What is a leveraged ETF?
A leveraged ETF targets a multiple of its benchmark index’s daily return. A conventional 1× index ETF seeks to track the index. A 2× or 3× fund targets twice or three times its daily return before fees and expenses. A 1% daily index gain implies a target gain of about 2% or 3%; a 1% decline implies a larger loss.
Funds use derivatives such as swaps and futures to create exposure larger than their net assets. A swap exchanges an agreed investment return with a counterparty. A futures contract sets a transaction for a future date. Funds combine these contracts with cash instruments and sometimes stocks. Derivatives also create counterparty and liquidity risks.
This article covers funds with daily targets. A daily multiple does not guarantee a multiple of the index’s ten-year return. QLD and TQQQ target 2× and 3× the daily Nasdaq-100 return before fees and expenses. An intraday holding period can also differ from the close-to-close target.
Why is a 3× ETF’s long-term return not three times the index return?
The fund adjusts exposure to its changed net assets each day. This is a daily reset. Today’s profit or loss changes tomorrow’s starting capital. Daily growth factors must therefore be multiplied, not added.
| Daily target | Day 1 | Day 2 | Return |
|---|---|---|---|
| 1× | 110.00 | 99.00 | -1.00% |
| 2× | 120.00 | 96.00 | -4.00% |
| 3× | 130.00 | 91.00 | -9.00% |
The 3× ETF earns 30 on day one, taking its wealth to 130. On day two it loses 30% of that larger balance: 39. Equal percentage gains and losses therefore produce unequal amounts. An index ending down 1% does not mean the 3× ETF loses only 3%.
Over ten years, each day’s return still applies to the wealth left by the previous day. Multiplying the index’s final return by three cannot capture that process.
What is volatility drag?
In the earlier example, the arithmetic average daily return was zero, yet wealth fell. This reduction in compound growth caused by fluctuations is called volatility drag. It occurs even when no fees are charged.
For an exact round trip of +25% followed by −20%, the index ends at 100. The 2× and 3× funds end at 90 and 70. With a smaller round trip of +5% followed by −4.7619…%, the 3× result is about 98.57. Both index paths end where they started, but larger swings cause a larger leveraged loss. Repeated round trips apply further losses to the remaining wealth.
Volatility drag does not mean the investment must finish at a loss. A strong upward trend can outweigh it. The index’s final return alone cannot explain the compounded result of a leveraged ETF.
Can a rising market deliver more than three times the index return?
Each row below applies the same daily index returns to all three multiples. Rows are separate two-day examples. They exclude costs and tracking differences.
| Index daily path | 1× | 2× | 3× |
|---|---|---|---|
| 10% → 10% | 21.00% | 44.00% | 69.00% |
| 25% → -20% | 0.00% | -10.00% | -30.00% |
| -10% → -10% | -19.00% | -36.00% | -51.00% |
In consecutive rises, 3× returns 1.3 × 1.3 − 1 = +69%, above three times the index’s +21%, or +63%. First-day gains earn further gains. In consecutive falls, the 3× loss is −51%, less negative than three times −19%, or −57%. It still loses more than half of starting capital. Doing better than that arithmetic comparison does not make the outcome safe.
What if the falling day comes first? Changing +10% then −10% to −10% then +10% still leaves the 3× fund at 91 in this model. The same numbers are multiplied in a different order. This assumes fixed daily leverage with no cash flows or costs. Interim wealth, drawdowns and recovery paths can differ. Contributions, withdrawals or changes in leverage can make order affect final wealth too.
Does that make all flat-ending markets equivalent? Two zero-return days and a +25%/−20% round trip both leave the index unchanged. But their 3× outcomes are 100 and 70. Reordering identical daily returns is a different question from changing the size of the swings.
Try two daily returns
Keep day one at +10%. Change day two from −10% to +10% and compare final wealth. How does another rising day change the 3× result?
Two rising days · Rise, then back to the start · Two falling days
| Daily target | Day 1 wealth | Final wealth | Return |
|---|---|---|---|
| 1× | 110.00 | 121.00 | 21.00% |
| 2× | 120.00 | 144.00 | 44.00% |
| 3× (selected) | 130.00 | 169.00 | 69.00% |
Formula and limits
Starting wealth is 100 units, not a currency amount; no FX conversion is applied. This daily-reset model excludes deposits, withdrawals, dividends, management expenses, financing costs and tracking differences. For daily decimal returns r1 and r2 and multiple L: final wealth = 100 × (1 + Lr1) × (1 + Lr2). Cumulative return = (final wealth / 100 − 1) × 100%. Daily factors are multiplied without intermediate rounding; displayed values are rounded to two decimal places. Over longer periods, multiply every daily factor (1 + Lr), then subtract 1 to get the cumulative return before costs. Returns are limited to −30% through +30% and L to 1, 2 or 3 to keep daily asset values positive. This teaching example does not model a fund closing, intraday rebalancing or a market crash beyond that range.
Why is recovery after a crash so difficult?
The loss applies to original wealth, but the recovery gain applies to what remains. After 100 falls to 20, recovering the missing 80 requires four times the remaining capital.
| Loss | Required gain |
|---|---|
| -50% | 100.00% |
| -80% | 400.00% |
| -90% | 900.00% |
The required gain is loss percentage ÷ (100 − loss percentage) × 100. An equal loss requires an equal recovery gain regardless of leverage. Leveraged funds can reach deep losses much faster. After a total loss, returns on the remaining zero balance cannot restore the capital.
Maximum drawdown measures the largest fall from a prior peak. Recovery time measures how long it takes to regain that peak. A high final return may hide a deep decline and a long period underwater. Use the MDD calculator and loss recovery calculator for these distinct questions.
How different were QQQ, QLD and TQQQ in practice?
What if you invested USD 1,000 in each of QQQ, QLD and TQQQ on the same date, then added USD 500 a month?
Final wealth was $57,861.38 for QQQ, $82,008.13 for QLD and $105,862.75 for TQQQ.
2021-09-22 to 2026-09-22 · USD 31,500 contributed to each fund · dividends reinvested · data as of .
These are ending balances for a monthly investment plan, not five-year buy-and-hold ETF returns.
The same contribution plan produced very different balances. Because the results include additional investments, the ratio between balances cannot be read as a multiple of long-term ETF returns. See the QQQ, QLD and TQQQ comparison for detailed performance and drawdowns.
Calculation assumptions for the actual ETF example
Amounts are in USD with no FX conversion. Personal tax and trading commissions are excluded. Internal fund costs remain reflected in actual prices and are not deducted again. The calculation allows fractional purchases, recognizes dividend cash on the ex-dividend date and reinvests it on the next trading day. This historical example uses matched dates and contribution plans; it does not predict future balances.
For semiconductors, see the SOXX and SOXL comparison. Official pages checked on October 29, 2026 identify the current benchmark of both funds as the NYSE Semiconductor Index. SOXL targets 300% of that index’s daily return before costs, not three times the daily price return of SOXX itself. Historical comparisons must check the benchmark used at the time, changes in benchmarks, costs, distributions and tracking differences. The SPY, SSO and UPRO comparison covers the S&P 500 family.
Is long-term investing in leveraged ETFs always worse?
In the two rising days above, the 3× fund earned more than a simple multiple of the index return. Sustained rises can also make compounding favorable over longer holding periods. Large swings and deep declines can reduce the gains accumulated along the way.
Actual funds also incur management expenses and financing costs. The former pay for operating the fund; the latter pay for extra exposure through instruments such as swaps. Financing costs depend on rates and contract terms and are not all necessarily included in the stated expense ratio. Both are separate from the compounding effect shown above.
In the QQQ and TQQQ backtest, keep the contribution plan and dividend setting fixed and move the start date to just before a major decline. Then try a date after the decline. Compare ending wealth and the time needed to regain the previous peak. One favorable period cannot establish a universally best multiple or holding period. The leveraged ETF effect calculator shows realized return multiples for a chosen period. Its price returns differ from dividend-reinvested total returns.
The practical leveraged ETF guide helps you choose a comparison and review drawdown, recovery and fund conditions.
Frequently asked questions
Does a daily reset mean I must trade the ETF every day?
No. The fund manager adjusts exposure inside the fund to maintain its target multiple. Investors do not need to trade the ETF every day to perform that reset. The risk of losses while holding it remains.
Will an intraday trade return three times the index return?
Not necessarily. The daily target generally runs from the previous market close to the current close. Your return between an intraday purchase and sale can differ from three times the index return over those same hours.
Is volatility drag the same as the expense ratio?
No. Volatility drag comes from fluctuations and compounding. It exists even without management or financing costs. Actual fund costs affect results separately.
Is SOXL three times SOXX itself?
No. Both official pages listed the NYSE Semiconductor Index as the current benchmark on October 29, 2026. SOXL targets 300% of that index’s daily return, not three times the market-price return of SOXX.
Official sources and dates
Structure and benchmark sources checked: 2026-10-29. Hypothetical calculations model daily targets without costs. The actual monthly investment example is dated 2026-09-22 and is not live performance.