The ETFIQ Quarterly, 2026-Q2

Corrected on Oct 3, 2026. In the Leveraged ETFs study, NVDB and TSLI were left out on an outdated name for their tickers and are now measured, so 349 funds are measured, not 347. The rows for AMDD, MARU, PLTD and TIPD carried another fund's prices: each now names the fund its ticker trades as. The data before and after.

1. 1A leveraged fund promises a multiple of one day and resets, so over a quarter the honest benchmark is that promise compounded. 289 of 349 came up short of their own, by a median of 2.80 points.
2. 2116 of 275 option-income funds paid out more cash than they made in total return. A yield is not a return and on this many funds the two pointed opposite ways.
3. 3How much of a theme a buyer already owned runs from 0.0 per cent of the fund to 72.8 across 227 funds sold under the same word, at a median fee of 0.56 percent that barely moves across that range.

What this is

Five studies of one closed quarter, one for each kind of fund ETFIQ measures, with the finding from each and a link to the whole of it. Every number here is calculated from a named public source, measured against the thing the fund's own prospectus names, over the same days for the fund and for what it is measured against.

A closed quarter is the only form of this work that can be quoted. Every other page on this site is measured to yesterday and moves tonight, which is right for a reader deciding today and useless to anybody citing it. These numbers are the same next year as this week, and each study publishes the file it was computed from, the funds it excluded and the test that excluded them.

Leveraged ETFs

289 of 349 behind their own stated multiple

A leveraged fund promises a multiple of one day and resets. Compounded through the days that happened, the median fund came up 2.80 points short of its own promise, and the sign flips by direction: long funds 3.45 behind, inverse 1.75 ahead, because a short position earns interest on its collateral.

Long: the 10 furthest above their own promise, of 272. Source: ETFIQ.

Inverse: the 10 furthest above their own promise, of 77. Source: ETFIQ.

Income ETFs

116 of 275 paid more than they made

The median fund paid out 4.66 percent of its own price and returned +6.55 percent in total, so the median arrangement worked. A yield is not a return, and on this many funds the two pointed opposite ways.

Writes on an index, an ETF or a basket: the 10 furthest ahead of their benchmark, of 196. Source: ETFIQ.

Writes on a single company: the 10 furthest ahead of their benchmark, of 79. Source: ETFIQ.

Thematic ETFs

32.5% of the median fund was already the index

Across 227 funds it runs from 0.0 percent at the tenth to 72.8 at the ninetieth, at a median fee of 0.56 percent that barely moves across that range. The word thematic tells a buyer nothing about the answer.

The 10 furthest ahead of the S&P 500, of 227. Source: ETFIQ.

Buffer ETFs

48.5% of the reference reached the holder

The reference rose for every one of the 114 funds and every one delivered less than all of it. The buffer engaged on 0 of them: the shape these funds promise arrives on the last day of the period and not before.

The 10 that delivered most of their reference, of 114. Source: ETFIQ.

Crypto ETFs

99.4% of its own coin kept by the median fund

A spot trust holds one thing and promises to hold it, so the only question is how much of the coin reached the holder. 15 of 41 finished behind theirs. Read the median and not the count: a coin trades all night and a fund prices at four.

The 10 furthest ahead of their own coin, of 41. Source: ETFIQ.

How to use this

Each line above is one category's answer to one question, and each links the study that shows every fund in the population it is drawn from. Nothing here is a view on any fund. The distances measured are properties of how each kind of fund is built, not verdicts on how any of them is run, and every study says so in its own words where the distance is largest.

Check any of it

Each category's study publishes the file every number in it was computed from, fund by fund, with the excluded funds and the reason for each. The data ETFIQ computed in them are free to use, including commercially, with attribution to ETFIQ and a link; issuer data stays under its owner's terms.

Cite this: ETFIQ, The ETFIQ Quarterly, 2026-Q2, Sep 11, 2026. https://etfiq.com/research/quarterly/2026-q2/