Comment by georgewsinger | Hacker News Reader

Comment by georgewsinger | Hacker News Reader

Studying "1-in-1,000 fastest growing new ventures" is completely misunderstanding tech. Startup returns are about power-laws and extraordinary outliers. If we instead study the age of founders of tech companies which actually generated almost all of the ecosystem gains over the past few decades:

1. Apple: $2.65T

  - Jobs: 21

  - Woz: 26

2. Alphabet: $1.73T

  - Larry/Sergey: 25ish

3. Microsoft: $2.22T

  - Gates: 19

  - Allen: 22

4. Amazon: $1.45T

  - Bezos: 29

5. Facebook: $0.84T

  - Zuckerberg: 19

Looking at just the age of founders instead of other substance is stupid. But if we're just looking at age and nothing else, then younger founders make for better outliers, and generating outliers is the whole point of this endeavor.

-----

EDIT: With its high market cap, Tesla belongs on this list, and actually had some older founders. Problem: the older founders left the company/were forcibly removed, didn't retain much equity, and aren't credited with much early success when they were actually in charge. Still, some interesting data points.

Tesla: $0.94T

  - Elon Musk: 32

  - J.B. Straubel: 28

  - Martin Eberhard: 43 (forcibly removed due to late/over budget Roadsters)

  - Marc Tappening: 39 (left Tesla in 2008)

No - it's statistically unsound to focus on extreme outliers, even if they account for most of the economy. Suppose a single company accounted for 99% of the economy and was founded by an 18-year-old. Would that mean that we should expect 18-year-olds to be the best entrepreneurs in general? Of course not. You can't base broad conclusions on a sample size of one, even if that one is disproportionately influential.

Look at the top 20, 50, and 100 to get a better idea of whether the top 5 observation is robust.

This is more of a problem with statistics than with focusing on outliers.

Per your example, if something has 0 explanation for whatever has 99% of influence, it is just not a useful explanatory mechanism. That doesn't mean there isn't a good explanation.

Statistics exists because people who think they have the explanation for a single important data point are very often wrong.

If you could recognize the causal networks behind a complex system the way a toddler can recognize a cat, after being shown very few examples, there would be no need for statistics.

Unfortunately, you can't. (But it doesn't stop lucky guessers thinking they can.)

Statistics is just one way to get a correct answer, no? Or it may not even be a way to get a correct answer at all. It seems more of a checking mechanism to filter out answers which are not correct.

A scientific hypothesis is correct or not independently of statistical analysis. It's true before it is verified. Statistics is just a social mechanism for demonstrating its correctness to other people (or yourself.)

It's entirely possible for someone to know they're right, actually be right, and it be uncheckable with statistics. The only difference between an outlier and a lucky guesser is knowledge. Blah blah blah anything that's measured ceases to be a good measure blah blah blah.

It's not statistically unsound to focus on extreme outliers. In fact, if your data set contains lots of outliers that your statistics can't explain, it's more of the case that your statistics are wrong (I'm not saying the math is wrong, I'm saying you're applying the wrong model to your dataset). This is basically the whole lesson behind books like Black Swan.

I agree with the comment you are replying to. You're commenting on a website of the most successful tech incubator of all time where, at least from a quick glance, all of the top companies were founded by people in their 20s. Given the researches intro sentence of "Many observers, and many investors, believe that young people are especially likely to produce the most successful new firms" I think their conclusion is really just using an invalid viewpoint of what "the most successful new firms" means.

The market cap of these companies IMO is possibly misleading.

Much of that gain came after going public, so we are looking at large corporate returns rather than startup returns.

If “startup investing” as opposed to “holding stock of a public company” you might want to measure gains to IPO at most.

This is an astute observation.

Additionally inorganic growth or 'buying other companies' has consolidated many other startups into these giants and it makes any kind of analysis on successful founders ages considerably harder.

Too much analysis of software giants. It's all industry specific.

2/3 Intel founders were about 40. Sam Walton did not own a store until at least age 35. Bell founded AT&T, maybe in historically relative terms the most successful tech company ever, at 37. Jensen founded Nvidia about 30. Warren Buffet bought up Berkshire in his late twenties.

This is a great response, especially RE Intel & Bell Labs.

Jensen and Buffett both count as "young" though, I think? In any event all of these founders are below the age of 45?

True. I'm just countering the tendency to think that pure youth is best. That seemed to peak with the Elizabeth Holmes mess. Had she gained five years of experience and a more grounded plan of attack with the same level of charisma and ambition it may have turned out more favorably.

This is a very optimistic view considering every part of her methodology was and always will be physically impossible without completely changing how we understand biology.

The implausibility was obvious even at the level of basic sciences. When I heard about Theranos, I checked the units on my last test from an established lab. Glucose and cholesterol -- which can be done with finger pricks -- measure in mg/dl. A standard blood test like a total thyroid test measures in ng/dl so that's one million times higher resolution. Free thyroid is in picograms/dl !!

Steve Jobs was 42 when he became Apple CEO the second time. I feel like we need to talk about Apple in terms of Jobs 1.0 and Jobs 2.0.

Right, a lot of Apple fans regularly forget that Apple was nearly bankrupt[1] in the late 1990's until the $150M deal with Microsoft and the return of Steve Jobs. It's almost Apple 1.0 and Apple 2.0. They are the same company pretty much in name only.

[1] https://www.investopedia.com/articles/personal-finance/05111...

> Tesla …

For what it’s worth, Elon Musk is not Tesla’s co-founder. He came in as a late investor and I believe one of the terms was that he’ll get the co-founder tag.

Founded in July 2003 by Martin Eberhard and Marc Tarpenning as Tesla Motors, the company's name is a tribute to inventor and electrical engineer Nikola Tesla. In February 2004, via a US$6.5 million investment, X.com co-founder Elon Musk became the largest shareholder of the company and its chairman. He has served as CEO since 2008. [0]

[0] https://en.m.wikipedia.org/wiki/Tesla,_Inc.

Reed Hastings founded Netflix when he was 37

It would be as stupid to base an investment strategy purely on a young age (as your text is implying). Good luck or good bye to your retirement funds this way..

I'd say identifying an underlying trend early (all examples you gave) and then picking the most promising candidate from companies building on that trend, is the key.

But still, there's also management quality, founders that grow as management personalities together with the business, and sheer luck.. point being, uber-successful investing is more an art than science. And surely, not just a young age..

I agree, and definitely don't endorse investing purely off of founder age (I even stated that in my original post).

The math on Apple, AWS, Microsoft, and Facebook are not that easy. They are now multi-LOB of companies, with significant revenue coming in from BU's created and run by people who aren't the original founders.

Ex: AWS appears to be Andy Jassy, who was probably 30 or 40. Similar and probably even "older" stories for GCP (Google) & Azure (MS). Their supporting "co-founders" may be even older, like Werner Vogels (63).

But there's plenty of other companies that are great investments that aren't FAANG. Your change of investing in one of these is almost zero and there's so few samples we can't tell the difference between noise and signal.

You misunderstand how tech works.

- Normally distributed endeavors: look at averages, ignore outliers.

- Power law distributed endeavors: ignore averages, focus only on outliers.

In tech, a few companies generate all of the returns. And it's fractal all the way down. E.g. there have recently been 900 U.S. unicorns minted in the past few years (IIRC), accounting for roughly $1T in total market cap. Yet 4 out of the 5 companies on this list are worth more than all of these new unicorns combined.

If you take the attitude "I don't have any chance of investing in one of these outliers", you'll only invest in "safe" things and go broke over time.

I think the point is that you still basically don't have a chance of investing in one of these outliers, so instead of going broke over time because you're playing the wrong game, you go broke over time because you just lose.

> If you take the attitude "I don't have any chance of investing in one of these outliers", you'll only invest in "safe" things and go broke over time.

This is an absurd conclusion from my comment. We still have unicorns and companies that grow 100x and 1000x without being FAANG. Those are great investments. I'd argue that they aren't safe either. I think you're finding a conclusion by looking at the answer. There's also only 4 companies there out of hundreds of thousands. You can't find the signal in the noise. As many others have pointed out here, companies you list also have another common factor: the internet was new. This offsets the industry experience factor that the paper discussed because there was no industry to speak of. This gives younger people an advantage. But there's not really something like that right now.

lol I love the "safe" things cause you to "go broke"

Don't invest in vtsax, you'll lose all your money.

Ask Warren Buffet - he’s never invested in safe things. /s

Though seriously, looking at Buffet - he invested in incredibly safe companies (ones generally valued at below their book value, at least for the first few decades) and used a bunch of leverage on them to make top-10 global wealth.

It bugs me (seems like it does for you, too) to hear the “you need risk for reward”.

Safe !== broke, and risky!== rich. But too many people equate ”safe” and bonds or index funds (which also won’t make you broke, they just won’t generate generational wealth for the most part).

Is your advice to buy stock in all companies founded by young people from 1976 (Apple) to 2004 (Facebook)? I guess not, you need many other factors that may make impossible to recognize these companies.

Is not possible that it gives you better return of investment if you choose less risky companies? It may be better to make good profit in 90% of your investment that incredible gains on 0.00000001 of it.

In the Apple founder list, you forgot Ronald Wayne who was 41 at the time of founding.

Facebook also had several cofounders who left early, but I think they were all Zuckerberg's generation (Sean Parker was not a founder).

Isn't it all about experience - most of those companies were starting from scratch in newish industries where no one had experience. But most companies are doing something where there are other companies doing similar things. For example most successful bio/chem/fin tech companies were formed by people with experience in other companies who understand the tech and market - some younger and many older.

Nowadays if you were looking to fund someones idea for a social media company you might prefer someone who had worked in a senior position in another - perhaps 29 or older rather than 19.

Two comments regarding your data points:

A) As far as I see all these companies were created in a different market, the most recent one from your list created 17 years ago:

1. Apple founded in 1976

2. Google founded in 1998

3. Microsoft founded in 1975

4. Amazon founded in 1994

5. Facebook founded in 2004

6. Tesla founded in 2003

I think it matters to take into consideration when we focus our analysis. The times were then and now. So analysis companies from 70' or early 2000's is different than analysing founded companies in 2020's.

I did not verified yet the dataset attached to the study, but I see that they say

> "Across the 2.7 million founders in the United States between 2007–2014"

so it appears they are studying 2007-2014.

So I have the same skepticism against the article. 2020 is a lot different than 2014. There are some similarities but also a lot of differences in the market, capital, know-how, competition, users expectations and more. What worked back then will not work now.

B) On one side Outliers can be fascinating to study. But for 99% of the population also unuseful in a way. Cannot replicate their pattern individually.

They are in a way a probabilistic "side-effect": they appear when there are enough interactions or data points, but nobody can pin point exactly from which initial data point they will appear. If this would be possible then with time everybody will invest early only companies that will become trillion dollars companies.

>Looking at just the age of founders instead of other substance is stupid.

I think looking at how much money (or) connections they got from their family (or) simply the privilege of having a safety-net as more accurate factor than the age for a successful($) startup.

Remember "Beleaguered Apple"? What about "If I was Steve Jobs, I would liquidate the company and return the money to investors."?

How old was Jobs when he came back to Apple in 98? That's right, 42 years old.

> Looking at just the age of founders instead of other substance is stupid. But if we're just looking at age and nothing else, then younger founders make for better outliers, and generating outliers is the whole point of this endeavor.

they're not really outliers at all anymore if you're forming distributions and considering the mean.

and generating outliers is the whole point of this endeavor

Tbh, I think that's just one definition of "the endeavor" in question. Not every start up needs to be a unicorn.

Also the ones you have picked have been around long enough that

1. Their growth has had surges and slows

2. The tech environment was quite different when they were founded.

"most successful firms" have to mean only FAANG?

well they are, FAAMG tbh

whatever! but that's not the point. What's the right benchmark?

I thought they were MAMA(A?) now (sorry Netflix).

Microsoft Apple Meta Amazon (Alphabet?)

Usually spelled GAFAM.

Why did you leave off Netflix?

Netflix is a lot smaller: $0.18T.

(Founded 1997. Reed was ~36, Marc was ~38.)

How about Uber

Uber founders were in there early 30s, but they had both already sold startups for many millions that they founded in their early 20s.

>Looking at just the age of founders instead of other substance is stupid.

Yeah, you get mostly founders who were young enough to live with their parents until they get enough connections to raise some capital.

Elizabeth Holmes is another example, emphasizing that the younger you begin, the more of a head start you have.

Some young people are just smarter. Zuckerberg is just not one of them.

Some young people have greater integrity too. Holmes is not one of those.

Reply on news.ycombinator.com