Skip to main content

Risk Management

What would guarantee failure — and are you accidentally doing it?

Risk is the probability of permanent loss, not volatility. Volatility is the price of admission. Loss of capital is the game over screen.

Position-Level Risk​

Controls applied to every individual position:

  • Thesis invalidation trigger — Before entering, define what would prove your thesis wrong. When it happens, exit. No renegotiation
  • Time stops — If a position hasn't moved toward thesis within the defined timeframe, re-evaluate. Dead money has opportunity cost
  • Stop losses — Mechanical exits at pre-defined levels. Set them when you're rational, not when you're panicking
  • Concentration limit — No single position above 7% of portfolio unless 5/5 conviction with multi-year track record

Portfolio-Level Risk​

Controls applied across the entire book:

  • Max drawdown limit — Define your personal pain threshold (eg. -25% portfolio). If hit, reduce exposure systematically, don't freeze
  • Concentration caps — No single sector above 30%. No single chain above 40%
  • Correlation check — Monthly review: are your "diversified" positions actually correlated? In crypto, everything correlates in a crash
  • Cash reserve — Maintain 10-20% dry powder. Opportunities come when everyone else is liquidating
  • Leverage ceiling — Only mild leverage during smooth price action. Leveraged positions get cut first in drawdowns

Behavioural Checklist​

Run weekly. Honest self-assessment prevents more losses than any stop-loss:

  • Overconfidence — Am I sizing positions based on how right I feel rather than evidence? Have I stress-tested my thesis recently?
  • Revenge trading — Am I trying to "make back" a loss? The market doesn't know or care about your entry price
  • Herd behaviour — Am I buying because others are buying? Check: can I explain this position without referencing what others are doing?
  • FOMO — Am I entering because the price already moved? The best entry points feel uncomfortable
  • Anchoring — Am I holding because of my entry price rather than current fundamentals? Would I buy this at today's price?
  • Sunk cost — Am I holding a loser because I've "already invested so much"? Past capital is gone. Future capital is what you control

Risk Quantification​

Simplified Kelly Criterion:

Optimal position size = (Win probability x Average win - Loss probability x Average loss) / Average win

In practice, use half-Kelly or quarter-Kelly. Full Kelly assumes perfect probability estimates. You don't have perfect estimates.

Expected value before every trade:

EV = (probability of win x size of win) - (probability of loss x size of loss)

If EV is negative, don't trade. If EV is positive but small, check if the opportunity cost justifies the attention.

The Nature of Risk​

From Howard Marks:

  • Risk is unquantifiable, both in advance and after the fact
  • More things can happen than will happen — the future is a range of possibilities
  • Lower prices make assets less risky. Rising prices make them riskier. This is counterintuitive but fundamental
  • An investment can appear safe for years before its flaws are revealed during a negative event
  • Risk is not a function of asset quality alone — high quality overpriced assets can be riskier than cheap low-quality ones

The Practice​

Weekly contrarian review:

Argue against your biggest position for 15 minutes. Write down the three strongest arguments for selling it. If you can't find three, you're not looking hard enough. If the arguments are stronger than your thesis, act.

Post-decision audit:

After every sell: Was the exit based on your framework or your emotions? Track your batting average on thesis-driven exits vs panic exits. The data will teach you what your gut won't.

Context​

Questions​

If risk is unquantifiable, what does "managing" it actually mean — and is the framework just a comfort blanket?

  • Which item on your behavioural checklist do you consistently fail at, and what has it cost you in the last 12 months?
  • When your stop-loss triggers, do you actually execute — or do you move the goalposts?
  • If you ran the contrarian review against your current largest position right now, what would you find?