The internet is drowning in trading-risk advice. Most of it is wrong. Not wrong in theory — wrong in practice.
The three rules
Rule 1: Risk a fraction, not a feeling
Pick R: your maximum loss on any single trade as a percentage of equity. For most retail accounts, 0.3%–1%. If new, 0.3%. If 3+ years profitable, up to 1%. Beyond that, you are not trading, you are gambling with extra steps.
Rule 2: The invalidation line comes first
Before you look at where you are going to buy, decide where the trade is wrong. Stop one tick beyond it, size at R / (entry − invalidation).
Rule 3: One portfolio, one risk budget
If you are already short NQ and you take a long on QQQ because “it is different”, you are not diversifying — you are doubling.
The two numbers
- Max weekly drawdown W. 3%–6% of equity. At 0.8 × W, cut R in half next week. At W, stop until Monday.
- Max concurrent positions. 3 soft cap, 5 hard cap. If you cannot explain each in a sentence, you have too many.
The one habit
Every Sunday night, 20 minutes, plain text file. Four lines per losing trade: thesis, invalidation, deviation, rule-that-would-have-helped. No winners. No deep dives. Over 12 months you build a personal rulebook that beats any trading book you can buy.
The clever frameworks that kill accounts
- Averaging down — turning a 2R loser into a 6R loser.
- Scaling into momentum — your R silently inflates by a factor of 3.
- Reverse martingale — biggest size deployed at most-euphoric mood.
- Hedging a loser with a correlated short — two trades to be wrong about.
What the engine does
OcuBit suggests an R-adjusted position size on every signal, refuses to loosen your drawdown rail inside a trading week, aggregates correlated exposure. Overrides are logged. These are not features. They are the floor.
Want sharper reads like this one?
Subscribe to the OcuBit briefing — twice a month, no fluff.