Position Sizing

Position sizing is the capital-allocation decision that connects an investor’s edge to actual portfolio results. E153.股神的牌局：复利公式 + 凯利公式 argues that even a positive-expectation idea can produce poor outcomes if the investor sizes too large, uses unnecessary leverage, or adds after the original edge has weakened.

E144.交易的艺术：不预测，统计优势，分散红利，随机波动 adds the No-Prediction Trading version: when the trader does not expect any single signal to be right, small initial exposure and broad participation become the mechanism that turns losses into feedback rather than portfolio damage.

E160.一个价值投资者的 20 年回顾：求积分，求胜率，求时间 adds the long-only value-investing version: position size expresses the intersection of business-quality conviction and price protection. A 10% position should require a strong Margin Of Safety and low permanent-loss probability, a 5% position can reflect either quality with weaker valuation protection or flaws offset by cheapness, and smaller positions can keep uncertain companies under live research.

泡沫的四个必要不充分条件 | 对谈经济学者朱宁教授 adds 朱宁 / Zhu Ning’s consequence-based version. When investors cannot know whether AI or another hot market is in a bubble, the sizing question becomes more useful than the prediction question: what happens if the position doubles, and what happens if it falls sharply.

Key Claims

• The episode treats position size as one multiplier in Compounding Growth Formula, alongside Investment Edge, opportunity density, and time.
• Kelly Criterion is the mathematical reference point, but the source favors fractional Kelly because real probabilities and payoffs are uncertain.
• A trade record can reveal that the investor’s directional judgment is acceptable while the sizing rule is damaging the equity curve.
• Single-position exposure should reflect the strategy’s actual historical win rate, payoff ratio, correlation, liquidity, and emotional tolerance.
• E144’s trend example suggests reducing single-position size and expanding the number of observed or traded instruments when the edge comes from repeated statistical exposure.
• In public-fund Value Investing, single-position exposure also reflects business durability, pessimistic-case return, holder liquidity, and whether the manager can explain the position during underperformance.
• Adding to a position should require new evidence, stronger payoff, or confirmed trend, not just a lower price or a desire to be right.
• Leverage raises the chance that normal volatility becomes forced exit, margin stress, or permanent loss.
• Position sizing is linked to Stop-Loss Discipline: a larger or added position needs a clearer invalidation point and a faster exit when wrong.
• Zhu Ning’s frame makes personal consequence part of sizing: a technically identical trade can deserve different size depending on income, liquidity needs, age, obligations, and ability to recover.
• In potential bubbles, sizing should respond to warning conditions without pretending to know the exact top.

Connections

• Kelly Criterion — formal repeated-game sizing rule.
• Investment Risk Management — broader portfolio discipline that includes sizing, leverage, diversification, and behavior.
• Pyramiding and Averaging Down — contrasting ways to change exposure after entry.
• Trend Following — framework where sizing can expand only after market confirmation.
• No-Prediction Trading — frame where sizing keeps repeated signals survivable.
• Passive Investing — alternative for investors who do not have enough edge to justify active sizing decisions.
• Margin Of Safety, Business Moat, and Value Trap — E160’s quality, price-protection, and cheapness-failure inputs to sizing.
• 朱宁 / Zhu Ning, Bubble Necessary Conditions, and AI Equity Valuation Risk — 42章经 interview extension around bubble uncertainty and personal risk capacity.