Funds · Specialised funds
Leveraged ETF decay: what holders of TQQQ, SSO and SOXL actually got
Sixteen years of TQQQ, SSO, SPXL and SOXL against their multiple of the index: how often holders beat it, how far they fell short, and the 2020 SEC rule.
On this page
A leveraged ETF promises a multiple of an index for one day. Anyone who holds it for longer gets something else: the product of several hundred one-day results, which can land well above or well below the multiple applied to the whole period. The shortfall is usually called decay. The daily arithmetic behind it is explained in Leveraged and inverse ETFs: daily reset math and volatility drag, and the formula that links it to volatility is in Leverage and volatility drag: why 2× risk is not 2× return. Neither is repeated here.
This lesson looks at the record instead. It takes four large funds with long histories, the 3× Nasdaq-100 fund TQQQ, the 2× and 3× S&P 500 funds SSO and SPXL, and the 3× semiconductor fund SOXL, and measures what a holder received over every 12-month period since each fund began. The results are less one-sided than the word decay suggests. Over the whole period these funds made far more than their indexes. Over a typical year they fell short of the multiple, sometimes by 40 points, and each of them at some point lost between 60% and 87% of its value.
All returns below are price changes. Dividends are left out on both sides, which makes little difference to the leveraged funds and slightly flatters them against index funds that pay about 1% a year.
The idea
The multiple in a leveraged fund's name applies to one trading day. ProShares states the objective of TQQQ as three times the daily performance of the Nasdaq-100, and the fund's page adds a sentence every holder should read: "For any holding period other than a day, your return may be higher or lower than the Daily Target." To keep the multiple constant the fund resets its exposure at each close, adding after a rise and cutting after a fall.
Over many days that reset compounds. A run of gains is compounded on a growing base, so the fund ends above the simple multiple. A rise and fall of equal size leaves the index roughly where it began and the fund lower. Which effect wins over a year depends on how far the index moved against how rough the path was. The measure used throughout this lesson is the gap between the two.
The gap has a second, steadier part. A 3× fund holds swaps and futures worth three times its assets, and the counterparties charge interest on the borrowed two-thirds. Add the fund's own expenses (TQQQ lists a gross expense ratio of 0.94% and SSO 0.88%) and there is a cost that grows with short-term interest rates and is paid whether the index moves or not.
How it works: sixteen years of TQQQ
TQQQ began trading on February 9, 2010. From the end of March 2010 to the end of October 2026 its price multiplied 214 times. QQQ multiplied 14.3 times. Expressed as a yearly rate that is 38.2% against 17.4%: a ratio of 2.2, bought with a fall of 79.2% along the way against 33.1% for QQQ.
The chart below plots every 12-month period in that history. Each dot is one period: QQQ's return across, TQQQ's return up. The line is three times QQQ.
Of the 188 periods, TQQQ beat three times QQQ in 74, or 39%. The median gap was 6.0 points below. The spread around that median is the important part:
- Worst gap: 40.9 points below, the 12 months ending April 2020. QQQ was up 15.5% after falling sharply and recovering in February and March. TQQQ was up 5.6%.
- Best gap: 80.7 points above, the 12 months ending March 2021. QQQ rose 67.6% with few interruptions. Three times that is 202.8%; TQQQ returned 283.5%.
- Same index return, different result. QQQ gained 34.6% in the year ending February 2014 and 34.2% in the year ending May 2020. TQQQ gained 139.7% in the first and 64.8% in the second. The first year was calm and the second contained the fall of March 2020.
- Index up, fund down. QQQ rose in 172 of the 188 periods. In 13 of those TQQQ fell.
Calendar years show the same pattern with fewer observations.
TQQQ beat the multiple in 2012, 2013, 2014, 2017, 2019, 2021 and 2023, all years in which QQQ rose at least 16% without a deep interruption. It fell short in 2011 (QQQ +2.5%, TQQQ −8.0%), 2015 and 2016 (two years of single-digit gains), 2018 (QQQ −1.0%, TQQQ −19.9%), 2020 (QQQ +47.6%, three times that +142.7%, TQQQ +110.1%), 2024 (+74.5% against +56.1%) and 2025 (+60.5% against +33.3%). In 2022 it lost 79.2%, which is less than three times QQQ's 33.1% loss only because a fund cannot lose more than everything. The daily reset cuts exposure as the price falls, and that is what kept the loss below 99%.
Using it: the other three funds
The same measurements for the four funds, each against the multiple of an index fund over the months both have traded:
| Fund (multiple, index fund) | Since | Growth of price | Index fund | Periods beating the multiple | Median gap | Worst gap | Deepest fall |
|---|---|---|---|---|---|---|---|
| TQQQ (3×, QQQ) | Apr 2010 | 214.0× | 14.3× | 74 of 188 (39%) | −6.0 | −40.9 | −79.2% |
| SSO (2×, SPY) | Aug 2006 | 14.7× | 5.9× | 102 of 232 (44%) | −0.5 | −12.6 | −82.5% |
| SPXL (3×, SPY) | Jan 2009 | 89.2× | 8.4× | 80 of 203 (39%) | −3.6 | −39.3 | −62.6% |
| SOXL (3×, SOXX as stand-in) | May 2010 | 103.1× | 28.1× | 63 of 187 (34%) | −11.1 | −100.2 | −87.0% |
Gaps are in percentage points over 12 months. SOXX is used as a stand-in for the semiconductor index SOXL follows. Both funds now track the NYSE Semiconductor Index (SOXL since August 2021); before that both followed the PHLX Semiconductor Sector Index. SOXX has its own fees and may have changed index on a different date, so SOXL's row is approximate.
Three things stand out.
Two times is much gentler than three. SSO's median gap is half a point. Its worst was 12.6 points, in the year ending September 2009, when SPY fell 9.0% and SSO fell 30.5%. The drag grows faster than the multiple: by the formula in the linked lesson a 3× fund carries three times the drag of a 2× fund on the same index.
The deepest falls came when the index fell by half. SSO lost 82.5% between May 2007 and February 2009 while SPY lost 52.2%. SSO's price did not return to its 2007 level until December 2013. In 2008 alone SPY lost 38.3% and SSO 68.3%.
A volatile index is the worst case. Over the last two years the semiconductor fund SOXX has been more than twice as volatile as SPY (45% a year against 17%). In 2024 SOXX rose 12.2% and SOXL fell 13.0%. In the 12 months ending October 2026 SOXX rose 53.8% and SOXL 61.3%, barely more than the unleveraged fund and 100 points below three times it. SOXL lost 87.0% between December 2021 and September 2022 and recovered only in April 2026. Yet in the 12 months ending May 2026, when SOXX rose 177.7% almost without pause, SOXL rose 1,287%.
The most recent two years of daily data confirm that the funds do what they say each day. From November 18, 2024 to November 17, 2026 TQQQ's daily moves were 2.97 times QQQ's on average, with almost no scatter. Over the whole two years QQQ gained 31.4% and TQQQ 36.8%, not 94.2%. SPY gained 24.5%, SSO 34.8% and SPXL 40.1%. The daily promise was kept, and the two-year result was barely better than the plain fund because the period included the sharp fall and rebound of spring 2025 and a financing cost that tracks short-term rates.
Where it fails or misleads
"Decay" suggests a steady loss, and that is wrong. There is no fixed percentage that leaks out each year. The gap is strongly positive in strong, calm years and strongly negative in choppy ones. A holder of TQQQ from 2010 made a fortune. A buyer in December 2021 was down 79% a year later and waited three and a half years to get back to even.
The long-run record depends on the period. Every fund in the table was launched within a few years of the 2009 low and has lived through one of the strongest stretches in the history of US shares. Funds that leveraged weaker indexes did not survive to appear in tables like this one. On March 20, 2020 Direxion announced that ten of its 3× funds, covering energy, gold miners, Brazil and Russia, would be cut to 2×, and that eight other funds would close.
Reverse splits hide what happened to the price. A fund whose price sinks toward zero combines its shares to lift the quote. Finabase's split records show eight reverse splits for the 3× inverse Nasdaq fund SQQQ between 2012 and 2025, at ratios of 4 or 5 to 1. One share today stands for about 160,000 of the original ones. Charts are adjusted for this, so the history looks like a smooth slide from a very high number. A Reverse stock split changes nothing about the value held, but a long series of them is a record of the losses that made them necessary. TQQQ's record is the mirror image: a series of forward splits.
Survivors make the product look safer than it is. A fall of one-third in the index in a single day would wipe out a 3× fund. Exchange trading halts make that unlikely for a broad index, but the protection is the exchange's, not the fund's.
The 2020 rule
The SEC adopted rule 18f-4 on October 28, 2020. It became effective on February 19, 2021, and funds had to comply by August 19, 2022. A fund that relies on the rule must keep its Value at risk (VaR) at or below 200% of the value-at-risk of a reference portfolio, normally its index. That is in effect a ceiling of 2× for new funds registered under the Investment Company Act, the law that covers almost all leveraged stock ETFs. The rule made one exception: a leveraged or inverse fund that was already operating on October 28, 2020 with a target above 200% may continue under conditions set out in the rule. This is why 3× funds such as TQQQ, SPXL and SOXL still trade, and why every single-stock fund launched since 2022 stops at 2× (see Single-stock ETFs: TSLL, NVDL, CONY and MSTY against their stocks).
Compared with its neighbours
| Way to get 2× or 3× exposure | Resets | Loss can exceed the stake | Main cost |
|---|---|---|---|
| Leveraged ETF | Daily | No | Volatility drag, swap financing, expenses |
| Margin loan | Never, until a margin call | Yes | Loan interest; forced sale after a fall |
| Futures | Never (marked to market daily) | Yes | Financing built into the price; roll |
| Call options | Never; expires | No | Premium and time decay |
A margin account's leverage rises as prices fall, which is the opposite of a daily-reset fund. That is why a margin investor is forced out at the bottom and a leveraged ETF holder is not, and also why the ETF recovers more slowly: after a fall it holds less exposure for the rebound. Exchange-traded notes with leverage carry the issuer's credit risk in addition (ETNs vs ETFs: an unsecured note is not a fund).
How Finabase does it
Every leveraged fund has a symbol page with a quote, chart and news. The comparison tab rebases the fund and its index fund to percent change from the first bar in view and tabulates price returns from one week to two years, which is enough to see the gap to the multiple over each of those periods. Dividends are not included there. The Dividends & splits tab lists every split, forward or reverse.
Finabase does not label a fund as leveraged or inverse, and it does not calculate decay or the gap to the multiple. The fund's name and its issuer's page are the place to confirm the multiple and the index.
Frequently asked questions
Do leveraged ETFs always lose money over time?
No. TQQQ, SSO, SPXL and SOXL all made several times their index's gain between launch and October 2026. What they did not do is deliver the stated multiple over a year with any reliability: they beat it in 34% to 44% of twelve-month periods and each lost more than 60% at least once.
How long can a leveraged ETF be held?
There is no limit in the rules; the funds can be held for years. The issuers design and describe them for one day. Over longer periods the result depends on the path: a strong, smooth rise produces more than the multiple and a volatile or falling market much less.
Why did TQQQ fall when QQQ rose?
Because the year contained a deep fall and a recovery. After a fall the fund holds less exposure, so the rebound restores the index before it restores the fund. That happened in 13 of the 172 twelve-month periods since 2010 in which QQQ gained.
Are 3× ETFs banned?
New ones are, in effect. SEC rule 18f-4 limits a fund's value-at-risk to 200% of its reference portfolio's. Funds above 200% that existed on October 28, 2020 may continue under conditions, which is why the existing 3× funds still trade.
Is the decay shown in the expense ratio?
No. The expense ratio covers management and operating costs. The effect of daily compounding is not a fee and appears nowhere on a fee table, and swap financing is a cost inside the fund's returns rather than a line in its expenses.
Summary
A leveraged ETF keeps its promise each day and makes no promise about a year. Measured over every 12-month period since 2010, TQQQ beat three times QQQ about four times in ten, missed by a median of 6 points, and ranged from 41 points below to 81 above. Calm, rising years produce more than the multiple; volatile years, falling years and high interest rates produce less, and a fall of half in the index has taken 80% or more from a 2× fund. The record of the survivors is spectacular, but it comes from a favourable period and from funds that were not closed, cut to a lower multiple or reverse-split into obscurity. Rule 18f-4 now caps new funds at about 2×.
Check your understanding
5 questions on this lesson. Each answer comes with the reason, and nothing is scored.
Over 12 months an index fund gains 15.5% and a 3× daily fund on the same index gains 5.6%. What is the gap to the multiple?
- −9.9 percentage points
- −40.9 percentage points
- +5.6 percentage points
Show the answer
B. −40.9 percentage points Three times 15.5% is 46.5%. The fund returned 5.6%, so it finished 40.9 points below the multiple. This was TQQQ against QQQ in the 12 months ending April 2020.
In which kind of year has a 3× fund tended to return more than three times its index?
- A year in which the index rose strongly with few interruptions
- A year in which the index ended flat after large swings
- Any year in which the index rose
Show the answer
A. A year in which the index rose strongly with few interruptions A steady rise compounds on a growing base. TQQQ beat three times QQQ in 2013, 2017, 2019 and 2023. Large swings in both directions do the opposite, and in 13 twelve-month periods QQQ rose while TQQQ fell.
TQQQ lost 79.2% in 2022 while QQQ lost 33.1%. Why was the loss not three times 33.1%?
- The fund suspended its daily reset during the fall
- The daily reset cuts exposure after each fall, so losses compound on a shrinking base
- The SEC limits yearly losses in leveraged funds
Show the answer
B. The daily reset cuts exposure after each fall, so losses compound on a shrinking base A fund cannot lose more than 100%, and resetting to three times a smaller asset base each day slows the loss in dollar terms. The same mechanism slows the recovery: TQQQ did not regain its December 2021 level until July 2025.
What does SEC rule 18f-4, adopted in October 2020, do to leveraged funds?
- It bans all leveraged and inverse ETFs
- It limits a fund’s value-at-risk to 200% of its reference portfolio’s, with an exception for funds above that level that already existed
- It requires leveraged funds to reset monthly instead of daily
Show the answer
B. It limits a fund’s value-at-risk to 200% of its reference portfolio’s, with an exception for funds above that level that already existed The relative value-at-risk limit is 200%. Leveraged or inverse funds with a higher target that were operating on October 28, 2020 may continue under conditions, which is why existing 3× funds still trade and new ones stop at 2×.
A chart shows an inverse leveraged fund starting from a price in the millions and sliding to $20. What most likely explains the starting price?
- The fund once traded at that price
- A data error
- A series of reverse splits, for which the chart’s history has been adjusted
Show the answer
C. A series of reverse splits, for which the chart’s history has been adjusted Charts restate old prices for later splits. A fund that has combined its shares many times shows very high adjusted prices in its early history. The number of reverse splits is itself a record of how much was lost.
Sources and further reading
- U.S. Securities and Exchange Commission — Use of Derivatives by Registered Investment Companies and Business Development Companies: a small entity compliance guide (rule 18f-4), February 4, 2021
- U.S. Securities and Exchange Commission — Release No. IC-34084, Use of Derivatives by Registered Investment Companies and Business Development Companies (adopted October 28, 2020)
- ProShares — UltraPro QQQ (TQQQ) fund page: objective, expenses, inception and holding-period warning, read November 18, 2026
- ProShares — Ultra S&P500 (SSO) fund page, read November 18, 2026
- Direxion — “Direxion Changes Objectives Of Ten Leveraged Funds To Address Extreme Market Conditions, While Also Closing Eight Funds Due to Limited Interest Since Launch,” press release, March 20, 2020
- Direxion Shares ETF Trust — Direxion Daily Semiconductor Bull 3X Shares (SOXL) summary prospectus, February 2026: the NYSE Semiconductor Index, and the PHLX Semiconductor Sector Index before August 25, 2021
- Finabase chart data — month-end prices of TQQQ, QQQ, SSO, SPXL, SPY, SOXL and SOXX through October 2026, daily prices from November 18, 2024 to November 17, 2026, and split records; price only, statistics computed by Finabase
