Pinning the safe ceiling of dynamic risk with 1-minute data

My previous analysis of risk control focused on daily closing prices, but prop firm rules are evaluated based on intraday equity swings. To bridge this gap, I rebuilt the current system (v1.5.0) using one-minute (M1) bars to see if increasing the risk multiplier (my “k” factor) would push the system past the critical 5% daily loss threshold during the trading day.

Intraday stress test results (11-year sample)

I used the reconstruct_intraday_daily method to track account equity from the start of each day to its lowest point, and then applied the multiplier (k) to those intraday losses.

Refining the safety ceiling

While the initial data suggested that a multiplier up to 2.0 was safe, further verification revealed this was overly optimistic. The reconstruction process had missed intraday volatility from specific components (like Connors or index-based sleeves), leading to an underestimation of the true worst-case scenario. When accounting for the actual peak-to-trough equity decline of 3.22% (verified against official closing-price equity data), the math changes. At that higher, more accurate baseline, a multiplier of 1.55 is the absolute limit for staying under a 5% loss. If I were to push to k=2.0, a repeat of the worst trading day would result in a 6.44% loss, which violates the threshold. In other words, while my initial recommendation of k=1.5 is robust and safe, there is no room to scale it higher. The boundary-based risk control logic is now fully verified through both Monte Carlo simulations and M1 intraday resolution, and I am ready to proceed with the MT5 implementation.

How this connects

This verification builds on earlier ones (what failed before and what I tried this time, comparisons between approaches).

• Scaling risk by distance-to-failure genuinely adds returns
• Attack harder as the profit cushion grows: verified