ETFIQ Research · rebuilt every trading night

The stated multiple and what the holder got

Over the three months to Sep 21, 2026, 400 of 532 leveraged and inverse ETFs finished short of their own stated multiple. The median fund finished -4.3 points from its stated multiple.

These funds deliver the multiple over a single day and reset. Daily results compound, so a run of days is not the multiple times the period's move, and the gap widens with how far the underlying travels rather than with where it ends up. 53 underlyings carry the same stated multiple from more than one issuer. Same mandate, same days, and the spread between the closest and the furthest is in the second table.

Short of the stated multiple, by holding period
One month
485 of 680
Three months
400 of 532
Six months
310 of 405
One year
205 of 314
Three years
62 of 124

400 of 532 leveraged and inverse ETFs finished short of their stated multiple over three months to Sep 21, 2026.

ETFIQ Narrative

Leveraged and inverse ETFs are built to deliver a stated multiple of their underlying's move, but only for a single day. Hold one past that, and daily results compound rather than add, so the fund's return over a stretch of time is not simply the multiple times the underlying's move over that stretch. The gap that opens up tracks how far the underlying travelled along the way, not just where it ended up. As of Sep 21, 2026, that gap shows up clearly across 532 leveraged and inverse ETFs measured over the three months to that date: 400 of them finished short of their own stated multiple, with the median fund landing 4.3 points short.

The size of the shortfall moves with the underlying's volatility, not with the fund's design alone. Grouped by three-month underlying volatility, funds tracking underlyings with 0% to 30% volatility had a median difference from stated of -1.8%, with 110 of 165 funds short. That widens to -5.5% for the 30% to 50% band, -7.9% for 50% to 80%, and -14.8% for underlyings at 80% volatility and above, where 105 of 123 funds finished short. Across holding periods, the one-month window shows a median difference of -1.5% against 1.9% median distance travelled and 45% median underlying volatility; by six months the median difference reaches -10.5% against 12.6% distance and 44% volatility. The one-year and three-year windows show smaller median differences (-8.3% and -0.8%) alongside markedly higher median distance travelled (14.3% and 39.3%) and lower median volatility (34% and 19%), reflecting a different mix of funds and underlyings in those longer-lived cohorts rather than a reversal of the pattern.

The data also lets the same trade be compared across issuers. 53 underlyings carry an identical stated multiple from more than one issuer, and even with the same mandate and the same days held, results differ. The widest spread appears in ETHA +2x, where four issuers priced against a +59.6% underlying move produced a 5.0% spread between the closest fund to stated (ETHU, -13.4) and the furthest (ETHT, -8.4). CPER +2x shows a 4.0% spread between CPXR (-1.7) and UCOP (-5.7) on a +4.8% underlying move, and XRPR +2x shows a 3.6% spread between XRPT (-12.7) and XXRP (-16.3) on a +32.1% move. Smaller spreads appear further down the table, including XLF +2x, where UYG (-1.8) and XLFX (-3.2) differ by 1.5% on a +4.5% underlying move.

Readers can pull up any individual fund on the live desk, check its stated multiple against its own underlying over the same dates, and see where it sits within these distributions.

Short of the stated multiple, by the underlying’s volatility
0% to 30%
110 of 165
30% to 50%
94 of 117
50% to 80%
91 of 127
80% and above
105 of 123

400 of 532 leveraged and inverse ETFs finished short of their stated multiple over three months to Sep 21, 2026, grouped by how much the underlying moved.

How far from the stated multiple, by how long it was held

WindowFundsShort of statedMedian differenceMedian distanceMedian underlying volatility
One month680485 of 680−1.5%1.9%45%
Three months532400 of 532−4.3%5.9%47%
Six months405310 of 405−10.5%12.6%44%
One year314205 of 314−8.3%14.3%34%
Three years12462 of 124−0.8%39.3%19%

The same trade from more than one issuer, widest spread first

OnIssuersUnderlying movedClosest to statedFurthestSpread
ETHA +2x4+59.6%ETHU +13.4ETHT +8.45.0%
CPER +2x2+4.8%CPXR -1.7UCOP -5.74.0%
XRPR +2x3+32.1%XRPT -12.7XXRP -16.33.6%
SPCX +2x7−1.8%SPCM -13.5SPCU -16.32.8%
PDD +2x2+1.5%KPDD -5.5PDDL -7.72.1%
XLE +2x3+16.2%ERX -0.9XLEX -2.71.8%
EWZ +2x3+11.1%BRZU -2.0BRZX -3.71.7%
SLV +2x2+1.2%AGQ -5.5USLV -7.21.7%
IBIT +2x4+34.3%BTCU +0.4BITU -1.31.7%
GLD +2x2+3.6%UGL -2.6UGLD -4.21.6%
KWEB +2x2+0.5%CWEB -2.9WEBX -4.41.5%
XLF +2x2+4.5%UYG -1.8XLFX -3.21.5%

Erosion against the underlying’s volatility, three months

Underlying volatilityFundsMedian volatilityMedian difference from statedShort of stated
0% to 30%16516%−1.8%110 of 165
30% to 50%11740%−5.5%94 of 117
50% to 80%12766%−7.9%91 of 127
80% and above12394%−14.8%105 of 123
Method

Every number is an ETFIQ calculation from Tiingo end-of-day prices with distributions reinvested, measured against the fund’s own underlying over exactly the same days. The stated multiple comes from the fund’s registered name; the underlying is mapped by hand and checked against the name the exchange gives it. Funds whose underlying has no investable tracker are listed on ETFIQ and excluded here, because there is nothing to measure them against. Rebuilt every trading night.

Data file: leverage.json. ETFIQ is an independent publisher and makes no recommendations. Standards →

Cite this page

ETFIQ Research, The stated multiple and what the holder got, data as of Sep 21, 2026. https://etfiq.com/research/2026-09-21/leverage-label

Open data →

Free to use with attribution; the underlying files are at Open data.