The True Formula for Profitable Trading

ELI5/TLDR

A risk manager at a day-trading firm gives a talk on the two things that separate traders who survive from traders who blow up: managing risk and having an edge. The core idea is that you don’t size every trade the same — you bet big on the rare, high-conviction setups and bet small on the everyday ones, because a tiny handful of trades make most of your money. The losses that kill you aren’t the planned ones; they’re the impulsive trades you never wrote down. And the boring homework — building a playbook, tracking your stats, reviewing your own screen recordings — is the price of admission, paid up front.

The Full Story

The speaker is Carlton (Carl), the risk manager at SMB Capital, talking to a room of traders. His job description is unglamorous: keep each trader inside their risk limits so they don’t become the cautionary tale on the right side of the slide. Watch for tail risk, dangerous concentration, the parabolic move that turns into a blowup. Prevent emotional trading. Hold people accountable for capital the firm has entrusted to them.

The whole talk circles two words.

Risk management and edge

If you’re missing these two components here as far as risk management or edge… if you don’t have any of that, more than likely you’re going to fail.

Edge is an informational, technological, or technical advantage — a reason your trades have positive expected value. Risk management is what keeps you alive long enough to express that edge. Neither works alone. Good risk management with no edge just loses money slowly; an edge with no risk management gets wiped out by one bad day.

Expected value, made concrete

The talk’s most useful stretch is a plain walk-through of expected value. A trade isn’t “good” or “bad” by whether it wins — it’s good if the math is positive over many repetitions. The example:

Reward if it works: $500, with a 40% chance of hitting.
Loss if it fails: roughly $200 drawdown, with a 60% chance.

You weight each outcome by its probability and net them. ($500 x 0.40) gives $200 of expected reward; ($200 x 0.60) gives $120 of expected loss. Net expected value: +$80. A winning trade — even though it loses six times out of ten.

Out of 10 times you’re only winning four times. A lot of people say, “Oh, that’s a losing trade.” Right? And you just walk away from it. But until you’re intimate with the stats itself and actually understand the concept, this is makes trading powerful.

The point that lands: you can’t compute any of this without tracking your own statistics. You have to know your average win, your average loss, your best setup, your worst. Treat it like a business or it isn’t one.

(Numbers in the talk wobble — he says $200 and $120 nets to $80, which works, though it doesn’t quite line up with the $500/$200 he opened with. The mechanism is right even where the arithmetic drifts.)

Study your winners, not just your losers

A counterintuitive note. Most traders obsess over losing trades. He argues you should spend at least as much time dissecting the winners — the outliers especially.

You really want to focus on your winners. You want to focus on where you have edge. Where you’re making money. What time you’re making money, how you’re making money.

Was it the volume? The catalyst? The sector? Were you just lucky? You need to know precisely how you win so you can do it on purpose. He’s heavy on AI here too — paste your P&L into a chatbot, ask which ten trades made the most money, what setup, what time of day, what ticker. Pattern recognition for free. He thinks traders who skip this will be “in huge trouble.”

The losses that actually kill you

There are two kinds of loss: expected and unexpected. Expected losses are the cost of doing business — you planned the trade, it didn’t work, you took the stop. Fine. Unexpected losses are the Achilles heel.

Those invented trades, those trades where you just sat down, you didn’t plan for it, and you just took it.

You get in, realize it was a mistake, and instead of getting out you tell yourself it’ll turn around. Then you’re praying. Those are the trades that blow up accounts and end careers. The fix is mechanical: follow the plan, every time. Write down your tilt triggers — the warm flush, the butterflies — and when you feel them, walk away and regroup.

He frames discipline with the math of drawdowns, which is the talk’s most sobering slide. Losses and recoveries aren’t symmetric:

Lose	Need to make back to break even
10%	11.1%
20%	25%
30%	~43%
50%	100%

Past 50%, the hole gets steep fast. Hence the obsession with preserving capital. The pilot analogy runs throughout: good risk management is a long runway. You’ll still make mistakes — they’re inevitable — but you want them to be controlled mistakes.

Asymmetric sizing — the heart of it

This is the actual “formula” in the title, and it’s the most worthwhile idea. You do not risk the same amount on every trade. You grade your setups and size exponentially:

A+ setups — happen maybe two to five times a year. Highest conviction. Deploy maximum risk; he says risk 100% of your daily stop, sometimes multiples. Here you should be risk-seeking, not fearful.
A setups — every one or two months. Risk ~25%.
B+ setups — about once a week.
B setups — the everyday trades. Risk ~5%.

The everyday trades are cash flow; they pad the P&L. The rare trades are where the money actually is. He calls this the 65/5 rule:

65% of all your profits… our A+ setups… they’re going to be probably 65% of our profit on the year… and that’s about 5% of all trades.

If you size every trade identically, the many small B losses eat the few big A wins, and you finish the month flat or red. The whole edge lives in betting big when conviction is highest and small the rest of the time.

Bumping and cutting risk

Risk allocation breathes. You earn the right to size up: two stops survived in a day, bump the stop ~20%; consistent low-variance P&L over a few weeks, bump ~10%. And you cut hard when things go wrong — hit a weekly stop, cut size 50%; hit it again, cut another 50%. When you’re in a losing cycle you shrink the bets until the curve turns, then scale back up. The risk dial only loosens when the results justify it.

Building the playbook — the heavy lifting

A playbook is a databased archive of setups: what they look like when they work, the time frames, the volume signatures, the nuances. Start from the firm’s morning game plan, watch how those stocks actually traded through the day — where they broke out, whether they held above VWAP, whether it was an opening-range break — and accumulate a collage of examples. He suggests 50 to 100 samples to call something a playbook, and you can pad that count using crypto and futures, not just equities.

His team runs only about six core playbooks. The job isn’t finding 3,000 patterns; it’s recognizing a handful of consistent ones. And patterns travel — when the metals went parabolic this year, the equity playbooks he’d built let his team trade silver futures across overnight and Asian sessions, a product they don’t normally touch (with risk dialed down because it was new).

The piece he says traders are abandoning, and the one he calls the core of SMB:

Screen recording review. Watching the tape. Understanding what happens before something breaks out.

Not the breakout itself — the before. The tight consolidation under a level that’s been tested a couple times, then the break. Watch that enough and it becomes muscle memory, so the next time it appears live you recognize it from two weeks ago.

He closes on routine. Pay now or pay later, his high-school math teacher told him, and paying later costs ten times more. Build the processes, the playbook, the rules, the accountability partner (a “pod” of other traders giving you more than your own two eyes) at the start of your career, when the lifting is heaviest. You might get away without them for a while. He’s just curious how long you’ll last.

Key Takeaways

The two non-negotiables are edge and risk management. Edge without risk control gets wiped out by one day; risk control without edge bleeds out slowly. You need both.
Expected value, not win rate, defines a good trade. Weight reward by win probability, subtract loss weighted by loss probability. A 40%-win trade can be highly profitable; a 70%-win trade can lose money.
You can’t compute EV without tracking your own stats — average win, average loss, best and worst setups, by ticker and time of day.
Drawdowns recover asymmetrically. Lose 50% and you need a 100% gain to break even. Lose 10% and you need 11.1%. The hole steepens past 50%, which is why capital preservation dominates.
Two kinds of loss: expected (planned, took the stop — fine) and unexpected (impulsive, unplanned — these blow up accounts). The unplanned “invented” trade is the career-ender.
Asymmetric sizing is the core formula. Grade setups A+ / A / B+ / B and size exponentially: ~100%+ of your stop on rare A+ setups, ~5% on everyday B setups.
The 65/5 rule: roughly 65% of annual profit comes from the ~5% of trades that are top-conviction A+ setups. Be risk-seeking precisely on those rare events.
Study winners as hard as losers. Most traders only autopsy losses; the edge is hiding in understanding exactly how the wins happened.
Risk allocation should breathe — bump size up after proven consistency, cut 50% (and again 50%) after hitting weekly/monthly stops.
A playbook is a databased archive of 50-100 setup examples; a small team needs only ~6 core patterns, and patterns transfer across equities, futures, and crypto.
Screen-recording / tape review of what happens before a breakout — the tight consolidation under a tested level — is how pattern recognition becomes muscle memory.
Write down your tilt triggers (the physical ones) and step away the moment you feel them.

Claude’s Take

This is a competent, well-organized version of advice that’s circulated in day-trading circles for years — Mike Bellafiore’s SMB has been preaching playbooks, A+ setups, and tape review since One Good Trade. Nothing here is wrong, and the asymmetric-sizing and 65/5 framing is genuinely the most useful idea in the talk: the insight that profitability is concentration, not consistency, that you survive on small bets and get paid on rare big ones. That generalizes well beyond trading.

The weak spots are worth naming. It’s a recruiting talk as much as a teaching one — the constant “raise your hand,” the networking pitch, the pod-trading encouragement all serve SMB’s funnel. The EV example’s numbers don’t fully reconcile ($500/$200 opens it, $200/$120 closes it), which is a small thing but telling in a talk about being intimate with your statistics. The AI section is enthusiastic to the point of hand-waving — “just paste your P&L into a chatbot” is fine for surfacing obvious clusters, less fine as the rigorous stat-tracking he otherwise demands. And the survivorship problem haunts the whole genre: the 13-year trader with a 2.23 Sharpe and a $2.3M day is shown as proof the system works, but firms like this onboard many and graduate few. The advice is real; the implied success rate is not.

Six out of ten. Solid, honest within its frame, and the sizing framework alone earns the time. Just read it as one capable practitioner’s operating manual, not a formula — the title oversells, as titles do.