Seven stops fit at $460 and fail at $650
The daily reference is 3.25%, or $3,250 on $100,000. Divide by seven and the cash unit is about $464. Round down to $460, which is 0.46%. Seven fills are $3,220, leaving $30 before the reference. At 0.35%, or $350, seven stops are only $2,450, which is comfortable and slow. At 0.65%, or $650, five stops are $3,250 exactly.
Use $460 if the goal is to be wrong seven times in a session and still be inside the day. Use $350 if you also want the 6.5% monthly VaR target, $6,500, to survive a second bad day. Open a Venasri account and store one unit, not a range you renegotiate after the second loss.
What $460 looks like on FX, gold, and an index
Majors pay about $10 per pip at 1.00 lot. A 20-pip stop at $460 needs 2.30 lots. Gold pays $10 per 0.10 move. A 50-pip stop needs 0.92 lots. A 70-pip gold stop needs about 0.66 lots. NAS100 at $1 per point needs 5.75 lots for an 80-point stop to risk $460. Round down when the platform cannot show the decimal you want.
The dangerous rounding is upward. 2.50 FX lots on a 20-pip stop are $500, already above the seven-stop unit. 1.00 lot of gold on a 50-pip stop is $500 for the same reason. Bot controls should compute cash from the stop distance, not from a favorite lot button.
The seed cares that tomorrow's lot matches today's
Risk Stability and Capacity are scored on the shape of the risk, not only on the percentage you meant. A week at 0.40 lots and a Friday at 3.00 lots can be inside $3,250 on the day and still cut how much DARWIN capital follows you. The seven-stop test is the loss budget. The sameness of the lot is the allocation budget.
Correlated trades share the unit. Two gold positions at 0.92 lots are a $920 idea if they can lose together. EURUSD and GBPUSD often do the same thing. Count them once. The 10% reference, $10,000, is far away if every day respects $3,250, and it gets close if one unstable day is allowed to be the whole story of the month.
Flatten on the fifth $500 or the fifth $460
Equity Guard at 2.50%, or $2,500, fires before the seventh $460 loss. Five losses at $460 are $2,300, so the guard and the seventh-stop plan meet in the same neighborhood. If you insist on the $650 unit, the guard must be the thing that stops you, because five losses already equal the firm's day.
None of this deletes the account by itself. It changes the D-Score and the seed weight, and it consumes the monthly VaR target the engine normalizes toward. The lot plan is how you stay copyable. The guard is how a bad morning ends while that is still true.