Holding Leveraged ETFs Long Term Isn't a Maths Problem. It's a You Problem.

I would have bought this.

That's the honest starting position. At twenty I put everything I had into one stock and it worked, which is the worst possible outcome, and the invoice for that belief turned up a few years later in a different currency. If a 2x version of that trade had sat one tap away, inside the same app as my boring money, I'd have tapped it. So when I read the 2026 arguments for holding leveraged ETFs long term, I don't hear a stranger being reckless. I hear a more articulate version of me at twenty-two, with better sources.

Which is where almost everyone writing about this goes wrong. They argue about the arithmetic, get it wrong in both directions, and stop before reaching the only variable that has ever decided anyone's outcome. The maths of leveraged index investing is genuinely two-sided. The behaviour is not. A 2x product does not double your return.

It doubles you.

What changed in 2026 was the packaging, not the leverage

In 2009 the entire US leveraged ETF category held about $30bn. ETF industry trackers put it at a record $198bn in June 2026, up from roughly $170bn at the end of 2025. That's trade data, not an official number, but the direction isn't in dispute.

The launch numbers made me sit up. 229 leveraged ETFs launched in all of 2025; by mid-August 2026 the count was 244, a record, and in June leveraged products were as much as half of every new US ETF launch. One startup issuer put out 127 of them this year, averaging $1m each.

Data: Morningstar Direct, as reported August 2026.

The loudest moment of the year was a 2x SpaceX product, SPCH, which recorded the largest first week of trading volume in ETF history: more than $4bn across four sessions. That's volume, not money raised — appetite for a doubled bet on a company that had barely reached public markets.

Compare that to how you used to buy leverage. You signed a margin agreement, you saw an interest rate printed on a statement every month, and a broker on the other end could liquidate you. The paperwork alone bought you time to think. All of that friction was accidental, and all of it was load-bearing. Now the same exposure sits in the same app, in the same list, one row below the world fund you buy every month. Same font. Same green button.

Amrita Nandakumar of Vident says issuers are now "scraping the bottom of the barrel," moving from Nvidia and Tesla toward smaller, untested names. Meanwhile 63 leveraged single-stock ETFs closed in the US this year against three in all of 2025, and half the category now holds under $7m. GraniteShares CEO Will Rhind puts the operating rule plainly: "If a fund is below break even or shows no real signs of adoption by the market, we'll close it." His firm liquidated its 2x Lucid fund after the shares fell about 51% in one session in July.

"Hold it for the long term" quietly assumes the product survives to the long term. A lot of these won't.

Can you hold leveraged ETFs long term? The manufacturer's own document says one day

I read a product page instead of a journalist's summary of one. WisdomTree's S&P 500 3x Daily Leveraged ETP is passported into fifteen European countries and listed on the LSE, Xetra, Euronext Paris and Borsa Italiana. Its own page states a recommended holding period of one day, and adds that WisdomTree "is not promoting or marketing" these products to retail clients. That's not a compliance hedge, it's the design spec. If your intended holding period and the manufacturer's differ by four orders of magnitude, the conversation is over before it starts.

European rules force issuers to put the reason in the name: ESMA requires funds that rebalance daily to carry the word "Daily."

Here's the mechanic, using the SEC's own worked example. An index falls 10%, 1,000 to 900, and the 2x fund falls 20%, exactly as promised. Next day the index rises 10% back to 990 and the fund rises 20% to 960, again exactly as promised. Over the two days the index lost 1% and the fund lost 4%. It kept every promise it made and handed you four times the loss instead of two.

That isn't a fee and it isn't a scandal. It's what resetting the multiplier every evening does to a bumpy path: the drag is a function of how choppy the road was, not which direction you faced.

The SEC and FINRA publish two real four-month cases. They're worse than any thought experiment.

Data: SEC Office of Investor Education & Advocacy / FINRA, Updated Investor Bulletin.

Now the correction most sceptical articles refuse to make.

Leveraged ETFs do not always lose money over time. In a smooth uptrend the daily reset compounds hard in your favour, and that is the entire appeal. The 3x Nasdaq fund fell 79.08% in calendar 2022, hit a maximum drawdown of 81.66%, took roughly two years to claw back to par, then returned 198% in 2023. Anyone who bought it at inception in 2010 and genuinely never touched it has beaten my boring approach by a margin no amount of decay commentary erases.

So the honest version has two clauses. The maths can work, and it charges you for every bump in the road. Some of that charge isn't even on the fact sheet: the fund costs 0.82% against 0.18% for a plain Nasdaq tracker, and on top sits the swap financing, which moves with short-term rates and never appears as a fee line. Leverage isn't free — it just doesn't send you a bill you can read.

The strongest case for leveraged ETFs long term is a case against these products

I'd rather not hand you a strawman, so here's the opposing case at full strength. It's real academic work, not a forum meme.

Ian Ayres and Barry Nalebuff at Yale make the lifecycle case. A 25-year-old with forty years of future savings already holds an enormous, low-volatility, bond-like asset: their own future contributions. They can't sell it and can't rebalance it, so their lifetime portfolio is overweight that bond-like thing and underweight equities. Cramming all their stock exposure into the last twenty working years isn't safe, it's a failure to diversify across time. The fix is roughly 200% equity exposure while young, ramping down toward 50% near retirement. Using US data back to 1871 they report expected retirement wealth around 90% higher than a target-date fund, at a lower standard deviation than a conventional 75/25 portfolio — numbers I'm taking from the book rather than a paper I read line by line.

I find it persuasive. Read carefully, it's also an argument against the products this article is about.

Ayres and Nalebuff cap the strategy at 2:1, and their vehicles are deep-in-the-money LEAPs or a plain margin account. Not daily-reset funds. The best academic case for lifetime leverage explicitly avoids the thing being sold to retail in 2026, because the daily reset is what turns a leverage ratio into a volatility tax.

The other serious paper lands in the same place from a different angle. Gayed and Bilello apply leverage only when the index sits above its moving average, and hold T-bills otherwise. Their finding is the honest statement of the arithmetic: volatility is the enemy of leverage, trends are its friend. Their conclusion isn't "buy and hold 3x," it's "hold leverage conditionally, on a mechanical rule" — another way of saying leverage works when a system removes you from the decision. Which is my own thesis arriving from the opposite direction.

Leverage doesn't multiply returns. It multiplies the investor.

Here's the number this article exists for.

Morningstar's Mind the Gap 2026 looked at ten leveraged single-stock ETFs over the three years to 31 December 2025. Apple, Amazon, Alphabet, Microsoft, Tesla, Nvidia, Meta, Coinbase. Doubled exposure, during one of the great bull runs in market history. The dollar-weighted return, meaning what the humans holding them actually earned, lagged the plain unleveraged shares by one percentage point a year.

They bought 2x exposure to the best stocks of the decade and finished behind simply owning the stocks.

The extreme case is almost funny. GraniteShares 2x Long COIN Daily returned 47.6% a year over that period. The average dollar invested in it lost more than 38% a year. Same fund, same window, an 86-point spread, and nobody was defrauded. The fund did its job perfectly. The gap is made entirely of when people bought and when they sold.

None of this is unique to leverage. It's just concentrated there.

Data: Morningstar, Mind the Gap 2026.

Across everything Morningstar tracks, the gap runs from 0.4 percentage points a year in the calmest quintile of funds to 2.1 in the most volatile. The progression isn't a clean staircase and I'm not smoothing it, because the messy version is the true one. The shape is unmistakable though, and volatility is what leverage manufactures by construction. Buy a 2x product and you've bought yourself a gap multiplier.

The category with a gap of exactly zero over ten years, incidentally, was large blend: average dollar 14.0%, funds 14.0%. I went through the full version of that gap separately.

Two pieces of honesty before I lean on any of it. Jeffrey Ptak, who wrote the Morningstar study, says himself that "even laudable practices like investing a portion of every paycheck or regularly rebalancing can open a gap between investor results and reported total returns," so some of the gap is structural rather than proof of stupidity. And there's a rebuttal in the Financial Analysts Journal this year from Fulkerson, Jordan, Riley and Yan, titled "Bad Timing Does Not Cost Investors 15% of Their Funds' Returns." I haven't read it in full and won't guess at its argument from a title. The behaviour gap is a contested measurement, not settled physics.

What I trust more is the panic-selling research, because it counts actions rather than returns. Elkind, Kaminski, Lo and colleagues went through 653,455 brokerage accounts and found that 30.9% of investors who panic-sell never return to risky assets. Not "buy back late." Never come back.

Then the finding that should be printed on the front of every one of these products. The people who freak out more often are older, married, have more dependents, and self-identify as having excellent investment knowledge. Confidence isn't protection. It's a risk factor. That study is describing me at twenty-two and I don't enjoy it either. The version of this written for the moment it actually matters is over here.

This is not an American argument

Almost everything published on this is written in tickers a European reader can't buy, which is a shame, because Europe sells the same machine.

Data: justETF ETF profile data, September 2026.

A 2x Nasdaq fund holding €1.27bn. A 2x S&P fund holding €550m. Leverage Shares hit a record $1.6bn in group assets in May, did $10.7bn of European turnover in the first half of 2026, then listed the first 3x SpaceX product on Xetra in July.

UCITS rules cap fund leverage well below 3x, which is precisely why the 3x products are sold as ETPs and ETNs: you're buying a debt security backed by a swap, not a fund holding shares. That distinction is invisible on a broker's search screen and extremely visible on a bad day.

The same WisdomTree page discloses a 20% intraday restrike threshold and a 25% overnight gap threshold: the levels at which the product stops being a position and becomes an event. If that sounds theoretical, the inverse-VIX note XIV held $1.9bn on a Friday in February 2018, fell 96.3% on the Monday, and was liquidated inside a fortnight. Its trigger was written down in advance too.

The behaviour crosses the border as cleanly as the product does. The Bank for International Settlements looked at the silver rush of early 2026 and put a central bank's name on the mechanism. Silver doubled over 2025, rose more than 50% in January 2026, then fell about 30% in a single day. The BIS found that leveraged ETF daily rebalancing and margin-triggered liquidations amplified the swings, that these funds do "predictable, momentum-like trading," buying as prices rise and selling as they fall, and that their rebalancing multiplier doubled over 2025. Retail drove the inflows; institutions held steady or trimmed.

Thousands of people thought they were expressing a view on inflation and the dollar. What they had bought, without ever deciding to, was a position in their own reaction time.

If you're going to do it anyway, contain it

I'm not going to pretend nobody reading this will buy one. That's how every article on the subject ends, and it's why none of them change anything.

So, without the moralising:

• Read the recommended holding period in the provider's own document before you read a single return. If it says one day and you're planning ten years, that's the whole conversation.
• Decide the size once, as a percentage of net worth, and write it down. That rule only means something if you know your net worth, which is roughly why I ended up building a tracker. Pick the number while calm and never top it up because it fell.
• No standing order into it. Automating into a boring fund automates you past your worst instincts. Automating into a daily-reset product automates you into the decay.
• Never money with a date on it. That 2022 recovery took roughly two years of continuous holding. If the money belongs to a deposit, a car or a child, the recovery window isn't available to you at any price.

Write the exit rule before you buy, in one sentence, and date it. The only real requirement is that the calm version of you wrote it.

Then count it as spending you might get back rather than as investing. That isn't semantics. It decides whether the money shows up in your plan's return assumptions, and so whether a 60% drawdown moves your retirement date or just ruins a weekend. Leave the boring core alone while you find out.

Where I think I'm wrong

Three places, and I'd rather say them than have someone say them for me.

"You can't handle it" is a claim about people, not about mathematics. Someone who held through 2008, 2020 and 2022 without selling has evidence about their own behaviour that a warning article doesn't have. If that's you, you know something about yourself I can't argue with from here.

I'm also already leveraged, and so is almost everyone reading this. We have a mortgage on the house we live in and we're slowly paying off a second one: a large, concentrated, illiquid position bought with borrowed money. Objecting to somebody's 2x sleeve while holding that would be a bit rich. The "diversified" world fund I buy is also more concentrated than its name suggests, which I argued out in why an index fund is less spread out than it looks.

The third one genuinely bothers me. My conclusion comes from a sample of one life, and it's a life where things went fine. I won on the stock, got out of crypto with my capital intact, and now write articles recommending caution. Somebody who lost early would write this same essay with the same confidence and the opposite moral. Which is why the argument rests on 653,455 accounts rather than on my twenties.

What I actually do on the first of the month

Nothing. That's the system.

A standing order fires, one accumulating world fund gets bought, and I'm not required to have an opinion that week. It has survived every hype cycle I've lived through for one unglamorous reason: it makes no promises I ever have to defend. Nothing to be right about, so nothing to double down on at the worst possible moment.

The other thing about holding leveraged ETFs long term, which nobody frames this way, is that it isn't a portfolio question. It's a plan question. A 60% drawdown on a tenth of your net worth moves your date by a knowable number of months, and seeing that number needs the budget, the net worth and the projection in one place rather than three apps that don't talk to each other. Until you've seen it you're not sizing a position, you're guessing at one. I sit down with mine maybe a dozen times a year. A position you check twelve times a day is not the same asset class as one you check twelve times a year, whatever the fact sheet says, and your checking frequency changes long before your returns do.

If you're up big on something with a multiplier in its name right now, I hope it keeps running. But write down today what you'd say the reason was. Then come back in three years and see whether the reason was ever really about you.