Episode 399: James Choi - Portfolio Theory in a Spreadsheet

What It Covers

Yale finance professor James Choi presents his paper "Practical Finance," which approximates the complex life cycle portfolio choice problem into a free Google Sheets tool. The model incorporates human capital as a bond-like asset to determine optimal stock-versus-risk-free allocation across different ages, wealth levels, risk aversion scores, and labor income characteristics.

Key Questions Answered

• •Human Capital as a Bond: Most portfolio advice ignores human capital, which functions like a bond in your total wealth portfolio. A 30-year-old with $2.2M in projected future earnings and only $500K saved should hold 91% equities in their financial portfolio—not because they need returns, but because their implicit bond position (human capital) already dominates their total wealth allocation.
• •Wealth Level Inverts Common Advice: Higher accumulated financial wealth relative to future human capital means you should hold *less* equity, not more. Two 40-year-olds with identical salaries should differ in equity allocation based on savings: the wealthier one holds a larger fraction of total wealth in financial assets, reducing the need for aggressive equity exposure in the portfolio.
• •Permanent vs. Transitory Income Risk: Labor income risk splits into transitory (layoffs, short gaps) and permanent (career trajectory, promotions). Permanent income risk meaningfully reduces optimal equity allocation; transitory risk barely affects it. Counterintuitively, high school graduates—who face more transitory but less permanent income risk—should hold more equities than college graduates, all else equal.
• •Welfare Cost of Common Rules: Following "100 minus age" in equities costs roughly 2% of lifetime welfare versus the optimal strategy. A fixed 60/40 allocation costs 3.75%. Being 100% cash costs 7.9%. Being 100% equities costs 11.8% on average—but only 0.56% for someone with risk aversion of 4, a realistic level, making all-equity portfolios defensible for many investors throughout working life.
• •Measuring Your Risk Aversion: Use this thought experiment to find your risk aversion score (1–10): imagine a coin flip between living on $100K or $50K for one year. Identify the guaranteed amount that makes you indifferent to the gamble. A risk-neutral person accepts $75K; more risk-averse individuals require less. The paper provides a lookup table converting that threshold amount into a risk aversion coefficient for the spreadsheet.
• •Approximate Solution Accuracy: Choi's model solves thousands of parameter combinations numerically, then fits simple regression-based formulas to approximate results. The average deviation from the true numerical optimum is 3–4 percentage points in equity allocation. Following the approximate strategy instead of the exact optimal produces less than 0.1% lifetime welfare loss—compared to 2–12% losses from common rules of thumb.