Run 3 Tripped the Weekly Kill Switch: −11.7%, Benched 72 Hours
Three days after the daily loss limit benched Run 3 for 24 hours, the bot came back, traded, lost more — and on October 21 the weekly kill switch fired. The trade log recorded one line: "WEEKLY kill-switch: −11.7%. Pausing 72h." Since then, every hourly run has done the same thing: checked the clock, seen the bench, skipped. No new entries until October 24.
Equity at halt: $888.20
P&L: −$111.80 (−11.2%)
Trades: 51 · Wins: 26 · Losses: 24
Fees paid: $25.71
Weekly drawdown: −11.7% → halted 72h
How it got there
Run 3's final session before the bench was all High-tier microcap trading: KIMCHI, MONEYBIZ, Pokecity, KYRO, GUZZLEGOAT — the usual suspects. The 1-hour fast exits kept producing mixed results: MONEYBIZ exited at +14.6% (+$3.37) and +14.7% (+$3.39) on consecutive cycles, while KIMCHI printed −29.1% (−$7.45) and Pokecity −17.1% (−$4.49) on the same rotation. Win rate stayed near 50% — 26 wins against 24 losses — but the average loss kept outweighing the average win.
One bright spot appeared in the Medium tier: its first-ever trade, a NEARUSD Donchian breakout entry, closed at +$0.56. Small, but it's the tier's only live data point so far.
Why a weekly limit exists
The daily guardrail is designed to stop a bad day from compounding. The weekly guardrail exists because the daily one wasn't enough: the bot was losing at a sustainable-looking rate — no single catastrophe, just a slow bleed that the daily limit kept capping, day after day. The weekly switch looks at the bigger picture: if the whole week is −11.7%, the problem isn't timing, it's the strategy. So trading stops for three days.
Both guardrails fired exactly as coded — the daily at −3.2%, the weekly at −11.7% — with nobody watching and nobody overriding. Risk limits that only work when you feel like enforcing them aren't risk limits.
A pause isn't a solution. 51 trades in, the honest verdict on the High tier is the same as it was at trade 7: the exits function, the entries are the problem, and real money stays off the table until the strategy earns it in public.
When the bots resume on October 24, they'll resume with the same rules, the same paper capital, and the same public log. If the losses continue, the next guardrail will fire too. That's the design.
Every skipped run, every trade, every fee — on the public dashboard.
Open Nova Trader →