edge. On 15-minute gold and on 8 markets at 30 minutes it lost clearly after costs. The highest win rate in any of the 14 versions was 45.5%.

Watch the 54-second version on YouTube

Is the 86% win rate MACD strategy real?

Not in our test. The 86% is in the video's title with no test shown behind it: no trade count, market or timeframe. Run on 413 gold 1-hour trades from 2020 to 2026, the rules reached their 1.5R target 42.1% of the time (95% range 37.5-46.9%).

What win rate does the MACD and 200 EMA strategy need?

With a target 1.5 times the risk, 40% wins breaks even before costs. After spread, slippage and financing on gold 1-hour it needs 44.1%; it won 42.1%.

Does it do better on 15-minute charts?

No. On 15-minute gold it won 38.5% of 1,509 trades and averaged -0.219R a trade after costs (-0.038R before).

Limits of this test

  • The video gives no timeframe, so 1-hour gold was chosen in advance as the main test, with 15-minute and 4-hour gold as checks. 86% is outside the 95% range in all nine gold versions.
  • The support and resistance step was not tested: it is a judgement call. Our mechanical stand-in (price back at a recent swing low or high) won 41.1% on gold 1-hour, -0.058R a trade after costs; it is our approximation, not the video's filter.
  • We could not download the video's transcript. The rules come from its description and two independent written summaries, which agree with each other and with a separate re-creation by another tester.
  • This checks a claim on price data already used in our earlier research; no held-back data was opened. Gold prices are Dukascopy's; costs use our own spread, slippage and overnight-financing model, not one broker's.
  • A variant with the stop at the last swing point instead of the moving average made +0.072R a trade after costs on gold 1-hour, but it lost in the first half of the period, failed our check against luck, and lost on 15-minute gold (-0.114R). It is recorded, not a recommendation.
  • Trading Rush published a separate hand test of a similar MACD and 200 moving average rule; this page is about TradingLab's video.

Measured results, not advice. We have no commercial link with TradingLab. If you made this tool and think a figure is wrong, write to corrections@trackedsignals.com; corrections are published within 7 days. How we test.

More strategy tests

All strategy tests