Active risk is re-answered every cycle, because the stop keeps moving.

Two risk fields

One is a historical fact and one is a live measurement.

Initial risk is entered once and never changes, because it records what was committed at entry and belongs to the trade's permanent history. Active risk is a separate field carrying the current remaining downside, and the manual is explicit that it does not automatically equal the entry figure after the stop has moved. Keeping both is deliberate: the historical figure is what per-trade sizing discipline is audited against later, while the live figure is what the throttle's capacity arithmetic consumes now. Collapsing them into one field would force a choice between an accurate audit trail and an accurate exposure reading, and the system needs both. The cost of keeping them separate is that one of them has to be maintained, and maintaining it is the discipline this brief is about.

The default is wrong

Carrying the entry figure forward overstates exposure in exactly one direction.

The failure has a consistent sign, which makes it worth naming precisely. Stops move toward break-even far more often than away from it, because that is what the management rules instruct, so a trade whose active risk has not been updated almost always shows more risk than it carries. The throttle then reads inflated open exposure, subtracts too much from the authorised pool, and authorises less fresh deployment than the account can actually support — possibly reducing the trade count below what a full cycle would allow. Nobody is harmed in the sense of a blown account, and the system has quietly become more conservative than its own doctrine, for a reason that appears nowhere and can be corrected in ten seconds. Conservatism arriving through neglect is not the same thing as conservatism by design.

The break-even case

Zero active risk with an open position is a real and frequently correct state.

The entry that most resists intuition is a trade sitting at break-even. Its status remains open, it still occupies one of the cycle's four slots, its floating profit or loss continues to move, and its active risk is zero — because if the stop is genuinely at entry, being stopped costs nothing beyond friction. The consequence is that the position stops burdening the pool entirely while remaining an open trade in every other respect. Operators frequently resist writing zero into a field for a live position, and the resistance is a category confusion: the field does not ask whether the trade is finished, it asks what it can still cost. This is also the mechanism by which advancing stops mechanically restores throughput, which is worth knowing as a reason to advance them promptly rather than as a reward for doing so.

The companion field

A dropdown records why the number is what it is.

Alongside the numeric field sits a risk-state note — still at original stop, moved to break-even, stop reduced, partial protected, excluded or closed — and its function is to make the numeric entry checkable. A bare zero in an active risk column is unverifiable after the fact; a zero paired with a break-even note is a claim that can be confirmed against the trade's own history. The pairing also makes the exposure audit tractable, because scanning the note column reveals immediately whether the ledger has been maintained: a set of open trades all reading still at original stop several days into their lives is either an unusual market or, far more likely, a ledger nobody has updated. That is a question the numeric column alone cannot raise.

When to answer it

The update happens before the cycle is read, not when convenient.

The operating checklist places the active risk update in the second step, before anything is read anywhere, and the placement is load-bearing. Every downstream figure the console produces — open exposure, remaining capacity, suggested fresh trades, the whole packet handed to the throttle — is computed from these fields at the moment they are read. Updating them afterwards does not correct the decision; it corrects the record of a decision already made on stale numbers. The practical rule is that the ledger is brought current first, as a distinct act, and only then is any output consulted. The question takes seconds per open trade and it has to be asked of each one individually, because there is no way to derive the answer from anything the workbook already holds.

• Initial risk is history and never changes; active risk is live and usually does.
• Stale active risk overstates exposure, so the error is silent under-deployment.
• Break-even means zero — an open trade that costs the pool nothing.

The key idea

A field phrased as a question stays maintained; one phrased as a value does not.

The difference between labelling a column risk and asking what would still be lost if the stop were hit right now is not presentational. The first invites a lookup — the number is already known, it was set at entry — and the second forces a fresh evaluation that can only be answered by examining the position as it currently stands. Since the honest answer changes over a trade's life and the recorded answer does not change on its own, the phrasing is doing the work of a reminder that arrives at the exact moment it is needed. It is a small piece of design carrying most of the accuracy of the exposure layer.