PayPal: A Digital Money Marketplace - [Business Breakdowns, EP.113]

Summary

• The guest calls PayPal a “chameleon” whose strategic strengths are neutrality, scale, and direct relationships with merchants and consumers globally — in the most valuable cases, on both sides of the network. He compares it to American Express, though PayPal does far more. Eighty-three percent of the largest 475 digital merchants accept PayPal versus 48% for Apple Pay, and PayPal works across platforms — for example, on both an iPhone and a Windows PC.
• Engagement is the central thesis: transactions per active account rose from 21 at the spin to 51.4, and the guest argues an incremental transaction from an existing user has 1.5 to 2 times the value of a new-user transaction because it carries no acquisition cost. His biggest criticism is that during the pandemic management prioritized network scale and a “super app” over engagement.
• The structural bear case is take-rate compression: the blended rate fell from roughly 2.80% at the split to about 1.88%-1.90%. The guest says every payments platform he has studied has seen take rates fall over time, so PayPal must make that up with volume, new use cases, and engagement. Competition at physical point of sale, leadership uncertainty, and value-destructive M&A are additional risks.
• Braintree is the one resounding M&A success: about $400B of TPV, roughly 30% of PayPal’s $1.4T, and growing at 40%-plus. The guest says it has been taking share through higher approval rates and lower instances of fraud, based on his merchant conversations. His open strategic question is whether Braintree should be separated to gain more perceived neutrality against Stripe and Adyen, which are each around $800B of TPV.
• Capital allocation has been mixed: more than $13B went to acquisitions and partnerships, while the guest says most growth and value came from internal innovation. BNPL required R&D rather than capital, TIO Networks was written down 100%, and Honey’s $4B price remains unproven. Against that, PayPal generated $7B from selling its credit book into a Synchrony partnership and spent $16B on buybacks; essentially all of this year’s $5B of free cash flow was expected to go to repurchases.
• Elliott Management’s involvement was followed by cost cuts, a public commitment to no large M&A, and — as the guest’s speculation — Schulman’s retirement. With Gabrielle Rabinovitch serving as interim CFO for nearly a year and no clear internal successor, his purely speculative “stalking-horse bet” was Cameron Zacky, the former PayPal executive who later helped build Adyen’s U.S. business and whose retirement Adyen announced the day before PayPal announced Schulman’s.
• The forward algorithm is revenue growth at or above e-commerce growth, mid- to upper-single-digit revenue growth near term, 18% bottom-line growth this year, and a return to high-teens margins. Optionality includes online/in-store convergence, possible access to Apple’s NFC under European antitrust pressure, monetizing Venmo’s 90M active accounts, and the $40B of customer balances supporting 4% savings accounts through Synchrony.
• The guest’s personal scorecard is candid: he was quoted in The Wall Street Journal after the 2016 Visa deal defending PayPal and was right, but says he mishandled his own investment during the COVID-era rise. “I wanted to get a double within five to ten years. I’m there, but the path has sucked.”

Deep dive

1. Two companies, a boardroom coup, and an accidental killer app

• The guest’s history: PayPal began as two firms — Peter Thiel’s Confinity and Elon Musk’s X.com — merged by their venture-capital backers in 2000, with no exact product vision. Musk wanted a “supermarket of banks” for the internet; Thiel’s crew had a PalmPilot experiment for sending money to friends that turned out to be valuable.
• Musk led the combined company and did not view PayPal as the core product, while Thiel’s team pushed it forward. A Sequoia-backed boardroom coup removed Musk as CEO and reinstalled Thiel, after which the company committed more fully to the payments use case.
• The eBay use case was “somewhat by accident”: users adopted PayPal to solve the marketplace’s biggest bottleneck — trust in payments between strangers. eBay repeatedly tried to disintermediate PayPal with its own Wells Fargo-partnered offering, but PayPal kept competition at bay. Max Levchin’s fraud-detection algorithms were a critical early differentiator in “the wild west of the internet.”
• The growth hack the guest still marvels at: $20 to open an account and $20 per referred friend — $60M-$70M of customer-acquisition spending by Musk’s estimate, mostly in 2000 and 2001, after the dot-com peak. People were making a couple thousand dollars by getting their peer groups to sign up. PayPal took the name in 2001, IPO’d in February 2002, and eBay offered to buy it for $1.5B by summer 2002 — the guest thinks July.

2. Inside eBay, PayPal was built upside down

• The guest’s signature framing: “PayPal existed upside down within eBay.” The corporate imperative was driving marketplace conversion, not building the world’s best payment app. eBay steered users to ACH for better transaction economics rather than giving them funding-source choice, and PayPal accumulated roughly ten years of tech debt while building marketplace features instead of payment features.
• The eBay era was not all downside: PayPal crossed 100M users inside eBay, and the Braintree acquisition, in 2012 or 2013 by the guest’s recollection, brought a mobile capability to a desktop-geared business and “a great leader” in Bill Ready.
• Carl Icahn took a stake in 2014 and agitated for separation. PayPal was growing while eBay’s growth had stalled, and the businesses were subsequently split, with PayPal becoming public again.

3. The scale today: metrics versus the split

• The numbers the guest cites: 435M users — 400M consumers and 35M merchants — plus 190M monthly active users. TPV reached $1.4T versus $288B at the split; revenue was nearly $28B versus $9B, and free cash flow was $5B versus $1.8B.
• Engagement rose from 21 transactions per active account at the time of the split to 51.4 by the end of the prior year. eBay fell from 17% of transaction volume and 40% of profits at the split to about 2% of each, or less for profits.
• The guest’s favorite statistics include roughly $180B of cross-border volume, which has more favorable economics for PayPal, and Braintree at about $400B of TPV — roughly 30% of the total and next to nothing at the split.
• Customer balances stored on PayPal rose from $12B at the split to $40B. The guest views that balance as evidence of friction in the system and as an opportunity to become more of a digital bank. Venmo has 90M active accounts, roughly the scale PayPal had not long before the split.

4. A chameleon with scale, neutrality, and network effects

• Where PayPal fits in payments: “they’re what I call a chameleon of sorts.” It participates in merchant acquiring, card issuing, credit, BNPL, remittances, cross-border payments, and banking through its Luxembourg bank. “It’s really hard to say what they are and where they fit,” which the guest sees as one of the problems investors have with the company.
• Its distinctive feature is a direct relationship with merchants and consumers globally. Unlike Visa and Mastercard, which are generally intermediated by issuers and merchant acquirers on both sides, PayPal is both in some cases — “not in many cases, but in the most valuable cases.” The guest compares that aspect with American Express.
• Neutrality is strategically valuable against Apple, Google, and Shopify. PayPal can work across an iPhone and a Windows PC, whereas Apple Pay is not equally usable across those platforms. The guest says neutrality, scale on both sides of the network, and the ability to act as a chameleon succinctly describe PayPal’s role.
• A share-loss correction: PayPal is not losing desktop or mobile e-commerce checkout share to digital wallets in the way the narrative often suggests. Apple Pay and Google Pay are expanding especially at physical point of sale, where PayPal has little presence, while the overall pie is growing.

5. Anatomy of a $100 transaction

• The blended take rate is almost 1.9%, including peer-to-peer transactions, but it is skewed by transaction type. Branded checkout is listed at 3.49% plus a fixed fee; in the U.S. the fee is 49¢, though large merchants may pay closer to 2.75%.
• Funding source determines how much PayPal keeps. If the customer uses stored PayPal funds, the guest says PayPal keeps the full fee. ACH creates roughly a 5¢ network charge. If the customer uses a Chase Visa card, roughly one-third of the economics may go to the issuer, with Visa taking a smaller share.
• Braintree economics are structurally thinner: 2.59% plus 49¢ when it provides the full stack — gateway and merchant acquiring — but only about 10¢ when it is gateway-only. In the guest’s Uber example, where Chase handles merchant acquiring, Braintree keeps only a dime of a roughly 2.6% fee while the remaining economics go to the acquirer and other parties.

6. Fraud detection is a compounding, hard-to-copy advantage

• The guest’s personal conversion story came after his card was compromised in the Home Depot hack, producing random charges at CVS stores in Rhode Island. He used PayPal more because the merchant does not receive the underlying card number: PayPal had “tokenization before it was called tokenization,” reducing the card’s points of exposure.
• On the merchant side, PayPal can better identify potentially fraudulent transactions, especially when it is the acquirer. Merchants bear the cost of fraud, and better detection can also mean fewer legitimate transactions are rejected.
• The guest says PayPal has the highest approval rates among processors. In his merchant conversations, higher approval rates and lower instances of fraud are key reasons Braintree has been taking share from Adyen and Stripe.
• Fraud mastery supports one-touch checkout across the open internet, passwordless login, and BNPL offered at traditional acquiring costs. The guest says each additional year of fraud-detection experience can put PayPal a year or more ahead of new entrants.

7. Engagement is the thesis — and management only recently “got religion”

• The guest’s core economic point: an incremental transaction from an existing user is worth 1.5 to 2 times a new-user transaction because there is no customer-acquisition cost against it. Engagement strengthens the network, makes PayPal more valuable to merchants, and supports the moat.
• His frustration is that during the pandemic PayPal prioritized expanding network scale rather than driving engagement. He argues the company should focus on how many people transact and how frequently each person transacts.
• The subscription opportunity is substantial: the average user transacts about 50 times a year, while the average person has three to five subscriptions. Four subscriptions on PayPal could account for a large portion of current engagement. The guest’s example is The New York Times: PayPal lets users cancel through Manage Your Wallet rather than calling and navigating a retention process.
• Rewards can also drive use. PayPal lets customers use card rewards as a more liquid payment instrument, including at merchants that do not directly accept those rewards. Honey built a rewards ecosystem on PayPal, though the guest thinks PayPal probably overpaid and took too long to develop the capabilities.
• The $40B of customer balances has an average duration of nine to twelve months and can be invested in short-term Treasuries. That interest is cash flow to the business even though it does not appear in EBITDA. PayPal has also begun offering 4% savings accounts through Synchrony, while the guest says Chase and Citi still pay him nearly zero.
• Venmo provides another engagement option. The guest’s example is paying for Chipotle with Venmo and sharing that activity with friends, creating a social overlay that could benefit both users and merchants.

8. Braintree: the one acquisition that worked — maybe too well to keep

• Braintree “has been a resounding success in a way that nothing they’ve done since has come close.” It moved PayPal from desktop to mobile, with Uber and Airbnb among the early customers that established it, and brought Bill Ready into the company.
• Braintree’s technology helped address PayPal’s monolithic, eBay-era tech debt. Using its capabilities, PayPal could deploy new products much faster, including rolling BNPL out across the network within a couple of weeks and launching crypto capabilities rapidly.
• Braintree also brought Venmo, which it had acquired a year before PayPal acquired Braintree. Venmo was originally part of Bill Ready’s vision of giving merchants access to a customer side, although the merchant-services opportunity has taken longer than hoped.
• The competitive scoreboard is about $400B of Braintree TPV versus roughly $800B each for Stripe and Adyen. Braintree is growing at 40%-plus; the guest says that is faster than Adyen on his comparison, whose reported growth benefited from euro translation, and says Braintree “has almost doubled off Stripe’s in the last year” without specifying the exact comparison.
• The company is reluctant to emphasize Braintree because its economics are weaker than those of branded checkout, but the guest thinks it should not shy away from the business.
• His open question is whether Braintree should be separated from PayPal. A standalone Braintree might appear more neutral to merchants wary of PayPal, while the company’s counterargument is that Braintree merchants can receive early access to Venmo. The guest thinks many of the two-sided benefits might survive a separation.

9. More than $13B of M&A, with mixed results; $16B of buybacks

• The guest’s verdict on capital allocation: PayPal spent more than $13B on acquisitions and partnerships, yet most growth and value came from internal innovation. BNPL required R&D and initiative rather than a major capital allocation.
• TIO Networks was written down 100% after PayPal found serious fraud-detection weaknesses. Xoom added nominal revenue relative to its $800M purchase price. Honey cost $4B and remains unproven. Hyperwallet may be one of the more valuable acquisitions because it enabled payouts in three-sided networks such as Uber’s payments to drivers.
• Paidy was PayPal’s second-largest acquisition and provided access to Japan, which the guest describes as the least digitally penetrated of the developed economies, along with the country’s largest installed base among such services.
• The Mercado Libre irony the guest laughs about: eBay took a similarly sized stake at the split and later sold it, while PayPal bought a stake at higher prices. A partnership expected within a year had still not appeared roughly five years later, in the guest’s view. He would sell the Uber and Mercado Libre stakes, worth nearly $2B.
• PayPal also received $7B from selling its consumer credit book into a capital-light Synchrony partnership. Synchrony supplies the balance sheet while PayPal receives a revenue share. The company spent $16B on repurchases, demonstrating how little capital the business needs to grow and generate cash.

10. Elliott, the super-app detour, and the succession question

• The super-app era is the guest’s sharpest criticism. PayPal committed to an expense base based on expected revenue growth near 20% just as e-commerce growth stalled, causing margin deleverage. His contrast with Amazon is that excess fulfillment capacity may eventually be used, while much of PayPal’s super-app spending may not have comparable value.
• His running joke was that he should drink every time Dan Schulman said “super app” at PayPal’s early-2021 investor day.
• After Elliott Management’s involvement became public, PayPal committed to expense reductions and disciplined capital allocation, including a public commitment to no large M&A. The guest estimated layoffs at a high-single-digit percentage of employees and speculated that Schulman’s retirement could be directly traced to Elliott’s involvement.
• John Rainey had left for Walmart, a hired CFO successor never really took over, and Gabrielle Rabinovitch had served as interim CFO for nearly a year. The guest described leadership at the top as unclear.
• His “stalking-horse bet” for CEO was Cameron Zacky, a former PayPal executive who spent about a decade there before helping build Adyen’s U.S. business. Adyen announced that Zacky would step down and retire the day before PayPal announced Schulman’s retirement. The guest emphasized that this was purely speculation.

11. Growth algorithm, risks, and the lessons of a bumpy double

• The expected algorithm is revenue growth at or above e-commerce growth. Near term, PayPal expected mid- to upper-single-digit revenue growth and 18% bottom-line growth, with at least one point of margin improvement. The guest sees a path from mid-teens margins back to the high teens and believes longer-term incremental margins could be in the mid-20s to 30% range.
• Upside options include the convergence of online and in-store transactions, possible access to Apple’s NFC capability under European antitrust pressure, and converting Venmo payments such as babysitting and rent into more commercial transactions.
• The guest worries about take-rate compression because every payments platform he has studied has seen rates fall over time. If PayPal cannot offset that through volume, engagement, and new use cases, margins and profitable growth will suffer.
• Other risks are competition at physical point of sale; Stripe’s ease of use for long-tail merchants; Adyen’s global capabilities for multinational merchants; leadership churn; a board with “no real payments people”; and value-destructive M&A. The guest believes PayPal does not need to become a super app and could instead focus on becoming a central online bank.
• His closing self-assessment is that being a generalist helped him evaluate the 2016 Visa deal without a technology- or payments-specialist bias. He also says he mishandled his own investment during the COVID-era rise. The lesson is to stay aligned with the company’s direction and focus on “engagement, engagement, engagement.” His honest scoreboard: “I wanted to get a double within five to ten years. I’m there, but the path has sucked.”