Measured over each fund’s own full life against the thing it owns, and measured as a simple yearly rate — the published column, an annualized arithmetic mean — 0 of 21 measured pairs kept pace with their underlying. The ruler has to be said out loud, because it decides one row: counted the way money compounds (geometric), 1 of 21 came out ahead — MSTY, by 3.9 points a year over a 2.3-year life with no bear market inside it — and on dollars its $1.17 beat its stock’s $1.07. §5 owns that row. But the sign is not the finding — the size is, and it is wildly dispersed. At the top of the ladder the cheapest call-writers cost between 2.2 and 2.8 points a year (TSPY, XDTE, QDTE), and a Goldman Sachs fund sitting just below that band, GPIX, gave up 3.3 points a year and grew net asset value 13.8% a year while doing it. At the bottom, CONY: the published figure is −48.4% a year — and because that column is an annualized arithmetic mean rather than a compounding rate, the honest way to state that fund is what a dollar actually became: $1.28, against $2.07 for the stock it sells calls on. In between sits JEPI, the $44.3B fund that defines the category, which paid a life yield of 8.8% a year and trailed SPY by 7.9 points a year across its whole 6.1-year life (7.1 points counted geometrically): the yield you were paid and the return you gave up are within a point of each other. Those are two independently measured quantities that came out near each other. We are not saying the yield causes the gap — this receipt does not license that, and §2 says so at length. In the form no annualization convention can touch: a dollar in JEPI became $1.83 and a dollar in SPY became $2.67, same 6.1 years, same money in.
Granted, not measured: that the cash is real and it arrives. These are large, regulated products from JPMorgan, Goldman Sachs, Global X, NEOS, Roundhill and Schwab, doing on the page exactly what their prospectuses say they do. We also grant, because our own table says so, that some of them are genuinely good: the cheapest call-writers on this page cost between 2.2 and 2.8 points a year, and the two Goldman funds just below that band sell calls on only part of what they hold. And we grant that sequence-of-returns protection is a real benefit — spending a distribution instead of selling into a falling market is worth something to somebody who is spending. None of that is what we measured.
Tested: one question, applied identically to every fund: did the fund’s total return over its own life keep pace with the thing it is made of? 24 funds; 21 of them have a named underlying on the source table and are measured against it over the identical window, month for month. Survivorship-free month-end price and distribution records through 2026-06-30, month-end net asset value plus distributions per share, split-consistent, the same method applied to all 24 so the columns are cross-comparable. No trading, no friction: this is an ownership comparison, not a strategy backtest.
Over each fund’s own life, did the fund’s total return keep pace with the thing it is made of — and if not, by how much?
Scar one: six of the 24 rows had no engine entry. The engine is hard-bound to its fund specification table, and the companion file the reference page’s own provenance block names is not on disk anywhere. The affected rows are CONY, NVDY, QYLD, RYLD, TSLY, XYLD — which includes QYLD, the longest and most load-bearing record on this page. We re-declared the 6 specifications from the source table’s own left-hand column and re-derived them from the same store: 24 of 24 master cells and 6 of 6 cap-cost cells reproduce. ⚠ That is a reconstruction of the spec, not a recovery of the file. It proves the figures are re-derivable; it does not prove the vanished file used this exact pairing. Agreement to print precision on 30 of 30 cells is the evidence that it did, and it is the only evidence there is.
Scar two: the columns labelled “CAGR” are not geometric CAGRs. Both — NAV CAGR and Total CAGR — are annualized arithmetic means, and those coincide with a geometric rate only at zero volatility. The total-return pair is printed both ways below; the NAV rate only as published, and no geometric NAV rate is on this page (§7). We found this in our own instrument, it is one of the two most important caveats on this page, and it has its own section (§5) rather than a footnote.
Scar three: the index leg was measured over one month too many, on every pair. Every one of the 21 fund-vs-index cells carried one extra month on the index side — the month before the fund’s first return, a month the fund did not exist for — and the reference page carried the same month, which is why the re-derivation reproduced it perfectly. Reproducing is not the same as being right. Fixed 2026-09-05: 21 of 21 pairs now carry the same count of monthly returns over the same month-ends on both legs. On the 17 index pairs the correction moved the published gap by up to 2.17 points a year; on the 4 single-stock pairs by up to 74.73 points; and on MSTY it flipped the sign of the geometric gap (§5). The corrected table is the one on this page, and the reference page was corrected the same day.
| Parameter | Value |
|---|---|
| Question | fund total return vs the total return of the thing the fund owns |
| Funds | 24 rows, parsed from the source master table (24 of 24 parsed) |
| Measured pairs | 21 — the rows that carry a named underlying; 20 of them call-writers |
| Window | each fund’s own full life, month-end, earliest 2011-10-31, all ending 2026-06-30; the index leg over the same month-ends, month for month (21 of 21 pairs asserted) |
| Longest / shortest record | QYLD 12.5 yr (151 mo) / LFGY 1.42 yr |
| Return definition | monthly total return = (change in net asset value + distributions per share) / prior net asset value, split-consistent |
| Cap cost | fund total return − underlying total return, over the fund’s own window — the instrument this whole page rests on |
| Ruler, stated honestly | the source table has two columns labelled “CAGR” — NAV CAGR and Total CAGR — and each is an annualized arithmetic mean, a simple yearly rate, not a geometric CAGR. Total return, the underlying and the cap cost are published both ways in the ledger below, with what a dollar became beside them; the NAV rate and the life yield are printed once, as published, and no geometric NAV rate is on this page (§7). Every rate spoken in the prose is the published rate unless it says geometric. |
| Survivorship | survivorship-free store; every fund measured over its actual life; no fund dropped for a bad record |
| Frictions | none — no trading, no slippage, no tax. This is an ownership comparison, not a backtest. |
| Third-party input | four columns — expense ratio, AUM, inception and distribution frequency — from stockanalysis.com as-of 2026-07-01, carried from the source table. Four more are the source table’s own labels (ticker, fund name, tier, the underlying pairing). Every measured column — 16 of the 24 published — is re-derived from the store. |
| # | Check — and the question it asked | Result | Clean? |
|---|---|---|---|
| 1 | POINTERS — do the four line anchors this report cites still land on the structure they name? | 4 of 4. Parsed 24 master rows and 21 cap-cost rows. | YES |
| 2 | RECOMPUTE — does the engine still produce the number the reference page prints, to that page’s own 0.1-point print precision? | 96 of 96 master-table cells across 24 of 24 funds, 0 diffs. Fund-only figures; row 3a is what this check could not see. | YES |
| 3 | CAP COST — does fund total return minus underlying total return, over the fund’s own window, reproduce the published cap-cost column? | 21 of 21 cap-cost cells, against the reference page as corrected on 2026-09-05. | YES |
| 4 | WINDOW (3a) — was the index leg measured over the fund’s own window, month for month? The reproduction check above could not ask this: both sides carried the same window. | Not until 2026-09-05. Every one of the 21 pairs carried one extra index month — the month before the fund’s first return — and the reference page carried the same month, so the cells above had reproduced a defective number perfectly. Fixed and re-derived: 21 of 21 pairs now carry the same count of monthly returns over the same month-ends on both legs. The correction moved the published gap by up to 2.17 points a year on the 17 index pairs and up to 74.73 points on the 4 single-stock pairs, and on one of them (MSTY) it flipped the sign of the geometric gap. See §5. | REPAIRED |
| 5 | ORPHANS — which rows can the surviving engine not address at all? | 6 of 24 (CONY, NVDY, QYLD, RYLD, TSLY, XYLD) had no engine entry: the companion file the reference page’s own provenance block names is not on disk. Re-declared here and re-derived — 24 of 24 master cells and 6 of 6 cap cells reproduce. | REPAIRED |
| 6 | VARIANCE DRAG — are the columns labelled “CAGR” actually geometric CAGRs? | No. Both (NAV CAGR and Total CAGR) are annualized arithmetic means. On total return, median gap across 24 funds 1.04 points a year; largest 30.37 points. See §5. The NAV rate’s gap was not measured: no geometric NAV rate is on this page. | NO |
| 7 | AUM — what does the tabled sample actually hold? | 24 of 24 AUM cells parsed; $139.5B across the 22 call-writing funds. A floor on the category, never the category — and the column is not ours. | FLOOR ONLY |
Three of the seven came back clean; three did not, and all three are ours. One is an engine file that vanished and had to be reconstructed before 6 of these rows could be re-derived at all. One is a column label on our own reference page that does not mean what it says — it changes what 6 of these funds may honestly be said to have done. And one is an index leg measured over one month too many on every pair — the most consequential of the three, because the reproduction check was structurally blind to it, and because it moved the bottom of the ladder and flipped one fund’s sign. The seventh is a qualification, not a defect: the AUM figure ships as a floor. None was repaired quietly.
| Figure | Value | What it means |
|---|---|---|
| Funds measured / pairs priced | 24 / 21 | 21 of the 24 rows carry a named underlying on the source table; three do not and are reported without a cap cost rather than paired with a guess. |
| Pairs that kept pace with their own underlying, measured as a simple yearly rate | 0 of 21 | On the published column, an annualized arithmetic mean. Counted geometrically, 1 of 21 (MSTY, +3.9% a year); on dollars 1 of 21 (the same fund, $1.17 against $1.07). The wrapper always cost something on the ruler the reference page prints, and §5 shows the one row where the ruler decides. |
| The dispersion, as published | −2.1% (SCHD) to −48.4% (CONY) a year | The whole finding. Across the 20 call-writers alone, −2.2% (TSPY) to −48.4%. Both ends are annualized arithmetic means; the bottom is read as a multiple below, not as an annual surrender. |
| The shape of the ladder | 4 / 12 / 5 | 4 of the 21 cost under three points a year, 12 sit between three and ten, 5 are worse than ten; median 5.1 points a year, which is what a typical wrapper charged. |
| The top rung — the cheapest call-writers | between 2.2 and 2.8 points a year | TSPY −2.2% · XDTE −2.6% · QDTE −2.8%. SCHD, a dividend fund rather than a call-writer, sits beside them at −2.1%. |
| The counterweight — GPIX, 2.7 yr | −3.3%/yr cap cost, +13.8%/yr net asset value | A Goldman Sachs fund that gave up 3.3 points a year — not a rounding error, nothing on this ladder is — and grew net asset value 13.8% a year while doing it. Its sibling GPIQ cost 4.8; both sell calls on only part of the book. A real product doing a real job. |
| JEPI life yield (6.1 yr, 74 mo) | 8.8% a year | Real, paid, arrived. Nobody disputes it, and this report does not try to. |
| JEPI vs SPY, same window | +11.0% vs +18.9% → −7.9 points/yr | Percentage points, not percent. −7.1% counted geometrically — between seven and eight depending on the convention, which is why the dollars below are the figure we hand you. Two independently measured quantities; the near-equality with the yield above is an observation, not a mechanism. |
| JEPI dollar test | $1.83 vs $2.67 | What a dollar became in the fund and in the index it owns. A multiple carries no annualization convention, so nothing here can be inflated by one. |
| QYLD dollar test (12.5 yr, 151 mo) | $1.76 vs $9.26 | The only call-writing fund on the table old enough to have lived through several bear markets. This is the through-the-cycle picture. |
| Highest life yield anywhere on the table (MSTY) | 144.9% a year → $1.17 | Against $1.07 for the stock it sells calls on: over the matched 2.3-year window the fund’s dollar beat the stock’s (§5), and a 144.9% yield still produced $1.17. The yield on the label did not tell you which fund you had bought. |
| Same yield, different ladder | JEPI 8.8% vs GPIX 8.5% | 0.3 points apart in yield, 4.6 points apart in what the wrapper cost. The fund this report opens on and the Goldman fund near the top of the ladder paid almost the same yield and sit nowhere near each other on it. The label does not point. |
| Funds with no bear market inside their life | 17 of 24 | A boundary on the evidence. The flattering young records are the least trustworthy rows on our own page. |
| What the tabled sample holds | at least $139.5B across 22 call-writing funds | A floor on the category, never the category. The AUM column is stockanalysis.com as-of 2026-07-01 — one of the four carried columns (expense ratio, AUM, inception and distribution frequency) that are not our measurement. |
17 of the 24 funds on this page have never seen a bear market. That is not a caveat riding under the table — it is the shape of the evidence. Seven funds have any real drawdown regime inside their life, as our reference page’s own regime caveat names them (QYLD, XYLD, RYLD, JEPI, SCHD, DIVO, TSLY): the three old buy-writes, JEPI, the two dividend funds, and one single-stock fund whose own stock crashed on its watch. The rest lived entirely inside the run that followed. Their positive net-asset-value records are bull-flattered, and a two-year record is a statement about the weather.
What real regime evidence exists on this tape: JEPI’s max drawdown −18.9% through 2022, and QYLD over 12.5 years with a max drawdown of −39.3% and a net-asset-value path of −2.0% a year. The two other old-guard buy-writes tell the same story: XYLD net asset value −1.3% a year over 5.8 years and RYLD −5.2% a year with a −44.0% drawdown.
The converse is ours to say as well: a young fund with a good number has not been disproved by anything here. It has simply not been tested. Those are different, and the reopen bounty in §8 is written against exactly that gap.
Each fund’s control is its own underlying, over its own window, computed by the same method. That is what makes the columns comparable and it is the entire instrument: SPY re-derives at +18.9% a year over JEPI’s 73 monthly returns and QQQ at +21.4% a year over QYLD’s 150. Change the window and both sides change together — which is exactly the property the extra index month in §1 broke, on every pair, until 2026-09-05.
The two dividend portfolios on the table are the second control. SCHD, which writes no calls, grew net asset value +10.5% a year over 14.7 years, and DIVO, which sells calls on only part of its book, +7.5% a year over 9.5 years, both through full cycles. If durable growth with a payout was the thing you wanted, that is where our own table points — and neither one is a high-yield product.
And the negative control on the metric: the four funds this report leans on hardest have almost no variance drag — JEPI 0.56pp · SPYI 0.62pp · QYLD 0.64pp · GPIX 0.68pp — so §5’s finding does not touch the spine of this page. It was checked before it was relied on, not after. That check covered each fund’s own return, not the gaps; JEPI’s gap counted both ways is in §2.
“Your headline column is not the statistic you think it is.” It is the best objection to this page, nobody raised it against us, and it is true. Two of the reference table’s columns are labelled “CAGR” — NAV CAGR and Total CAGR — and the engine computes each as an annualized arithmetic mean of monthly changes. An arithmetic mean and a geometric rate coincide only at zero volatility, and for a violently volatile fund the published figure sits well above what a holder actually compounded.
We measured the gap on the total-return column — the one the cap cost rests on — for every fund rather than arguing about it. Median across the 24 funds: 1.04 points a year. Every fund whose gap is two points or more — all 6 of them single-stock or concentrated option-income funds — is here in full, with the published figure, the geometric figure, and what a dollar actually became:
| Fund | Years | Total return (published, annualized arithmetic mean) | Total return (geometric) | Drag | Cap cost (published) | Cap cost (geometric) | $1 became | Its underlying |
|---|---|---|---|---|---|---|---|---|
| MSTY | 2.33 | +37.3% | +7.0% | 30.37pp | −17.5% | +3.9% | $1.17 | $1.07 |
| CONY | 2.83 | +27.3% | +9.1% | 18.26pp | −48.4% | −20.1% | $1.28 | $2.07 |
| TSLY | 3.58 | +21.1% | +9.8% | 11.34pp | −24.0% | −13.1% | $1.40 | $2.09 |
| LFGY | 1.42 | +13.6% | +7.5% | 6.08pp | — | — | $1.11 | — |
| NVDY | 3.08 | +57.9% | +52.3% | 5.54pp | −27.7% | −22.0% | $3.66 | $5.55 |
| ULTY | 2.33 | +7.4% | +4.7% | 2.62pp | — | — | $1.11 | — |
Two things this does not mean. It does not mean the annualized arithmetic mean is the wrong statistic in general: for a distribution rate — cash paid out, never compounded — it is arguably the honest one, which is why the life-yield column is left exactly as published. And it does not touch the spine of this report: the four funds it leans on hardest carry drags of JEPI 0.56pp · SPYI 0.62pp · QYLD 0.64pp · GPIX 0.68pp. We found the defect in our own instrument, we published the size of it, and we changed how 6 rows may be spoken about. We did not delete them.
The mechanism is not in this receipt. This page measures a gap. It does not prove that selling calls is what caused the gap, because nothing here reads a fund’s books. What we hold instead is a contract-level prior from our own options lab, published in an earlier episode, and it is a prior, not an answer: writing weekly calls on QQQ at the bid over six years collected $327 per share in premium and handed back $429 per share in intrinsic value on capped weeks, and at the five-delta limit the overlay finished +192.9% against +193.7% for simply holding, winning from 7 of 52 start dates. That is a self-managed overlay on one underlying. It contains no net asset value, no distribution, no return of capital, no expense ratio and no fund. Reasoning from it to a fund is reasoning, and it is labelled as reasoning wherever it appears.
The numbers here that are not ours. Four columns are carried from stockanalysis.com as-of 2026-07-01 — expense ratio, AUM, inception and distribution frequency — and none of them can be re-derived from a price-and-distribution store. The AUM column is the one that reaches the prose: it sums to at least $139.5B across 22 call-writing funds (JEPI $44.3B · JEPQ $39.0B · QQQI $12.5B), and it ships as a floor on the category rather than as the category: 24 funds is not the universe, and we did not count the universe.
The three outside sources the episode cites are witnesses, not specimens — and their figures are deliberately absent from this report. The episode credits an income educator who publishes his own book, a retail holder who publishes his dashboard, and an independent analyst who reconciled one fund’s published trade files for one month. Every figure of theirs is theirs, is credited to them on screen and in the description, and is not a measurement of ours — so none of it is carried here, where the whole point is that every figure on the page came out of the same instrument. The subject of this test is a fund structure. It is not a person, and no person appears in this report’s title, verdict or headline.
And the honest scars on the sample. Six of the 24 rows had to be re-declared from the source table’s own left-hand column before they could be re-derived at all (§1). They reproduce to print precision on 30 of 30 cells, which is the evidence, and it is a reconstruction of the spec, not a recovery of the file. And every fund-vs-index cell on our reference page carried one extra index month until 2026-09-05; this page carries the corrected cells, and says so.
ticker · fund · tier · inception · window_start · window_end · months · years · life_yield_pct_yr · nav_rate_pct_yr_published_arithmetic · total_return_pct_yr_published_arithmetic · total_return_pct_yr_geometric · variance_drag_pp · underlying · underlying_pct_yr_published_arithmetic · underlying_pct_yr_geometric · cap_cost_pct_yr_published_arithmetic · cap_cost_pct_yr_geometric · max_drawdown_pct · dollar_became · underlying_dollar_became · aum_usd_bn_stockanalysis_2026_07_01 · expense_ratio · distribution_freq. Two build-time provenance columns — a private row index and an engine-source note — are not published. Check our arithmetic: every fund figure on this page is a cell in the CSV below, or arithmetic on one; the one figure quoted from elsewhere is the contract-level receipt in §6, which belongs to an earlier episode.| # | Ticker | Fund | Tier | Life | Window | Life yield (published — annualized arithmetic mean) | NAV rate (published — annualized arithmetic mean) | Total return (published — annualized arithmetic mean) | Total return (geometric) | Drag | Owns | Underlying (published — annualized arithmetic mean) | Cap cost (published — annualized arithmetic mean) | Cap cost (geometric) | Max DD | $1 became | Its underlying | AUM |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | SPYI | NEOS S&P 500 High Income | Anchor·NEOS | 47 mo | 2022-08-31 → 2026-06-30 | 12.8% | +2.7% | +15.8% | +15.2% | 0.62pp | SPY | +20.9% | −5.1% | −4.4% | −9.7% | $1.72 | $1.99 | $6.2B |
| 2 | QQQI | NEOS Nasdaq-100 High Income | Anchor·NEOS | 30 mo | 2024-01-31 → 2026-06-30 | 15.4% | +6.8% | +23.1% | +22.3% | 0.85pp | QQQ | +28.9% | −5.8% | −4.9% | −10.5% | $1.63 | $1.79 | $12.5B |
| 3 | IWMI | NEOS Russell-2000 High Income | Anchor·NEOS | 25 mo | 2024-06-30 → 2026-06-30 | 15.6% | +3.9% | +20.0% | +19.0% | 1.08pp | IWM | +24.4% | −4.3% | −3.4% | −20.4% | $1.42 | $1.50 | $0.9B |
| 4 | CSHI | NEOS Enhanced Income Cash Alt | Anchor·cash | 47 mo | 2022-08-31 → 2026-06-30 | 5.6% | −0.2% | +5.4% | +5.4% | 0.00pp | — | — | — | — | −1.1% | $1.23 | — | $1.3B |
| 5 | TSPY | TappAlpha SPY Growth & Daily Income | Anchor·TappAlpha | 23 mo | 2024-08-31 → 2026-06-30 | 14.2% | +2.1% | +16.6% | +15.6% | 0.94pp | SPY | +18.8% | −2.2% | −2.2% | −15.2% | $1.31 | $1.35 | $0.3B |
| 6 | JEPI | JPM Equity Premium Income | Anchor·JPM | 74 mo | 2020-05-31 → 2026-06-30 | 8.8% | +2.0% | +11.0% | +10.5% | 0.56pp | SPY | +18.9% | −7.9% | −7.1% | −18.9% | $1.83 | $2.67 | $44.3B |
| 7 | JEPQ | JPM Nasdaq Equity Premium Income | Anchor·JPM | 50 mo | 2022-05-31 → 2026-06-30 | 11.6% | +6.6% | +18.8% | +17.7% | 1.06pp | QQQ | +27.0% | −8.2% | −6.8% | −17.1% | $1.95 | $2.45 | $39.0B |
| 8 | GPIX | GS S&P 500 Core Premium Income | Anchor·GS | 33 mo | 2023-10-31 → 2026-06-30 | 8.5% | +13.8% | +23.4% | +22.8% | 0.68pp | SPY | +26.7% | −3.3% | −3.0% | −9.4% | $1.73 | $1.84 | $4.4B |
| 9 | GPIQ | GS Nasdaq-100 Core Premium Income | Anchor·GS | 33 mo | 2023-10-31 → 2026-06-30 | 10.7% | +17.5% | +29.8% | +28.7% | 1.14pp | QQQ | +34.6% | −4.8% | −4.1% | −10.4% | $1.96 | $2.13 | $4.6B |
| 10 | ISPY | ProShares S&P 500 High Income | Anchor·ProShares | 31 mo | 2023-12-31 → 2026-06-30 | 9.2% | +7.9% | +17.7% | +17.0% | 0.71pp | SPY | +21.9% | −4.2% | −4.0% | −13.4% | $1.48 | $1.61 | $1.3B |
| 11 | SCHD | Schwab US Dividend Equity | Div-growth | 177 mo | 2011-10-31 → 2026-06-30 | 3.2% | +10.5% | +13.9% | +12.9% | 1.02pp | SPY | +16.0% | −2.1% | −2.0% | −22.3% | $5.95 | $7.68 | $95.2B |
| 12 | DIVO | Amplify CWP Enhanced Dividend Income | Div + CC | 115 mo | 2016-12-31 → 2026-06-30 | 5.6% | +7.5% | +13.5% | +12.6% | 0.88pp | SPY | +16.7% | −3.2% | −2.7% | −20.7% | $3.09 | $3.87 | $7.2B |
| 13 | XYLD | Global X S&P 500 Covered Call | Buy-write | 71 mo | 2020-08-31 → 2026-06-30 | 11.5% | −1.3% | +10.0% | +9.6% | 0.47pp | SPY | +16.9% | −6.8% | −5.9% | −24.6% | $1.70 | $2.31 | $3.2B |
| 14 | RYLD | Global X Russell-2000 Covered Call | Buy-write | 87 mo | 2019-04-30 → 2026-06-30 | 12.4% | −5.2% | +6.7% | +5.4% | 1.24pp | IWM | +13.4% | −6.7% | −5.2% | −44.0% | $1.46 | $2.06 | $1.4B |
| 15 | QYLD | Global X Nasdaq-100 Covered Call | Buy-write | 151 mo | 2013-12-31 → 2026-06-30 | 7.5% | −2.0% | +5.3% | +4.6% | 0.64pp | QQQ | +21.4% | −16.1% | −14.9% | −39.3% | $1.76 | $9.26 | $8.2B |
| 16 | QDTE | Roundhill Nasdaq-100 0DTE | Booster | 28 mo | 2024-03-31 → 2026-06-30 | 44.8% | −14.2% | +24.9% | +23.4% | 1.53pp | QQQ | +27.6% | −2.8% | −2.5% | −40.3% | $1.60 | $1.68 | $0.9B |
| 17 | XDTE | Roundhill S&P 500 0DTE | Booster | 28 mo | 2024-03-31 → 2026-06-30 | 31.7% | −11.7% | +16.7% | +15.8% | 0.87pp | SPY | +19.3% | −2.6% | −2.5% | −30.8% | $1.39 | $1.46 | $0.3B |
| 18 | RDTE | Roundhill Russell-2000 0DTE | Booster | 22 mo | 2024-09-30 → 2026-06-30 | 46.6% | −20.2% | +17.8% | +16.3% | 1.56pp | IWM | +21.7% | −3.9% | −3.5% | −40.3% | $1.30 | $1.37 | $0.2B |
| 19 | NVDY | YieldMax NVDA Option Income | Juicer | 38 mo | 2023-05-31 → 2026-06-30 | 80.8% | −13.3% | +57.9% | +52.3% | 5.54pp | NVDA | +85.5% | −27.7% | −22.0% | −55.9% | $3.66 | $5.55 | $1.4B |
| 20 | CONY | YieldMax COIN Option Income | Juicer | 35 mo | 2023-08-31 → 2026-06-30 | 129.2% | −46.8% | +27.3% | +9.1% | 18.26pp | COIN | +75.7% | −48.4% | −20.1% | −92.8% | $1.28 | $2.07 | $0.4B |
| 21 | TSLY | YieldMax TSLA Option Income | Juicer | 44 mo | 2022-11-30 → 2026-06-30 | 89.2% | −37.6% | +21.1% | +9.8% | 11.34pp | TSLA | +45.1% | −24.0% | −13.1% | −86.9% | $1.40 | $2.09 | $0.8B |
| 22 | MSTY | YieldMax MSTR Option Income | Juicer | 29 mo | 2024-02-29 → 2026-06-30 | 144.9% | −46.4% | +37.3% | +7.0% | 30.37pp | MSTR | +54.8% | −17.5% | +3.9% | −92.8% | $1.17 | $1.07 | $1.0B |
| 23 | ULTY | YieldMax Ultra Option Income | Juicer·diversified | 29 mo | 2024-02-29 → 2026-06-30 | 119.8% | −53.6% | +7.4% | +4.7% | 2.62pp | — | — | — | — | −84.4% | $1.11 | — | $0.9B |
| 24 | LFGY | YieldMax Crypto Industry & Tech | Juicer·diversified | 18 mo | 2025-01-31 → 2026-06-30 | 79.8% | −38.3% | +13.6% | +7.5% | 6.08pp | — | — | — | — | −59.3% | $1.11 | — | $0.1B |
A yield means nothing until it is standing next to the thing the fund owns. That comparison is the entire instrument on this page, it is subtraction rather than opinion, and it is the one number no fund’s own marketing page is responsible for printing.
The wrapper always costs something; the only question is how much. Measured as a simple yearly rate, 0 of 21 kept pace. The cost ran from 2.1 points a year to a fund where a dollar became $1.28 against $2.07 in its own stock, and the headline yield did not tell you which one you had bought. The cheapest call-writers cost between 2.2 and 2.8 points a year; the two Goldman funds just below them gave up 3.3 and 4.8, and both sell calls on only part of the book — the modest version of this product, on our own table.
Ask what a dollar became, not what the rate says. A multiple carries no annualization convention, so it cannot be inflated by one — and we needed that rule because of a defect in our own column, not in anybody else’s. It also decided one row the rate got wrong (§5).
A fund with no bear market inside its life has a record about the weather. 17 of these 24 have never been in one. 11 of the 24 grew net asset value while paying you; 8 of those 11 did it entirely inside a rising market.
What the wrapper actually is, on this evidence: a paid drawdown dampener and a bounded decumulation tool. If you are spending the distributions rather than reinvesting them, that is worth something real. It is not a wealth engine, and it was never built to be one.
Everything we make is free — the reports, the ledgers, all of it. The only thing that costs us real money is the market data.