“The PayPal Mafia” is rightfully famous for producing a slew of valuable companies — SpaceX ($33 billion), Palantir ($26 billion), Tesla ($42 billion market cap plus or minus a few billion dollars depending on the most recent press releases and late-night CEO tweets), LinkedIn (acquired for $26 billion), YouTube (acquired for $1.65 billion and worth as much as 100 times that).
Collectively, these companies are worth over 100 times PayPal’s $1.5 billion acquisition price.
Why aren’t there other similarly successful mafias? There are plenty of former Google and Facebook employees who have gone on to start successful companies, but the network isn’t as dense. In fact, most PayPal-sized exits don’t produce anything remotely like the PayPal Mafia. What’s going on?
The best way to look at this is to consider the closest analogy to the PayPal Mafia: the Tiger Cubs. Coatue, Tiger Global, Lone Pine, Maverick, Samlyn, Steadfast, and Viking — once again, the companies founded by former Tiger employees manage vastly more money than Tiger did at its peak.
Tiger’s alumni network isn’t quite as distinctive as PayPal’s; in finance, there’s a longer tradition of spinoffs, perhaps because it’s easier to start a company that competes with your former employer without competing head-to-head. The Goldman Sachs arbitrage desk produced several successful funds over the years (Perry Capital, ESL, Och-Ziff, and TPG-Axon), and Commodities Corporation seems to have produced a disproportionate share of the 1980s and early ’90s best macro traders — Paul Tudor Jones, Louis Bacon, and Bruce Kovner¹.
Hedge funds naturally produce spinoffs because of their structurally high turnover. The two things most hedge fund managers can’t stand for long are portfolio managers who lose money and portfolio managers who make more money for the fund than the fund manager does.
If you want your organization to produce a really effective alumni network, you absolutely need to shut it down at the wrong time.
Still, in terms of pure numbers, it’s striking what a huge fraction of PayPal’s and Tiger’s alumni ended up running companies. Some of the parallels between them are obvious, but a few are surprising. The ingredients to a business mafia seem to be these:
- Elitist recruiting with an unusual skew
- Relatively low headcount and long working hours
- An organizational unwind that leaves most employees with enough money to start a company but not enough to retire
- An engineered soft landing from the founder
Recruiting the best from a specific pool
One simple explanation for why PayPal and Tiger had good alumni is that they had good employees. And why was that? Both of them recruited very specific kinds of people.
In PayPal’s case, recruiting started in concentric circles from the founders’ alma maters: Tech meant people Max Levchin knew at the University of Illinois at Urbana-Champaign, and the business side was populated by people Peter Thiel knew at Stanford.
Tiger, like plenty of other hedge funds, recruited from elite schools, but they tended to prefer hiring from the best schools in the South to hiring from the best schools in the country. Four years at the University of Virginia plus two years at Morgan Stanley seems like the canonical background.
Finance has always had an elitist tinge to it — being a member of the upper crust doesn’t just mean you’re richer than average; it also means you can draw on your family’s reputational balance sheet. By deliberately hiring from a slightly different pool, Tiger was able to select employees who (1) had something in common and (2) felt that this job was their lucky break.
Both companies also lucked out in terms of who they hired and when. PayPal raised money right at the peak of the dot-com bubble and was basically the only company other than Google heavily competing for talent after the bust. Tiger kept its focus on dowdy industrial stocks throughout the late ’90s, so if you were the kind of person who thought “dot-com” was a four-letter word, you had a home there.
This doesn’t imply that every single Tiger analyst hailed from below the Mason-Dixon Line or that everyone at PayPal came from one of two schools. But it doesn’t take a very big critical mass to establish cultural norms and to establish those norms as self-reinforcing.
When you hit Dunbar’s number, exit
That number — 200 — is an interesting number. Anthropologist Robin Dunbar noticed something about monkeys a few decades ago: A correlation existed between relative neocortex size and the size of the monkeys’ social groups. Monkeys with small neocortexes have small troops; the ones with bigger neocortexes have bigger groups. Plot out the points on a graph, and you could estimate that human social groups should be somewhere between 100 and 250 members.
If Dunbar’s theory is true, most people will fill out their Dunbar quota with friends, family, and a few coworkers.
But consider a peculiar set of people: They’re young, they’re likely to have moved to a new city, and they spend 80+ hours per week at work. They haven’t had time to accumulate non-work adult friends, they don’t have time to maintain non-work friendships, and many of them got hired through close friends from school.
At 200 employees, you don’t necessarily know everyone, but you almost certainly know someone who knows everyone. That makes the social network as dense as it possibly can be — hiring and fundraising happen lickety-split when every intro is, at most, two hops on the graph.
Of course, a generic network doesn’t do any good. A long list of people who have nothing in common is not that useful. Fortunately for both alumni networks, what everyone had in common was that they worked on solving basically the same set of problems. In Tiger’s case, that was the two-sided problem of (1) making trades that were contrarian enough to be profitable but (2) presenting yourself to investors as someone conventional enough that they wouldn’t look stupid for cutting you a check. In PayPal’s case, the company also had two key problems to solve: using network effects to get a lot of users (a la Yelp, LinkedIn, and YouTube) and winning a technical arms race against sophisticated fraud networks. I wouldn’t say that SpaceX navigating the military’s procurement system is exactly like outsmarting Russian organized crime, but there may be some commonalities.
Choose the wrong time to get hit by a bus
If you want your organization to produce a really effective alumni network, you absolutely need to shut it down at the wrong time.
PayPal sold to eBay in 2002, when equity valuations were compressed and internet valuations were especially bad. PYPL was spun off in 2015 and is now worth $120 billion, so 98% of the company’s wealth creation happened after the sale.
There’s a reason for this. Before the sale, PayPal had basically two businesses it processed transactions for:
- Gambling sites
That second one, eBay, was big and growing but risky. It’s received wisdom now that you can’t get rich building on someone else’s platform, and it was certainly something people were aware of at the time. The other business was PayPal’s BATNA: It could produce revenue, and eBay couldn’t shut it down, but it wasn’t exactly a stable foundation for the business.
According to this Quora post, PayPal’s team realized that eBay’s business was slowing down, which would have crushed PayPal’s multiple. Meanwhile, eBay was trying to get merchants to switch from PayPal to eBay’s own offering. On a 10-year scale, it wasn’t the right time to sell out, but a company without 10 years of runway doesn’t think in 10-year terms.
So they sold.
The Tiger story is also a story of mistiming. Tiger had a stellar record in the 1980s and early ’90s but faced some missteps in the late ’90s. They made a few large, ill-timed macro bets and avoided the internet bubble. In modern terms, you could say they were long value and short momentum and had bad factor timing, but in more prosaic terms, their performance just wasn’t up to its previous standard.
Eventually, Julian Robertson decided the insane internet values weren’t showing any signs of getting saner, so he shut the fund down in March of 2000. Astute market historians may recall that March of 2000 was literally the peak of the bubble. (This was also when PayPal raised what turned out to be a lifesaving $100 million round. Timing, then, is hard but hardly impossible.)
Importantly, these weren’t bad exits; the companies didn’t crash and burn. Plenty of crazy internet companies completely fell apart in the early 2000s, but since early employees were wiped out, they weren’t interested in risking everything in a new venture. If their bosses cashed in early, that was actually worse. While there’s income inequality between a founder who makes tens of millions and a junior engineer who makes tens of thousands, the income inequality between any number and a net worth of zero is infinite. Famous hedge fund failures also haven’t produced robust alumni networks. Long-Term Capital Management alumni have worked on some interesting stuff (including, hilariously, a relative-value fixed-income fund that did well until there was a liquidity crunch, at which point it blew up; it’s straight out of Pet Sematary), but nothing as big as LTCM. Likewise for Amaranth².
They also exited at a time when they hadn’t won. PayPal could plausibly be described as a story of lucky timing — starting in 1998, raising a lot of money at the peak, and being a thorn in the side of eBay, which happened to be the rare dot-com whose business model worked through the bubble and bust. Tiger Global also looked questionable. Sure, they had good returns in the ’80s and ’90s, when stocks went from cheaply priced to normally priced, but in an era where growth and momentum matter most, value might be obsolete.
It sets up an interesting counterfactual: In 2006, Yahoo offered to buy Facebook for $1 billion at a time when the company had fewer than 10 million users and was far behind MySpace. Facebook didn’t sell. (Allegedly, the discussion went like this: A board member told Mark Zuckerberg to think of all the things he could do with that kind of money, and he said that with that much money he’d probably start a social networking site and that he liked the one he had.) If Facebook had exited then, Yahoo probably would have lost to MySpace, Orkut, or some random new company. And there would be a Facebook Mafia of early employees all ready to build something new.
Both PayPal and Tiger exited with around 200 employees, most of whom were young and rich enough to start their own companies but not rich enough to retire. They clearly weren’t losers, but they arguably weren’t yet winners either. That’s a good combination: the means, motive, and opportunity to come back for round two.
It gets better.
Engineering a soft landing
Robertson shut down the fund, but he kept his lease, he still had over a billion dollars, and he kept his analysts close. Several funds started in Tiger’s offices, seeded by Robertson. To this day, that office building is stuffed with Tiger Cubs — call a company a “101 Park stock,” and people know what you mean. (The office also has some decorative choices; some of the conference rooms have little tiger pawprint patterns on the windows.³)
The PayPal team went similarly: Early employees with small payouts started new companies, and early employees with bigger payouts staked them. This later got formalized through Founders Fund, and other key PayPal employees joined Sequoia and Greylock.
This combination — lots of people with enough money to experiment and a few people with enough money to retire but no interest whatsoever in doing so — kept the network tight for longer. It didn’t hurt that PayPal alumni were starting internet companies after that category had vaporized a few trillion dollars of shareholder wealth and that Tiger Cubs were forming hedge funds at a time when investors were still licking their wounds.
Essentially, these exits engineered the existence of a venture capital fund and a hedge fund seeder with an unmatched deal flow at an opportune time. And since both of those businesses rely heavily on maintaining personal networks, these businesses happened to have the social center of gravity that kept the network alive.
Making a mafia
Whenever people talk about unusual success, there’s a tedious debate over how much of it is skill and how much is luck. With a small sample size, you can’t definitely argue for either one, but you can argue persuasively against either.
What both PayPal and Tiger show is that the key to an inordinately successful alumni network is both. You need the skill to assemble a good team and to grow it to the right size. And then you need the bad luck to quit too early.
- A cynic might say that these are “famous” portfolio managers because they got profiled in Market Wizards, a book written by one of their former colleagues. But they also number among the best fund managers of all time.
- Incidentally, the single funniest thing Jim Cramer has ever written is his article making fun of Julian Robertson and George Soros for retiring too late. Both of them made most of their money after that article came out. In the meantime, Cramer has been on TV a lot and operated a Mexican restaurant. Slamming someone as a has-been is a reputational naked short position, which may be unwise.
- Enron is a really interesting case study, on which I don’t have enough data. Anecdotally, alumni from Enron’s energy trading business did spectacularly well at several funds, with Amaranth bête noire John Arnold as the most notable example. And Enron’s pipeline assets were the foundation for Kinder Morgan. Both of those were good businesses, weighed down by Enron’s lethal combination of financial engineering and aggressive accounting.