Elon Musk didn’t build up SpaceX and Tesla by tinkering in his garage. He grew them by investing large sums of cash into each:
The conventional wisdom of the time said to take a deep breath and wait for the next big thing to arrive in due course. Musk rejected that logic by throwing $100 million into SpaceX, $70 million into Tesla, and $10 million into SolarCity
All three companies compete in capital intensive industries, which is part of the reason there was historically so little innovation in these industries.
Where did he get hundreds of millions? From selling PayPal. How did he build PayPal (then called X.com)? He built it partly by investing millions he made from from Zip2:
He actually plowed the majority of the money he made from Zip2 into X.com. There were practical reasons for this decision. Investors catch a break under the tax law if they roll a windfall into a new venture within a couple of months. But even by Silicon Valley’s high-risk standards, it was shocking to put so much of one’s newfound wealth into something as iffy as an online bank. All told, Musk invested about $12 million into X.com, leaving him, after taxes, with $4 million or so for personal use.
How did he build Zip2? He built it by investing a lot of time and some money to building the software, funded partly by his father:
Errol Musk gave his sons $28,000 to help them through this period, but they were more or less broke after getting the office space, licensing software, and buying some equipment.
To summarize:
Compaq bought Zip2 in 1999 for $307 million. Musk made $22 million from the deal and poured almost all of it into his next venture, a start-up that would morph into PayPal. As the largest shareholder in PayPal, Musk became fantastically well-to-do when eBay acquired the company for $1.5 billion in 2002… The conventional wisdom of the time said to take a deep breath and wait for the next big thing to arrive in due course. Musk rejected that logic by throwing $100 million into SpaceX, $70 million into Tesla, and $10 million into SolarCity
Similarly, Steve Jobs invested $12 million he made from Apple to found Next and another $5 million to buy Pixar. He invested more in them over time.
These cases of large concentrated bets are obvious, but large un-diversified risks are unavoidable, especially among successful entrepreneurs. From Zero to One:
The power law is not just important to investors; rather, it’s important to everybody because everybody is an investor. An entrepreneur makes a major investment just by spending her time working on a startup. Therefore every entrepreneur must think about whether her company is going to succeed and become valuable. Every individual is unavoidably an investor, too. When you choose a career, you act on your belief that the kind of work you do will be valuable decades from now.
The easiest way to amass great wealth is through compounded returns and some kind of competence. You could also be lucky, but luck doesn’t persist over time (otherwise it would be competence). So multiple games will ensure you revert to the mean.
Consider a coin flipping where you start with some balance, have some probability to double your money. You wager a fixed percentage of your wealth over every year
def play(
start: float,
prob_win: float,
wager: float,
num_rounds: int,
num_games: int,
seed: int = 42
):
random.seed(seed)
res = []
for _ in range(num_games):
cur_balance = start
for _ in range(num_rounds):
if cur_balance <= 0:
break
bet = min(cur_balance, max(1, wager * cur_balance))
results = bet if random.random() < prob_win else -bet
cur_balance += results
res.append(cur_balance)
print(f"Max {int(max(res)):,}")
print(f"Num wins: {len([r for r in res if r > 100]) / num_games}")
If you have no skill (50% chance of winning) you could get lucky but over time you’ll lose, especially if you’re making large bets. Here you bet 50% of your fortune with 50% odds of doubling your money:
>>> play(start=100, prob_win=0.5, wager=0.5, num_rounds=100, num_games=10000)
Max 65,821 (658%)
Num wins: 0.0045
Out of 10,000 people, only 0.45% made money and the luckiest made a 658% return over 100 rounds
But if you have competence (e.g. 60% chance of winning), some will do incredibly well:
>>> play(start=100, prob_win=0.6, wager=0.5, num_rounds=100, num_games=10000)
Max 3,886,686,899 (38,866,868%)
Num wins: 0.2351
Here 23% of people made money, and the best return was 38 million percent. But note, that majority still lost (76.49%). That’s because the bet size was too big given their probability of winning. The optimal bet size for a 10% edge and 100% payoff is 20%, but obviously the trade off is felt at the upper end with a considerably smaller max return:
>>> play(start=100, prob_win=0.6, wager=0.2, num_rounds=100, num_games=10000)
Max 737,947 (7379%)
Num wins: 0.8249
And this is the pattern you see with most successful entrepreneurs: concentrated repeated bets with odds in their favor.
Conspicuous consumption is still a relatively small amount of what successful people spend their money on. Even if you look at lavish purchases like Larry Ellison buying an island in Hawaii, the sums are relatively small: $300 million for someone worth $44 billion at the time. Extreme wealth is almost predicated on compound returns, and compounded returns only work when you invest capital in concentrated plays. No one becomes a billionaire through diversification.