LIVE FORWARD RESULTS SINCE JANUARY 2024 · 2.0 PRODUCTION TRACKING SINCE MARCH 2026
Every strategy. Tracked daily. Published monthly.
The full ATS 2.0 stack, forward tested with actual executed trades in live market conditions and updated after each month closes. Profit curves, month by month numbers, and downloadable daily P&L and trade log files for every configuration. Backtests reach back to January 2024. Losing months included.
This page is educational. Nothing here is financial, investment, or trading advice, and nothing here is a recommendation to buy, sell, or allocate capital in any specific way. Futures trading involves substantial risk of loss and is not suitable for all investors. Past performance, simulated or actual, does not guarantee future results. Conduct your own analysis and consult a qualified financial professional before making trading decisions.
Jan 2024 · Zeus enters live forward testing · TradingView / NinjaTrader
Jun 2025 · Underlying strategy engines testing begins
Mar 2026 · 2.0 production strategy tracking begins
NT / TV legacy algos running throughout
Live forward results since January 2024. The legacy Zeus algos have traded on TradingView and NinjaTrader without interruption since then, and they are still running. The 2.0 strategies were built on the Zeus engine and others, applying lessons learned from those live years, and entered production tracking in March 2026.
Backtest months
26
Forward test months
5
Forward tracked configs
8
Stack P&L, as tracked
+$1,424,554
The chart combines all 8 forward tracked configurations at their listed contract sizes in one continuous curve: the simulated 2.0 backtest through February 2026 flows into actual forward trading results from March 2026 at the magenta rule, and the population is identical across the whole span. The underlying strategy engines began testing in June 2025. Every configuration is in the matrix and strategy sections below. Data through the last completed month; updated 2026-08-08.
THESE RESULTS ARE BASED ON SIMULATED OR HYPOTHETICAL PERFORMANCE RESULTS THAT HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE THE RESULTS SHOWN IN AN ACTUAL PERFORMANCE RECORD, THESE RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, BECAUSE THESE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THESE RESULTS MAY HAVE UNDER-OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED OR HYPOTHETICAL 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 PROFITS OR LOSSES SIMILAR TO THESE BEING SHOWN.
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS.
2026 month by month
Completed months only; the current month is published after it closes. Every figure is colored by outcome, blue for profit and orange for loss. January and February are backtest months, tagged in the header. Live forward tracking began March 1, 2026. Losing months stay on the board; that is the point.
All figures are P&L in USD at each configuration's listed contract size, blue for profit and orange for loss. January and February come from the simulated backtest record for every configuration (tagged in the header), as does every month of the three default configurations, which are shown from the backtest record until their tracked forward results are published here. March onward is actual forward trading results for the other eight. A dash means no tracked result for that cell. Row totals sum the months shown. The Master Stack row combines the eight forward tracked configurations; the three default rows are not part of it.
THESE RESULTS ARE BASED ON SIMULATED OR HYPOTHETICAL PERFORMANCE RESULTS THAT HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE THE RESULTS SHOWN IN AN ACTUAL PERFORMANCE RECORD, THESE RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, BECAUSE THESE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THESE RESULTS MAY HAVE UNDER-OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED OR HYPOTHETICAL 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 PROFITS OR LOSSES SIMILAR TO THESE BEING SHOWN.
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS.
Omega Stack
Four strategies in one. The Omega sleeves trade NQ, ES, RTY, and YM as a single diversified stack, momentum driven day trading built on the Zeus architecture, with each sleeve tuned to its own instrument. Omega is published as a stack only.
Four strategies in one: both records are composites of the four Omega sleeves (NQ · ES · RTY · YM).
Backtest · Jan 2024 to Feb 2026
Net P&L
+$256,344
Max drawdown
-$8,263
Sharpe
4.56
Best month
+$31,878
Worst month
-$3,806
Forward test · Mar to Jul 2026
Net P&L
+$18,862
Max drawdown
-$15,474
Best month
+$12,910
Worst month
-$12,750
Green days
53%
Forward test months
+$7,604
Mar
+$6,570
Apr
-$12,750
May
+$4,528
Jun
+$12,910
Jul
THESE RESULTS ARE BASED ON SIMULATED OR HYPOTHETICAL PERFORMANCE RESULTS THAT HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE THE RESULTS SHOWN IN AN ACTUAL PERFORMANCE RECORD, THESE RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, BECAUSE THESE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THESE RESULTS MAY HAVE UNDER-OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED OR HYPOTHETICAL 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 PROFITS OR LOSSES SIMILAR TO THESE BEING SHOWN.
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS.
Poseidon
Structure based NQ day trading built on the Zeus architecture. Rebound mode trades snap-backs at stretched levels. The custom configuration runs the full model with dynamic sizing.
Out of the box settings, forward running since March 2026. The tracked forward record is being added to this page; the chart shows the backtest until then.
Backtest · Jan 2024 to Jul 2026
Net P&L
+$97,218
Max drawdown
-$18,560
Sharpe
2.10
Best month
+$10,405
Worst month
-$12,396
THESE RESULTS ARE BASED ON SIMULATED OR HYPOTHETICAL PERFORMANCE RESULTS THAT HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE THE RESULTS SHOWN IN AN ACTUAL PERFORMANCE RECORD, THESE RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, BECAUSE THESE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THESE RESULTS MAY HAVE UNDER-OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED OR HYPOTHETICAL 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 PROFITS OR LOSSES SIMILAR TO THESE BEING SHOWN.
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS.
QLiquidity
Orderflow and liquidity driven NQ execution. Development used data reaching back to 2022, and a manual trading version of the strategy has been traded since early 2023. Two custom configurations with different aggression profiles, plus the out of the box default.
QLiquidity charts reach further back than the rest of the page: the replay starts May 2022, where the archived daily settings record begins. A manual trading version has been traded since early 2023.
Developed on data reaching back to 2022; a manual trading version has been traded since early 2023. The replay starts May 2022, where the archived daily settings record begins.
Developed on data reaching back to 2022; a manual trading version has been traded since early 2023. The replay starts May 2022, where the archived daily settings record begins.
Developed on data reaching back to 2022; a manual trading version has been traded since early 2023. The replay starts May 2022, where the archived daily settings record begins. Out of the box settings, forward tested since March 2026 like the rest of the stack. The tracked forward record is being added to this page; the chart shows the backtest until then.
Backtest · May 2022 to Jul 2026
Net P&L
+$488,214
Max drawdown
-$27,718
Sharpe
3.34
Best month
+$31,055
Worst month
-$10,647
THESE RESULTS ARE BASED ON SIMULATED OR HYPOTHETICAL PERFORMANCE RESULTS THAT HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE THE RESULTS SHOWN IN AN ACTUAL PERFORMANCE RECORD, THESE RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, BECAUSE THESE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THESE RESULTS MAY HAVE UNDER-OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED OR HYPOTHETICAL 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 PROFITS OR LOSSES SIMILAR TO THESE BEING SHOWN.
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS.
QORB
Opening range breakout on NQ with confidence gating and a daily loss limit.
THESE RESULTS ARE BASED ON SIMULATED OR HYPOTHETICAL PERFORMANCE RESULTS THAT HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE THE RESULTS SHOWN IN AN ACTUAL PERFORMANCE RECORD, THESE RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, BECAUSE THESE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THESE RESULTS MAY HAVE UNDER-OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED OR HYPOTHETICAL 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 PROFITS OR LOSSES SIMILAR TO THESE BEING SHOWN.
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS.
Zeus
The original architecture, and the longest live record in the stack. Zeus has run live on TradingView and NinjaTrader since January 2024 and never stopped; QZeus 2.0, the production rebuild, began 2.0 based forward testing in June 2025. Omega and Poseidon are built on the Zeus architecture.
The backtest portions of these records are QZeus 2.0 replays of each configuration's legacy preset. QZeus 2.0 entered 2.0 based forward testing in June 2025.
Backtest segment is a QZeus 2.0 replay of the equivalent legacy preset with the Direction Classifier enabled at Low confidence, a 2.0 feature. The forward record ran on TradingView with NinjaTrader execution, without the classifier.
The original QZeus record on NQ, running on TradingView with NinjaTrader execution since January 2024 and still running today. Actual trading record; martingale ladder sizing from 1 up to 16 contracts; commissions not included. The record ends the window at its equity high.
TradingView record · Jan 2024 to Jul 2026
Net P&L
+$162,510
Max drawdown
-$20,420
Sharpe
2.31
Trades
411
Best month
+$21,025
Worst month
-$11,215
THESE RESULTS ARE BASED ON SIMULATED OR HYPOTHETICAL PERFORMANCE RESULTS THAT HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE THE RESULTS SHOWN IN AN ACTUAL PERFORMANCE RECORD, THESE RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, BECAUSE THESE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THESE RESULTS MAY HAVE UNDER-OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED OR HYPOTHETICAL 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 PROFITS OR LOSSES SIMILAR TO THESE BEING SHOWN.
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS.
Advanced strategies
Two strategies in the 2.0 lineup are built for power users. Both expose a deep configuration surface, so results depend on how you set them up. Import a profile, or tune your own.
QPilot 2.0
Automated index futures execution with wide control over sizing, session windows, and exit behavior. Highly configurable, made to be shaped to your own risk profile.
QKronos EVO
The newest strategy on the platform. Timing and risk parameters are fully tunable for traders who want complete control of how it behaves.
Overfitting check
An overfit system memorizes its development data. The backtest looks clean, then the edge decays as soon as real trading starts, because the patterns it learned were noise. This record splits into a development window (January 2024 through February 2026) and a live window (March through July 2026), so the question can be tested directly on the daily P&L of the same six configurations. If the stack were overfit, the live window would be measurably worse than the window it was built on. It is not.
Live mean daily P&L vs dev
102%
Welch t-test p (means)
0.94
KS test p (distributions)
0.37
Sharpe retained live
86%
Development backtest Jan 2024 to Feb 2026
Live forward run Mar to Jul 2026
Business days
564
110
Days traded
530
106
Total P&L
+$1,070,150
+$213,440
Mean daily P&L
+$1,897
+$1,940
Daily std dev
$4,808
$5,736
Sharpe (daily, √252)
6.26
5.37
Winning days
73%
64%
Avg winning day
+$3,963
+$4,972
Profit factor
3.29
2.71
Max drawdown
-$30,326
-$24,239
Both windows are the same six configurations at their tracked contract sizes, daily P&L zero filled on business days. Sharpe is computed the same way in both columns.
Every contiguous 110 business day window of the development backtest, totaled (455 windows), with the live forward window marked. A record that decayed out of sample would sit in the left tail. The live window lands at the 56th percentile, near the middle of what the development record produced.
Welch's t-test asks whether the live mean daily P&L differs from the development mean. At p = 0.94 there is no detectable difference; the live mean came in at 102% of development. A two-sample Kolmogorov-Smirnov test compares the full shape of the two daily P&L distributions. At p = 0.37, no detectable difference there either. The live window kept 86% of the development Sharpe (5.37 against 6.26). Published studies of backtest overfitting find that strategies tuned to their development data typically give back half or more of their backtest Sharpe once they trade out of sample (Bailey, Borwein, López de Prado and Zhu, 2014; Harvey and Liu, 2015). This stack gave back 14%. Ranked on Sharpe, the live window sits at the 22nd percentile of the development windows, below the median and inside the ordinary range.
The match is not perfect, and the mismatches are reported on purpose. Daily volatility ran 19% higher live (Levene test p = 0.04), the share of winning days fell from 73% to 64%, and profit factor fell from 3.29 to 2.71. The daily edge held anyway because live winning days averaged +$4,972 against +$3,963 in development. Fewer, larger wins with the same mean change the texture of the equity curve and leave the edge intact.
Second check, same dates on both records
From March through July 2026 the six 2.0 native configurations also carry two records over identical dates: the live forward run, tracked day by day, and the backtest engine replayed across the exact same window. Correlating the two daily P&L streams measures whether the engine that produced the development record behaves like live trading.
Daily P&L correlation (r)
0.96
R²
0.92
Shared trading days
106
Total P&L gap
-3.4%
2.0 configurations, identical population in both records: forward +$213,440 vs backtest +$221,062 over the window.
Backtest, replayed over the same datesLive forward run
Each dot is one trading day: how the stack scored in the backtest that day (x) against the live forward run (y). Days on the dashed line matched exactly.
Correlation by configuration
Omega Stack
0.83
Poseidon NQ · Custom
0.94
Poseidon NQ · Rebound
0.78
QLiquidity NQ · Config A
0.97
QLiquidity NQ · Config B
0.97
QORB NQ · Custom
1.00
Method: daily P&L from both records, zero filled on days either record did not trade, Pearson correlation across the union of trading days in the shared window. The backtest trade logs covering this window are downloadable on the configuration cards.
THESE RESULTS ARE BASED ON SIMULATED OR HYPOTHETICAL PERFORMANCE RESULTS THAT HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE THE RESULTS SHOWN IN AN ACTUAL PERFORMANCE RECORD, THESE RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, BECAUSE THESE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THESE RESULTS MAY HAVE UNDER-OR OVER-COMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED OR HYPOTHETICAL 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 PROFITS OR LOSSES SIMILAR TO THESE BEING SHOWN.
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS.
How to read these numbers
One timeline, one curve. Each record chart is a single continuous curve. The record behind it combines a backtest, replayed on the same engine that powers customer backtests inside ATS 2.0, with actual forward trading results tracked from March 1, 2026. Backtests are simulated; the forward records are actual executed trades. Backtests cover January 2024 through February 2026, except QLiquidity, whose replay reaches back to May 2022, and Poseidon Rebound, whose replay starts March 2024. Each backtest runs the configuration's own settings, and where a backtest's settings differ from what its forward record ran, the card note says so: the Zeus Intermediate backtest adds the Direction Classifier at Low confidence, a 2.0 feature the TradingView era record did not have. The charts do not split the records visually; the hero chart and the timeline under it mark the switchover and the earlier platform eras, and the chip clusters on the cards give each record's stats separately. Zeus has been forward tested since January 2024, on TradingView and NinjaTrader before QZeus 2.0. Backtest trade logs cover the full window through July 2026, including the months where forward tracking ran in parallel, so the two records can be compared over the same dates.
Blue is profit, orange is loss. Every figure in the month by month table is colored by outcome. The header tags and the table footnote identify which cells come from the simulated backtest record: the January and February months that predate live tracking, and the default configurations awaiting published forward results.
The overfitting check is the proof section. Above this section, the daily P&L of the six 2.0 native configurations (Omega Stack, Poseidon Custom and Rebound, QLiquidity Config A and B, and QORB Custom; Zeus is excluded because its forward record ran on another platform) is tested for out-of-sample decay: development window against live window, plus the live run against the engine replayed over the same dates. The statistical tests and their p-values are shown there, not asserted.
Contract sizing. Daily P&L is tracked at the configuration's listed contract size: 1 lot except where the configuration name says otherwise, such as the Zeus 4 lot ladders. No compounding, no reinvestment.
Settings are copy and paste. Every configuration shown ships as a settings profile that imports into ATS 2.0 with a copy and paste code. No manual parameter entry, and no strategy internals are exposed.
Test on demo first. Run any strategy on a demo account for at least 60 days before trading it live, whether it is new to the community or new to you.
Run the same stack
Every configuration on this page imports into ATS 2.0 in a couple of clicks. Setup takes one onboarding call.
