The risk officer's line-by-line critique of the day's strategy review.
- 4-hour long-only / short-only paper accounts: the proposed 1% risk per trade breaks the team's 0.5% limit, so the officer recommended 0.5% risk and ≤10% notional. Historical replays of the short-only version were negative, so it was reclassified as a control group.
- Gold 1-minute test account (priced off a forex broker quote feed, demo only): at the minimum lot size (0.01 lot = 1 oz), a $3 stop already risks about 3% of a $100 virtual account, 6× the limit. It is labelled an "over-limit experiment". Its −4.92 result at the time excluded commission, so it was optimistic.
- Data reliability: funding costs and liquidation weren't simulated in some tests; a "4-week" evaluation probably won't reach 30 trades, so that would read as "not enough data", not a pass.
- On the multi-year EMA study: almost all of the long-only gain came from the first 149 days (+247.9%); later regimes compound to about −43%. Funding paid (50.3% of notional) exceeded trading costs (38.4%).
- Self-correction: the officer corrected their own attribution. Losses beyond 0.5% came from funding plus exit fees, not funding alone.