The cap sets the size. The drawdown sets how many.
Classic position-sizing maths says how much to commit when you have an edge. Under a half-percent profit ceiling it almost never gets a say. The question that replaces it is how many full losses your drawdown can hold.
Most traders who think carefully about sizing have met the Kelly criterion. It comes from a 1956 paper by John Kelly at Bell Labs, “A New Interpretation of Information Rate”, published in the Bell System Technical Journal, and it answers one question: given an edge on a repeated opportunity, what fraction of your capital should you commit to maximise long-run growth?
For a contract that pays $1 or nothing, the answer has a clean form. If you buy YES at price p and your honest probability is q, the Kelly fraction is:
f = (q − p) ÷ (1 − p)
It is a good formula. Inside a Your Call evaluation it almost never gets to make the decision, and seeing why changes what the sizing question actually is.
The ceiling binds long before Kelly does
Maximum profit from any single event is 0.5% of account size. On a 50,000
account that is $250. At entry price p, the outlay that reaches that ceiling
exactly is 250 × p ÷ (1 − p). At 50 cents, $250.
Now put an edge on it. Say you believe a contract priced at 50 is really a 60.
Kelly says commit (0.60 − 0.50) ÷ 0.50 = 20% of the account. That is $10,000.
The ceiling makes everything past $250 capital that can lose in full and earn
nothing extra.
The two cross at a strikingly small edge. Divide one by the other and the ceiling is the tighter limit whenever:
q − p > 0.005 × p
At 50 cents, that is any edge larger than a quarter of a cent. Traders who deliberately use a quarter of Kelly to reduce volatility hit the ceiling once the edge passes a cent. For any view worth acting on, the cap is what sizes the single position. Max Profit Per Event, explained covers why sizing above it is a pure error.
Kelly also assumes there is no floor
There is a second mismatch, and it matters more. Kelly maximises growth over an unlimited sequence and assumes you can sit through whatever path that growth takes. Its paths are rough. Even in the idealised model, a full-Kelly account will very likely spend time below where it started, and often far below it.
An evaluation has a hard floor. The daily drawdown limit is 3% of account size, resetting at 5PM Eastern, and crossing it ends the evaluation immediately. The maximum drawdown is 6% of account size, trailing an equity high-water mark that never locks. A sizing rule that tolerates deep dips is answering a question this structure does not ask.
The question that replaces it
If the cap sets the size of each position, the decision left is how many you hold at once. The useful unit is the full loss: what one capped position costs if it resolves against you, including the 1% commission. The commission is charged on cost basis — what you actually pay for the shares, not what they would be worth at settlement.
On a 50,000 account, measured from a fresh equity peak:
| Entry | Loss per capped position | Full losses inside the 3% daily limit | Full losses inside the 6% maximum |
|---|---|---|---|
| 20 cents | $63.13 | 23 | 47 |
| 35 cents | $135.96 | 11 | 22 |
| 50 cents | $252.50 | 5 | 11 |
| 65 cents | $468.93 | 3 | 6 |
| 80 cents | $1,010.00 | 1 | 2 |
Read the bottom row slowly. At 80 cents, one capped position going against you uses two thirds of a day’s limit. A second on the same day ends the evaluation. At the bottom of the band, the same daily limit holds twenty-three.
The ratios hold at every account size, because the ceiling and both limits are all percentages of account size.
Three things the table does not show
It assumes independence. If one release decides several of those positions, they are one loss, not several. Twenty events is not twenty risks is about exactly that.
It is measured from the peak. Every limit is equity-based, so open gains raise the high-water mark and the floor with it. After a good run, the budget is counted from the new peak, not from where you started.
It ignores the path. Positions are marked continuously, so a position can spend part of the budget on its way to resolving, even if it then resolves in your favour.
Why the low rows are not free
It would be easy to read the table as an argument for cheap contracts. It is not one. A contract at 20 cents usually resolves NO, and you need to be right more than one time in five just to break even. A hit rate on its own tells you nothing works through why the entry price is the threshold you have to beat.
What the table does show is that entry price sets how many mistakes you can survive at once, independently of how good your selection is. That matters because the work is a volume problem. Against a 0.5% cap and a 10% Equity Growth Target, twenty capped winners is the theoretical floor. With entries around the middle of the permitted range, it is roughly 50 events at a 70% hit rate, roughly 100 at 60% and roughly 200 at 55%. At 50% it is not achievable. The rulebook sets out the full table.
A process that needs a hundred resolved events has to survive the losing streaks that come with a hundred events. The count of full losses your drawdown can hold is how you check that before the first order, rather than finding out during one.
The short version
The ceiling decides the size of each position, and it binds for almost any edge. What is left to decide is how many capped losses you have open at once, how correlated they are, and at what entry prices. Count those in full losses against the daily limit and the trailing maximum. It is a smaller number than most people expect, and at the expensive end of the band it is very small indeed.
Evaluations are conducted on simulated accounts. Not financial advice.