The Accidental Star Manager
...that earned nearly twice the NASDAQ's return, without even trying
There are times intuition fails you. Plenty of times, actually. But with funds you kind of form heuristics based on patterns you’ve observed repeatedly. For example, fund companies have had a tendency to launch narrow or theme-based products late in a cycle, after that area or idea has largely played itself out.
Well, it doesn’t get any narrower than single-stock leveraged ETFs. And so I reasoned the firms launching those ETFs had done so at inopportune times, just before the stocks in question rolled over.
And…I could not have been more wrong: This is the growth of $10,000 invested in a hypothetical equal-weighted portfolio consisting of the stocks that leveraged single-stock ETFs have referenced since the first such ETF launched in August 20221.
This portfolio—which was not leveraged as these are simply the stocks those leveraged ETFs reference—would have gained 38.9% per year through April 10, 2026, more than tripling in value over that span. That would have been almost double the NASDAQ’s return over that period.
Moreover, the strategy would have been very consistent, topping the NASDAQ in all but one rolling 52-week period since the start, the average outperformance margin being more than 37 percentage points.
What drove performance? I ran a simple attribution and while the portfolio eventually came to hold 179 stocks (corresponding to the 179 unique single-stock leveraged ETFs that had launched by March 2026), these were the five names that made the biggest contribution to returns.
Nvidia | Purchased 12/12/22 | 7.6% avg. weight | 895.6% return
Meta | Purchased 12/12/22 | 7.9% avg. weight | 402.6% return
Coinbase | Purchased 8/8/22 | 9.0% avg. weight | 78.1% return
Palantir | Purchased 9/4/24 | 1.0% avg. weight | 378.2% return
Strategy | Purchased 8/14/24 | 1.2% avg. weight | -4.3% return
(Cumulative returns shown. Data as of Mar. 31, 2026. I’ll explain Strategy further below.)
Accidental Star Manager
I wouldn’t say the stocks these firms selected for these ETFs, or when they chose to launch the ETFs that reference them, was random.
The reference stocks tend to be streaky and boast deep options markets while striking the kind of profile that would resonate with retail-trader types. As for timing, firms typically launched daily leveraged single-stock ETFs after the reference stock had gone on a run.
To illustrate, the chart below plots each reference stock’s return in the six-months leading up to the ETF’s launch (x-axis) as well as its annual returns since then (y-axis). About two-thirds of the stocks had notched a gain in the period before the ETF’s launch, the average return being 78% (19% median).
All told, nearly half the stocks referenced by a leveraged single-stock ETF went on to earn a positive return after the ETF launched, with the average winning stock gaining more on average (40.6%) than the average losing stock lost (-26.0%). Remarkable numbers, the stuff of a star manager2.
Sad Trombone
That said, there also would have been some luck involved. Consider that only nine leveraged single-stock ETFs launched in 2022, none in 2023, and 13 in 2024. That means the portfolio would have been concentrated initially in a small number of stocks which, it so happens, excelled: 19 of 22 earned a positive return after the ETF referencing it launched, with the average name gaining almost 30% per year.
Not only that, the portfolio would have held its biggest weightings in these stocks in the 2022 - 2024 period, when they were on a tear, and smaller weightings when returns moderated in 2025 and this year. Was that prescience? No. It was mechanics — as more ETFs launched, the portfolio would have added more stocks (i.e., the names those new ETFs were referencing), diluting the original stocks’ weightings.
Strategy is a good example of how fortuitously things would have worked out. The portfolio would have bought the stock in Aug. 2024 (8.3% weight) when the first leveraged MSTR ETF incepted. That purchase would have come just before the stock went parabolic, soaring 250% over the next three months.
As we know, Strategy got crushed subsequently but that selloff didn’t begin in earnest until July 2025, by which time the portfolio would have cut its stake (1.4% weight). Why? Because by then many other leveraged single-stock ETFs had been minted and, thus, the stocks they referenced would joined the portfolio, claiming their share of its weight and reducing Strategy’s.
In addition, while the portfolio was a model of consistency, the stocks these ETFs referenced got iffier as time went on. For instance, 60% of the stocks referenced by leveraged single-stock ETFs that launched last year and thus far in 2026 have lost money in the time since the ETF launched, with the average name up only 1.9%.
You’d think this would have sunk this hypothetical portfolio last year. But good fortune would have smiled upon it once again: The strategy would have added quantum-computing play D-wave Quantum on Apr. 24, 2025 (1.9% weight), just before it soared to a 528.1% gain over the next five-plus months.
That stock has since lost nearly two-thirds of its value but that wouldn’t have been as costly to the portfolio as it could have been since it would have already trimmed its stake by the time that drawdown began in October 2025. Again, the reason being other single-stock leveraged ETFs had launched by then, siphoning weight away.
Bottom-line
Was it all just a fluke? I regressed the portfolio’s returns against a four-factor model. While it wasn’t the tightest of fits (0.82 adjusted r-squared), it was still revealing: The portfolio would have had a positive, but statistically insignificant alpha, with most of its return explained by big loadings to high-beta (1.29), high-momentum (0.54) stocks.
Bottom-line: The sponsors of these leveraged single-stock ETFs have shown a flair for picking stocks about to pop. Yet, it’s questionable whether that’s repeatable3. Which in a way makes it not all too different from actual star managers.
Other Reading
I appreciate everyone who follows this Substack. I also write for my employer, Morningstar. You can find all my Morningstar articles here.
Here’s a short run-down of recent pieces I’ve written, fwiw.
“You Can Beat the Stock Market by Avoiding Its Worst Days. But You Won’t”
“The Biggest Active Stock Funds Picked the Right Stocks. They Still Lagged.”
The views and opinions expressed in this blog post are those of Jeffrey Ptak and do not necessarily reflect those of Morningstar Research Services or its affiliates.
Here’s the full list of ETFs, reference stocks, and ETF inception dates (each reference stock was added to the hypothetical portfolio on its ETF’s inception date). When there were multiple leveraged single-stock ETFs that referenced the same stock, I chose the one with the earliest inception date.
It technically wasn’t a single manager, but rather nine different firms/brands: Leverage Shares (49 ETFs), Tradr (42), Defiance (32), GraniteShares (20), T-REX (16), Direxion (14), KraneShares (4), ProShares (1), and REX (1).
In case you’re a daring type and wondering whether there are any investible analogs for the hypothetical portfolio I constructed, the short answer is no: When I regressed the entire ETF universe against the hypo’s monthly returns, the closest matches were VanEck Social Sentiment ETF (BUZZ), which had an 85 r-squared, and SoFi Social 50 ETF (SFYF), which had an 83 r-squared. These ETFs performed well, nearly matching the NASDAQ’s return over the Aug. 2022 - Apr. 2026 period, but gained only about half as much as the hypo.