Hacker News new | past | comments | ask | show | jobs | submit login
Bubble talk (2015) (samaltman.com)
138 points by Miraries 21 days ago | hide | past | web | favorite | 78 comments



I dont think he won the first bet did he? Airbnb, Palantir, Dropbox, Uber, Spacex, and Pinterest together are less than 200 bn and falling. All still pretty good companies (not you Uber), doing interesting things though.

bet 2 - I have not done the calculation, but just stripe puts it way beyond the 27 bn figgure.

bet 3 - plenty of good companies in that batch. Razorpay was valued at half a billion dollars in last round. Overall seems like he was right. https://www.ycdb.co/batch/w15


Regarding the first bet, VC Fred Wilson talks about something related in item 2 of his "What Happened in the 2010s" post: https://avc.com/2019/12/what-happened-in-the-2010s/

He starts out: "The massive experiment in using capital as a moat to build startups into sustainable businesses has now played out and we can call it a failure for the most part."

That slice of startup-land is always what felt most bubbly to me, WeWork being the biggest example, where even SoftBank admits losses of something like $9 billion. In contrast, Bubble 1.0's biggest loser was WebVan, at $1.2 billion.

(And as a side note to some of the discussion: that there are successes this time around doesn't prove it's not a bubble. E.g., Amazon and Google made it through, but it was still a bubble back then.)


> The massive experiment in using capital as a moat to build startups into sustainable businesses has now played out and we can call it a failure for the most part.

The issue is that the value of money has been decreasing each year for the last decade. I don't remember anyone predicting that this was going to happen as far back as 2010, although it was already obvious that it was happening by 2013 or so.

That's why you'd be hard pressed to name a single VC-funded tech startup that was founded within the last five years, because these days VCs are just funding small businesses and not startups.


Bet 2, 3 are (more or less) bets that VC will continue to overvalue companies in the presence of cheap credit. Bet 1 is what the rest of the market thinks. Altman is a decent VC, so no surprise he'd "win" 2 & 3.


Uber: 50.73 bln Spacex: 33.3 bln? Airbnb: 35 bln? Palantir: 26 bln ? Pinterest: 10.4Bln Dropbox: 7.44 Bln

They are far from $200bln, so he certainly lost the first one. Might have won the other 3 though.


To me, the outcome is just another example of the power of diversification and that we really don’t know anything about which stocks or companies are best. Stripe alone made Bet 2 worth it.


I'm inclined to read it as early opportunities having the highest average growth potential. Big companies are more likely to be already around their actual ceiling.


20% off. I’d say that’s a pretty decent guess in 5 years.


bet 3 - gitlab was in that batch and is alone worth 2.7 bn, so yeah total batch valuation is well north of 3bn.


It checks out if you move the much larger valuations in bet 2 to bet 1 as he really underestimated the companies in bet 2. Things are a bit random, so I'd say he arguably won the bet in spirit even though he didn't do it according to his own exact rules.


Aka he lost the bet :)


Sam Altman's bet isn't interesting in itself, as it looks today the market was significantly undervalued in 2015. Even Sam Altman's pretty aggressive estimations undervalued it, he just guessed wrong on which players would become big. So it definitely wasn't a bubble.


And over 10 years ago, people were joking about bubbles in the context of companies like Facebook. https://www.theguardian.com/technology/blog/2007/dec/19/here...

Whatever is to come with e.g. the decreased interest in the public markets for loss-making companies with no apparent path to profitability, the last 10 years have apparently not been a bubble for any reasonable definition of the term.


Definitely. If we get to wiggle things in retrospect, then we're not talking about the bet itself, but the set of all bets that 5 years later somebody on HN would see as plausible wiggling of the terms to match current conditions. That's surely a lot of bets, among which we'd expect to find several winners.


I suspect the mistake in Bet 1 was betting on Pinterest rather than Instagram or WhatsApp (though IG had been acquired at the time of writing)


Bet 3: GitLab, Atomwise, Razorpay. So yes, I suspect he won that one.


Datapoint for reference: if you invested in an SP500 fund during the same time you would have roughly doubled (around 180%) your money (an annualized return of 11%).


That's the proper comparison.

Investing into startups and unicorns should provide significant yield over diversified and less risky investment. Getting the same result from taking more risk is not a good investment.

from the end of March 30, 2015

+56% SP500 Price index

+70% SP500 Total return index

+83% Nasdaq price index

Uber, Palantir, Airbnb, Dropbox, Pinterest, and SpaceX doubling their valuations would give only +3.2% annualized return over Nasdaq. I don't see the risk/reward ratio of a good investment if the price just doubles in that period.

(FAANG is another good comparison).


How is investing in companies which are already 100 bn in valuations more risky than SP500?

I am sure all investors who put money in these companies when they were risky, made very good returns, far more than SP500.


It's not the size but the way they operate:

* plan is to forego profits to gain market share, negative earnings

* free cash flow is negative

* lots of debt.

Dropbox -40%, Uber -28%, Pinterest -23% since the IPO.


Are you saying that all one needs to do is put money into the risky companies that will succeed and avoid the risky ones that won't?


Regardless of the fact that he missed by a bit, this type of writing is way more valuable than the standard NYT/Bloomberg "OMG IS IT A BUBBLE???" articles?


I would say so. The writer is giving analysis with money and reputation on the line using specific metrics.

It gives us something to measure and review. It also adds/removes credibility when we read another article with his analysis.


>Of course, there could be a macro collapse in 2018 or 2019, which wouldn’t have time to recover by 2020.

Which hasn't happened...

There was another thread about bubble and calling me out. Just to point out, If you bought S&P Index in 2015 you would now have roughly 200% return.

But this is in the context of tech. We forget Apple, Microsoft, Google are not just doing fine, they are doing great. And not just in US, TSMC and Samsung representing SemiConductor industry are all doing great, thanks to those ridiculous profits from NAND and DRAM; Samsung is reinvesting everything into making DRAM NAND and Foundry more competitive. Which will hopefully fence off those State sponsored Chinese Competition.

In terms of Startup, While Sam might have loss on first proposition ( I think he won the other 2 ), but they are operating just fine with similar valuation if not more, just not meeting the 200B target. Unlike other Startup that cashes in from IPO and then devalued into nothing.

Will we see a correction? Very likely. Are we in a bubble? No, at least not yet. ( Unless you like to redefine any correction as bubble burst )


> Are we in a bubble? No, at least not yet.

I agree with everything you said except for this. Not because I am definitively saying we are in a bubble, but just pointing out it's impossible to know if we are. Very few people saw the last financial collapse coming.


https://imgur.com/a/7ntUkKC

That is a simple chart of recent 30 Yrs S&P 500 P/E.

> Very few people saw the last financial collapse coming

While media points the last financial collapse blame to CDOs, the fundamentals of it was housing prices. May be it is new to many Americans, but for those who have lived and seen through Japan, South Korea, Hong Kong housing prices collapses in the 90s, it was rather obvious. I remember the moment when I mention it in 2005 and 2006, but no one believe market would collapse, the idea of housing prices would fall was literally unheard of.

https://imgur.com/a/0jPNE26

Now there are concerns of Cooperate Debt reaches new height, which is often the argument of bubble. The trouble is you cant look at cooperate debt alone and judge it. Amazon is 25B in debt, but they have 45B in cash. Apple are 100B in debt, but also ~250B cash. Amazon were issuing at 10 years and yields about 3.2%, for reference US 30 Yr treasuries has been floating at around 2.4 The vast majority of companies are swimming in cash. ( There is less of it now because of buy backs, which is what pushes those stock prices up )

So while the market, and general economy may not be healthy, I argued it isn't in a bubble either.


> So while the market, and general economy may not be healthy, I argued it isn't in a bubble either.

You're completely missing the point. You can make arguments either way. Hindsight is 20/20. You're making an educated guess based on data which is informative, but it's theoretically impossible to predict a Black Swan event. And even without a Black Swan event, your analysis could be wrong. It's not like all analysts agree on much (if anything) - so how do you decide who's right when nobody has a perfect track record? You're making the best guess you can and I respect your analysis, but I don't assume it's correct.

My point is simple: nobody knows for sure.


I sort of agree with what you said.

>My point is simple: nobody knows for sure.

Absolutely. I am suggesting to those who consistently points to doom's day needs better explanation rather than just debt alone.

Edit: Actually it was my understanding that all bubbles, by definition must have some form or irregularities or speculation. After all that is what bubbles means, turns out when I was going to quote wiki for this reply, I discover there are whole set of economic papers suggesting bubbles completely unrelated to its intrinsic value and economic fundamentals. Will Read those paper when I have time.


The last financial collapse had many companies claiming paper valuations of assets. Companies like AAPL/MS/GOOG/FB have significant net income with significant moats around their products that still have plenty of room to grow as they all move to subscription income.


It's always easy to come up with justifications for why "this time it's different" and people will be doing that even after the stock prices are already declining.

It's impossible to foresee an unknown unknown.


No macro collapse, and the first proposition didn't work out: Uber: 50B, Palantir < 40B, Airbnb < 40B, Dropbox < 10B, Pinterest ~10B, SpaceX < 40B. So total 160-180B market cap. Close, but not quite.


Biggest difference was that the public market valued a bunch of these companies waaay less than VCs did.


So it was a bubble then?


Not sure if it's the right nomenclature, since bubbles are supposed to be pop. Instead they just grew in value slower than they were.


The muffins didn't rise properly.


Perhaps, but dealt with via smaller adjustments rather than catastrophic collapse.

The underlying principles of the companies is sound, just priced a bit too high. The market fixed it (as it is supposed to do)


“Fixed” is too definitive of a word, isn’t it? Who knows if the market is correct right now.

Repriced, adjusted, revaluation. Those seem more appropriate for the timeframe.

More time is needed to tell if it was fixed or not.


Although Sam would have lost this bet, I respect that he had the courage to make a forecast with a specific number and a specific future date.

Too many pundits make grand but imprecise predictions like "AI will become dominant in the next decade". The book 'Superforcasting' tackles this problem in detail: https://www.amazon.com/Superforecasting-Science-Prediction-P...


If I recall correctly, the bet was accepted.

EDIT: Ah, I found documentation:

"Michael de la Maza, a Boston-based investor and TechStars mentor, has taken my recent bet on valuations."

https://twitter.com/sama/status/582292425449730049


Thanks for using Twitter to remind Mr. Altman of the bet.


I'm actually surprised that he didn't win this bet since it wasn't really based on non-bubble metrics in the first place. A more interesting bet to me would have have involved profitability, P/E ratio, or some other long-term indicator that current high-rolling unicorns are actually sustainable businesses. I think WeWork has demonstrated pretty clearly that billions of dollars of VC money can prop up unsustainable companies for more than 5 years (i.e. the market can remain irrational longer than the term of this bet).


His bet was a parlay so it lost, but not by a lot. Had he been able to include Snapchat it probably would have done worse.

I think what we’ve seen in the last five years is that there’s some bubble-like irrationally exuberance but overall it’s far from 2000 all over again. Companies have high valuations but this time also often have financials and growth prospects to go with them.


Matt,

When will portfolio 1 reach $200B?

Michael


I certainly don’t know.


People were projecting that he will lose last year: https://news.ycombinator.com/item?id=18792890


He technically has until March of 2020. I still think there’s a chance that he pulls #1 off (SpaceX pulling off human flights in Q1 and re-raises, Airbnb going public as a profitable and defensible business, Uber and rationalization in ridesharing and delivery + continued execution)


...no? He said it himself, he has until January 1st of 2020. He lost the bet.


So... Did he win? I’m especially curious about part 3 :)


Part 1 didn't quite make it. Eyeballing it I think it's around $150B, which I think still substantially backs his point, but it loses him the bet.

Part 2 was an easy win from Stripe alone.

I scanned the YC '15 Winter list and I don't think I recognize any of them. Someone could track it better via Crunchbase, but I rather doubt that he bet correctly there either.

EDIT: Oh, GitLab.


Part 3: Gitlab alone is almost $3B


Ah, cool. Missed them somehow.


Also Atomwise is in that group.


I just googled the company values, and it looks like he came up a bit short for number 1 but got close, easily beat number 2 with Stripe alone.


Sucks that Sam lost his bet because as far as I'm concerned he "won" his point (I.e. there was clearly not a bubble and all these 1,2,3 companies did amazing). Still, it's great that it's going to a charity


Wonder who took the other side and where they chose for him to donate to.

Edited due to ambiguity.


According to a previous HN thread, Michael de la Maza took the other side.

https://news.ycombinator.com/item?id=18792890


Well, the other side won. So the other side will not need to donate. Albeit that it was close.


Sorry for my poor wording. "He" was referring to Sam.


Seems odd to me to deny bubble talk, then base "success" on market cap. Outsized market caps are associated with bubbles. There are a few companies in there for which survival is questionable.


The idea of the bet was that if there was a bubble in 2015 it would have deflated by now.


Bet 1 should be less than 200bn. Bet 2+3 was spot on and easily won.


Many here are claiming he lost the bet on #1, but consider the details:

Predicted Aggregate value: $200B

Actual value: $145B (Math in [1])

His prediction is off by 25%. Technically a lost bet, but I'd say a prudent bet as a VC.

As for #2, #3, he wins them outright. Overall, being 25% off the most optimistic prediction isn't too bad a loss. I'd say he won the argument convincingly.

I will note that anyone above average intellect in his position (as an investor with gobbles of analytics, best in class prediction models and advisors, and access to thousands of high quality pitches to see down the pipeline) would likely predict as he did, so while he won, he wasn't clairvoyant, he was merely following the data.

[1]: Math: Uber: $50B

SpaceX: $35B.

Airbnb: $35B

Palantir: $25B

Pinterest: $10B

Dropbox: $7.5B


2015, March.


Original title lacked a date -- and comment pointing out the omission gets downvoted. Such downvotes are destructive to the quality of information on HN. @dang?


The market can stay irrational longer than you can stay solvent.


Yeah, a better bet would be about those companies actually delivering profits and "growing into their vaulations", as people politely put it. None of those companies have earned a single dollar for shareholders. Not a single dollar.


> None of those companies have earned a single dollar for shareholders.

Is that what shareholders are asking for, right now? Don’t they want fast growth more than they want profits?


I suppose, yes, and this is a little scary - the world being in such a state that you can wait 10 years for a company to make profits and it is still considered OK. Cheapening money.

But also how much longer can we wait? Google, Apple, Microsoft and Facebook were all turning profit before the IPO! Find the old S-1s if you don't believe me. And back in those days IPOs happened much much earlier.


I think Amazon is one of the rare examples where it paid off. They were making losses in exchange for extreme growth in many sectors.


Correct me if I'm wrong, but Amazon was strongly cash flow positive as of 2002. Moreover, their operations were throwing off cash the year they went public (1997).

Everyone should be considering this a bit more because it says something about the irrationality of the current market. Large VC funds were investing under the thesis that some of the current crop of (now) less than impressive unicorns could follow the Amazon model: invest in growth at the expense of net income. These unicorns haven't simply failed to produce positive net income by intentionally following the Amazon model, rather they've simply burned through mountains of cash following less than impressive business models. They never followed the Amazon model to begin with.


Amazon made very controlled losses, and it was clear the core business was good and they were investing in new ideas. Sort of like alphabet except much bigger.


Didn’t early stage investors in Uber and Pinterest made money when they IPO’d?


I said "companies delivering profits", not "people making money trading the stock". I'm sure speculation can be profitable on such volatile securities, after all volatility is everything for traders. But there's nothing here (so far) for a fundamentals based long term investor to write home about. Zero.


[flagged]


I don't think that the trade war had disproportionate effects on Altman's prop 1 companies.


Altman's bet was about absolute values, not proportions. The trade collapse impacts those as much as anything else.


The prop 1 companies underperformed Altman's expectations in a way that the market as a whole, and the prop 2 and prop 3 companies, did not. Explaining that via recourse to a trade war is difficult.


I would argue that Trump is the worst president in history but can we please leave names like “Cheeto-in-chief” on reddit?


I'm sorry if you took offense at any part of my comment, and I've tried to streamline it a little bit. But I think that some things should not be left unsaid - Trump has been called "the worst president in history" for all sorts of reasons, many and perhaps most of which are just about partisan bias rather than substance. Focusing on some of the actual policies he's taken credit for, and their real-world effects is probably more helpful.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: