Episode 397: Hendrik Bessembinder - Constant Leverage & Measuring Investor Outcomes

Episode Summary

Executive Summary: This episode explores Hank Bessembinder’s research on leveraged single-stock ETFs and new measures of investor outcomes. He argues that volatility alone is not the issue; serial correlation, fees, borrowing costs, and daily rebalancing determine outcomes. The second half shifts to a critique of conventional return metrics and introduces sustainable return and proportional sustainable return as more practical, cash-flow-aware ways to measure long-term investing success.

Main Topics: Leveraged single-stock ETFs and constant leverage (Priority: 5/5): Bessembinder explains how constant-leverage products work, why their daily rebalancing creates momentum-like effects, and why these products are especially risky on individual stocks versus broad indexes. Volatility, skewness, and the mean-median gap (Priority: 5/5): He distinguishes between volatility as a driver of the gap between average and median outcomes and volatility as a possible amplifier of gains or losses depending on reversals versus continuations. Benchmarking and decomposition of ETF underperformance (Priority: 5/5): The paper benchmarks levered ETFs against a frictionless buy-and-hold leveraged position and decomposes underperformance into rebalancing effects and implementation frictions. Risks and behavior of levered single-stock ETFs (Priority: 4/5): The discussion emphasizes overconfidence, speculative use cases, and the possibility that some products can exceed -100% target losses, especially in extreme moves. Measuring long-term investor returns (Priority: 5/5): Bessembinder critiques arithmetic means, geometric means, and log returns as imperfect for long-horizon investor experience, arguing they do not truly capture what investors earn over time. Dollar-weighted returns, IRRs, and modified IRRs (Priority: 4/5): He favors cash-flow-aware measures more than standard IRRs, noting that modified IRRs better respect actual reinvestment assumptions and cash flows in/out of portfolios. Sustainable return and proportional sustainable return (Priority: 5/5): He introduces new ex post measures that ask how much could be withdrawn while preserving ending capital, aiming to better connect finance research to retirement and spending needs.

Key Arguments: Volatility does not automatically hurt constant-leverage ETFs; its effect depends on whether underlying prices tend to continue or reverse after rebalancing. Daily rebalancing is not inherently bad. It can help in continuation regimes and hurt in reversal regimes. For levered products, the more important real-world drags are often fees, borrowing costs embedded in swaps, and trading frictions. Single-stock leveraged ETFs are especially dangerous because individual stocks have higher volatility and more skewness than indexes. Past research has been too focused on arithmetic means, geometric means, and factor alphas, which do not directly measure a long-term investor’s lived outcome. Dollar-weighted and modified IRR measures are better aligned with actual investor cash flows than simple average returns. Sustainable return reframes performance around how much cash flow an investment could support without reducing principal. The aggregate investor is not a buy-and-hold investor because dividends, new issues, and repurchases mean cash moves in and out of the market. Overconfidence and the desire for lottery-like payoffs make many investors vulnerable to complex leveraged products. The expected sustainable return is numerically very close to the expected geometric mean, but sustainable return is framed around spending and ending wealth preservation.

Data Points: Leveraged single-stock ETF sample: 35 funds - Current study sample with reasonably long time series; more funds were launched later and were not yet included. Long levered ETF underperformance: 0.79% per month - Average underperformance versus the frictionless leveraged benchmark for long single-stock ETFs. Short levered ETF underperformance: 1.0% per month - Average underperformance versus the frictionless leveraged benchmark for short single-stock ETFs. Annualized long underperformance: More than 9 percentage points per year - Benjamin Felix highlighted the annual implication of a 0.79% monthly shortfall. Annualized short underperformance: About 12 percentage points per year - Benjamin Felix highlighted the annual implication of a 1.0% monthly shortfall. Long ETF underperformance decomposition: ~0.25% per month rebalancing; >0.5% per month frictions - For long products, most underperformance came from implementation frictions rather than rebalancing. Short ETF underperformance decomposition: ~0.75% per month rebalancing; ~0.25% per month frictions - For inverse products, rebalancing accounted for most of the shortfall. Hypothetical leveraged ETF blow-up frequency: About 5 times per day on average - In a 50-year historical simulation across stocks, a -100% or worse target return would have occurred frequently under hypothetical leverage on all stocks. Aggregate response to high volatility: High volatility often coincided with reversals - Bessembinder found empirically that high-volatility periods tended also to be reversal periods, which worsens leveraged ETF outcomes. Expected sustainable return: Almost precisely the expected geometric mean - The paper’s sustainable-return framework yields an expected value nearly identical to the geometric mean. Proportional sustainable return formula: Geometric mean / (1 + geometric mean) - Defined as a constant percentage withdrawal rate that preserves ending capital in expectation.

Pivotal Quotes: "volatility is not my favorite expression here because it has this aura of inevitability about it" — Hank Bessembinder: He is explaining why he avoids treating volatility as the sole or inevitable cause of poor leveraged-product performance. "the daily rebalance doesn't necessarily hurt you" — Hank Bessembinder: Central point of the leveraged-ETF discussion: daily resetting can help or hurt depending on whether returns continue or reverse. "we've been searching under the econometric Streetlight" — Hank Bessembinder: His critique of academic finance metrics: researchers often measure what is easy, not what best reflects long-run investor outcomes.

Implications: Investors should be cautious with leveraged single-stock ETFs and with return metrics that ignore cash flows. The episode pushes finance toward spending-aware, investor-centric measurement and suggests future research should compare index and single-stock leverage side by side.