Leveraged ETFs delivered big index gains, but single-stock funds face steeper hurdles

Leveraged ETFs are neither uniformly destructive nor uniformly beneficial, according to a preprint by Chris Murray and Marco Sammon of Harvard Business School. Their analysis finds substantial aggregate gains in leveraged equity-index funds, alongside higher costs and more difficult return hurdles for single-stock products.

The distinction matters because both can advertise the same daily leverage target. A fund seeking twice an index’s daily return and one seeking twice a volatile stock’s daily return share a basic objective, but their longer-term economics can differ sharply. The study, posted on arXiv, examines historical outcomes rather than establishing what either category will earn next.

Measuring gains against an alternative

The researchers assembled daily-reset leveraged and inverse funds and exchange-traded notes covering 2006 through September 2025. They combined CRSP returns with Bloomberg shares-outstanding data and public fund disclosures. Products had to reach at least $30 million in assets in one calendar month. Weekly- and monthly-reset products, and those using options as their primary leverage mechanism, were excluded.

Their aggregate profit-and-loss measure multiplies each fund’s beginning-of-day assets by its daily return, then sums across funds and days. This weights performance by the dollars exposed when gains or losses occurred; it is not simply an average fund return.

Long leveraged equity-index ETFs generated $117.8 billion in aggregate investor gains. Single-stock leveraged ETFs generated $13.3 billion. Across all long leveraged products, gains totaled $114.8 billion, with losses in commodity and volatility-linked products offsetting some gains elsewhere.

To assess performance against ordinary exposure, the authors constructed hypothetical investments in the underlying assets. These received identical dollar inflows on identical dates, while withdrawals removed the same percentage of assets as redemptions from the leveraged fund.

In the subset with matched underlying-return data, equity-index leveraged ETFs generated $107.6 billion, versus $61.1 billion for the unlevered counterfactual—a $46.5 billion difference. Single-stock products generated $13.1 billion against $11.7 billion. Thus, single-stock funds were profitable in aggregate, but their incremental gain over unlevered exposure was much smaller.

Why volatility changes the calculation

A daily leverage target does not promise the same multiple of a benchmark’s annual return. Daily gains and losses compound, making the sequence of returns important.

The paper illustrates this with an asset that rises from $100 to $105 and then returns to $100. Before fees and financing costs, a twice-leveraged fund gains 10% and then loses about 9.52%, leaving it roughly 0.48% down even though the asset finishes unchanged.

The authors emphasize that volatility drag comes from compounding leveraged returns, not uniquely from resetting leverage every day. Their approximation shows that the drag term increases with the square of underlying volatility. Trending markets can nevertheless produce favorable compounding; volatility drag does not guarantee a loss.

In an illustrative calculation assuming a 4% annual risk-free rate, a 1% expense ratio and a 0.5% financing spread, an asset with 5% daily volatility must gain about 98.4% over a year for a twice-leveraged fund merely to match an unlevered investment.

Financing adds another distinction. Expense ratios averaged around 1% for both equity-index and single-stock leveraged funds. Estimated annual funding spreads above the risk-free rate were below 1% for equity-index products, compared with 6.8%–9.8% for twice-leveraged single-stock products, depending on weighting. These spreads were inferred from return shortfalls after accounting for leverage targets, risk-free financing and disclosed fees, rather than directly observed borrowing charges.

Products and trading behavior matter

For eligible single-stock launches from February 2023 onward, the median underlying stock ranked near the 94th percentile for volatility and the 92nd percentile for previous six-month returns among the largest 1,000 stocks. More volatile underlying stocks were also associated with greater fund assets after controlling for company size and fund age.

The authors interpret these patterns as consistent with sponsors catering to demand for attention-grabbing, lottery-like exposures. The associations do not establish investors’ motives or prove that volatility caused greater demand.

Flow regressions found that long leveraged funds generally attracted capital after recent losses and lost capital after gains, unlike the return-chasing pattern in the unlevered comparison sample. However, those flows showed no stable positive relationship with subsequent returns. A separate decomposition also found negative within-fund timing effects.

Historical gains are not a forecast

The sample benefited from strong equity returns, while single-stock funds had a much shorter history, beginning in 2022. Their results therefore cannot be compared as if both categories experienced identical market conditions.

The asset threshold also limits coverage of small products. Fund-level creations and redemptions do not reveal every investor’s trading, and the calculations omit investor taxes and intraday execution costs. Nor does the counterfactual compare alternative ways to obtain leverage, such as margin or futures.

The central finding is conditional: leverage helped in favorable equity-index conditions, but extending it to volatile stocks raised the hurdle investors needed to clear.