EXP-009 design — Connors RSI(2) 5-minute cost and slippage confirmation

EXP-009 design — Connors RSI(2) 5-minute cost and slippage confirmation

• Work week: 2026-W36
• Run date: 2026-09-04
• Status: run completed · hold
• Baseline: EXP-006 5-minute pilot, EXP-007 cost normalization review

Purpose

Without mixing daily results into this experiment, check whether the Connors RSI(2) signal already moved to 5-minute bars keeps the same direction under changes in per-product costs and slippage. Instead of repeating the same data and signals again, the slippage sensitivity that the existing EXP-006/007 could not separate is recorded under a new experiment ID .

Fixed contract

• Input: CREON A005930, A069500 5-minute bars, 2021-07-16~2026-08-26
• Signal: the same 7 A/B variants as the existing EXP-006
• Execution: the open of the next valid 5-minute bar after the signal bar's close is confirmed
• Capital: an independent 3,000,000 KRW account per variant and symbol
• Quantity: 95% of available cash just before entry, integer quantity rounded down, sequential compounding, no leverage
• Periods: train 2021~2023, validation 2024, locked OOS 2025-01-01~2026-08-26
• Existing EXP-006/007 results and ledgers are not overwritten

Cost sensitivity

A069500 is a domestic equity ETF with a 0bp sell tax; A005930 uses the EXP-007 KOSPI per-period transaction tax schedule unchanged. The 1.5bp commission is provisional until actual CREON fills are checked, and the slippage range is a sensitivity range, not a value verified against real quotes.

Decision rules

• Check the direction of both symbols and the main tracks together under actual_proxy and actual_proxy_slip2.
• Do not adopt on the best return of any single cell.
• Even if OOS results are positive, do not raise to paper_candidate before passing concentration checks across both symbols, years and times of day and a ledger recalculation.
• This result is a pilot of about 5 years and 2 symbols, so the default is hold, not advance.

Outputs

• Run: scripts/backtest/run_connors_rsi2_m5_confirmation.py
• Summary: reports/ledgers/EXP-009-summary.csv
• Trade ledgers: reports/ledgers/EXP-009-m5-*-trades.csv
• Static results: reports/experiments/EXP-009-connors-rsi2-m5-confirmation.md
• Charts: reports/experiments/EXP-009-connors-rsi2-m5-confirmation.html