Most trading rules are preferences wearing costumes — held sincerely, checked never. These ten are different in one respect: every one of them is enforced somewhere in machinery that runs whether we feel like obeying it or not.
Long-dated positions in companies you would want without the trade attached. Everything else here is machinery built on top of that foundation — and none of it repairs a bad one. If you would not hold it uncapped, it is not an income position; it is a capped bet.
Two of the only reliable forces in options. Selling premium harvests both — but only where the premium is genuinely rich against that name's own history. Thin vol is not a discount; it is the market declining to pay you.
the cap is not being paid for and leans you lighter.
Institutions cannot position quietly; option chains record it — a date being paid up for, a strike accumulating, volatility bid with no news attached. You cannot know their intent. You can know where they put their money, and when.
A setup that wins seven times in ten will still end you on the other three if the size is wrong. The first job of position size is to guarantee you are still trading next month — the second job is returns, and it is a distant second.
Panic and greed are the source of every edge worth having, and "buy the dip" without measurement is how accounts die. Fade fear when the structure underneath is intact — dealers holding their ground, volatility calm, the selling mechanical rather than urgent.
Declining pays invisibly, which is why almost nobody counts it. Traders tally what they lost and never what they avoided. Keep the record of your refusals; it is usually your best-performing book.
Every rule above reverses sign somewhere. Premium selling is superb in chop and bleeds in trends; breakouts pay in trends and shred you in ranges. Strategy is weather-dependent, and the weather is knowable before you commit.
Spreads, roll debits, assignment, taxes, slippage on illiquid wings. An edge smaller than its friction is a loss wearing a costume — and the friction is always knowable in advance, which makes paying it a choice.
Memory keeps the survivors: the one enormous winner, the alert that called it. The ledger keeps everything, including the forty that went nowhere. Where memory and the ledger disagree, the ledger is right — and this is the only rule that can audit the other nine.
A good decision can lose; a bad one can win. Neither result means anything alone. Below roughly twenty trials a number is a rumour; past it, it begins to be testimony. Never let one outcome overturn a rule the sample supports — and never let a rule survive a sample that says no.
still on trial — and means it.
You can trade these ten rules with a spreadsheet and enough discipline — traders have done it for decades. What breaks is not the rules; it's the bookkeeping. Rule 9 is the one nobody keeps, and without it the other nine slowly become whatever you felt like doing.
Greeksmith exists because one trader got tired of losing that argument with himself. The products are just these rules, made mechanical enough that skipping one takes effort.
Institutional positioning — detected, dated, checked against the calendar, and scored in public every Friday.
Your own rules, made executable — and an honest ledger that grades both your strategy and your overrides of it.
The one-page version, for printing or pinning: download the creed (PDF)