What a buffer ETF delivers mid-period, 2026-Q2

1. 1The reference rose for every one of the 114 funds, and all 114 delivered less than the whole of it. The median fund passed on 48.5 percent of its reference’s rise.
2. 2The more the reference rose, the less came through. Funds on QQQ, up +26.17%, passed on a median 37.6 percent; funds on EFA, up +7.00%, passed on 58.6 percent.
3. 3The buffer did nothing. On 0 of the 114 was the reference below the level where its buffer starts at the quarter’s close, which is what protection looks like in a quarter that goes up.

The problem

A buffer ETF promises a shape. You get the reference asset’s return, capped above at a level struck when the period began, and protected below down to another one. Every part of that is true, and all of it arrives on a single day: the last day of the outcome period. On every other day of the year what you own is a basket of options that have not matured, and a basket of unmatured options does not have the shape it will settle into.

So the fund moves a fraction of the reference in both directions, the fraction depends on how much time is left and where the reference sits between the two ends, and for most of the period neither end is doing anything at all. None of that is hidden. It is simply not measured anywhere, and a holder checking a price in month four has no way to tell whether what they are looking at is the fund working or the fund failing.

What happened in the second quarter of 2026

The reference rose for every one of the 114 funds whose period ran through the whole quarter. Every one of them delivered less than the whole of that rise. The median fund passed on 48.5 percent, the tenth 31.3 and the ninetieth 67.1, and the median gap to the reference was 7.48 percentage points.

That is the instrument working. A fund that had passed on all of a rise in month two would be a fund with nothing left to give at the cap, and the ones passing on least are mostly the ones whose reference has run furthest toward it.

Median share of the reference’s rise that reached the holder, by reference, 2026-Q2. Source: ETFIQ.

The ordering is the point. QQQ rose +26.17% and its funds passed on a median 37.6%; EFA rose +7.00% and its funds passed on 58.6%. The further a reference runs toward a cap, the less of each further point comes through.

41 of the 106 funds with a stated cap had their reference above it at the quarter’s close, measured from each fund’s own period start. Those are the funds with the least left to give.

0 had the buffer engaged. The protection every one of these funds was bought for did nothing for three months, which is exactly what it should do in a quarter that rises.

None of these is what a holder finally gets. That is settled on the period’s last day, and for most of these funds that day is months away.

What the fraction is made of

Three things, and the first two are not costs. Time: the options that produce the cap and the buffer are worth less than their settled value until they settle, so a fund holding them prices between the two ends rather than at them. Where the reference sits: near the middle of the range a fund moves close to one-for-one with it, and as the reference approaches the cap each further point is worth less to the fund because the fund has sold that point away. And the fee, which is the only part of the three a holder pays for and the only one that does not reverse by the last day.

The first two reverse. A fund at 48 percent of its reference in June is not 52 percent behind for good; if the reference finishes the period where it is, the fund converges on the shape it promised as the options mature. What this study measures is the middle of that journey, which is where every holder who checks a price actually stands.

Every fund, 2026-Q2

All 114 funds, least of the reference passed on first. The cap and the buffer are as struck at the start of each fund’s own outcome period. Returns are over the quarter’s days. Numbers in percentage points.

The two ends, captured.

How this was computed

Each fund’s total return across the quarter on split and distribution adjusted closes, against its own reference asset’s total return over the same two dates, from Tiingo. The share captured is the fund’s return as a percentage of the reference’s, computed only where the reference moved more than half a point, because a share of nothing is a division by zero wearing a percent sign.

Whether the cap or the buffer was binding is measured against each fund’s own period start and at the quarter’s close, on the reference’s price rather than its total return, because that is the level the terms are struck against. The Buffer page publishes the same quantity as of tonight, which is a different number and a fact about tonight: 41 of these funds have their reference past the cap today, and that says nothing about a quarter that closed in June.

What was left out, and why

An outcome period is a year and a quarter is three months, so the two do not line up. A fund that reset inside the quarter has two sets of terms inside it, and ETFIQ carries only the newer, so measuring it against those would be measuring most of the quarter against a cap and a buffer that did not exist yet. Those funds are counted out here rather than published.

Funds excluded and why, 2026-Q2. Source: ETFIQ.

Where the quarter starts

Each return here runs from the first close inside the quarter to the last, so the move on the quarter’s first trading day is left out. An issuer’s quarterly number starts at the prior quarter-end close, so the two can differ by that day’s move, which for a leveraged or single-stock fund can be several points.

Check it yourself

Every number above is in the file below, fund by fund, with the excluded funds and the reason for each. The data ETFIQ computed is free to use, including commercially, with attribution to ETFIQ and a link; the terms a fund row restates are its issuer's and stay under its owner's terms.