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

What It Covers

Hank Bessembinder joins Rational Reminder to examine constant leverage ETFs for single stocks, breaking down how daily rebalancing, volatility, and reversals drive underperformance of 0.79%–1.0% per month versus a frictionless benchmark, then introduces two new return measures—sustainable return and proportional sustainable return—to better capture long-horizon investor outcomes beyond arithmetic and geometric means.

Key Questions Answered

• •Volatility decay misconception: Volatility does not inevitably drag down mean returns in constant leverage ETFs. Daily rebalancing functions as a momentum trade—buying after gains, selling after losses. Whether this helps or hurts depends entirely on whether the underlying asset experiences return continuations or reversals. High volatility amplifies whichever pattern dominates, meaning the "volatility decay" label overstates the certainty of harm from daily resets.
• •Leveraged single stock ETF costs: Bessembinder's sample of 35 single-stock leveraged ETFs launched since 2022 underperforms a frictionless leveraged buy-and-hold benchmark by 0.79% per month for long funds and 1.0% per month for short funds—equivalent to roughly 9.5 and 12 percentage points annually. For long funds, two-thirds of underperformance comes from frictions (fees, swap borrowing costs above the federal funds rate); one-third from rebalancing.
• •Short vs. long leverage asymmetry: Inverse leveraged ETFs face larger rebalancing trades than long ETFs because the required trade size scales with the leverage ratio squared minus the leverage ratio (L²−L). For negative leverage ratios, this formula produces larger values, meaning short funds incur greater rebalancing costs. Empirically, about three-quarters of short fund underperformance stems from rebalancing, versus only one-quarter from frictions like fees and swap rates.
• •Arithmetic means and alpha limitations: Arithmetic means, Sharpe ratios, and factor alphas all derive from single-period CAPM logic and make no attempt to capture multi-period investor outcomes. The arithmetic mean only accurately represents returns for an investor who rebalances every period to maintain a constant dollar position—a strategy almost no one uses. Investors evaluating long-horizon performance should treat these metrics as incomplete rather than definitive measures of wealth accumulation.
• •Sustainable return framework: Bessembinder introduces the sustainable return: the constant periodic withdrawal, expressed as a percentage of initial investment, that leaves terminal portfolio value equal to starting value—essentially how much could be spent without eroding capital. Mathematically, the expected sustainable return nearly equals the expected geometric mean. The proportional variant—withdrawing a fixed percentage of current portfolio value each period—equals the geometric mean divided by one plus the geometric mean, eliminating ruin risk.
• •Dollar-weighted returns for real investors: Dollar-weighted returns (IRRs) better capture actual investor experience than buy-and-hold geometric means because they account for the timing and size of cash flows in and out. Modified IRRs improve further by avoiding the unrealistic assumption that interim cash flows are reinvested at the IRR rate. Bessembinder advocates for more academic research using these measures, noting they are already used in practice by Morningstar's Mind the Gap series and Vanguard account reporting.