Find Your Next Winning Strategy In Minutes, Not Months
We Used Auto Strategy Builder To Pass A Real $50K Apex Eval.
Sam built a strategy from scratch in ASB and ran it for 30 days on a live Apex 50K evaluation account. The bot took every trade, no manual entries. Sam re-optimized once a weekend (a few minutes' work) so the strategy kept up with market conditions. Account passed. Here is the full walkthrough.
"I had blown three evals in a row hesitating on entries. I built one strategy in ASB, set the weekly re-opt, and just let it run. No screen time, no tilt. Funded in 30 days."
Sam, co-founder, TradeDevils
Watch the full webinar above before you commit. No credit card to trial.
Most evals get blown by the trader, not the strategy.
You find a setup that looks profitable. You scroll back two weeks, count winners, convince yourself it's an edge. You take it live, hesitate on the first entry, miss the overnight move, take the revenge trade after a loss, and bleed the account. The strategy wasn't bad. The hands on the mouse were the bottleneck.
Auto Strategy Builder removes those hands. You point it at any indicator you already trust, it builds and stress-tests thousands of rule combinations, kills the fragile ones, and exports a native NinjaScript bot that runs the result for you. No coding, no babysitting fills, no revenge trades at 2 a.m.
There's also a second way evals die: the strategy was fine, but the firm's rules killed it. A trailing drawdown that locks at balance, a consistency cap that penalizes one big winner, an EOD flatten you forgot to wire. As of v1.2.6.6, you tell ASB which firm and account you're on and the optimizer rejects anything that would have blown those rules. The leaderboard you see is the leaderboard you can actually trade.
That is what Sam used to pass the Apex eval above. The full process is in the recording.
Why hand-tuning indicators rarely holds up live.
The settings that look perfect on yesterday's chart usually break by next month. Three reasons that happens, and what ASB does about each.
You can't test enough
A real edge needs thousands of parameter combinations across years of data, then validated on data the optimizer never saw. Doing that by eye on a chart is impossible. ASB does 40,000 combinations in roughly 10 minutes on a normal laptop.
You curve-fit without knowing it
Settings that look perfect for the last 30 days usually break in week two of live trading. ASB runs walk-forward analysis (train on the first part of history, test on data the optimizer never touched) and Monte Carlo (shuffle trade order, add random slippage) so you find this out before you risk capital, not after.
You can't tell fragile from real
If your strategy breaks when you change a 20 EMA to a 22 EMA, it was never an edge. ASB shifts every parameter by plus/minus 25 percent across 20 variations and gives you a Rock-Solid to Fragile score. Only the robust ones survive to the export step.
Build, test, and deploy without writing a line of code.
Strategy leaderboard and equity curves, generated by ASB in one run.
Works with any indicator you already own.
If it loads in NinjaTrader, ASB can use it. EMA, MACD, your favorite paid order-flow tool, an indicator your developer wrote you two years ago. It reads the plots, tests thousands of entry and exit combinations against the data, and ranks the survivors on a leaderboard.
- Built-in, custom, and third-party NT8 indicators.
- Triggers on specific plot values (lines, dots, bands).
- Up to 40,000 rule variations per run, finished in minutes.
Tell it what kind of trader you want the bot to be.
Pick a behavioral profile before you run the optimizer and ASB constrains the rules to match. Useful when you already know your risk tolerance and how often you want the bot in the market.
- Trend following: larger wins, wider trailing stops, fewer trades.
- Mean reversion: high win-rate snap-back trades, tighter targets.
- Scalping: short hold times, lots of trades, small targets and tight stops.
From leaderboard to live in one click.
When you've picked a strategy, ASB writes it as a native NinjaScript file. You don't get a black-box plugin or an obfuscated wrapper. You get a real, readable C# strategy you can drop on a chart, run in the strategy analyzer, or hand to a developer if you want to tweak it.
- Clean, readable, execution-ready NinjaScript.
- Session timing and prop-firm trade rules handled automatically.
- You own the code. Cancel your subscription, the strategies you generated still run.
The strategy passed every backtest. Then the trailing drawdown killed you.
You found something that looked clean. Sharpe was good, Monte Carlo was green, equity curve climbed. You loaded it on your Apex 50K and on day two a normal losing afternoon dragged the trail to balance and the account was done. The strategy wasn't the problem. The firm's rules were never in the fitness function.
Tell ASB which firm and which account size you're trading. The optimizer stops chasing pretty equity curves and starts rejecting anything that would have busted you on the actual rules: trailing drawdown, daily loss limit, consistency rule, EOD flatten, contract caps, automation policy. The leaderboard you read is the leaderboard you can actually trade.
- Works with Apex, Topstep / TopstepX, MyFundedFutures, Bulenox, Tradeify, TradeDay, and 9 more, all account sizes, rules refreshed as the firms update them.
- Same firm-aware optimization is wired into ASP's ATM tuner. The strategy you trust keeps trusting itself when you move it to a different funded account.
- Stop building strategies that beat the market and lose to the rulebook.
How Auto Strategy Builder Compares
A side-by-side look at the most popular auto/visual strategy tools for NinjaTrader traders.
| Capability | Auto Strategy Builder OURS | Ninza Strategy Builder | Bloodhound / Blackbird | StrategyQuant X | Build Alpha | NinjaTrader Strategy Builder |
|---|---|---|---|---|---|---|
| Runs natively inside NinjaTrader 8 | Yes | Yes | Yes | NoStandalone desktop app | NoStandalone desktop app | Yes |
| Auto-generates strategies (Genetic Algorithm) | YesUp to 40,000 combos in ~10 min | NoManual rule builder | NoVisual logic designer, not auto-gen | Yes | Yes | No |
| Rejects strategies that would blow your prop firm's rules ONLY ON ASB | Yes15 firms (Apex, Topstep, MFF, Bulenox, Tradeify, TradeDay, +9 more). Trailing DD, daily loss, consistency rule, EOD flatten, contract caps, automation policy — all enforced in the GA, and rechecked against every Monte Carlo permutation so a single reshuffle that breaches survival rules fails the candidate. | No | No | NoGeneric risk filters, no firm-specific rules | NoGeneric risk filters, no firm-specific rules | No |
| Walk-Forward Analysis | Yes | No | No | Yes | Yes | No |
| Monte Carlo robustness testing | Yes8-metric 0–100 dashboard | No | No | Yes | Yes | No |
| Selection-bias filter (Vs-Random edge test) ONLY ON ASB | Yes500 same-risk random strategies per candidate, real must beat the top 20% to survive. Rejection, not a report. | No | No | PartialStatistical reporting, not a rejection gate | PartialStatistical reporting, not a rejection gate | No |
| Parameter stress test (±25% shift) | YesRock-Solid → Fragile scoring | No | No | PartialVia optimizer, manual setup | Yes | No |
| Out-of-sample incubation period | YesUnseen data held out by default | No | No | Partial | Yes | No |
| Style-based generation (trend / mean-reversion / scalping) | YesBehavioral scoring per style | No | No | No | No | No |
| Works with any NinjaTrader indicator (built-in, custom, 3rd-party) | YesAuto-detects params & plots | Partial | Yes | NoIts own indicator library | NoIts own signal library | Yes |
| Exports NinjaScript (C#) ready for live trading | Yes | Yes | Yes | PartialVia plugin, often rewrite needed | Partial | Yes |
| Built-in news-event blackout in exported strategies ONLY ON ASB | YesAuto-skips entries around scheduled high-impact news, configurable window before and after. Critical for prop firm accounts that prohibit news trading. Baked into the exported .cs — no external dependencies. | No | No | PartialNews filter in their app, not in the NinjaScript export | PartialNews filter in their app, not in the NinjaScript export | No |
| 11 optimization metrics (Sharpe, Sortino, MAR, K-Ratio, …) | YesCustom weighted combos | No | Partial | Yes | Yes | No |
| Automatic overfitting / quality filters | YesOOS/IS ratio, WF variance, duplicate detection | No | No | Yes | Yes | No |
| No-code interface | Yes | Yes | Yes | Yes | Yes | Yes |
| Learning curve | LowPick indicators, click Start | Medium | HighNode-graph logic | High | High | Medium |
| Pricing | $149 / monthCancel anytime | ~$400 one-time | ~$800–$1,200 + add-ons | ~$990–$3,990 | ~$2,499–$3,999 | Free with NT8 |
Comparison reflects publicly documented features at time of writing. Pricing for third-party products is approximate and may change. Check vendor sites for current figures. Trademarks belong to their respective owners.
Other builders ship the best-looking backtest. ASB ships the one that survives.
Strategy generators are not new. SQX, Build Alpha and EA Studio have been around for years and they do real work. Where ASB is different: we treat robustness as a hard rejection gate, not as an optional report you read after the fact. If a candidate strategy can't pass these checks, you never see it on the leaderboard.
Walk-forward, not just backtest
Every candidate is trained on one slice of history and then tested on a slice the optimizer never saw. If it only worked on the data it learned from, it gets rejected. No exceptions, no manual override.
Monte Carlo trade shuffling
Run the same trades in a different order with random slippage and the equity curve has to still hold up. This catches strategies that only profited because the wins happened to land before the losses.
Parameter stress, plus or minus 25%
Every input gets shifted by up to 25 percent across 20 variations. If a 20 EMA only works as a 20 EMA, the strategy was never an edge, it was a coincidence. Fragile candidates are flagged and dropped before export.
Out-of-sample incubation lock
You can ring-fence the most recent slice of data (typically the last 10 percent) so the optimizer literally cannot touch it. The bot then has to perform on that untouched window before it gets approved as a finalist.
Golden Ticket Score
A composite robustness score that combines mean return, performance versus buy-and-hold, equity-curve linearity (R-squared), and downside deviation, multiplicatively. If any one component is bad, the whole score collapses to zero. Lucky outlier trades, lumpy equity curves, or strategies that can't even beat sitting in the index all get caught here.
Duplicate strategy penalty
Every candidate's daily equity curve is correlated against everything already on the leaderboard. If the correlation crosses 70 percent, fitness is penalized up to 30 percent. The leaderboard ends up with genuinely different edges instead of twenty mutations of the same strategy with slightly tweaked lookbacks.
Vs-Random edge test
For every candidate we generate 500 synthetic "random" strategies with the same trade sizes but random direction. The real strategy has to land in the top 20 percent of that distribution to survive. If it can't beat a coin flip with its own risk profile, it gets cut. Catches the curves that look like edge but are statistically noise.
Prop firm rules vs every Monte Carlo path
The prop firm survival check doesn't just run on one trade order. It re-runs on every Monte Carlo permutation. A single trailing-drawdown or daily-loss breach in any reshuffled run flips the candidate to fail. So the leaderboard only contains strategies that survive the firm's rules across every plausible sequence the live market could throw at them, not just the lucky historical one.
A strategy must pass every one of these before it shows up on the leaderboard. If it doesn't, you never see it. That isn't gatekeeping. It's the whole point of building strategies you can actually risk capital on.
The Nerd Vault
Want every parameter, every metric, every option spelled out?
The Auto Strategy Builder doesn't just run simple backtests. It utilizes computational stress testing to programmatically quantify edge stability.
- Walk-Forward Analysis (WFA): Slices historical data into discrete training and testing windows, ensuring the generated logic works on "blind" Out-Of-Sample (OOS) data before rendering a positive score.
- Monte Carlo Sequence Randomization: Runs hundreds of simulations shuffling the precise order of winning and losing trades, proving that account survival isn't dependent on a mathematically "lucky" sequence.
- Parameter Stress Distribution: Shifts every optimized input variable by +/- 25% simultaneously. If the profit curve collapses, the strategy is flagged as curve-fit and instantly rejected.
- Incubation Isolation: Optionally locks the most recent percentage of market data (e.g., 10%) as a final, untouchable proving ground for the generated bot.
-
v1.3.0.0 (Latest)
- New Half-Hour PnL tab for every leaderboard strategy: 48 columns covering 00:00 to 23:30, green for profit and red for loss, so you can see which half-hour windows your strategy actually makes money in.
- News blackout filter on every exported strategy: skip entries around scheduled high-impact news events, with a configurable window before and after. Stops your deployed strategies from getting chopped up by news-driven volatility.
- Vs-Random filter: each candidate is now tested against random strategies with the same risk profile. Only strategies with a real, measurable edge over noise make the leaderboard, so lucky-looking equity curves that won't replicate live get filtered out before you ever see them.
- Prop Firm verdict now checked against every Monte Carlo permutation, not just the baseline run. A single survival-rule breach (trailing drawdown, daily loss limit) in any permutation flips the candidate to fail, so the leaderboard only contains strategies that survive prop firm rules across the full distribution of plausible trade orderings.
- Underwater chart — how deep and how long the equity sat below its highest point. Critical for prop firm trailing-drawdown accounts: catches strategies that look profitable overall but spend weeks at a time in a deep hole.
- Walk-Forward Efficiency chart — side-by-side IS vs OOS profit factor per window. Instantly see which training windows held up on unseen data (green) and which collapsed (red), without reading three pages of metrics.
- Rolling Sharpe + Profit Factor charts — 30-trade sliding window of both metrics across the strategy's lifetime. A trending-down line means the edge died mid-period even if the aggregate numbers still look fine.
- Trade P&L Histogram — distribution of per-trade profits. A broad mound = many small wins (robust). A tall spike at the right = one or two giant trades carried everything (fragile).
- Vs-Random Fan chart — your equity drawn on top of the percentile bands of 500 fake strategies with the same trade sizes but random direction. Above the blue band = a real edge; inside it = indistinguishable from luck.
- Calendar Heatmap — PnL grouped by weekday and hour. Spots concentrated edges ("all the profit is Tuesday morning") that decay fast once the regime shifts.
- MAE/MFE Scatter — per-trade max pain vs max gain before exit. Green dots in the upper-left = you let winners run and cut losers fast. Lots of dots near the stop = stops parked in the noise.
- Strategy Correlation chart — Pearson correlation between the selected strategy and every other leaderboard entry. Low bars = unique diversifier. Tall bars = you have duplicates in your top 100.
- Plain-English grades on every new chart — each of the eight new charts now shows a Poor → Outstanding rating with a one-line interpretation of the numbers, using the same color scale as the existing Monte Carlo and Parameter Stress Test charts. No quant background needed to read them.
- End-of-session boundary fix: trades held to session close no longer mis-stamp into the wrong half-hour bucket.
- v1.2.6.6 - Prop Firm Mode: pick your firm and account size, the optimizer rejects anything that would have busted the trailing drawdown, daily loss limit, consistency rule, EOD flatten or contract cap. 15 firms covered (Apex, Topstep, MFF, Bulenox, Tradeify, TradeDay and more), all account sizes, kept current as the firms change rules. Same engine wired into the ASP ATM Optimizer. Also: long optimization runs no longer leak memory across generations.
- v1.2.6.0 - Added Parameter Stress Test to catch curve-fitting. Implemented Leaderboard filtering, advanced Enum parameter support, and 2.5x faster generation via Pre-sampling optimization.
- v1.2.5.1 - Introduced Strategy Styles (Trend Following, Mean Reversion, Scalping) with unique exit mechanisms. Added Trailing Stop logic and 0-100 Style Scoring.
- v1.2.5.0 - Added Profit Consistency Monte Carlo metric, mandatory/optional indicator constraints, and same-bar trade prevention logic. Fixed decimal handling for standard deviations.
- v1.2.4.6 - Initial core release featuring Genetic Algorithm engine, Walk-Forward Analysis, and NinjaScript export.
Frequently Asked Questions
The questions traders ask most before buying ASB. If yours isn't here, ask us in Discord.
About Auto Strategy Builder
Using ASB
Strategies & Real Results
Since that webinar, we've gone one step further. In v1.2.6.6 you tell the optimizer which firm and account size you're trading (Apex, Topstep, MyFundedFutures, Bulenox, Tradeify, TradeDay and 9 others), and it rejects any strategy that would have busted the actual rules: trailing drawdown, daily loss limit, consistency rule, EOD flatten time, contract caps, automation policy. The leaderboard you read is the leaderboard you can actually trade on your funded account.
Trial, Pricing & Licensing
Ready to let the bot do the trading?
Stop guessing. Start executing.