Now You See It, Now You Don’t: The Magic Behind Triple-Digit ETF Yields

There is a category of exchange-traded fund that advertises annual distribution yields of fifty, eighty, sometimes more than one hundred percent. To an investor scanning a list of income funds, these numbers are difficult to ignore. A fund that appears to pay back the better part of your investment every year, in cash, seems almost too generous to pass up.

We want to take that generosity apart and show you precisely what is inside it. The conclusion we will reach, and prove with the funds’ own audited financial statements, is that for many of these funds the headline yield is not income in any meaningful sense. A large part of it is your own money handed back to you, with a management fee deducted on the way out. We will also show you why the documents the fund sponsors point investors toward are structurally incapable of revealing this, and why the one document that does reveal it is the one they rarely mention.

We begin with the strategy these funds are built on.

What a covered call actually is

Imagine you own one hundred shares of a company. You are content to hold them, but you would not mind collecting some cash while you wait. So you strike a deal with another investor. For a fee paid to you today, you grant that investor the right to buy your shares at a fixed price within a set window of time. The fixed price is called the strike, and the fee you collect is called the premium.

Two things can happen. If the share price stays below the strike, the other investor has no reason to pay an above-market price for your shares, so the right expires unused. You keep your shares and you keep the premium. If instead the share price climbs above the strike, the investor exercises the right and buys your shares at the strike. You keep the premium, but you have surrendered every dollar of gain above the strike.

That is a covered call. You are renting out the upside on a stock you own in exchange for cash today. The trade has a definite shape. In a flat or gently rising market it works pleasantly; you pocket the premium and the strike is never reached. In a sharply rising market you leave a great deal of money on the table, because your gains are capped at the strike while the stock runs past it. And in a falling market the premium softens the blow slightly, but you still own an asset that is losing value, and the premium you collected is rarely enough to cover the loss.

Why we do not consider covered calls a sound foundation for income

Here we will part company with much of the financial-marketing industry, and we want to state our view plainly so you can judge it for yourself.

A covered call does not generate income in the way an income investor should want income generated. When you buy shares of a profitable business and collect its dividend, that dividend is paid out of the company’s earnings. It is a share of real profits produced by real commerce. It tends to be stable, it tends to grow over time as the business grows, and crucially it does not depend on the moods of the stock market from one month to the next.

Option premium is a different animal entirely. The amount of premium a covered call generates depends on two things, both unpredictable and outside anyone’s control: the future price of the underlying stock and the future level of its implied volatility. When markets are calm, premiums shrink. When a stock falls, its implied volatility typically spikes, and for a moment the premium on each dollar of exposure actually rises. But the position that generated the premium has lost money, and over any sustained decline the shrinking value of the holding means smaller dollar premiums going forward. The income is not drawn from the productive output of a business. It is extracted by selling away the uncertain future gains of a volatile asset, and the size of that income rises and falls with conditions no one can forecast.

An income stream that shrinks alongside its own asset base precisely when markets are difficult, that depends on volatility staying elevated, and that is funded by surrendering the very upside that would let a portfolio recover from its losses, is not, in our view, a reliable income stream. It is a trade dressed in the costume of income. For an investor who intends to live on the cash a portfolio produces, we consider it closer to a gimmick than to a sound strategy.

What the YieldMax funds are

The YieldMax lineup runs a covered-call-style strategy against a range of underliers. Most of its funds target a single, highly volatile stock, such as Tesla, Nvidia, or Strategy (the company formerly known as MicroStrategy), though some target baskets of stocks or even other YieldMax funds. We will focus throughout on the Tesla fund, TSLY, for two reasons. A single-stock fund is the easiest to follow, with one clear underlier to reason about. And TSLY happens to be a vivid example of the problem we are describing, with a great deal of the destruction we will document.

The yields these funds advertise are the centerpiece of the marketing. The natural question, the one every careful person should ask, is whether a fund can truly generate eighty percent of its value in income year after year, or whether something else is going on. To answer it, we first have to understand how these funds are actually built, because the construction is not what most investors imagine.

Inside the machine: how a YieldMax fund actually works

We examined the prospectus and the actual daily holdings of TSLY as of July 17, 2026. What they reveal is a three-layer machine, and seeing the layers makes the whole strategy intelligible.

The first layer is collateral. Nearly the entire portfolio sits in short-term U.S. Treasury bills and a government money-market fund. On the date we examined, against net assets of about nine hundred eight million dollars, the fund held roughly eight hundred nineteen million dollars in Treasury bills and government obligations. The fund does not own a single share of Tesla.

The second layer is synthetic Tesla exposure. Rather than buying Tesla stock, the fund manufactures the economic equivalent of owning it using a pair of options. As the prospectus describes it, the fund buys call options on the stock and simultaneously sells put options at strike prices approximately equal to the current share price, with one-month to six-month terms. A purchased call profits when the stock rises. A sold put loses when the stock falls. Combined, this pair behaves almost exactly like owning the stock itself: it gains when the stock gains and loses when the stock loses, very nearly dollar for dollar. The prospectus states the goal directly, which is to “synthetically replicate 100% of the price movements of the underlying security.” On the date we examined, this synthetic position amounted to 20,869 option contracts, representing roughly 2.1 million shares of Tesla exposure.

The third layer is the income engine. Against that synthetic exposure, the fund writes short-dated call options, with terms of one month or less under the prospectus and in practice often a single week, at strikes described as zero to fifteen percent above the current share price. It collects the premium and rolls the position again and again. The premium is the cash that funds the distributions.

A note on the precise form of those written calls, because the fund does not always sell them outright. Often it sells a call spread, meaning it sells a call at one strike and buys another call at a higher strike, which is the structure we found in the actual holdings. A plain sold call caps the fund’s gains above its strike permanently. A call spread softens that somewhat: the fund forfeits gains in the band between the two strikes but participates again above the higher strike, which lets it capture part of a very large rally. This is a modest improvement on the upside cap, not an escape from it. In the far more common case of a moderate rise, the fund still surrenders the gains in the band, and as we will see, the spreads do nothing to change the fund’s destruction of capital.

Now we arrive at the structural flaw. On the day we examined, the calls the fund had sold against its position totaled 20,869 contracts, matched by 20,869 contracts of higher-strike calls bought to form the spreads. That number is identical to the size of the synthetic long position. The fund had written calls against the entirety of its Tesla exposure. Once Tesla rises into and through the strikes of those sold calls, the fund stops participating in most of the gains on its whole position. Its downside, however, is not capped at all. The prospectus is candid about this, warning that the sold calls limit the fund’s participation in the stock’s appreciation, while the sold puts expose it to the full extent of any decline in the share price.

That is the trap in a single sentence: full exposure to the stock’s declines, sharply limited participation in its advances, repeated every week. Over a full market cycle, a strategy shaped this way cannot keep pace with the stock it is built on. It collects steady small premiums and absorbs occasional large losses, and the premiums are surrendered upside, not free money.

There is a further implication worth drawing out, because it undermines the entire rationale for owning such a fund. Suppose you want the distributions to hold steady rather than decay. For that to happen, the fund’s net asset value must hold up, and for the net asset value to hold up, the underlying stock must rise enough to offset both the cap on the fund’s upside and any over-distribution. But in precisely that scenario, a rising stock, the covered-call structure underperforms simply owning the stock, because it has sold away the gains. The only conditions under which the fund’s income does not decay are the same conditions under which you would have done better owning the stock directly. And if you owned the stock directly and wanted cash, you could simply sell a few shares whenever you needed it. Selling your own shares produces cash on demand without capping your upside and without paying a fund a yearly fee to cap it for you. In a rising market the do-it-yourself approach wins because you keep the full appreciation. In a falling market it wins because you control the size and timing of your withdrawals rather than having a fund hand your capital back to you on a fixed schedule. It is extremely unlikely that any one of these volatile stocks will move in just the right way for the covered-call fund to come out ahead of simply holding the shares yourself while also sustaining the income that is its whole reason for existing. The one scenario in which the premiums reliably cushion the fund relative to the stock, a sustained decline, is precisely the scenario in which its distributions and its net asset value are collapsing together. And the fund must clear that bar while also charging an expense ratio of roughly one percent a year, a headwind the do-it-yourself approach never faces.

The synthetic structure is not a source of free yield

We must now address a claim that appears throughout the marketing of these funds, and that many investors quietly believe: the idea that holding hundreds of millions of dollars in Treasury bills, earning interest, while simultaneously holding full Tesla exposure through options, amounts to free extra yield on top of the stock exposure. It does not, and the reason is one of the most reliable relationships in all of finance.

First, the honest reason these funds use the synthetic structure at all. It is not to capture Treasury income. It is to satisfy a tax rule. To qualify as a regulated investment company, and thereby avoid being taxed at the fund level, a fund must meet the asset-diversification requirements of Subchapter M of the Internal Revenue Code. A fund cannot simply hold a single stock as nearly its entire portfolio and still qualify. The prospectus spells out how the funds navigate this: each fund aims to keep the value of the options it holds below twenty-five percent of its total assets at the close of every quarter. By holding most of their assets in Treasury securities and keeping the options below that threshold, the funds maintain their tax status while still obtaining roughly one hundred percent exposure to a single stock through those very options. The Treasuries are there to thread a regulatory needle, not to enrich the shareholder.

Now to the deeper point. The reason the Treasury income is not incremental return is captured by a relationship called put-call parity, which holds by simple no-arbitrage logic. For a stock paying no dividend, the prices of a call and a put at the same strike and expiration are bound together by:

where C is the call price, P is the put price, S is the current stock price, K is the strike, r is the risk-free interest rate, and T is the time to expiration. The term e^−rT is simply the present-value discount factor. Strictly speaking, the equality is exact for European-style options, which cannot be exercised early; the American-style contracts these funds often trade add a small early-exercise wrinkle to the sold puts, and it changes none of what follows.

This equation says that when you build a synthetic long position by buying a call and selling a put at the same strike, your net cost to establish it is not zero. It is S − K · e^−rT, which is the stock price minus the discounted strike. That difference is precisely the financing cost embedded in the options, and it equals, very closely, the risk-free interest on the position over the holding period. The options market has already priced in the risk-free rate. You cannot escape it.

A worked example makes this concrete. Suppose Tesla trades at four hundred dollars, the risk-free rate is four and a half percent, and the options expire in three months. The discount factor is about 0.9888, so the discounted strike is about $395.53, and the net cost to establish the synthetic long is about four dollars and forty-seven cents per share. Meanwhile, the fund holds roughly four hundred dollars per share in Treasury bills, which over three months at four and a half percent earns about four dollars and fifty-three cents. The two figures are essentially identical. The Treasury interest the fund earns is consumed, almost to the penny, by the financing cost embedded in the at-the-money options used to create the stock exposure.

We can state this more rigorously. If you hold the synthetic long and also hold a Treasury bill with face value equal to the strike, the total cost today is exactly the current stock price, and the total value at expiration is exactly the stock price at that time, whether the stock has risen or fallen. The combination replicates owning the stock outright, with no extra return whatsoever. The practical conclusion is important. The synthetic-plus-Treasuries package is economically identical to simply owning the stock. The only thing that makes a YieldMax fund’s return differ from owning the stock is the covered-call overlay, the selling away of the upside, which is the very feature that almost guarantees the fund will trail the stock over time. The Treasury income, presented as a contributor to the distribution, adds nothing the financing cost has not already taken away. There is no free lunch here, because put-call parity does not permit one.

We will add an editorial note, because the point deserves one. The diversification requirement these funds work around exists for a reason: to keep funds from concentrating investors’ money in a single security and the outsized risk that comes with it. Using options to manufacture one hundred percent exposure to a single volatile stock, while parking just enough in Treasuries to satisfy the letter of the rule, honors the form of that protection while defeating its purpose. The result is a product that passes a diversification test on paper but behaves like a leveraged bet on one stock, sold to retail investors who may have no idea they are taking single-stock risk of that magnitude. In our view this is a structure regulators should look at hard, and at a minimum it places unsophisticated buyers in far more risk than they are likely to realize.

The two kinds of return of capital

We can now turn to the heart of the matter. When a fund sends you a distribution, that distribution receives a label for tax purposes. It might be labeled as income, as capital gains, or as something called return of capital, abbreviated ROC. The name alarms people, and sometimes it should. But return of capital comes in two very different forms, and confusing them is the single most common error in this entire subject.

The first kind we will call non-destructive return of capital. Here the fund genuinely earned the money it paid you, but for tax reasons the distribution is classified as return of capital rather than as income. This happens routinely and innocently. Real estate funds do it because depreciation deductions shelter income from tax even though the buildings are throwing off real cash. Certain well-run options funds do it because of the way their realized losses and unrealized gains line up, which we will explain shortly. In these cases the fund earned the distribution through its strategy, the net asset value holds steady, and the return-of-capital label is nothing more than a quirk of tax accounting. No principal has been touched.

The second kind we will call destructive return of capital. Here the fund pays out more than it actually earned and makes up the difference by reaching into principal. It is handing you back the very money you invested and calling it a distribution. When this happens, the fund’s net asset value falls as a direct consequence. This is the dangerous form, and it is the form we will prove is at work in TSLY.

The inconvenient fact is that the label on your tax form does not tell you which of the two you are looking at. A perfectly healthy fund and a fund in the process of liquidating itself can both report their distributions as return of capital. To tell them apart, you must look past the tax label to the economics, and to do that you have to understand that funds keep two entirely separate sets of books.

Two sets of books: tax accounting and financial accounting

This is the conceptual key to the whole subject, so we will take it slowly.

Every fund maintains its records under two different accounting systems that answer two different questions.

The first is financial accounting, sometimes called book or GAAP accounting. This is the system used to prepare the fund’s financial statements and to calculate its net asset value each day. It marks every holding to its current market value, which means it recognizes gains and losses as they occur, including gains and losses that exist only on paper and have not yet been locked in by selling. This system answers the question: what is the fund actually worth, and what did it actually earn or lose, economically, over the period? The net asset value is the product of this system, and it is the honest ledger of the fund’s economic reality.

The second is tax accounting, which governs the character of distributions. It follows a different and more rigid set of rules built around a concept called earnings and profits. Critically, tax accounting generally counts only realized gains and losses. A loss that exists on paper but has not been crystallized by closing the position reduces the fund’s market value, and therefore its net asset value, but it does not reduce the fund’s earnings and profits for tax purposes. Tax accounting also carries forward accumulated earnings from prior years, and applies special rules to certain instruments. This system answers a narrower question: how should each distribution be characterized on a tax form? Return of capital, in the tax sense, is simply the portion of a distribution that exceeds the fund’s current and accumulated earnings and profits.

These two systems can diverge enormously, and the gap between them is where the confusion lives. Let us work two examples to make the distinction unmistakable.

A worked example of non-destructive return of capital. This is, in fact, roughly how a well-run options fund such as the NEOS funds generates its return of capital, and it is worth understanding because it is the benign case people most often mistake for the harmful one. Suppose a fund holds a basket of stocks and sells call options against them. Over the year the stocks it holds appreciate, but because the fund continues to hold them, that gain is unrealized. Meanwhile the call options it wrote lose money as the market rises, and those losses are realized as the options are closed and rolled. The fund may even deliberately harvest losses for tax purposes. Now look at the two sets of books. Under financial accounting, the unrealized stock gains more than offset the realized option losses, so the fund’s total economic result is positive and its net asset value holds steady or rises. But under tax accounting, only the realized results count, and the realized results are dominated by the option losses, which push the fund’s earnings and profits down toward zero. When the fund distributes its cash, that distribution exceeds its earnings and profits, and the excess is classified as return of capital. An investor reading the tax form sees a large return-of-capital figure and panics. Yet the fund earned every dollar it paid out, the net asset value is intact, and the return-of-capital label is purely a tax artifact arising from realized losses lining up against unrealized gains. This is non-destructive return of capital. The annual report would show total economic earnings comfortably covering the distribution and a stable net asset value.

A worked example of destructive return of capital. Now picture a different fund. It begins the year with a net asset value of fifty dollars per share and collects five dollars per share of option premium, which is genuinely earned income. But the underlying stock falls, and the fund’s position loses eight dollars per share. Its total economic result for the year is five dollars of premium minus eight dollars of losses, a net loss of three dollars. The eight-dollar loss is a market movement; it would have reduced the net asset value whether or not the fund paid any distribution, exactly as any stock fund’s net asset value falls in a down market. The question is what the manager chooses to distribute. Had the fund distributed its five dollars of earned income, it would have been passing through what it actually earned, and the only reason the net asset value fell would be the market loss on the position. That is not a return of principal; it is a normal fund passing through income while the market moves against it. Instead, suppose the manager chooses to distribute twelve dollars. The seven dollars above the fund’s earned income now comes out of capital the strategy never produced. The manager was under no obligation to do this. He did it because he wanted to sustain the advertised yield. The market handed the fund a loss, and the manager chose to compound it by handing investors back their own principal on top. That choice, not the market, is what makes this destructive.

Hold on to the lesson from these two examples: the tax label does not tell you whether a distribution was earned. Two funds can report identical return-of-capital figures while one is perfectly healthy and the other is bleeding. To know which is which, you have to compare what the fund earned against what it paid out, and for that you need the financial statements.

How to detect destructive return of capital

The financial-accounting truth lives in a single audited table called the Financial Highlights, found in every fund’s annual report, which is filed with the Securities and Exchange Commission on a form called the N-CSR. This table presents the complete economic story of one share over the year, and it has been examined by an independent auditor. We will walk through the lines that matter and then give you a method for using them.

Net asset value, beginning of period. What one share was worth at the start of the year.

Net investment income. The income the fund earned during the year in the strict accounting sense, less the fund’s expenses. One point here surprises nearly everyone, and it matters: option premium does not appear on this line. Under the accounting rules that govern funds, the premium collected from written options is not treated as income at all; it flows through the realized and unrealized gain-or-loss line below when the options are closed out or expire. For a YieldMax fund, net investment income is therefore essentially the interest on its Treasury bills and money-market holdings, minus the management fee, the very interest that put-call parity told us the financing cost of the synthetic position has already claimed.

Net realized and unrealized gain or loss on investments. The change in the market value of the fund’s positions, both the gains and losses it locked in by closing positions and the gains and losses that exist on paper at year-end. For a YieldMax fund, this line captures how the synthetic stock position performed. This is the market-movement component.

Total from investment operations. The sum of the previous two lines. This is everything the fund produced economically over the year, income plus market performance.

Distributions, broken into sub-lines. The cash paid out, separated by tax character: distributions from net investment income, from net realized gains, and from return of capital. These sub-lines are the tax classification, and as we have seen, they do not measure economic earning.

Total distributions. The sum of those sub-lines, every dollar paid out.

Net asset value, end of period. The beginning value, plus total from investment operations, minus total distributions, with a small adjustment for shares being created and redeemed.

Now for the method, and here we have to be careful, because the obvious approach is wrong.

The obvious approach would be to subtract total from investment operations from total distributions and call the difference destructive return of capital. In a single year that calculation is misleading, and a single fund shows why. Consider a healthy dividend fund in a down market year. Its stocks fall in price, so its total from investment operations is negative, yet it still pays its ordinary dividend out of the dividend income it collected. Subtracting a negative total-from-operations from a positive distribution would produce a large apparent “return of capital,” and you would wrongly conclude the fund had handed back principal. It had not. Its distribution was covered by real income, and its net asset value fell only because the market fell, a decline that reverses when the market recovers. The naive subtraction has mistaken a market movement for a return of principal.

The correct method has two steps.

Step one, the primary test: compare total distributions to net investment income. Net investment income is the income the fund genuinely earned. If a fund’s distribution does not exceed its net investment income, it is paying out only what it earned, and it is not returning principal, regardless of how the market moved that year. This is why a healthy dividend fund never shows destructive return of capital even in a losing year: its distribution stays within its dividend income, and any decline in net asset value is market-driven. A fund that distributes far more than its net investment income, on the other hand, is the warning sign that sends you to step two. We should add one caveat for the options funds specifically. Their true economic earnings come from option premium, and as we noted above, premium never touches net investment income at all; it lives in the realized and unrealized line, tangled together with the gains and losses on those same option positions. For a YieldMax fund, then, net investment income is just collateral interest minus the fee, and the test cuts only one way: a distribution that dwarfs net investment income tells you the payout is not funded by anything the accountants call income, but only the cumulative test below can tell you whether the strategy as a whole earned what it paid.

Step two, the magnitude: compare cumulative distributions to cumulative total from investment operations over the fund’s entire life. A fund can only pay out, in total, either what it has earned in total or your principal. So sum everything the fund has produced economically since inception, and sum everything it has distributed since inception. If the cumulative distributions exceed the cumulative total from operations, the difference is principal returned, the true destructive return of capital, and it will closely match the decline in net asset value from the fund’s starting price. We use the cumulative figures, never a single year, precisely because a single year’s market swings, especially unrealized ones, distort the picture, while over the full life those swings resolve into the strategy’s actual economic result.

With this method in hand, we can examine the two funds.

What healthy distributions look like: SCHD

For the model of a fund whose distributions are fully earned, we turn to SCHD, a large dividend fund, using its audited Financial Highlights from its most recent annual report as of this writing, for the fiscal year ending in August 2025. The figures below are per share and reflect the fund’s three-for-one split.

Look first at the relationship between what SCHD earned in income and what it distributed, across five fiscal years:

Apply step one. In every single year, the distribution was fully covered by net investment income, the real dividend cash the fund collected, and in most years it was slightly less. And in every single year, the return of capital was zero. SCHD distributes what it earns in dividends, and not a penny more.

Now the demanding test, the year the market fell. In the fiscal year ending August 2022, SCHD’s net realized and unrealized line was negative one dollar and seventy-nine cents per share, and its total from investment operations was negative ninety-five cents. The fund lost money that year; its total return was negative 3.74 percent. This is exactly the case where the naive subtraction misleads: distributions of eighty-one cents minus a negative total-from-operations of ninety-five cents would suggest a dollar and seventy-six cents of “return of capital.” But the fund returned nothing of the sort. It paid its eighty-one-cent distribution entirely out of its eighty-four cents of net investment income, with zero return of capital, and its net asset value fell only because the stock prices fell. Step one gives the right answer where the naive formula gives the wrong one.

This is the structural feature that matters, and it is the precise point on which SCHD and TSLY diverge. SCHD’s income comes from the dividends paid by the roughly one hundred companies it holds, and those dividends are paid out of those companies’ business earnings. They arrive at the fund regardless of what SCHD’s own share price or net asset value is doing. A company does not cut its dividend because a fund’s net asset value fell. SCHD’s income source is external to the fund and independent of its own value. A YieldMax fund’s income, by contrast, is generated from the fund’s own portfolio, by selling options whose premium depends on the size and volatility of that very portfolio. As the portfolio erodes, the income capacity erodes with it. The income is internal and self-consuming. That single difference is why one fund’s distributions are durable and the other’s are corrosive.

The proof: TSLY’s own audited numbers

We now apply our method to TSLY, using the audited Financial Highlights from the fund’s annual report for the period ending October 31, 2025. Every figure below is the fund’s own reported, audited number, on a per-share basis.

One housekeeping note before the numbers, so that nothing here looks inconsistent with a price chart. TSLY has executed two reverse splits, one-for-two in February 2024 and one-for-five in December 2025, and the Financial Highlights restate the fund’s entire per-share history onto the current share count, ten of today’s shares to one original share. Every per-share figure in this section is on that post-split basis. That is why the record shows a fund that began near two hundred dollars per share when it originally listed near twenty. The economics are identical on either basis; only the denomination of a share changes.

Begin with step one. In the 2025 fiscal year, TSLY earned one dollar and forty-eight cents per share in net investment income and distributed forty-two dollars and twenty-two cents. The distribution was roughly twenty-eight times the earned income. That alone is the warning sign. And remember what that one dollar and forty-eight cents actually is: almost entirely Treasury interest, the same interest the financing cost of the synthetic position has already consumed.

Now step two, the full record:

Read the bottom row slowly. Over three years, the strategy produced about twenty-two dollars per share of genuine economic return, combining all its income and all its market gains and losses. Over those same three years, the fund distributed nearly one hundred eighty dollars per share. Applying step two, the cumulative destructive return of capital is one hundred seventy-nine dollars and fifty-seven cents minus twenty-two dollars and one cent, which is one hundred fifty-seven dollars and fifty-six cents per share.

That figure did not come from the strategy, because the strategy did not produce it. It came from principal. And we can confirm this independently. The fund launched in November 2022 at a net asset value of two hundred dollars per share on the post-split basis described above. By the end of this period it stood at forty-two dollars and fifty-eight cents, a decline of one hundred fifty-seven dollars and forty-two cents, within fourteen cents of our calculated destruction. The small remainder is not a discrepancy. It is the per-share transaction fees that market makers pay into the fund when they create and redeem shares, which the Financial Highlights reports as its own small line, plus a penny of rounding.

So of every two hundred dollars an original investor placed in this fund, about one hundred fifty-eight dollars was eventually handed back to them as their own capital, dressed as a distribution, with a management fee skimmed off each year along the way. The investor was not being paid income. The investor was being returned a shrinking pile of their own money and charged for the service.

It is worth pausing on the 2025 column, because it is the year a defender of the fund would point to. Tesla rose substantially that year, and the strategy genuinely produced twenty-five dollars per share of economic return, its best showing. And still the fund distributed forty-two dollars, overshooting its earnings by more than seventeen dollars and eroding the net asset value even in a good year. When the best year still destroys capital, the structure is the problem.

Why the funds do this on purpose

None of this is an accident. Over-distributing is a choice. No rule requires a fund to pay out more than it earns. The fund’s board sets the distribution policy, and it could just as easily distribute only the modest amount the strategy genuinely produces. It chooses not to, and the reason is plain.

The headline yield is the product. A fund advertising an enormous yield draws enormous attention and enormous inflows, and those inflows become assets under management, on which the sponsor earns its fee. The incentive could not be clearer: maximize the advertised yield to maximize the assets to maximize the fee income. Whether the distribution is real income or returned principal is, from the sponsor’s commercial vantage point, beside the point.

There is a self-destructive twist to this, and it follows directly from the difference we drew with SCHD. Because the YieldMax income is generated from the fund’s own portfolio rather than from outside it, over-distributing erodes the very asset base that produces the income. A smaller base generates smaller premiums, which produces a smaller income capacity, which forces either an honest cut to the distribution or a deeper raid on principal. The machine consumes its own fuel. This is why the per-share distributions of several of these funds have collapsed over time as their net asset values have withered. The yield that attracts investors is the very mechanism that destroys what they invested.

Not every YieldMax fund is equally destructive: NVDY

We owe you a piece of intellectual honesty here, because an argument is only as strong as its treatment of the cases that seem to cut against it.

Not every YieldMax fund has destroyed capital to the degree TSLY has. The Nvidia fund, NVDY, is the instructive counterexample. Its structure is identical to TSLY’s, but its underlying stock behaved very differently over the period. Nvidia rose relentlessly, and that powerful, sustained rally handed the strategy enough genuine gains to nearly cover its distributions. Here are its audited per-share figures.

Across its three fiscal years, the first of them a partial year from the fund’s May 2023 launch, NVDY earned about thirty-two dollars per share and distributed about thirty-five. The cumulative destructive return of capital was only about three dollars and thirty-one cents per share, and the net asset value fell modestly, from twenty dollars to about sixteen dollars and seventy-two cents. By the standards of TSLY, that is a mild outcome.

The lesson is not that NVDY is a good fund. The lesson is that the destruction is not guaranteed in every single year, but the structural vulnerability always is. When the underlying stock cooperates with a strong and sustained rally, the strategy can nearly keep its head above water. When the underlying stock does not cooperate, the structure does to an investor’s capital exactly what it did to TSLY’s. A buyer of these funds is not making a bet on income. They are betting that the underlying stock will rise fast enough and steadily enough to outrun a strategy deliberately built to fall behind it. That is a very different proposition from the one the marketing presents, and it is a poor one for anyone seeking income they can rely on.

The tax form understates the damage

The NVDY example lets us demonstrate, with real audited numbers, a point we proved in the abstract earlier: that the tax figure understates the economic destruction.

In NVDY’s most recent fiscal year, the fund’s total from operations was four dollars and eighty-six cents, while it distributed thirteen dollars and three cents. The economic shortfall was eight dollars and seventeen cents per share. But the amount formally classified as return of capital on the fund’s tax accounting was only six dollars and fourteen cents. More than two dollars of real economic destruction went unlabeled.

The divergence flows from the two features of tax accounting we described. First, tax accounting counts only realized losses, so any losses still sitting on paper at year-end reduced the net asset value without reducing the fund’s earnings and profits, allowing more of the distribution to be characterized as something other than return of capital. Second, Nvidia’s enormous prior year had let the fund bank earnings and profits, and those accumulated earnings could absorb part of the current year’s distribution for tax purposes, even though, economically, the current year’s payout plainly exceeded the current year’s production.

The practical consequence bears repeating, because it is the crux of why these funds are so easily misunderstood. The return-of-capital figure on a tax form is the floor of the damage, not the ceiling. By the time the tax accounting fully catches up to economic reality, which it must once those paper losses are eventually realized, the net asset value has already been spent.

Why the documents YieldMax points to cannot reveal destructive return of capital

This brings us to a point that deserves emphasis, because it explains why so many investors in these funds have no idea what is happening to their capital. The fund sponsors, in their educational materials on return of capital, direct investors toward three documents: the Form 1099-DIV, the Section 19(a) notice, and Form 8937. Each is a legitimate document. And each, by its very design, is incapable of revealing destructive return of capital. We will take them one at a time, and we have verified the description of each against the governing rules.

Before we do, one fact worth holding in mind, because it dissolves a common misconception. Every dollar of return of capital, whether destructive or not, carries the same tax treatment, and every dollar of it reduces your cost basis in the fund. Non-destructive return of capital reduces your basis, and so does destructive return of capital. The tax character is identical. That is exactly why these three documents, all of which speak the language of tax character and cost basis, cannot distinguish the two. A document that tells you your basis was reduced has told you nothing about whether the fund earned the money.

The Form 1099-DIV is the year-end tax form you receive. It reports your distributions sorted by tax character, with return of capital appearing in Box 3 as a nondividend distribution. Every figure on this form is a product of tax accounting, the earnings-and-profits system that counts only realized results and carries forward prior-year balances. The Box 3 figure is therefore the tax figure, the one that ignores unrealized losses and understates economic destruction. The 1099-DIV tells you how to report what you received and how much to reduce your basis. It does not tell you whether the fund earned it.

The Section 19(a) notice, required under Rule 19a-1 of the Investment Company Act, accompanies a distribution whenever part of it comes from a source other than net investment income. It provides an estimated breakdown of the distribution’s sources. Three features make it unfit for measuring destruction. It is explicitly an estimate, revised as the year progresses, and the funds themselves state it is not for tax purposes. Its source categories are built around realized results, so its headline return-of-capital figure does not reflect unrealized losses. And it characterizes the sources of a single distribution rather than comparing the fund’s total economic earnings to its total distributions. The rule’s purpose, as the SEC describes it, is to keep shareholders from mistaking a return of capital for investment income, which is a worthy aim, but the notice was never designed to measure whether a strategy is eroding principal, and it does not.

Form 8937, titled Report of Organizational Actions Affecting Basis of Securities, is filed when a distribution reduces shareholders’ cost basis. For a return-of-capital distribution it reports the per-share amount that reduces your basis. As the filings themselves state, a return of capital here is the amount of a distribution paid in excess of the issuer’s current and accumulated earnings and profits. That is the tax definition once again. Form 8937 exists to adjust your cost basis. It compares distributions to earnings and profits, not to economic earnings, and so it too obscures destruction.

Notice what all three share. Every one of them reports figures derived from tax accounting or from realized-only estimates. Not one incorporates the unrealized losses that are eroding the net asset value, and not one places total economic earnings beside total distributions. To see the real economic sources of a fund’s returns, and to measure them against what the fund actually paid out, you must do what we did in this piece: open the annual report, find the Financial Highlights, and compare what the strategy produced against what the fund distributed over its life. That is the only document that shows the economic truth, and it is conspicuously absent from the return-of-capital education these sponsors provide. We will let readers draw their own inference about why the one document that would settle the question is the one not mentioned.

What YieldMax says in its defense

YieldMax maintains an educational page devoted to return of capital, and its central claim is that ROC is, in its words, “a tax concept, not an economic concept.” The page goes on to describe return of capital as something often misunderstood as a sign of performance when it is in fact beneficial for most investors, and to assert that the classification does not indicate a fund is returning investor principal due to losses.

We should give the central claim its due, because in a narrow sense it is correct. The specific label that appears on a tax form is indeed a creature of tax accounting, and as the NEOS example showed, the label by itself proves nothing bad. But the argument performs a quiet substitution. From the narrow and true observation that the label is a tax concept, the reader is invited to conclude that the underlying economic event, a fund paying out more than it earned and eroding its net asset value, is not real or not a cause for concern. That conclusion does not follow. Returning capital in the economic sense, distributing principal because the strategy did not generate enough to cover the payout, is unmistakably an economic event, whatever it happens to be called on a form.

As for the assurance that the classification does not indicate a fund is returning investor principal because of losses: for funds whose distributions exceed their economic earnings, that assurance fails, and the funds’ own audited financial statements are the proof. TSLY distributed about one hundred eighty dollars per share against about twenty-two dollars of economic earnings. The roughly one hundred fifty-eight dollar difference was investor principal, returned precisely because the strategy did not produce enough to cover the distributions. The statement is true of the benign kind of return of capital and false of the destructive kind, and the page does not equip the reader to tell the two apart. It is telling that the defense rests entirely on the tax framing and the tax documents, and never invites the investor to open the annual report.

A precise word about NAV erosion

The phrase NAV erosion is used loosely in this corner of the market, and because precision is what makes an argument durable, we want to be exact about it.

Mechanically, NAV erosion is simply a decline in net asset value per share over time. But a decline can arise from several causes, and only some of them are meaningful. The kind of erosion that should concern an investor is the kind that comes from deliberate actions by the fund’s managers, not from market movements in the holdings. We see three such sources, in order of importance.

The first and largest is destructive return of capital, the subject of this entire piece: distributing more than the fund earned, which converts principal into distributions and pulls the net asset value down year after year. This is a choice the managers make, and for the YieldMax funds it is the dominant driver.

The second is above-average fees. A fund’s expenses are deducted from its assets, so a high expense ratio erodes the net asset value steadily and relentlessly, compounding against shareholders over time. These funds charge roughly one percent a year, well above what a plain index fund costs, and that drag never stops.

The third, and least important, is the trading cost of a high-turnover strategy. Rolling options every week generates enormous portfolio turnover, and every trade crosses a bid-ask spread and incurs transaction costs that are paid out of fund assets. The prospectus itself acknowledges that the practice of rolling options may result in high turnover. This is a smaller leak than the other two, but it is a real one, and it runs constantly.

What ties these three together is that each is a controllable, manager-driven reduction of the net asset value, as opposed to the ordinary rise and fall of the fund’s holdings with the market. And the reason NAV erosion matters so much, beyond the immediate loss, is forward-looking. A shrinking net asset value is a shrinking asset base, and a smaller asset base can generate less future income, because there is less capital left to put to work. Erosion is therefore self-reinforcing: the more principal a fund pays out, the less it can earn, which pressures it to pay out still more principal to maintain the appearance of a high yield. This is the doom loop we described earlier, stated in the language of the net asset value.

We want to distinguish this carefully from a sloppier claim we sometimes hear, that covered-call funds cause NAV erosion simply by underperforming the underlying stock. That is not erosion in any meaningful sense; it is merely the opportunity cost of a strategy that caps its upside, and by that loose logic any fund that fails to track the single best-performing asset would be said to erode. The erosion that concerns us, and that we have documented, is the part the managers cause: paying out capital the fund did not earn, charging high fees, and churning the portfolio. We draw the line carefully so the claim cannot be dismissed with the obvious reply that any fund’s value can fall when its holdings fall.

The takeaway

A high yield is not a high return. The two are not the same thing, and the gap between them is where investors lose money without quite noticing. A distribution is income only if the fund earned it. If the fund did not earn it, the distribution is your own capital, handed back to you, with a fee subtracted on the way out.

The covered call, in our view, is not a sound foundation for income to begin with, because its proceeds depend on stock prices and volatility that no one can forecast, and because it funds its payouts by surrendering the upside that would let a portfolio recover. The YieldMax funds take that shaky foundation and compound the problem by distributing far more than even the strategy produces, financing the difference out of principal. They manufacture their stock exposure synthetically, not to earn free yield, which put-call parity forbids, but to satisfy a tax rule, and they point investors toward tax documents that by design cannot reveal the erosion underway.

The good news is that the tool to check is public and free. Open the annual report and find the Financial Highlights. First, see whether the distributions stay within net investment income; if they do, the fund is paying out earned income. If they tower over it, set cumulative distributions beside cumulative total from investment operations over the fund’s life. If the distributions have exceeded everything the strategy ever produced, you are not looking at an income fund. You are looking at a fund returning your own capital to you, slowly, and charging you for the privilege.

We hold no animus toward genuine income investing, which we practice and respect. Our objection is narrow and, we believe, fully demonstrated. A fund that systematically distributes more than it earns is consuming its investors’ capital, and no amount of careful language on a tax-education page changes what the financial statements plainly show.

See it yourself: the annual report is here. To find TSLY’s Financial Highlights, scroll a little more than halfway down until page numbers appear in the bottom-right corner and find page 95, or simply use your browser’s find function to search for “Financial Highlights” — TSLY is labeled “YieldMax TSLA Option Income Strategy ETF.”

Tools we use — see our full toolkit (some links earn us a commission).