Stay-in-the-Game Position Sizing
Size every bet so one mistake cannot end your investing career
- Difficulty
- Moderate
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 98%
This position-sizing framework treats investing as a repeated game in which survival matters more than winning every round. For each opportunity, estimate the odds and plausible downside using current information, then cap exposure so an adverse result cannot destroy the portfolio or end the manager's ability to continue. The same rule applies even when conviction is high, because confidence does not eliminate uncertainty. Portfolio-level stress testing should also account for correlations that can turn several apparently separate positions into one concentrated bet. Decisions are evaluated based on the information and sizing logic available beforehand, not on regret after a profitable position proves too small. Preserving capital, track record, and client confidence keeps future opportunities available.
Origin
Paul McCulley credited Bill Gross with teaching him to avoid gambler's ruin through disciplined bet sizing at PIMCO.
Core principles
- 01Winning more often than losing is enough
- 02No single position should threaten survival
- 03Calculate odds with current information
- 04Judge sizing before learning the outcome
- 05Longevity compounds future opportunities
How to run it
- 1
Frame the repeated game
Treat the opportunity as one of many future decisions rather than a single chance to prove yourself.
Pro tip Define success as durable compounding, not winning every trade.
Watch out A one-shot mentality encourages excessive concentration.
- 2
Estimate the odds
Assess potential outcomes, probabilities, and what evidence could change them in real time.
Pro tip Use ranges rather than a single forecast.
Watch out Do not convert strong conviction into false certainty.
- 3
Quantify survivable loss
Determine how much capital, alpha, liquidity, or client confidence the portfolio can lose without permanent impairment.
Pro tip Include business consequences as well as mark-to-market loss.
Watch out A financially survivable loss may still be career-ending if it violates mandates.
- 4
Set the position cap
Choose a size whose adverse outcome remains below the survival threshold.
Pro tip Size from downside first, then examine upside.
Watch out Do not size positions solely from expected return.
- 5
Stress the whole portfolio
Test whether correlated positions could fail together and collectively exceed the loss limit.
Pro tip Group holdings by shared economic driver.
Watch out Nominal diversification can conceal one large macro bet.
- 6
Evaluate without hindsight
After the result, review the assumptions and process rather than regretting that a winner was not larger.
Pro tip Keep a contemporaneous sizing memo.
Watch out Outcome bias can train managers to take ruinous risk after a lucky win.
In the wild
A bond manager sees an attractive relative-value trade but recognizes that fixed-income alpha is small and client mandates are sensitive to drawdowns. The position is capped so a wrong call reduces performance without destroying the track record or undermining the next client competition.
→ The manager preserves longevity and the ability to earn future alpha.
An investor believes a technology company is substantially undervalued but limits it to a survivable allocation after modeling a severe earnings miss and correlation with other growth holdings. The stock falls sharply, yet the portfolio remains intact and can rebalance at lower prices.
→ A mistaken thesis produces a manageable loss rather than permanent impairment.
Common mistakes
Sizing from confidence alone
High confidence does not remove uncertainty, and a concentrated error can eliminate the chance to recover.
Regretting a small winner
A position that later succeeds was not necessarily undersized if a larger loss would have threatened survival.
Ignoring correlated exposure
Several modest positions can combine into one ruinous bet when they share the same underlying driver.
Is it for you?
Best for
It is best for portfolio managers, traders, allocators, and entrepreneurs making repeated uncertain bets.
Not ideal for
It is not ideal for guaranteed obligations that cannot be diversified or resized.
From the transcript
“I don't have the objective of winning all the time, and I want to calculate the odds in real time. Uh. And also I want…”
“And Bill Gross was the guy who fundamentally drove that into my head just in sizing of bets, of sizing in a portfolio.”
“No, you shouldn't have made it bigger because that because if you were wrong, it had taken you out.”
From the episode
An Interview With Paul McCulley: Masters in Business (Audio)