The importance of having true Conviction when Investing
Conviction = you’ve mapped power + incentives, identified who can change the rules, and sized a position as if those actors will keep doing what pays them. It is not faith. It’s alignment math.
What conviction is (and isn’t)
Conviction = you’ve mapped power + incentives, identified who can change the rules, and sized a position as if those actors will keep doing what pays them. It is not faith. It’s alignment math.
Lens: incentives > ideals; control > fairness; stability > truth.
Rule: always read revealed preferences (budgets, defaults, standards, AUPs, procurement, index rules), not speeches.
Why conviction matters
Markets are “efficient to constraints”, not to truth. Pricing lags where career risk, optics, and mandates prevent sizing. Conviction lets you buy where others can’t write the memo yet.
Volatility taxes the unprepared. Big winners travel through deep drawdowns. Weak hands sell to the strong because they never built the why or the when to be wrong.
Edge compounds only if held. If you’re right on the structure but exit on optics, you donate your edge to indexers later.
How to build real conviction (not cope)
1) Map the power stack
Who sets the dials? (regulators, app stores, clouds, banks, exchanges)
Which knobs exist? (ID mandates, provenance, blacklist APIs, issuance mix, margin rules)
What’s the ratchet? (temporary “pilots” that won’t sunset once behavior locks)
If your thesis survives knob turns, it’s structural. If it needs “they won’t dare”, it’s hopium.
2) Demand revealed-preference proof
Follow money: budget lines, multi-year RFPs, standards with verbs: attest, revoke, trace, rollback, prove.
Follow defaults: what ships on by default wins (security/compliance toggles, wallet types, custody).
Follow harmonization: same rule appearing across jurisdictions (policy synchronization) = global TAM.
3) Define invariants and falsifiers
Invariants (what must stay true): e.g., “States prefer programmable money over cash”, “Admissible AI beats raw accuracy”, “Identity lineage becomes mandatory”.
Falsifiers (hard kill-switches): e.g., “Major democracies pass binding laws preventing cross-domain data fusion”, “App stores explicitly protect non-KYC wallets long-term”, “Budgets pivot from audit/ID to tax cuts for multiple cycles”.
Write them. Date them. Re-check quarterly.
4) Establish base rates
Study prior ratchets (Patriot Act, e-invoicing/CTC, C2PA-like provenance pushes, Online Safety regimes). The base rate is tightening, not loosening.
Applying the lens (concrete)
Palantir (state-embedded AI)
Incentives > ideals: buyers need admissible, auditable decisions (policy lineage, rollback), not leaderboard accuracy.
Control > fairness: procurement favors fast Authorization to Operate, pre-baked compliance, not “best model”.
Stability > truth: cross-jurisdiction templates beat bespoke deployments.
Revealed preference: growing classified/health/defense renewals; standards drift toward prove/attest/trace.
Conviction triggers: new mandates on AI lineage; e-gov identity expansions; crisis RFPs; app-store/bank Acceptable Use Policy tightenings.
Falsifiers: credible open alternative with court-grade lineage wins gov at scale; Authorization to Operate advantage erodes; budgets rotate away from ID/audit.
Microsoft (compliance default)
Defaults: Entra/Defender/Compliance center ship on; the control plane is the product.
Incentives: CIOs minimize audit pain; spend flows to the checkbox that passes regulators by default.
Falsifiers: regulators mandate open identity that bypasses hyperscalers; OS-level attestation barred.
Bitcoin
Incentives: states prefer Medium-of-Exchange containment + Store-of-Value toleration; paperization dampens upside vol.
Revealed preference: ETF approvals, bank custody, Travel Rule, KYC perimeters.
Conviction structure: self-custody for policy-shock convexity; paper BTC for carry; buy fear/sell clarity, keep core.
Falsifiers: multiple big economies legalize non-KYC Medium-of-Exchange, protect self-custody by law, app stores enshrine non-KYC wallets.
Conviction ≠ stubbornness: the three failsafes
Stop asking “am I right?” — ask “what would change this buyer’s incentive?”
E.g., if courts start rejecting AI outputs without provenance, your non-admissible AI longs are dead.Track PSC/LPI weekly
PSC (Policy Synchronization Coefficient): rate identical rules spread across allies.
LPI (Legibility Pressure Index): density of attest/rollback/prove in RFPs/laws.
Rising PSC/LPI → your control-aligned longs are safe; falling → reassess.
Run “control flips”
What lever (app store, bank, cloud, pool) could flip against you overnight?
If one lever kills your thesis, resize. If your thesis uses the lever, size up.
Why most people can’t hold conviction (and how to outcompete them)
Career risk > truth. Portfolio Managers hug benchmarks; consultants punish weirdness.
Optics > outcomes. They can’t say “technocracy substrate”, so they can’t size it.
Willpower > design. They rely on feelings not frameworks; volatility ejects them.
Your edge: rules, not vibes — pre-committed entries, adds, exits keyed to PSC/LPI, net-liquidity shifts, and falsifier triggers. You don’t need to read headlines.
Red flags that your “conviction” is actually cope
You need public approval to hold.
You can’t state three falsifiers that would make you sell.
The thesis relies on “they won’t dare”.
You’ve never mapped who can change the rules.
You confuse truth with pricing; markets price power first.
Bottom line
Conviction is structural alignment with how power actually behaves, not how it should. Use incentives > ideals; control > fairness; stability > truth as your guardrails. Read defaults, budgets, and standards; ignore speeches. Pre-write invariants/falsifiers and size to survive. Then hold through the noise while the rest of the street waits for permission to agree with you.
None of this should be considered investment advice.