Operator brief · 440

The goal is not zero risk of ruin. It is a number you have seen and accepted.

The key idea

The uncomfortable premise

Positive expectancy does not eliminate the possibility of ruin.

A system can have a genuine edge, be executed correctly on every trade, and still reach its shutdown boundary — because outcomes arrive in a sequence, and some sequences are bad enough to end the account before the edge has room to express itself. This is not a flaw in the system or a failure of discipline. It is a property of taking risk with a finite bankroll, and it is present in every strategy that has ever been deployed. Refusing to look at it does not lower it. What refusing does accomplish is to move the moment of discovery from the planning stage, where the figure is an input, to the event itself, where it is merely an explanation.

The three ways it gets hidden

Ruin risk is usually not measured, not stated, or not read.

Most trading plans obscure this number in one of three ways. Some never compute it at all, and treat the absence of a figure as the absence of a risk. Some compute it and report only the favourable half of the distribution, so the reader sees median outcomes and never the tail. And some report it correctly but position it as a technicality beneath the headline results, where it is read as a disclaimer rather than a finding. Only the first is ignorance; the other two are presentation choices.

FigureThe full distribution, including the half nobody quotes

Schematic shape of simulated outcomes. Marketing quotes the top band and stops.

What acceptable means

Acceptable is a decision, and it has to be made by a person.

There is no formula that converts a lock probability into a verdict, because the answer depends on what the capital is for. A figure that is fine for a deliberately-sized speculative account is not fine for capital that has obligations attached to it. The system's job is to produce the number honestly and put it in front of the operator before deployment; the operator's job is to say yes or no to it. What the architecture refuses to do is make that decision quietly on the operator's behalf by never showing the figure.

Priced, not just accepted

Accepting the number means changing something in response to it.

Acceptance that produces no change is not acceptance, it is acknowledgement. A ruin figure that has been genuinely priced shows up in the configuration: position size chosen with it in view, the gate ladder's thresholds set knowing how often each will be visited, capital committed at a level where the tail outcome is survivable outside the account. If reading the number changed nothing about how the system is run, the number was not used — it was merely seen. The test is straightforward: name the specific configuration decision that would have been different had the figure come back twice as large. If nothing would have changed, the number was decorative.

Why the figure is directional

It is a model output, and models are wrong in known ways.

The simulated lock probability rests on assumptions: that the branch probabilities hold, that execution matches the profile, that the trade sequence is drawn the way the model draws it. Live conditions will differ from all three to some degree. The figure is therefore best read as an order of magnitude and a relative instrument — useful for comparing configurations and for knowing whether the risk is in the low single digits or considerably higher, not as a precise forecast of one account's fate. Treating it as precise is its own kind of wishful reading.

The reframe

The number's job is to be looked at before it matters.

Every trading account has a ruin probability whether or not anyone has calculated it. The only thing simulation changes is when the operator finds out. Computed in advance, the figure is an input to sizing, to gate thresholds, and to the decision about how much capital belongs in the account at all. Discovered afterwards, it is not a figure — it is an explanation, arriving to describe something that has already happened and can no longer be configured around.

Connected inside MARS

Monte Carlo Survivability →

Monte Carlo Lab →

Risk-Tier Performance Bands →

Risk Gate Map →

Every brief documents the same shipped system.

The complete MARS package — eleven workbooks, three TradingView indicators, the full manual library — $497.

Get MARS Pro — $497 Explore the Platform