📦 Bilbo Box — Pareto Hill-Climb

Multi-objective optimization of the Bilbo Box breakout mechanics: maximize mean P&L while minimizing max drawdown, over 153 evaluated configurations on paired zero-drift Monte Carlo panels.

Forward window 2026-09-25 to 2026-10-22 · 40 screening paths/seed 1234 · 150 validation paths/fresh seed 2233 · 98.7 min compute

⛔ NOT FUNDABLE

The hill-climb pick

$281mean P&L / path (screen)
11.6%mean max drawdown
45%P(profit)
3.2trades / path

Winning configuration (diffs vs the author's baseline):

Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.15 ATR run mult=0.75 DTE target=21 Time cap (days)=10

Honest read: zero-drift mechanics optimization cannot manufacture edge. A fundable verdict requires positive out-of-sample expectancy on real data. This optimization tunes mechanics under a random walk — entry/exit/sizing geometry — not predictive edge. Every configuration, including the pick, was evaluated under zero drift.

Pareto frontier

Non-dominated configurations across all 153 evaluations. No frontier point is better on both axes than any other.

9.5%$-52711.1%$-29612.7%$-6514.3%$16615.9%$397 mean max drawdown % (lower is better →) mean P&L $ per path (higher is better ↑)
hill-climb pick frontier point author baseline

Frontier table

Config (vs baseline)Mean P&LMedianMean DDP(profit)Kill-swTrades
Trend gate=band Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.15 DTE target=21 Time cap (days)=10$347$8112.5%50%38%3.6
Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.15 ATR run mult=0.75 DTE target=21 Time cap (days)=10$281$-8211.6%45%30%3.2
Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.15 Strike ATR mult=1.0 DTE target=21 Time cap (days)=10$257$011.3%48%22%3.1
Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.5 DTE target=21 Time cap (days)=10$207$3510.5%52%18%3.3

Fresh-seed validation (150 paths)

Top-8 frontier points + baseline re-run on unseen panels. If the pick only won on the screening seed, it shows up here.

Config (vs baseline)Mean P&LMean DDP(profit)Kill-sw
Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.5 DTE target=21 Time cap (days)=10$-25210.5%23%17%
Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.15 Strike ATR mult=1.0 DTE target=21 Time cap (days)=10$-25411.4%21%23%
Trend gate=band Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.15 DTE target=21 Time cap (days)=10$-27713.6%25%39%
Kill-switch DD %=15.0 Signal end (ET)=14 Trail keep frac=0.15 ATR run mult=0.75 DTE target=21 Time cap (days)=10$-31011.9%19%28%
author baseline$-47714.9%19%7%

Author's published results (reference)

Milkman's live paper log as of 2026-09-24 — the real-money-shaped anchor this research is measured against:

+66.79%since $10k start (Jul 9)
41%win rate (46 closed)
1.72profit factor
-18.25%max drawdown
50/128signals taken

Source: milkmantrades.com/bilbo-paper.html daily log. Our zero-drift simulation cannot reproduce his edge — it only stress-tests mechanics.

Parameter space searched

Trend gatestrict, off, slope, band, halfsize
Target premium %2.0, 3.0, 4.0, 5.0, 6.0
Kill-switch DD %15.0, 20.0, 25.0, 30.0
Spread cap0.03, 0.04, 0.05, 0.06
Signal start (ET)9, 10
Signal end (ET)14, 15, 16
Trail keep frac0.15, 0.25, 0.35, 0.5
ATR run mult0.75, 1.0, 1.25
Strike ATR mult0.5, 0.75, 1.0
DTE target21, 28, 35
Time cap (days)10, 14, 21

Greedy Pareto walk from the author baseline: each round evaluates every single-parameter neighbor on identical panels; moves only to a Pareto-dominating neighbor (noise epsilons: $10 P&L, 0.2pp DD). Climbed 6 rounds.

Assumptions (stated, not hidden)

⬇ Full run JSON (configs, all evaluations, frontier, validation)

Research artifact · engine: bilbo.engine.Engine (episode-corrected 2026-09-22) + config tuning knobs · generated 2026-09-29 00:55 ET · not investment advice