Jim Simons: 5 Principles for a Statistically Sound Stock Portfolio

Factor-Based Investing

• 💡 Diversification is about independence, not just owning many things; focus on independent sources of return.
• 🎯 Stop picking stocks based on stories or narratives, as these are not an edge or alpha.
• 📊 Select stocks based on statistical properties that predict future returns, such as value, momentum, quality, and size.
• ✅ Portfolios built on these measurable characteristics consistently outperform those based on compelling narratives, on average.

Prudent Position Sizing

• ⚠️ Accept that you will be wrong constantly and build a portfolio designed to survive your ignorance.
• 🔑 Position sizing is more critical than position selection; avoid high concentration in any single idea.
• 📉 Never let any single position exceed 5% of your portfolio at cost, with 2-3% being even more rigorous.
• 🛡️ Spreading risk across many independent bets ensures that being wrong about individual holdings doesn't destroy your overall portfolio.

Systematic Rebalancing

• 🔄 Rebalancing is a mathematical necessity, as markets drift and unbalance portfolios over time.
• 📈 Emotional investors often fail to sell winners, passively increasing concentration in positions with lower expected returns.
• 🎯 Systematically rebalance based on deviations from target weightings, trimming positions that exceed their allocation and adding to those that fall below.
• ✅ This disciplined approach forces you to sell strength and buy weakness, compounding returns over decades.

True Diversification

• 🧩 Correlation is paramount; owning stocks in different sectors or countries doesn't guarantee diversification if they move together.
• 📉 During periods of market stress, correlations surge, causing seemingly diversified portfolios to collapse in unison.
• 🔍 Measure the actual statistical relationships between your positions to ensure genuine independence.
• 🚀 Seek exposure to multiple statistical factors (e.g., trend following, value, short-term momentum, credit risk) for real diversification, rather than just owning many stocks.

Long-Term Probabilistic Mindset

• ⏳ You must ignore short-term results and optimize for long-term probability, accepting periods of underperformance.
• 🧠 Understand that probability-based investing means expecting to lose a certain percentage of the time (e.g., 40% for a 60% win rate).
• 🛑 Avoid the temptation to abandon a sound strategy during losing months or years and chasing whatever is currently performing well.
• ⚙️ Treat your portfolio as a machine that runs on probabilities, not a scoreboard measuring your intelligence, to achieve long-term outperformance.