The bid was the floor. Then the bidder walked.

We put 5% on the deal collapsing. It collapsed. The stock is indicated down 13%. This note is about what the $60.50 bid still proves, what it never proved, and why the second question matters more than the first.

01 / The tapeWhat happened overnight.

Bloomberg reported late on 27 August that the consortium of Advent International and Stripe has abandoned its pursuit of PayPal Holdings. The report cites people familiar with the matter, notes that the situation "remains fluid," and states that the consortium could return with a revised offer if circumstances change. PayPal, Stripe and Advent all declined to comment. Reuters recirculated the report. As of this writing there is no company statement, no 8-K, and no on-the-record confirmation from any of the three parties — a fact worth holding onto, because every stage of this six-month sequence has been sourced the same way.

PYPL closed the regular 27 August session at $61.47. It traded down to roughly $53.66 overnight, a decline of about 12.7%. That is the price this note underwrites. It is a pre-market indication, not a close.

The full sequence, because the dates carry the argument:

02 / Marking our own bookWe assigned this 5%.

On 15 July we published The Bid Is the Floor. We argued the board would be forced by Revlon duties into a competitive process, we put 55% on a clearing price of $75–$95, and we put 5% on "deal collapses; stock re-rates to standalone fundamentals, ~$50."

The 5% branch is the one that hit. We should be precise about what we got wrong, because it is not what it looks like.

• We were right that the board would not accept $60.50. It did not. It reportedly held out for something near $70.
• We were right that the process would continue. It did — talks reheated in mid-August at a price above the original bid, and the stock traded through $62.
• We were wrong about the auction. No competing bidder emerged. Not Alphabet, not Amazon, not Apple, not KKR or Silver Lake or Blackstone. Six weeks of a publicly known, fully financed $53B offer sitting on the table produced zero counter-bids. That silence is the single most important new datapoint in this situation, and it is not a friendly one.
• We under-weighted walk-away risk. We wrote that a withdrawal was "unlikely given the $50B in committed financing already lined up." Committed financing is a condition of a bid, not a commitment to bid. Sponsors walk. We priced that at 5% and it should have been 15–20%.

03 / What the bid still provesThe datapoint does not disappear with the bidder.

The core of the constructive case is simple, and it is the reason we think a 13% single-day markdown is an overreaction rather than a correction.

Between March and August 2026, two of the most sophisticated buyers in payments — a strategic operator that runs the competing rail and a private equity firm with four decades of financial-services underwriting — spent months on this asset. They arranged approximately $50 billion of committed bank financing. They submitted a formal, all-cash proposal at $60.50. They did not walk when it was rejected in July; they came back in August to discuss a higher number. That is not a tyre-kick. That is a fully underwritten view that PayPal's equity is worth more than $53 billion to an owner who can hold it.

A financial sponsor underwriting a take-private at $60.50 is not buying at their estimate of fair value. They are buying at a price that clears a required return — typically a 2.0–2.5× multiple of invested capital over five years. Whatever Advent's internal exit case was, it was materially above $60.50. That number does not evaporate because they declined to pay $70 for it.

The mistake the tape is making this morning is treating the withdrawal as new information about the business. It is not. It is information about a negotiation. Nothing about PayPal's cash generation changed between the 27 August close and the 28 August indication. What changed is that a near-term catalyst was removed — and the market repriced roughly $6.9 billion of equity value in response to the removal of a catalyst, not the arrival of a fact.

04 / The arithmeticWhat you are buying at $53.66.

PayPal guides FY2026 to ~$5.38 non-GAAP EPS, ~$15.6B of transaction margin dollars ($14.5B excluding interest on customer balances), $6B+ of adjusted free cash flow, and ~$6B of share repurchases. It closed Q2 with $15.3B of cash and investments against $13.4B of debt — a net cash position of roughly $1.9B. On ~882M diluted shares:

Three things follow. First, at $53.66 the company trades at ten times earnings and 7.6× EV to free cash flow, with net cash, for a business that processed $486 billion of payment volume last quarter across 439 million active accounts. Second, the rejected bid — the number a financed buyer was willing to pay — implies only 8.6× EV/FCF. The buyer was not paying a rich multiple; the buyer was paying a private equity multiple, which is precisely the point we made in July and which remains true. Third, the board's reported $70 ask implies 13.0×, which is roughly PayPal's own three-year average P/E. The board was not asking for a fantasy. It was asking to be paid its own historical multiple.

The buyback is the mechanical part of the argument. At a $47.3B market capitalisation, a $6B annual repurchase retires about 12.7% of the shares outstanding every year. That is not a rounding item; it is the dominant driver of per-share value at this multiple. The company has already retired ~111M shares over the trailing twelve months. Every dollar the stock falls makes that programme more accretive — which is the one genuine mechanism by which a failed deal is better for a patient holder than a completed one at $60.50.

A takeout at $60.50 converts a 12.7% free cash flow yield into cash at 11.2%. Continued ownership at $53.66 converts it into 12.7% of the share count, annually, for as long as the discount persists. The second is worth more — but only if the cash flow holds.

05 / What the bid does not proveThe bear case, stated properly.

We would be doing the reader a disservice if we left it there. The "rejected bid is a floor" argument has four serious holes, and an honest note names them.

• A leveraged buyer's price is not a public shareholder's value. Advent and Stripe were underwriting an asset they could recapitalise with ~$50B of debt, run without quarterly guidance, and strip of cost. PayPal itself has just announced a target of at least $1.5B of gross run-rate savings over two to three years — roughly 25% of non-GAAP operating income, and precisely the sponsor playbook. But a public shareholder has no mechanism to compel that outcome, no control premium, and no exit date. The $60.50 is a floor for a controlling owner. It has never been a floor for a minority holder, and this morning is the proof.
• Informed buyers walking is a ceiling datapoint, not only a floor datapoint. Two parties with months of access declined to pay $65, let alone $70. The most economical reading of the withdrawal is that the consortium's own model did not support the board's number. If $60.50 was near the top of what a levered, synergy-taking, cost-cutting buyer would pay, then the unlevered public value is plausibly below $60.50, not above it.
• The absence of a counter-bid is damning. A $53B all-cash offer for the second-largest online payments platform in the world was public for six weeks. Not one strategic or financial party stepped forward. Our July note assumed the auction dynamic; the auction did not exist. Either the antitrust and integration problems are worse than we modelled, or the industry's assessment of the terminal value of a legacy checkout button is worse than ours.
• The core is not compounding. This is the number that matters most and it is not good. In Q2, revenue grew 5% (3% currency-neutral) and TPV grew 10% — but transaction margin dollars grew 1%, and 3% excluding interest on customer balances. Non-GAAP operating margin fell 248 basis points. GAAP net income fell 11%. Active accounts were flat at 439 million. Branded checkout TPV grew 2% currency-neutral. Volume is growing at low-margin Braintree and Venmo; the high-margin branded button is not. Non-GAAP EPS of ~$5.38 against 2025's $5.31 is a ~1% increase delivered on roughly an 8% reduction in average diluted share count — which means underlying earnings declined.

06 / What has to be trueFor the market to be right at $53.

We find it useful to invert. For the bear case to be correct — for PayPal to be worth $45 rather than $65 — the following must hold:

• Branded checkout must roll over from low-single-digit growth to outright decline within 24 months, as agentic commerce, native platform wallets and stablecoin rails route around the button. This is the entire thesis. Everything else is second order.
• The $1.5B cost programme must fail to reach the transaction margin line, or be competed away in take rate. If it lands, transaction margin dollars can grow mid-single digits even with a flat branded business.
• Venmo and Braintree must remain structurally low-margin. Venmo TPV grew 14% in Q2; if monetisation improves materially, the mix argument reverses.
• The buyback must be cut or redirected. At a 12.7% annual retirement rate, even zero earnings growth produces high-single-digit per-share compounding. The bear case requires the capital return to stop.

We think the first is a genuine, live risk and the reason this position is medium- and not high-conviction. We think the last three are unlikely to all hold simultaneously.

07 / ScenariosWhere this clears.

Our revised probability-weighted view, on a 6–12 month horizon, against the ~$53.66 indication:

That weights to approximately $59, or ~10% above the indicated price — before the buyback. It is a materially thinner edge than the 30–40% we underwrote in July, and it should be. The free option we were holding has expired. What is left is an ordinary value proposition with an unusual amount of downside protection and one genuine terminal risk.

08 / PositionWhat we are doing.

AMA Invest holds a long position in PayPal (PYPL) at the time of publication. We initiated on the 15 July bid tape as a medium-conviction event-driven position. That characterisation no longer applies: with the consortium withdrawn, this is no longer an arbitrage. We are re-underwriting it as a value position with optionality, at unchanged size, with a materially longer intended holding period.

We are not adding into the gap. The discipline we would apply to anyone else's note applies to our own: we assigned the collapse a 5% probability, we were wrong about the auction dynamic, and the appropriate response to being wrong about the mechanism is to re-underwrite at the same size rather than to average down into a thesis that has just lost one of its two legs.

We would add on evidence — not on price. Specifically: a Q3 print showing transaction margin dollars ex-interest accelerating above 3%, branded checkout holding low-single-digit growth, or an explicit board commitment to an enlarged repurchase at these levels. We would exit on a Q3 print showing branded checkout turning negative.

The July trade was "buy PayPal because the process will produce a higher clearing price." That trade is over and it did not work. The trade now is narrower and duller: buy the second-largest payments platform on the internet at ten times earnings, with net cash, retiring an eighth of itself every year, at a 12% discount to what a fully financed buyer offered eight weeks ago — and accept that you are being paid to carry a real risk that the checkout button is a decaying asset. We think that is a fair trade. We do not think it is a free one.