The short answer
Not the direction. On about 5,700 forecasts every two hours on gold and EURUSD, the Kronos foundation model called the direction right 49.7% and 51.0% of the time, a coin flip, and trading its calls lost -0.03R and -0.06R after costs (slightly better than random calls). Its volatility forecasts were better than a simple average-range estimate on gold and worse on EURUSD.
What we tested
- Markets
- XAUUSD (gold), EURUSD
- Period
- Year 2 (after the model's release)
- Kronos-small (24.7 million parameters, open source), 256 bars of context, one forecast path, every two hours.
- Direction trades at 1:1 after costs, against random directions at the same moments; volatility forecasts against 4 x the M15 average true range.
Results
| Gold | EURUSD | |
|---|---|---|
| Direction hit rate | 49.7% | 51.0% |
| Direction trades, after costs | -0.033R (n 1,674) | -0.059R (n 1,562) |
| Random direction, same moments | -0.046R | -0.082R |
| Volatility rank correlation: Kronos vs average range | 0.439 vs 0.391 | 0.419 vs 0.439 |
What it means
- RezSync Algo's AI does not forecast prices: it reads the market's structure, the account and the rules, and decides whether a trade is allowed. This test is why we do not sell price prediction.
Limits of this test
- The small model on CPU; larger versions were not completed. Stopped after gold and most of EURUSD.
Frequently asked questions
Can AI predict the forex market?
The open Kronos market model we tested called gold and EURUSD direction about 50% of the time and lost after costs; we found no evidence that it predicts direction.
How every test on these pages works
- Prices: the broker's own bars from an FTMO cTrader account. Fills are simulated minute by minute on one-minute bars; when one minute touches both the stop and the target, the stop is assumed to fill first.
- Costs: the broker's bid/ask spread measured by hour of day, commission per market, and the broker's own overnight swap rates for anything held past the daily rollover.
- Periods: year 1 = 24 Oct 2024 to 23 Oct 2025, year 2 = 24 Oct 2025 to 24 Oct 2026. Daily rules also use Dec 2020 to Oct 2024 ("2021-24"). Rules are written down before testing and are not tuned on the period that judges them.
- Controls: every rule is compared with entries that have no reason behind them (random sides, random days, or the same rule at fake levels), so a result has to beat chance, not just be positive.
- R is the trade's risk (entry to stop). An average of +0.10R means the average trade made a tenth of what it risked; -0.10R means it lost a tenth.
The full results for the rules RezSync Algo trades, by market and risk profile with every trade downloadable, are in the backtest report.
Related research
Research, not financial advice. These are simulated results on historical prices: they involve no real money, are produced with hindsight of the rules, and past results do not predict future ones. Costs and fills are modelled, and your broker's prices and costs will differ. Trading leveraged products carries a high risk of losing money; test on a demo account first.