Inconclusive · Sizing · 2026-10-10

What would a 1%-risk paper account have done?

What this means

Taking every 1h setup in the whole universe at 1% risk each 'ruined' the paper account by October 2022. The saved summary shows why: so many trades were open at once that median open risk was 84% of equity. Scaled down to the live 50/20/20 mix, the sim compounds at 33% a year with a −24% worst drawdown, or 16% and −35% after a cost of 0.1R per setup. Exit dates are reconstructed and drawdowns use closed trades only, so these figures are rough.

Explanation checked against the data.

By timeframe

Timeframe	Verdict	n	Result
15-minute	Not tested: account-level sizing simulation of the live 1h/4h/1d mix, not a per-timeframe signal test
1-hour	Not tested: account-level sizing simulation of the live 1h/4h/1d mix, not a per-timeframe signal test
4-hour	Not tested: account-level sizing simulation of the live 1h/4h/1d mix, not a per-timeframe signal test
Daily	Not tested: account-level sizing simulation of the live 1h/4h/1d mix, not a per-timeframe signal test

Hypothesis: Taking every setup at 1% risk compounds to a usable return with tolerable drawdowns.
Method: Own Python sim on TradingView tranche outcomes; exits dated by walking 1h/daily prices to each level (97.9% dated); 10,000 start; 1% per setup; no costs (variant: -0.1R per setup).
Result: Whole 1h universe at 1% per setup: ruined (2022-10). 'Universe scaled' to the live 50/20/20 mix: CAGR 33%, max DD -23.6%, 73% months up, worst year 2022 -11%; with 0.1R cost: 16%, -35.4%, worst year -24%. In-sample live basket: 181% CAGR (not a fair estimate).
How much to trust it: Low: reconstructed exit dates; closed-trade drawdowns; ~4x median notional leverage; survivorship
Did it change the live rule?: No.
Notes: Not in the repo.

Scripts and data for this study will be linked here when the research code is published on GitHub.

← All experiments

Education and research only. Not financial advice. Backtests use today's index members (survivorship bias).