digest
Short Summary
Main recommendation: Ben Felix argues that dividends are irrelevant for explaining differences in expected returns, so portfolios should not be built or spending planned around dividend yield or dividend income. Instead, investors should target the drivers of expected returns (market beta, size, relative price, profitability, investment) directly and use a total-return spending plan.
Key Points
- Fama and French's 1993 three-factor model showed, and Felix's own update using the five-factor model confirmed, that after controlling for market beta, size, relative price, profitability, and investment, high-dividend portfolios have alphas statistically indistinguishable from zero.
- Investors who treat dividends as "free money" (the "free dividends fallacy," per Hartzmark and Solomon) overpay for dividend stocks when yield is in high demand, reducing expected returns by roughly 2 to 4% per year.
- Dividend-based portfolios tend to be underdiversified: only about 38% of US stocks paid dividends at the end of 2021, and dividend strategies often underweight small companies, where Blitz and Hanauer (2020) find value and profitability premiums tend to be larger.
- Dividend-focused portfolios create uncompensated sector bets, and domestic dividends' tax advantages make global diversification more costly for dividend investors.
- Consumption tied to dividends is excessively sensitive to dividend receipts, and dividends are an arbitrary, corporate-determined spending rule rather than one tailored to individual preferences.
- Dividends are not a reliable inflation hedge: US dividend stocks show no strong evidence of better real performance during high inflation or rising rates.
Details Worth Noting
- Concrete instructions: Rather than building around dividends, target the underlying drivers directly: value and profitability premiums, with attention to small-cap exposure, and diversify globally and across industries. For spending, define a desired consumption path, calculate how much can be spent sustainably from the total portfolio, and spend that amount regardless of whether returns come as dividends or capital gains. If you want variable spending without relying on dividends, one example given: start at 4% of a $1 million portfolio ($40,000). If the portfolio falls to $800,000, cut spending by no more than 5% year over year to $38,000, rather than cutting to $32,000. The ceiling and floor can be adjusted to personal preference.
- The five-factor model used by Felix includes profitability and investment, and the dividend-based portfolios he tested were an international developed high dividend index and a US high dividend index over their full histories.
- In a simulation of rolling 30-year withdrawal periods using the top 30% highest dividend yield US stocks (1927 to March 2022) compared against 1,000 variable spending strategies, spending dividends produced lower ending net worth for a given consumption level, lower ending net worth for a given consumption variability, and roughly equivalent total consumption for a given consumption variability—making it typically an inefficient way to meet a planning objective.
- Small-cap dividend growth funds tend to have higher fees and turnover without delivering expected returns unavailable from lower-cost small-cap value funds.
- Felix notes that some dividend-paying stocks have attractive characteristics, which can be targeted directly without a dividend filter, and that the optimal spending strategy depends on individual preferences, which corporate dividend policies ignore.
Bottom Line
Dividend yield is a noisy, costly proxy for the factors that actually drive returns, so investors who want those returns should target the factors directly and plan retirement spending on total returns.
