Compounding Process, Not PnL Streaks

Compounding process vs PnL streak means your journal should stack evidence about whether you kept constraints—budget, horizon, shape, monetization—not a highlighted win chain. Streaks make vivid stories. Process metrics make reusable lessons. ConvexClub sides with the duller object because asymmetric learning dies when a green week becomes identity.

Direct answer: treat win/loss streaks as unreliable teachers; log whether the five process fields held. Psychology literature on the "hot hand" shows people over-read dependence in sequences. Use that as a warning label on streak screenshots, not as proof markets are basketball.

In this article

1. Why streaks feel like skill
2. What the hot-hand studies actually claimed
3. What should compound in a journal instead
4. How streak culture corrupts notes
5. Process metrics you can actually count
6. Hot hand vs gambler's fallacy in the same week
7. Desk risk language without cosplaying a bank
8. A short anti-streak ritual
9. Try this yourself
10. Frequently asked questions

Why streaks feel like skill

Short sequences look patterned. A few wins feel like a regime change; a few losses feel like a curse. Tip chats and PnL cards exploit that fluency. The human brain prefers a story with momentum over a spreadsheet with constraints.

Hartford Funds' investor-behavior note on hot-hand and gambler's fallacies (citing the same research line) frames the investing translation as extrapolation bias: expecting the recent path to continue. Whether or not any particular market series has real serial correlation is not the point of a process journal. The point is that belief without measurement is how size creeps up after three greens.

What the hot-hand studies actually claimed

Thomas Gilovich, Robert Vallone, and Amos Tversky's classic 1985 Cognitive Psychology paper, "The hot hand in basketball: On the misperception of random sequences," tested whether hit probability truly rose after hits among professional and controlled shooting samples. Analyzing Philadelphia 76ers shooting logs, Boston Celtics free throws, and Cornell varsity experiments, they reported that players and fans believed in streak dependence far more than the shot data supported—a misperception of chance in short sequences. Later work (for example Ayton & Fischer, 2004) discusses hot-hand and gambler's fallacies as related faces of subjective randomness.

Markets are not free throws. Treat the citation as a cognitive warning: people confidently narrate streaks; measurement often disagrees. Your journal should assume that warning applies to you until process metrics say otherwise.

What should compound in a journal instead

Compound the boring adherence checks:

• Did budget as percent of liquid NW stay inside the rule? (budget post)
• Did horizon rewrites happen when calendars slipped?
• Did risk–return shape match the instrument?
• Did monetization stay executable?
• Did lessons get written in past tense after messy outcomes?

A year of constraint adherence—even with mediocre marks—is a compoundable artifact. A year of undocumented streak mythology is entertainment.

Related: process edge vs information edge myth and tip groups vs process journals.

How streak culture corrupts notes

• Upsizing after wins without rewriting risk–return shape.
• Deleting losing tickets so the streak screenshot stays clean.
• Calling randomness "being in flow."
• Treating one regime's greens as portable skill across domains.

Past-tense antidote: log the streak belief itself as a field ("I upsized because of three wins") so future you can audit the heuristic.

Process metrics you can actually count

Streaks are easy to count and hard to trust. Process metrics are slightly harder and much more reusable:

• Percent of experiences with a completed five-factor packet before entry.
• Percent with horizon rewrites logged within 72 hours of a calendar slip.
• Percent where max-loss language matched instrument shape.
• Number of deleted or rewritten notes after losses (integrity signal).
• Ratio of past-tense lessons to forward-looking pep talks.

None of these guarantee PnL. All of them compound into a record you can audit. That is the product point of ConvexClub: lived experience density, not a casino loyalty card.

Hot hand vs gambler's fallacy in the same week

People oscillate. After wins: hot hand ("keep pressing"). After losses: gambler's fallacy ("due for a win") or cold-hand narrative ("everything is broken"). Ayton and Fischer discuss these as related faces of how humans read randomness. Your journal can capture the switch—the moment you changed size because of a streak story—regardless of which fallacy costume you wore.

Desk risk language without cosplaying a bank

Public bank risk manuals, where available, obsess over limits, breaches, and escalation—not celebratory streaks. Credit Suisse's Archegos materials (see the budget guide) are extreme, but the pattern is familiar: process breaches stacked while profits looked fine. Retail streak culture is a miniature of that blindness.

You do not need a risk committee. You need a rule that forbids upsizing solely because the last three tickets were green—unless a new packet rewrites budget and shape for an independent reason.

A short anti-streak ritual

1. After three consecutive wins, force a blank day or a packet rewrite—your choice, but pick one rule.
2. Screenshot the streak belief in the note ("I feel unstoppable") so future you can cringe productively.
3. Grade the week on constraint adherence first, marks second.

If that ritual feels anti-fun, good. Journals are not content feeds.

Try this yourself

Open your last green run. Write whether size, horizon, or shape changed because of the streak. Log that confession as a past-tense experience on ConvexClub. The goal is not humility theater; it is measurement.

Frequently asked questions

Does compounding process raise expected returns?

We do not claim that. We claim undocumented streak narratives make learning accidental. See CLAIMS mindset in product copy: no tip performance promises.

Are all streaks illusions?

No. Some series have real dependence. The journal still needs constraint fields—because even "real" streaks do not automatically justify unbounded size.

What if my community ranks PnL streaks?

That is a social object, not a process object. Prefer circles and notes that discuss logged experiences—see Partners—over leaderboard cosplay.

Next steps

Start free · Pro · Five factors