Download PDF ↓
Signal accuracy · measured, not promised
January 1 – July 16, 2026 · 600,000+ measured signal outcomes across 448 markets
Every figure in this report comes from systematic, timestamped measurement of the signals our engines produced — including the asset classes where results were flat. Nothing is cherry-picked; section 10 exists on purpose.
win rate on confirmed forex intraday signals — July 2026
n = 2,393 signals · 91.8% including timeouts · measured exactly as delivered in the product
Across the full reporting period (January 1 – July 16, 2026), confirmed forex intraday signals won 87.7% of decided outcomes across 19,929 signals.
What we measured. Every signal produced by our two engines — a daily scan and a continuous intraday engine — was assigned a fixed take-profit and stop-loss at a 1:4 distance ratio and tracked to first touch. At that geometry a win returns +0.25R and a loss −1R, so the breakeven win rate is exactly 80%. Numbers above 80% represent genuine positive expectancy, not a generous scoreboard.
What we found. Signals that pass our AI confirmation layer materially outperform the unfiltered signal stream in every asset class. In the flagship case — forex intraday — confirmation lifted the win rate from 81.8% to 92.7% in July 2026 and roughly 7× the expected value per signal, while keeping about one signal in nine.
What we're honest about. ETF signals were flat over the period. Crypto and stock intraday signals cleared confirmation with roughly breakeven expectancy. Those results are printed in full in sections 5, 6 and 10 — a report you can trust on its best numbers has to be equally clear about its worst.
TrendTrader Pro is a signals platform. It does not hold funds or execute trades — it tells you which way each market is trending, and how confident it is.
Scans every covered market once per day and issues a directional signal — buy, sell, or between zones — with a day count since the trend last flipped. Built for swing traders and investors who want a clean daily read without staring at charts.
Runs continuously against 1-minute market data across the same universe, reacting to trend flips within minutes rather than the next day. Available on the Pro plan; this is the engine behind the flagship forex numbers in this report.
The signal stack. Both engines are built on a proprietary trend-detection framework. On top of it sits an AI confirmation layer: an independent filter that evaluates each raw signal and passes only those meeting its confirmation criteria. Subscribers see confirmed signals highlighted in the product; this report measures both populations — the full signal stream and the confirmed subset — so the value of confirmation is visible rather than asserted.
The universe. 448 markets across five asset classes: forex (63), stocks (190), crypto (118), ETFs (60) and commodities (17), with market data from an institutional-grade provider.
One measurement rule, applied identically to every signal, in every market, in every asset class.
January 1 – July 16, 2026 · all covered markets · 1:4 take-profit : stop-loss · decided outcomes
| All detected signals | Confirmed signals | ||||||
|---|---|---|---|---|---|---|---|
| Asset class | Win % | EV (R) | n | Win % | EV (R) | n | OOS EV (R) |
| Commodities | 80.3 | 0.00 | 507 | 92.9 | +0.16 | 28 | +0.17 |
| Crypto | 80.7 | +0.01 | 3,598 | 87.3 | +0.09 | 165 | +0.13 |
| Forex | 84.3 | +0.05 | 1,728 | 86.5 | +0.08 | 133 | +0.07 |
| Stocks | 79.1 | −0.01 | 4,808 | 84.4 | +0.06 | 347 | +0.06 |
| ETFs | 77.9 | −0.03 | 1,673 | 79.1 | −0.01 | 134 | −0.10 |
Win % on decided outcomes. EV in R-multiples per signal (win = +0.25R, loss = −1R). OOS EV = out-of-sample only, confirmed arm. Confirmed commodities (n = 28) and forex (n = 133) are small samples — directional, and printed as such.
Confirmation clears breakeven in four of five classes. The unfiltered daily stream hovers at or below the 80% breakeven everywhere except forex. After confirmation, commodities, crypto, forex and stocks all sit above breakeven — and hold positive EV out-of-sample, which is the stricter test.
Commodities led on rate, crypto led on evidence. Confirmed commodities printed the period's best daily win rate (92.9%) on a small sample of 28 signals; confirmed crypto delivered +0.09R across 165 signals with the strongest out-of-sample EV (+0.13R) at meaningful size. We treat the commodities number as directional and the crypto number as the more bankable daily result.
ETFs did not clear the bar. Confirmed ETF signals were roughly flat in-sample (−0.01R) and negative out-of-sample (−0.10R). We publish them anyway: the measurement framework doesn't get to only report the classes it flatters. See section 10.
January 1 – July 16, 2026 · 599,374 measured outcomes · 1:4 take-profit : stop-loss · decided outcomes
| All detected signals | Confirmed signals | ||||||
|---|---|---|---|---|---|---|---|
| Asset class | Win % | EV (R) | n | Win % | EV (R) | n | OOS EV (R) |
| Forex | 79.2 | −0.01 | 175,061 | 87.7 | +0.10 | 19,929 | +0.12 |
| Commodities | 77.7 | −0.03 | 40,287 | 83.1 | +0.04 | 3,998 | +0.03 |
| Stocks | 79.0 | −0.01 | 73,133 | 80.3 | 0.00 | 5,633 | 0.00 |
| Crypto | 77.5 | −0.03 | 280,446 | 80.0 | 0.00 | 27,805 | +0.01 |
| ETFs | 78.9 | −0.01 | 23,333 | 79.8 | 0.00 | 1,878 | 0.00 |
Win % on decided outcomes. EV in R-multiples per signal (win = +0.25R, loss = −1R). OOS EV = out-of-sample only, confirmed arm. Sample sizes here are large enough that the forex and commodities lifts are far outside statistical noise.
Forex is the intraday flagship — at scale. Confirmed forex intraday signals won 87.7% of decided outcomes across 19,929 signals over six and a half months, with out-of-sample EV (+0.12R) above the in-sample figure — the opposite of what overfitting looks like. This is the population behind the July product-matched 92.7% on page 11.
Commodities replicate the pattern at smaller scale. Confirmation lifted commodities from 77.7% to 83.1% across 3,998 signals and held positive out-of-sample (+0.03R).
Stocks, crypto and ETFs: confirmation helps, but the edge is thin. All three classes land at or a hair above breakeven after confirmation. The honest read is that our intraday edge is currently concentrated in forex and commodities — and that is exactly how we describe the product.
A month-deep study measuring signals precisely as a subscriber experiences them: from the moment the signal appears, held until the take-profit or stop-loss resolves.
| July 2026 (product-matched) | Win % | Incl. timeouts | EV (R) | n |
|---|---|---|---|---|
| Forex — confirmed | 92.7 | 91.8 | +0.16 | 2,393 |
| Forex — all detected | 81.8 | — | +0.02 | 20,956 |
| Commodities — confirmed | 91.0 | 91.0 · 0% timeouts | +0.14 | 100 |
| Commodities — all detected | 77.7 | — | −0.03 | 2,372 |
Confirmed commodities is a small arm (n = 100, zero timeouts) — directional; the durable commodity figure is the full-period 83.1% on 3,998 signals (section 6).
The AI confirmation layer is a filter, not a promise. Here is what it paid for in July 2026, forex intraday, both arms in the open.
The unfiltered stream is roughly breakeven. All detected forex signals in July decided at 81.8% — barely above the 80% breakeven, worth about +0.02R each. That is the honest baseline of raw trend detection, and we print it, because the product's value lives in the gap above it.
Confirmation concentrates the edge. The one-in-nine signals that pass confirmation carried virtually all of the expectancy: +0.16R per signal, roughly seven times the unfiltered stream. The same pattern repeats across the full period (section 6: 79.2% → 87.7% on 19,929 confirmed signals) and in commodities (77.7% → 91.0% in July; 77.7% → 83.1% full-period).
Selectivity is the product. A subscriber acting only on confirmed signals sees fewer opportunities — by design. Filtering nine signals down to one is exactly the work the AI layer exists to do, and the two-arm measurement above is how we verify it keeps earning its place.
Statistical claims are only as good as the volume and breadth behind them.
| Asset class | Markets | Intraday outcomes measured |
|---|---|---|
| Forex | 63 | 175,061 |
| Stocks | 190 | 73,133 |
| Crypto | 118 | 280,446 |
| ETFs | 60 | 23,333 |
| Commodities | 17 | 40,287 |
Per-class rows are the production-configuration measurement arms of section 6 (confirmed arms are subsets). The corpus totals above count every event measured in the study, including measurement runs outside the production arms shown; the daily engine's 12,975 measured outcomes are tabled in section 5.
Cadence. The daily engine scans the full universe once per day; the intraday engine evaluates continuously against 1-minute data. Every signal each engine produced in the period was measured — there is no discretionary selection anywhere in the pipeline.
A performance report you can trust has to be specific about its own weak points. These are ours.
TrendTrader Pro is a software-as-a-service analysis platform. All content in this report and on our platform is provided for educational and informational purposes only and is not intended to be, and should not be construed as, financial, investment, or trading advice. Decisions to buy, sell, hold, or trade in securities, commodities, currencies, digital assets, and other investments involve risk and are best made based on the advice of qualified financial professionals. TrendTrader Pro does not connect to brokerage accounts, does not execute trades, and does not provide personalized investment advice.
Trading and investing are inherently risky. Many people lose money in connection with trading and investing activities. The risk of loss can be substantial, and you may lose more than your initial investment in leveraged products. You should carefully consider whether trading is suitable for you in light of your financial condition, and never risk capital you cannot afford to lose.
All performance figures in this report are derived from systematic backtesting of signal outcomes against historical market data. Backtested performance results have inherent limitations. Unlike an actual performance record, backtested results do not represent actual trading, and since no trades were executed, the results may have under- or over-compensated for the impact of certain market factors, including lack of liquidity, spread, slippage, commissions, and fees. No representation is being made that any account will or is likely to achieve profits or losses similar to those shown. Win rates, expected values, sample counts, and all other statistics herein are presented for informational purposes only and are not indicative of future results.
Market data underlying this report is supplied by third-party providers and is used "as is." Figures may differ from those produced on other platforms due to differences in data vendors, time-stamping, rounding, and evaluation conventions. The reporting window ends July 16, 2026, the date of the final data fetch used in this study.
Past performance, whether actual or backtested, does not guarantee future results. Market conditions change, and a method that was profitable in one regime may be unprofitable in another. TrendTrader Pro, LLC accepts no liability for any trading or investment decisions made in reliance on this report.
See the signals behind the numbers
Start at trendtrader.pro — 7-day money-back guarantee, conditions apply