Results
Two kinds of results, kept apart: a backtest with costs (hypothetical) and a forward test on a demo account (in progress). Losses are shown next to profits.
Hypothetical backtest, not real trading.
HYPOTHETICAL OR SIMULATED PERFORMANCE RESULTS HAVE CERTAIN LIMITATIONS. UNLIKE AN ACTUAL PERFORMANCE RECORD, SIMULATED RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, SINCE THE TRADES HAVE NOT BEEN EXECUTED, THE RESULTS MAY HAVE UNDER-OR-OVER COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED TRADING PROGRAMS IN GENERAL ARE ALSO SUBJECT TO THE FACT THAT THEY ARE DESIGNED WITH THE BENEFIT OF HINDSIGHT. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFIT OR LOSSES SIMILAR TO THOSE SHOWN.
Live forward test
Forward test, demo account
Not published yet. Version 1.00 started its forward test on 16 October 2026 on a demo account: MNQ, netting, default settings, 1 contract, no manual intervention. The first report will appear here once there is a full month to report, with a link to the monitored MQL5 signal.
- Every trade will be listed, losing trades included, with entry and exit time in New York time.
- Each month we run the backtest on the same period and publish the differences.
- If anything is done by hand on the account, for example after an error, it is written in the report.
- If the forward test does not confirm the backtest, we will say so here and keep the record public.
| # | Strategy | Direction | Entry (New York) | Exit (New York) | Result (USD) | Cumulative |
|---|---|---|---|---|---|---|
| The trade log will be published with the first monthly report. | ||||||
Backtest summary
Hypothetical backtest, not real trading
Opening Bell Nasdaq MT5 version 1.00, default inputs, 1 MNQ contract, $50,000 starting balance. MetaTrader 5 Strategy Tester with 100% real ticks, on the 11 quarterly MNQ contracts from June 2024 (MNQM24) to December 2026 (MNQZ26), each tested in its own front-month period. Test period: 1 April 2024 to 9 October 2026. All amounts in US dollars.
| Measure | No costs | With costs ($2.00 per round turn) | Costs + slippage ($3.00 per round turn) |
|---|---|---|---|
| Net profit | 24,695.50 | 20,941.50 | 19,064.50 |
| Profit factor | 1.26 | 1.21 | 1.19 |
| Maximum drawdown (closed trades) | -2,375.50 | -2,761.50 | -2,954.50 |
| Trades | 1,877 | 1,877 | 1,877 |
| Winning trades | 40.6% | 39.6% | 39.4% |
| Average trade | 13.16 | 11.16 | 10.16 |
| Losing months | 9 of 31 | 9 of 31 | 10 of 31 |
| Losing days | 349 of 645 | 353 of 645 | 354 of 645 |
| Largest daily loss (12 June 2026) | -690.00 | -696.00 | -699.00 |
| Longest losing streak | 18 trades | 18 trades | 18 trades |
| Worst run of consecutive losses | -1,352.50 | -1,376.50 | -1,388.50 |
| Total costs deducted | 0 | 3,754.00 | 5,631.00 |
The longest losing streak, 18 trades in a row, ran from 16 to 25 September 2024. With costs, the largest single loss was -$424.00 and the largest single win +$3,359.50. The maximum drawdown is measured on closed trades; the drawdown while trades are open is larger.
Month by month, with costs
Hypothetical backtest, not real trading
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2024 | -438.50 | +880.50 | -110.50 | +2,619.50 | -1,036.00 | +87.00 | +377.50 | +116.50 | +2,477.00 | +4,973.00 | |||
| 2025 | -776.50 | +1,171.50 | +1,017.00 | +5,422.50 | -861.00 | -1,317.50 | +253.50 | +923.00 | +91.50 | +94.00 | +1,932.00 | -537.50 | +7,412.50 |
| 2026 | -934.50 | +3,114.50 | +1,216.50 | +59.50 | +982.00 | +2,290.00 | +12.50 | +170.00 | +1,837.50 | -192.00 | +8,556.00 |
By year
Hypothetical backtest, not real trading
| Year | Net, no costs | Net, with costs | Net, costs + slippage | Profit factor (costs) | Max drawdown (costs) | Trades | Losing months |
|---|---|---|---|---|---|---|---|
| 2024 (from 1 April) | 6,089.00 | 4,973.00 | 4,415.00 | 1.24 | -1,664 | 558 | 3 of 9 |
| 2025 | 8,924.50 | 7,412.50 | 6,656.50 | 1.20 | -2,762 | 756 | 4 of 12 |
| 2026 (to 9 October) | 9,682.00 | 8,556.00 | 7,993.00 | 1.22 | -2,135 | 563 | 2 of 10 |
The two strategies separately
Hypothetical backtest, not real trading
| With costs ($2.00) | Net | Profit factor | Max drawdown | Trades | Winning | 2024 | 2025 | 2026 |
|---|---|---|---|---|---|---|---|---|
| Strategy 1 (trend) | 8,130.00 | 1.19 | -2,177.50 | 1,296 | 42.3% | -578.00 | 2,496.50 | 6,211.50 |
| Strategy 2 (opening range) | 12,811.50 | 1.24 | -2,611.00 | 581 | 33.6% | 5,551.00 | 4,916.00 | 2,344.50 |
Costs weigh more on Strategy 1: it trades twice as often and its average trade is small. With costs, Strategy 1 lost money in 2024 (-$578.00, -$961.00 with slippage) and had 13 losing months out of 31. Strategy 2 was positive in every year. The two strategies never hold a position at the same time.
Distribution of trades and risk figures
Hypothetical backtest, not real trading
One day matters a lot
On 9 April 2025, the day the US paused new tariffs and the Nasdaq-100 moved sharply, Strategy 2 made one trade of +$3,359.50. That day produced +$3,655.00 in total, about 17% of the net profit with costs. Without that day, the backtest would show roughly $17,300 instead of $20,941.50. Days like that are rare and you should not count on them.
How the backtest was made
- Data. MetaTrader 5 Strategy Tester, "Every tick based on real ticks", on the tick history of a futures broker (AMP). Each quarterly contract is tested in the period when it was the front month, from MNQM24 to MNQZ26.
- Costs. Measured on real account statements of the same broker from May to October 2026: $1.95 per round turn per contract, all fees included. We used $2.00. The cost is deducted from every closed trade.
- Slippage scenario. One tick ($0.50 on MNQ) on every entry and every exit, $3.00 per round turn in total. It is a cautious assumption: exits at the take profit are limit orders and do not usually slip against you.
- Settings. Default inputs, 1 contract, no optimization per contract.
- In-sample. The rules and their settings were developed during research that used market data from this same period. These results are therefore in-sample. The forward test above is the out-of-sample check.
- Limits. Slippage is modelled as a fixed cost; it does not change when a stop is hit. At a broker that charges commission inside the deal, the daily target and the pause after winning trades can trigger slightly differently. This is not tested.
NAS100 CFD: in testing
The NAS100 CFD profile is not supported yet. We fixed the criteria before testing, so that we cannot pick them after seeing the result:
- the same settings as on MNQ, in index points, with no optimization per broker;
- a profit factor of at least 1.15 after spread and commission, over the whole period;
- positive in every year tested;
- a maximum drawdown no larger than 1.5 times the MNQ drawdown at the same exposure;
- passing at at least 2 of 3 brokers.
Status: 1 broker of 3 tested. It passed, narrowly (profit factor 1.16). Strategy 2 behaved as on futures; Strategy 1 was weaker on the CFD, where the overnight spread is wider. We will publish the full NAS100 results once the second broker is done. If NAS100 does not pass, the product stays MNQ and NQ only and we will say so here.