Risk and Variance

Risk isn’t just “bad things happening.” It’s the range of possible outcomes—and your exposure to the ones that matter.

Key Distinctions

Risk vs. Uncertainty

Frank Knight’s foundational distinction (Knight, 1921):

TypeDefinitionExample
RiskKnown probabilitiesCoin flip: 50/50
UncertaintyUnknown probabilitiesWill AI take my job?

You can calculate risk. You can only estimate uncertainty.

Variance vs. Expected Value

Expected value: The average outcome if you repeat something many times. Variance: How much outcomes scatter around that average.

Same expected value, different variance:

  • Option A: Always get $100
  • Option B: 50% chance of $0, 50% chance of $200

Both have $100 expected value. Option B has higher variance. Which is “better” depends on context.

Tail Risk

Most outcomes cluster around the average. Tail risks are the rare extremes that can be catastrophic.

DomainTail RiskMitigation
HealthCancer, accident, chronic diseaseScreening, insurance, safety habits
WealthJob loss, market crash, lawsuitEmergency fund, diversification, insurance
SocialRelationship breakdown, reputation damageMaintenance, boundaries, integrity
MeaningBurnout, depression, crisis of purposeRest, connection, professional help

Rule: Avoid ruin. You can recover from bad outcomes. You can’t recover from catastrophic ones (Taleb, 2012).

Cross-Domain Risk Framework

Evaluate Before Acting

  1. What’s the expected outcome? (average case)
  2. What’s the variance? (range of outcomes)
  3. What’s the worst case? (tail risk)
  4. Can I survive the worst case? (ruin check)
  5. Is the upside worth the downside? (asymmetry)

Seek Asymmetric Bets

Good asymmetry: Limited downside, unlimited upside

  • Learning a skill (time cost, career upside)
  • Starting a side project (small investment, potential big return)

Bad asymmetry: Unlimited downside, limited upside

  • High-leverage speculation
  • Ignoring health for work

Decision Criteria

Take more risk when:

  • You’re young (time to recover)
  • Downside is capped
  • You have margin (emergency fund, health buffer)
  • Upside is large relative to downside

Take less risk when:

  • Failure means ruin
  • You’re responsible for others
  • Recovery time is limited
  • You’re already exposed

The goal isn’t to eliminate risk. It’s to take the right risks—ones with asymmetric upside and survivable downside.

Knight, F. H. (1921). Risk, Uncertainty and Profit. Houghton Mifflin.
Taleb, N. N. (2012). Antifragile: Things That Gain from Disorder. Random House.