Dynamic 7-Tier Benchmark · The Two Questions
Deviation classes
5
Comparison bands
P5–P95
Review slot
MONTHLY
How MARS uses this
Both benchmark questions — inside the envelope? and why? — are answered on this overlay. Position inside the band answers the first at a glance; the shape of how live equity tracks, hugs, or diverges from the median is where the second question begins.
How it benefits you
You stop grading months by feel. A flat quarter inside the band stops causing panic edits, a hot streak at P85 stops justifying size increases, and genuine model breaks get caught by classification instead of by pain.
Question one is position. Question two is the shape of the tracking — both live on the same overlay.
The key idea
In practice
Equity percentile fan across 50,000 simulated paths. Violet is P10-P90, blue is P25-P75, the gold dash is the median, and the green line is live equity drawn against the envelope.
The interpretive boundary
The standard names the qualifier directly: speed is not a virtue when it arrives with excessive drawdown, higher lock risk, casual override behaviour, or abnormal tier deployment. This is the sentence that makes the second question necessary at all. Without it, the comparison collapses into an equity race, and an equity race has exactly one winner every month regardless of how the win was purchased.
Order of cross-checks
The workflow reads position before conclusion. Live month-end equity, monthly return, and max drawdown come out of Compliance Panel 3 and get placed against the comparison bands. Then drawdown is cross-checked against its own distribution, gate dwell against expected time in each gate, tier behaviour against the modelled deployment, and open exposure and overrides against the stress tables. Any one of those four can turn a strong equity reading into a governance finding.
Provisional readings
03The QA checklist before any benchmark use requires that live data is complete for the period, that live labels match benchmark periods, that drawdown sign convention is consistent, and that target-capped equity is understood. Many high percentile values flatten near the target because paths stop at completion, so a P95 band that looks suspiciously level is the cap, not a modelling error. Comparing three weeks to a monthly band without saying so is the most common way a review reaches a confident wrong answer.
Where the answer is written
Each review ends by recording a verdict in one of a fixed set: within benchmark, outperforming, underperforming, drawdown adverse, throttle inefficient, or model mismatch. A closed vocabulary matters more than it looks. It makes verdicts countable across months, so a run of four consecutive underperforming verdicts becomes visible as a pattern rather than four separate months that each felt survivable at the time.
Reference
| Live pattern | Benchmark read | Correct response |
|---|---|---|
| Equity above median, drawdown normal | Clean outperformance | Continue. Document which branches, regimes, or throttle states are adding it |
| Equity above median, drawdown worse than P90 adverse | Fast but risky | Do not celebrate. Audit sizing, overrides, gate compliance, open exposure |
| Equity below median but above P25, drawdown normal | Normal underperformance | No panic. Review EV, fees, branch mix. Avoid forced catch-up |
| Equity P10 to P25, drawdown elevated | Weak and risky | Diagnostic review. Reduce aggression if gate, throttle, and SDE support caution |
| Equity below P5 | Outside the expected band | Formal audit: journal accuracy, branch labels, fees, adherence, assumption mismatch |
| Drawdown worse than the 95th adverse band | Capital-defence warning | Respect gate contraction. No override to recover |
| Extended dwell in Buffer or Floor | Defensive dwell exceeds benchmark | Check for clustered losses, hostile volatility regime, or loose tier usage |
| T6 or T7 used frequently | High-tier behaviour abnormal | Confirm structural permission; reduce tier aggression if unjustified |
Before you go deeper
Connected inside MARS
Go deeper
Deep dive — 01
Inside the envelope or not — and if not, what kind of gap? Everything else is supporting evidence.
Read the full brief →
Deep dive — 02
Which branch, which exit style, which failure mode, which throttle state — four tables, no adjectives.
Read the full brief →
Deep dive — 03
The upper bands flatten at target. Locked paths are not slow paths. Sign conventions are not cosmetic.
Read the full brief →
One price. Eleven workbooks, three TradingView indicators, and the full manual library — $497.
