Leveraged ETFs Explained: Daily Reset, Volatility Decay and Why 3x Isn't 3x Over Time
A leveraged ETF promises a multiple of an index’s return for one day. A 3x fund on the Nasdaq-100 aims to rise 3 percent when the index rises 1 percent on a given day, and to fall 3 percent when it falls 1 percent. An inverse fund does the same with a minus sign. The key words are “for one day”. Over a week, a month or a year the fund can return far more or far less than three times the index, and in a choppy market it can lose money while the index goes nowhere. Anyone who holds these products for longer than a session needs to understand why.
How the daily reset works
Each fund targets a fixed multiple of the index’s return from one close to the next. To keep that multiple constant, it has to rebalance every afternoon. When the index rises, the fund’s net assets grow, so it must add exposure to stay at 3x. When the index falls, the fund’s assets shrink and it must cut exposure. The fund buys after up days and sells after down days, every day, mechanically.
That rebalancing is the reason the multiple only holds for one day. Each day’s return is applied to a new starting value, so returns compound. Here is the simplest case: the index rises 10 percent, then falls back to where it started.
| Index | 2x fund | 3x fund | -3x fund | |
|---|---|---|---|---|
| Start | 100.00 | 100.00 | 100.00 | 100.00 |
| Day 1 (index +10%) | 110.00 | 120.00 | 130.00 | 70.00 |
| Day 2 (index -9.09%) | 100.00 | 98.18 | 94.55 | 89.09 |
| Two-day return | 0.00% | -1.82% | -5.45% | -10.91% |
The arithmetic on day 2: to get from 110 back to 100 the index falls 10/110, or 9.0909 percent. The 3x fund falls three times that, 27.27 percent, so 130 × (1 − 0.2727) = 94.55. The 2x fund falls 18.18 percent: 120 × 0.8182 = 98.18. The -3x fund rises 27.27 percent from 70: 70 × 1.2727 = 89.09. The index is flat. Every leveraged version is down, and the more leverage, the bigger the loss.
Stretch it out and the effect grows. Suppose the index spends ten days alternating between +5 percent and a drop back to its starting level (a fall of 4.76 percent). After ten days the index is exactly where it began. The 2x fund is down 2.4 percent, the 3x fund is down 6.9 percent and the -3x fund is down 13.5 percent.
Compounding cuts both ways, though. In a steady trend it helps. If the index rises 2 percent five days in a row, it gains 10.41 percent. Three times that is 31.22 percent, but the 3x fund gains 33.82 percent (1.06 to the fifth power), because each day’s gain builds on a larger base. Leveraged funds beat their stated multiple in smooth trends and fall short of it in choppy ones. What decides the outcome is the path the index took, and the same start and end points can produce very different fund returns. That is what people mean by path dependence.
Volatility decay, in numbers
The drag from choppy markets has a neat approximation. Over a period, a fund with leverage L tends to return about L times the index’s log return, minus a penalty of (L² − L) ÷ 2 × σ² × t, where σ is the index’s annualised volatility and t is time in years. Fees and financing come on top.
The penalty grows with the square of volatility and faster than linearly with leverage. For a 2x fund the multiplier is 1; for 3x it is 3; for -1x it is 1; for -3x it is 6. That last figure explains why triple inverse funds bleed so badly.
Here is what the penalty means over one year when the index finishes exactly where it started, before any fees:
| Index volatility | 2x fund | 3x fund | -3x fund |
|---|---|---|---|
| 15% (a calm year for the S&P 500) | -2.2% | -6.5% | -12.6% |
| 30% (a rough year, or a typical year for chip stocks) | -8.6% | -23.7% | -41.7% |
| 45% (stressed semiconductor market) | -18.3% | -45.5% | -70.3% |
| 60% (single volatile stock) | -30.2% | -66.0% | -88.5% |
A computer simulation of thousands of random one-year paths with a flat ending index gives the same answer: a 3x fund at 30 percent volatility ends around 76 cents on the dollar, and at 45 percent volatility around 55 cents. So “3x isn’t 3x over time” is plain arithmetic. A volatile underlying with no trend is the worst possible setting for a leveraged fund, and it can drain one while the index chart looks harmless.
This is also why the VIX matters to anyone holding these funds. Implied volatility is a decent read on how much drag the market expects over the next month.
What’s inside the fund: swaps, futures and costs
Leveraged ETFs rarely hold three dollars of stock for every dollar of assets. Most of the exposure comes from total return swaps with banks, plus index futures where a liquid contract exists. ProShares UltraPro QQQ (TQQQ), the largest of the group with about $40 billion in assets, holds Nasdaq-100 swaps with several bank counterparties alongside E-mini Nasdaq-100 futures. The fund’s cash sits as collateral in Treasury bills and money market instruments.
Swaps carry two costs that never show up in the expense ratio. The first is financing. A bank that gives a fund three times its assets in index exposure charges a floating rate on that notional, plus a spread. The fund earns interest on its cash collateral, but only on 1x; it pays on 3x. The net cost is roughly two times the short-term rate plus the bank spread. With the federal funds target at 3.75 to 4.00 percent after the Federal Reserve’s September 2026 hike, that works out to about 7.5 to 8 percent a year before the spread, a quiet headwind that rises and falls with the short end of the yield curve. The second is counterparty exposure to the banks on the other side, which funds manage with collateral and by spreading swaps across several dealers.
Then there are fees. Direxion Daily Semiconductor Bull 3X (SOXL) charges a 0.75 percent net expense ratio, and its inverse twin SOXS charges 1.00 percent. Daily rebalancing also adds trading costs, and big funds trading at the close can move prices in the final minutes of the session.
Single-stock, thematic and inverse funds
SOXL and SOXS, both launched on March 11, 2010, track three times the daily move of the NYSE Semiconductor Index, up and down. Chip stocks are volatile on their own, and an index of them still swings far more than the S&P 500. That makes SOXL the textbook illustration of both sides: it can more than triple in a strong chip year, and from its late-2021 peak to its October 2022 low it fell roughly 90 percent while the semiconductor index lost less than half. Readers following the AI chip trade (see our explainer on high bandwidth memory) will recognise the names inside that index.
Single-stock leveraged ETFs arrived in the US in mid-2022 and now cover names such as Nvidia, Tesla and many others, mostly at 2x long or -1x short. One stock is more volatile than an index, so the decay table’s bottom row applies. A 2x fund on a stock with 60 percent volatility gives up about 30 percent a year to drag in a flat market.
Inverse funds deserve their own warning. A -1x fund looks like a simple short position, but it isn’t. A short seller’s exposure shrinks as the stock falls; an inverse ETF resets to full -1x every day. Even at -1x the drag multiplier is the same as a 2x long fund. At -3x it is six times the variance, double the 3x long figure. Holding SQQQ or SOXS for months as a “hedge” has historically been a reliable way to lose money even when the hedge was directionally right for part of the time. The February 2018 collapse of the XIV inverse volatility note, which lost more than 90 percent of its value overnight and was then shut down, is the extreme version of an inverse product meeting a volatility spike. Spikes like that also tempt traders into buying 3x funds at the bottom, a timing bet our look at the VIX “buy when it spikes” rule tests against the data.
The SEC’s rules: Rule 18f-4 and the 2x line
US funds that use derivatives fall under Rule 18f-4, adopted by the SEC on October 28, 2020, with compliance required from August 19, 2022. Its main test is value at risk (VaR), a statistical estimate of how much a portfolio could lose on a bad day. A fund’s VaR generally can’t exceed 200 percent of the VaR of its designated reference portfolio, usually the unleveraged index or asset it tracks. In practice that caps new leveraged funds at about 2x. A 3x fund’s VaR is about three times its index’s, so it fails.
Funds already operating at more than 2x on October 28, 2020 were allowed to continue, on condition that they don’t change their underlying index or increase their leverage. That is why TQQQ, SOXL, SQQQ and the other established 3x funds still trade, while new 3x launches in the US have stalled.
The SEC drew the line more firmly in December 2025, when it sent nearly identical letters to nine issuers, including Direxion, ProShares, GraniteShares and Tidal, about proposed funds offering more than 2x exposure on single stocks, indexes and crypto. The staff told them to revise the filings to comply or withdraw the products. As of October 2026, 2x remains the practical ceiling for new US leveraged ETFs.
When it adopted Rule 18f-4, the SEC also considered “sales practices” rules that would have required brokers to vet retail investors before letting them buy leveraged and inverse funds. It dropped them, relying on Regulation Best Interest and existing broker duties instead. FINRA has warned since 2009 that these products are generally unsuitable for retail investors who plan to hold them longer than a day, and many brokers make customers acknowledge the risks before a first trade. Margin rules matter too for active traders using them; see our note on FINRA ending the pattern day trader rule.
Who they suit, and who they don’t
Leveraged ETFs work as designed for:
- Day traders and short-term tactical traders who want amplified exposure without a margin account or options, and who close positions within days.
- Investors with a strong view on a trend over a short window, who accept that sideways chop will cost them.
- Professionals hedging for a day or two, who size the hedge and roll it deliberately.
They tend to hurt:
- Long-term holders who treat a 3x fund as “the index, but more”, especially on volatile underlyings.
- Anyone using inverse funds as a long-running portfolio hedge.
- Buyers who average down into a falling leveraged fund, since the fund itself has already been cutting exposure into the decline.
Some investors do hold 2x or 3x broad index funds for years and do well, because the Nasdaq-100 has spent long stretches in low-volatility uptrends where compounding helps. That record came with drawdowns near 80 percent along the way, and it depends heavily on which years you picked. If you want a quick read on market mood before taking on leverage, the CNN Fear & Greed Index and market breadth indicators are reasonable places to start. For the jargon traders throw around these products, see our Wall Street trading slang guide.
The practical test before buying one: estimate the underlying’s volatility, plug it into (L² − L) ÷ 2 × σ², and ask whether your expected move over your holding period is large enough to clear that drag plus 7.5 to 8 percent a year in financing and the expense ratio. If the answer depends on the market going straight up, size the position accordingly, and check it every day, because the fund resets every day whether you’re watching or not.