Position sizing and risk

The arithmetic

Three steps. It is deliberately boring, and that is the point.

1. Risk budget = account × risk %. A $25,000 account at 1% gives $250 of acceptable loss.
2. Risk per contract = (entry − stop) × 100, because one standard equity option contract controls 100 shares. Entry $3.40, stop $2.55 → $85 per contract.
3. Contracts = budget ÷ risk per contract, rounded down. $250 ÷ $85 = 2 contracts.

The calculator does this live, including the percentage of account actually at risk once rounding is applied.

Why losing streaks are longer than you think

A strategy that wins 55% of the time is a good strategy. It also produces losing runs that feel like the system has broken. The probability of a run of six consecutive losses appearing somewhere in a hundred trades at that win rate is substantial — closer to a coin flip than to a rare event.

What that streak costs depends entirely on size:

Risk per trade	Cost of 6 straight losses	Gain needed to recover
1%	~5.9%	6.3%
2%	~11.4%	12.9%
5%	~26.5%	36.1%
10%	~46.9%	88.3%
20%	~73.8%	281.5%

The right-hand column is the one that ends accounts. Losses and gains are not symmetric: down 50% requires 100% to get back to flat. At 1% risk a bad month is an inconvenience. At 10% the same sequence of trades — the same edge, the same decisions — requires nearly doubling the remaining capital just to return to where you started.

Why "% of account" is not enough on options

Stock traders size off a stop and generally get filled near it. Options add three complications:

• The stop may not be reachable. On a fast move or a wide spread, the fill can be materially worse than the level. Your modelled 1% becomes a realised 1.8%.
• Premium can go to zero. With short-dated contracts, total loss is a normal outcome rather than a tail event. Size 0DTE as though the whole premium is at risk, because frequently it is.
• Spreads change the maths. Defined-risk structures have a known maximum loss. Size those on max loss, not on entry price.

Kelly, and why you should use a fraction of it

The Kelly criterion computes the bet size that maximises long-run growth given an edge. Traders discover it, calculate something like 15%, and lose most of their account learning why that number is theoretical.

Kelly assumes you know your win rate and payoff precisely. You do not — you have an estimate from a small sample, in a market that changes. Overestimating your edge slightly produces a Kelly figure that is dramatically too large, and full Kelly generates drawdowns most people cannot sit through even when the maths is right. Fractional Kelly, typically a quarter or a half, gives up a little theoretical growth for a very large reduction in variance. In practice, that lands most people back at 1–2% per trade, which is where the desks were already.

Rules the desk runs on

• Fixed fractional risk. The same percentage on every trade. No sizing up because a setup feels better — conviction is not calibrated, and the trades that feel best are not reliably the ones that work.
• Stop set before entry. Derived from the structure that justified the position. If price invalidates the level, the reason is gone.
• Round down, always. Two contracts, never "nearly three."
• Daily loss limit. A fixed number of stop-outs ends the session. Revenge trading has cost more accounts than any bad setup.
• Correlated positions count as one. Three long SPX calls at three strikes is one bet, sized accordingly.

The honest summary

Position sizing is unglamorous and entirely within your control, which makes it the highest-leverage thing most traders are ignoring. You cannot control whether a level holds. You can control whether being wrong about it costs 1% or 15%. Follow every alert we publish at the wrong size and you can still lose money on a profitable sequence of trades — which is exactly why the record shows the trades and not a promised return.