Choose the asset class, then choose the tickers.
Choose the market and symbols you want to study, then apply them to Atlas. Market data stays protected, and Atlas uses it only for the research you request.
Asset class
universeFutures contracts
2 selectedSelected universe
selectionRegime command matrix
expectancy x drawdown pressure| Context | Trades | Expectancy | PF | Max DD | Verdict |
|---|---|---|---|---|---|
| VR aligned + risk-on | 318 | +$91 | 1.86 | -4.2% | Keep |
| NQ continuation | 221 | +$124 | 2.08 | -3.8% | Core |
| Liquidity expanding | 286 | +$58 | 1.52 | -5.1% | Trade |
| Afternoon reversal | 176 | +$9 | 1.06 | -6.4% | Resize |
| Volatility expansion | 144 | -$37 | 0.78 | -8.9% | Avoid |
| VR not aligned | 103 | -$62 | 0.64 | -9.8% | Block |
High expectancy, clean profit factor, and the shallowest simulated drawdown pocket.
Small trade count, but a large share of left-tail outcomes and challenge breach risk.
Expectancy falls from +$42 to +$31, but breach risk improves. That argues for real edge with sizing fragility.
Risk containment and regime filtering are higher leverage than adding complexity.
ATLAS can explain computed results, but the tool runner owns the numbers. Research output only, not financial advice.
A command center for OSIRIS, ZENITH, and APEX: compare behavior, run one what-if across all models, inspect fit, and jump into sealed scenario tools. Atlas gets aggregate outputs only: no signals, entries, exits, proprietary logic, or future predictions.
OSIRIS leads win rate and per-trade return; ZENITH is the steadiest.
For absolute per-trade return, OSIRIS and APEX lead. For drawdown control, ZENITH has the tightest tail. APEX carries the highest average return but the widest variance.
I'm Atlas — your research and model-building partner.
Tell me what you want in normal words; you do not need trading or coding jargon. I can explain the terms, propose clear rules, and show exactly what would be tested. Nothing runs or saves until you approve it. Checking whether backtesting is available for this account. Research only — not financial advice.
Turn a market idea into a tested model.
Choose a verified source, then describe the strategy in plain English. Atlas can shape and save any clear idea; backtest-ready templates can return trades, P&L, an equity curve, and validation.
- Source
- Plan
- Implement
- Backtest
- Validate
Choose and configure a source in Data Vault, then press Apply to Atlas.
Review your closed-beta access
A completed backtest is required before this option can appear.
Bring Atlas your trades or model, or start from a blank brief.
Start with evidence you already own. Atlas calculates results from your data before explaining them.
Day trade pack: intraday expectancy, time-of-day edge, loss streaks, stop discipline, session heat, and rule breach pressure.
Swing trade pack: overnight gap risk, holding-period expectancy, market regime alignment, weekly drawdown pressure, and event exposure.
Multi-hour pack: 2-8 hour holds, session transitions, afternoon risk, trend persistence, signal decay, and drawdown tolerance.
Prop account pack: pass probability, trailing drawdown risk, consistency rule exposure, daily loss cap, and payout-readiness plan.
What to trade, when to stop, and how to limit damage.
Buckets trades by session, result, streak, and rule pressure.
Stops weak sessions from becoming account-level drawdowns.
Max trades, stop time, size rule, and next validation window.
Day trade decision map
session x loss-cap impactRule impact ladder
deterministic what-ifQuestions worth asking
ranked by decision valueStyle packs change the questions, charts, and risk assumptions without letting the AI invent metrics or write production paths.
The same expectancy survives, but bad paths arrive in streaks that need daily containment.
Compare expectancy cost against breach reduction before the user risks capital.
Trades resampled i.i.d. from your 1,248-trade history. A model, not a guarantee — not financial advice.
| Preset | Median | P5 | Breach | Action |
|---|---|---|---|---|
| Base strategy | $27.3k | $12.1k | 17% | watch |
| 2.4R daily cap | $26.8k | $14.6k | 9% | run |
| Block VR-off vol | $25.9k | $15.1k | 8% | run |
| Aggressive size-up | $34.2k | $8.4k | 24% | avoid |
Challenge path simulator
target / guardrail / breach route| Rule | Current | Pressure | Fix | Impact |
|---|---|---|---|---|
| Max drawdown | -9.8% | High | Block VR-off vol expansion | +6pp |
| Daily loss | -3.8R | Medium | 2.4R daily stop | +8pp |
| Consistency | 44% top decile | Medium | Trim NQ top decile | tail |
| Target pace | 31 days | Low | Do not increase size yet | stable |
Estimated from the resampled Monte Carlo against the stated rule set — not financial advice.
NQ continuation when VR aligned
Core edge bucket: +$124 expectancy, 2.08 profit factor, and shallowest simulated drawdown.
Afternoon reversal tags
Positive but thin. Drop to 0.65x size unless the broader regime is aligned.
VR-not-aligned volatility expansion
This bucket creates a small count of trades but dominates left-tail losses and challenge breach odds.
Unseen month and walk-forward week
Do not increase size until the NQ continuation edge holds on fresh data outside this upload.
Improvement impact map
before / after tail containment| Decision | Rule | Impact | Next test |
|---|---|---|---|
| Keep | NQ + VR aligned only | core edge kept | Monitor breadth |
| Resize | Afternoon reversal to 0.65x | less tail drag | Session heat |
| Block | VR-off volatility expansion | -8pp breach | Monte Carlo rerun |
| Validate | 30d unseen window | required | Walk-forward |
The first improvement is not a new signal. It is containment: keep the strongest regime bucket, remove the worst regime bucket, and add a loss cap before paying for an evaluation attempt.
Risk rule frontier
pass odds vs breach cost| Rule set | Expectancy | Pass odds | Breach | Verdict |
|---|---|---|---|---|
| Current | +$42 | 71% | 17% | risky |
| Daily cap only | +$40 | 79% | 11% | good |
| Cap + block VR off | +$37 | 82% | 9% | best |
| Aggressive size-up | +$58 | 54% | 24% | avoid |
Challenge calculations are simulations from uploaded trades. Research only, not financial advice.