Twenty events is not twenty risks

The per-event cap settles the strategy question before you start. Maximum profit from any single event is 0.5% of account size, the Equity Growth Target is 10%, and those two numbers put a floor of twenty capped winning events under any completed evaluation. At a 70% hit rate it is around fifty events. At 60%, around a hundred. At 55%, around two hundred.

So you go looking for events. That is the correct response, and it is where a second problem starts, because the rule counts events and your account counts dollars, and those are not the same unit.

The rule’s definition is administrative

An event, for the purposes of the cap, is a grouping on the platform. Several markets under one election aggregate against a single ceiling — that much is settled and it closes off the obvious workaround of splitting a view across candidates.

What the rule does not do, because it cannot, is look at two genuinely separate events and notice that the same fact decides both.

A scheduled statistical release can be the resolution input for markets on the release itself, on a central bank decision that follows it, and on a threshold being crossed by year end. Three events, three caps, three tickets. One number, published at one moment, moves all three the same direction.

The platform sees three. Your equity sees one.

Why this is worse here than in ordinary position sizing

Correlation is an old problem and every trading education covers it. Two features of this structure make it bite harder than usual.

The cap is on the upside only. If a correlated cluster of ten positions all resolve your way, each one is individually limited to 0.5% of account size. If the same cluster all goes against you, each one loses its full outlay, and nothing limits that. Concentrating into one underlying fact therefore produces a payoff that is truncated when you are right and complete when you are wrong. You are taking the full risk of a concentrated position and the capped reward of a diversified one.

Everything is measured on live equity, continuously. There is no balance-based calculation and no distinction between open and closed positions. A cluster that moves together moves your whole account at once, intraday, against a daily drawdown limit of 3% of account size.

Put numbers on that. On a 50,000 account the cap is $250 an event and the daily limit is $1,500. At a 40-cent entry, the outlay that produces exactly the cap is 250 × 0.4 ÷ 0.6 = $166.67. Ten positions sized to the cap is $1,666.70 of outlay. If one release moves all ten against you and the cluster gives up about ninety percent of its value, you are at the daily limit — from a single piece of news, on positions you sized correctly one at a time.

Ten genuinely independent events do not do that. One headline moves one of them, and the other nine are unaffected. The difference between those two portfolios is invisible in the position list and decisive in the outcome.

The test

There is no clean way to compute this. Most of these contracts have no return history worth regressing, and the ones that do are short and sparse. Anyone offering you a correlation matrix for a book of event contracts is offering you a number with very little behind it. So it stays a judgment exercise, and the judgment reduces to one question:

What single fact, if I learned it right now, would resolve several of these the same way?

If there is such a fact, those markets are one exposure for risk purposes no matter how many caps they occupy. Size them as one. The cap counts them separately; nothing obliges you to.

A few patterns worth noticing when you run that test:

• Markets whose resolution text names the same source. A single agency print, a single scheduled announcement, a single scoreboard.
• Markets on different stages of one process — a nomination, then a confirmation, then a first act in office.
• Markets that are arithmetically linked: a threshold at one level and a threshold at a higher level on the same series cannot resolve independently.
• Markets whose resolution dates cluster on one day. Even without a shared cause, simultaneous resolution removes your ability to learn from the first before committing to the second.

That last one is worth dwelling on. Breadth is not only about what moves together, it is about sequencing. A hundred events that resolve one after another give you a hundred chances to update. A hundred that resolve in the same week give you one.

What this costs you to fix

Less than it sounds. The remedy is not to avoid correlated markets — often the area you know best is exactly where your events will cluster, and trading outside your competence to chase statistical independence is a worse trade than the one you were trying to avoid.

The remedy is to size the cluster rather than the ticket. If four markets turn on one release, decide what the whole group is worth risking and divide, rather than sizing each to the cap because the cap permits it. The permitted size is not a recommended size. The cap tells you the most the rules will pay you for an event; it says nothing at all about how much you should have at risk in it.

And note the direction of the error. Undersizing a correlated cluster costs you a little of a capped upside. Oversizing one can end the evaluation in an afternoon. Those are not symmetrical mistakes, and the structure of the rules is what makes them asymmetrical.

Evaluations are conducted on simulated accounts. Not financial advice.