Sam Altman's Bubble Talk Bet Is 1 Year Old | Hacker News Reader

Sam Altman's Bubble Talk Bet Is 1 Year Old

blog.samaltman.com

132 pointsivankirigin69 comments

Here are his three bets reproduced for comments:

1) The top 6 US companies at http://fortune.com/2015/01/22/the-age-of-unicorns/ (Uber, Palantir, Airbnb, Dropbox, Pinterest, and SpaceX) are currently worth just over $100B. I am leaving out Snapchat because I couldn’t get verification of its valuation. Proposition 1: On January 1st, 2020, these companies will be worth at least $200B in aggregate.

Seems he is on track for that one.

2) Stripe, Zenefits, Instacart, Mixpanel, Teespring, Optimizely, Coinbase, Docker, and Weebly are a selection of mid-stage YC companies currently worth less than $9B in aggregate. Proposition 2: On January 1st, 2020, they will be worth at least $27B in aggregate.

Hmm... not as confident but I would not bet against it.

3) Proposition 3: The current YC Winter 2015 batch—currently worth something that rounds down to $0—will be worth at least $3B on Jan 1st, 2020.

http://techcrunch.com/2015/03/23/here-are-the-companies-that...

http://techcrunch.com/2015/03/24/y-combinator-demos/

Might be on track as well, though I am even less confident.

If I were a betting man, I would say he is more right than wrong, and on track for all three.

I think this underscores how long 5 years is in internet time.

Basecamp is now a $100 billion dollar company, according to a group of investors who have agreed to purchase 0.000000001% of the company in exchange for $1.

Source: https://m.signalvnoise.com/press-release-basecamp-valuation-...

This is a trite complaint, and in my view, misses what is often bogus about the lofty valuations - liquidation preferences provide a lot of downside protection.

For instance, if I invest $1bn in Uber at a $10bn post and a 1x preference, I get the first $1bn of any sale. Yes, I invested at a $10b valuation, but I don't actually lose any money unless the valuation sinks below $1bn.

This is a much better deal than simply paying $1bn for 10% of Uber, and in my view, actually does not imply a $10bn valuation at all.

You're right on, and there may in fact be a half dozen more material economic terms which also adjust down the implied $10bn number. Add into that the fact that each round often has it's own set of terms, and you begin to realize that the "valuation" numbers thrown around mean absolutely nothing.

They should have sold it for $10 to make it an even trillion.

You do know that post is one big "valuations are high and we are in a bubble" joke, right?

> I think this underscores how long 5 years is in internet time.

You're absolutely right.

The 10 biggest tech stories of 2011 http://edition.cnn.com/2011/12/26/tech/web/tech-news-2011/in...

I'm not sure I understand, those stories don't seem to bolster your point - a few of them are one-time events (Steve Jobs' death), and most of them are still ongoing concerns - hackers, facebook, etc.

I probably agree that 5 years is a long time in tech, but some variation on almost all those stories could easily be published today.

In that vein, World War 2 was merely a variation on the Roman invasion of Gaul. The reason I linked to the article is that things which are commonplace and obvious today - a multitude of cheap Android-powered tablets, social media being used to facilitate political action - were newsworthy just five years ago.

A single bad IPO from that list of firms (Which will likely happen in a climate of pessimism, even if we don't hit a recession) will be disastrous to their aggregate valuation, even if they stay private.

I feel that the only way Sam will win the bet is if none of the firms in #1 or #2 will IPO between now and 2020.

True, but conversely, one standout that does well could alone hit the target. If you believe that the successful companies follow a power law distribution, you'd expect something like that.

That would be the case if valuations were independent of each-other (Which is a bit of a stretch even in a truly efficient market, like the NYSE.)

On the other hand if, say, AirBnB and Palantir have a dismal IPO in 2018, that will likely bring valuations down across the board.

I think #1 is actually the toughest bet. I can see Palantir, Dropbox and Pinterest either remaining static or even dropping from their current valuations. Even Uber and Airbnb don't have a clear path to huge growth in front of them, Airbnb is the safer bet for growth because of the lack of competition. Uber still seems a strong business but I'm not sure how well they continue to grow. X-factor for Uber is how it well it expands into complementary product verticals like self-driving cars, deliveries etc.

SpaceX could end up carrying this category.

> Uber and Airbnb

Imagine a world in which any object (human, animal, package, food) can be moved from one place in a city to any other place in the city using a smartphone. An electric, self-driving vehicle picks it up and drops it off.

That world is very nearly possible, and Uber is on the tip of the spear that's creating it. There's an absolutely massive upside. Uber is probably very under-valued.

Airbnb is less world-changing, but they still have the rest of the hospitality industry to replace. I personally never stay at hotels anymore, and I don't understand why anyone would. People will continue to use hotels less and Airbnb more.

I hear that argument in favor of Uber a lot, but what barriers to entry exist for potential competitors when there are autonomous vehicles?

Currently, you can argue that Uber's competitive advantage and big barrier to entry is the driver network. When you don't need a pool of humans and software that makes the human's job easier, why would Uber be better suited than another company to utilize self-driving vehicles?

Also note that in this hypothetical world, Uber suddenly has to make gigantic capital investments in fleets of automobiles, whereas now they don't have that limiter on their growth.

I find that a LOT of people go from "I envision a world in which rides-on-demand serves massively more rides" to "and thus Uber does very well as a for-profit company" with no steps in between.

Since 1970, the number of passenger miles traveled by air in the US has grown about 6-fold [1]. And airlines have, as an entire industry, overall run a loss. There's a cautionary tale in there for Uber bulls.

(BTW: Driverless cars don't necessarily mean that rides-on-demand becomes the dominant car transportation model, either, though it's certainly within the realm of the possible).

[1] http://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/pub...

That's one model, but you could also have owners of autonomous cars allow Uber to use their car in the fleet.

Basically you can pay people a lot less to be 'drivers' if they don't need to be driving.

But then Uber has to share the profits with the owners of the cars.

Also: if very large numbers of people are buying autonomous cars, then there's probably not a lot of increase in the rides-for-hire market.

> why would Uber be better suited than another company to utilize self-driving vehicles?

People used to say the same about Amazon. Amazon was leading a race to the bottom in terms of prices, and they were selling the same products as everyone else.

Amazon succeeds today because they embraced their competitors and they're aggressively pro-consumer (even to the detriment of their own employees). They're also way ahead of the curve in terms of operational investments. For example, Target can't match the Prime service even if it tries.

Is that exactly the model Uber will follow? I doubt it. My point is just that even in commoditized industries, there can be a dominant player.

I wouldn't bet on Uber based on self-driving cars. See https://en.wikipedia.org/wiki/Autonomous_car: most auto companies and most big tech companies are working on them, and I wouldn't bet that Uber can beat everyone from Audi to Google to the punch. It's more likely that they'll buy self-driving cars from whoever develops them first -- or that whoever develops them will create an app cheaper to use than Uber. (It's not like Uber would be particularly hard to replicate, at least not if you're Google.)

Really? I prefer hotels over Airbnb. I appreciate the fact that I can instantly book any hotel room, rather than having to contact a host first. I've also not found a huge savings from Airbnb when compared to similar levels of privacy and convenience through Airbnb. Of course, each its own.

> Airbnb is the safer bet for growth because of the lack of competition

Airbnb may not face huge direct competition like Uber does with Lyft, but there are still lots of alternative products going against Airbnb.

I look at Airbnb as the eBay for short-term rentals. It's a big marketplace with a lot of inventory and has a huge network effect helping it, but it's hard to serve every niche perfectly. So, just like companies like Reverb have flanked business from eBay, I think specialty companies can flank business from Airbnb as well.

Uber has zero stickiness outside retail car hailing. Their easy and seamless transactions are great compared to taxis, but are par-for-the-course in deliveries, etc.

And the only thing Uber has going for it is being the only choice in most areas. When there's a competitor (as in SF) everyone has all the apps on their phone and shops around.

Also, Uber's huge innovation is zero-capital rollouts. They use your car. When they have to buy a self-driving car they're back to square one and will also own all the liability.

Uber has also benefited from the fact that they offered a service in a well-established market--getting a driver to take you somewhere without reserving in advance. But a market that in many, maybe most, places was poorly served by existing incumbents for all sorts of local monopoly and regulatory capture reasons.

It's easier to displace incumbents when they offer a widely used service but they basically suck at it.

Things like food and package delivery, on the other hand, work pretty well where the economics have worked and it's unclear that uberizing, say, grocery delivery makes it attractive and financially viable where it wasn't before.

Cruise (YC W14) was recently acquired by GM for $1B, so if he picked one earlier class the $3B bet would be attainable in my mind. Not to slight any W15 companies, I just haven't seen much.

I'm really curious whether that $1B is truly $1B, or if that's the sticker price based on some crazy metrics they have to hit, and the actual value of the deal is in the $150-$250m range (not that that's anything to sneeze at either), plus unlikely incentives.

For readers unfamiliar with this type of term: https://en.wikipedia.org/wiki/Earnout

I think sam's predictions will hold even if we are in a bubble, because those companies (not sure about uber and palantir) seem likely to survive bubbles. There's also enough diversity (zenefits has died, e.g.) that there's some cushioning.

Nonetheless, if you think we're in a bubble you should be prepared to liquidate everything and start from scratch at the bottom.

Regarding #1, this means that all six will need to double their value, or some of them will need to grow by more than 100% to compensate for those which don't. Five years is a long time, but 100% is also a lot of growth. There aren't too many tech companies which I can think of who could provide adequate comps, but if we use GOOG as a generic "tech company":

- from 2012-present they went from $350 - ~$725

- from 2005-2012 they went from ~$125 - $350, if you skip a short lull at the end of 2008 when everything was down

One doubling on a 7 year scale, then next one on a ~4 year scale, using the today as the arbitrary starting point and measuring back.

If we ask the question a different way, which of these six companies have 100% growth ahead of them, i.e. their valuation is huge, and either their current market penetration is low or there are large enough new markets available to them. To me, Uber and SpaceX are the only obvious ones on that list. Palantir and Airbnb are less obviously well-positioned, but I would push them both to yes if I were personally to take this bet. Dropbox and Pinterest seem to be clearly in the other side, and if they maintain their current values, then each of the other companies will need to split $25B in growth four ways in order for this bet to hold.

Distributing that evenly:

- Palantir was last valued at $20B, so this is a 100%+ growth target over 4 years.

- As of 2015, Airbnb was valued at $25B so it's a similar situation for them.

- As of the beginning of this year, Uber was valued around $50B, so they would need to grow by 50%+

- SpaceX is valued somewhere around $12-15B, so this would be a ~200% growth for them!

Yet, we all know that unicorn gains are not distributed evenly. Even among the unicorns this is true, so a more likely scenario is that one or two of these companies will experience outsized growth while the others will experience impressive, but "modest", increases.

It's a fun, but probably pointless, exercise to try and pick which of these will be the unicorns among unicorns. (But I'd say Uber and SpaceX still)

Just to contextualize all this -- it was 2009, 7 years ago, when the DJI was last at half its value today. It was also 2009 when the S&P 500 was half today's value. We're talking about companies doubling on 4-5 year scales, 33-50% faster than the rest of the market.

I don't think they will grow evenly though. So the target is $200 Billion and Uber alone is "valued" (I agree that this includes preferences) at $68B. So if Uber "just" tripled its value and everyone else went to 0, he would still hit his prediction.

You need one big standout for him to be right.

Another comment in this thread claims to have done the same analysis and come to the opposite conclusion. But neither of you shared your math. Could you show the analysis that led to your conclusion, so we can compare it to the other one?

I concede i didn't do a lot of math, and I agree with the points raised here:

1. That the valuation preference is more relevant than the value cap. Altman is well aware of this, and I think himself might have brought it up in an interview (with Kara Swisher?). However, for our purposes, I am using the value cap as a measure of valuation for ease of use.

2. I agree with the other comment that one or two companys in the pool are doing the heavy lifting. There is a power law distribution.

So for the first preduction, i just looked at Uber's last value (again, I didn't account for the preference so this is misleading) at $68 Billion and think yes, Uber + these other companies will probably get up to $200 in another 4 years.

Similarly, I'll just pick the top 1 or 2 companies in a pool and guess that if we get over 20% of the value in one year then I just extend linearly.

Its not detailed, its just a quick mental short cut.

I feel better about this now than I did a year ago. Proposition 1 is up about 50% in the first year, and proposition 2 is up about 65%.

The main point remains--we spend far too much time talking about whether or not startups are in a bubble. It's boring and it gets in the way. Sometimes it will be true and sometimes it won't, but the stories claiming a huge bubble for the last 10 years have been generally wrong.

It was telling to me that, even with the fever pitch of VCs calling for the end of the world last year, only one investor (a TechStars mentor) would take this bet.

The goal of this bet was to deflate the bubble conversation, and that seemed somewhat effective.

> The goal of this bet was to deflate the bubble conversation, and that seemed somewhat effective.

Is it the bet which deflated the bubble conversation, or a deflation (or speculation of such) of the bubble itself? I ask because if you take a look at layman-accessible headlines or even some of the conversations which take place here, there's still quite a lot about the "downturn" in the bay. That alone's likely enough to get people to stop talking "bubble" even if the only downturn was in the rate of acceleration of growth, not in actual growth itself.

Still, I appreciate the bet for putting the focus on entrepreneurial efforts as a force of worldwide change rather than as a force of wealth creation.

Hmmm. I think at least some of these companies are worth less now than a year ago. Zenefits, coinbase, palantir, mixpanel, etc. am I wrong? Agree on the bubble issue. It's a boring discussion

That's why the bet is structured as a portfolio. In startup investing, it doesn't matter if some of the portfolio loses big (or everything) as long as there's at least one winner that expands many-fold.

Sure. But I'm sceptical of the claim that the whole portfolio is worth 50% more than last year. I don't know for sure, but as an example, Palantir valuation is down

http://www.wsj.com/articles/t-rowe-price-marks-down-most-of-...

Do you think that perhaps there were crickets because the people with the money/position to be able to take the bet have good reason to share your bullish view because they also stand to benefit from you being right? So they have to not only be bearish but be really confident about being bearish to take your bet?

I believe we are in a bubble but I'm not a millionaire I'm just a regular developer, so I don't have $100k to gamble. Maybe I should have offered to take your bet at $100 a year ago? :)

To be honest, though, even though I believe we are in a bubble I don't feel that confident it will burst by 2020. There's too many factors at play that I can't accurately predict a timeline. I just think we're in a bubble because I look at valuations of companies like Uber and AirBnB and they seem inflated to me.

I question your valuations... they're neither public market numbers nor completed acquisitions. You're using the bubble to validate your assumptions that you're not in a bubble.

Well, that's kind of what you have to do. The definition of a bubble is contingent on the burst at the end. If it never bursts, it wasn't a bubble. So you take today's valuations (which you assume are bubble-inflated) and then compare them to the valuation at some future date. If the valuations are a lot lower, then there was a burst and thus a bubble. If not, no bubble (or it's still ongoing).

I was curious, so I looked up what the press has reported on each company's most recent valuation (and therefore, these numbers could be off, because they don't factor in reported write downs / markups from individual investors [i.e. Fidelity], financing rounds not reported by the press, huge gains/downturns in revenue over the last year, etc.)

1.) The top 6 companies mentioned are now valued at $146.5b.

2.) The 9 midstage YC companies mentioned in the article are valued at $15.43b.

Did anyone take the bet?

---

EDIT: OK, found out that "Boston-based venture capitalist Michael de la Maza" took him up on it. http://money.cnn.com/2015/03/30/technology/sam-altman-tech-s...

Lame. It was just someone who wanted to work with Sam. Not someone who actually believed we are in a bubble, and was "Putting his money where his mouth was"

Previous discussion from longer than 1 year ago: https://news.ycombinator.com/item?id=9271679

My comment(s) then:

Ironically a basic statistics class indicates that cherry picking companies that deliver 2x, 3x and ... whatever the fuck that third pick is ... as a guaranteed return over 5 years is indicative of the overenthusiastic hype that historically surrounds bubble valuations.

#3 is a die roll. #2 is the killer. And I might take the bet on just #1.

I admire Sam's balls but the externalities here are immense. The greatest financial mind of our time built Berkshire Hathaway to $350B over 50 years. GE is worth $250B. Microsoft $340B.

To believe Sam's motley list of companies can either hold onto valuations approaching those "real" companies for five more years, let alone actually generate viable earnings and go public (even at goofy P/E multiples) in line with what GE, Microsoft, or Buffett's candy, ketchup and mac'n'cheese subsidiaries alone make seems ... optimistic at best.

If he loses, might I suggest the book title? "Oops! Brands Aren't Businesses!" by Samuel H. Altman.

I would not call tech industry a bubble by any means. However, since my work took me away from Silicon Valley - I have to say there are almost no startups that have really changed my life in any noticeable way. All I have watched happen is my new smartphones and laptops get thinner, and the apps on them switch to flat design language. Glass was neat, more novelty though. Cardboard was cool, also a novelty - plus I don't game so VR doesn't effect me (yet) and on top of that, neither of these were made by a startup - but a huge corporate entity. I thought back in 2010 I would be experiencing my entire tech life through Augmented reality by at least 2015. Needless to say, I am not. :(

"The gleeful anticipation of a correction by investors and pundits is not helping the world get better in any meaningful way."

Probably the most important line of the post. Hate that someone invested $<X> M in an app that just sends "Yo" to other people? Great, don't use it, don't invest your money in it. Yelling on the sidelines about how crappy it all is, seems counter-productive. Writing self-fulfilling prophecies to get clicks seems even more egregious.

Uh oh, it looks like someone has skin in the game and is worried. :'(

More like broke-ass engineer annoyed at my fellow ilk on the peanut gallery. Though in a way, I don't want to return to the days the Blue-Chips ran the show, if that is what you want to call "skin in the game".

So far, I don't think he's right on all 3

This would be a much more useful comment if you explained why.

If you take the one year out results, and using those growth rates project out to end of bet period, he would not be right on all 3. But it's only been one year and tech doesn't grow linearly. That's why I said so far.

http://www.wsj.com/articles/t-rowe-price-marks-down-most-of-...

I think it's too early to tell. There only needs to be one winner in each bet for them all to win & it still seems possible to me. Hard to tell until a couple of years from now though.

Betting on valuations of private companies is a bit silly.

TLDR: Venture capitalist claims the tech sector is not in a bubble.

Exactly. To be fair, it's pretty hard to remain objective in his spot. If he spent a year anywhere else in the world, doing something different, he would certainly have a different perspective. But that's fine. He should feel the way he feels about this.

I was going to make a similar bet but in the other direction. My instincts say that 90% of the companies mentioned would be out of business by then and the remaining in the "still-going-but-not-relevant/MySpace" category. The coming Tech Winter will be devastating to the unicorns, precisely because so much is expected from them.

But I don't want to be one of those people. A community is built by people supporting each other, up or down, Tuesday or Wednesday. Maybe Sam is right and maybe I am, but at the end of the day, let's innovate and let the future unfold how it will.

"Bubble" is a bit of a loaded word. It implies both overvaluation and general instability; but these things aren't necessarily linked.

Flip side: Journalist writes article with 'bubble' in the title.

It's not a claim. It's a bet with $100,000 backing it.

Protip: Bubbles always last 18 months longer than you think they can.

I figured the Canadian real-estate bubble would pop like 5 years ago. Not so much.

If the 'bubble' pops now, prices will still be higher than they were 5 years ago. That makes your prediction extra bad.

Unless you normalize by currency, in which case its just a fairly bad prediction.

Well... It popped in Fort McMurray.

Edit: And that was way before this year's wildfire.

He'll likely be right on all 3 but only because 1 company in each will disproportionately skew. His bet is overly broad.

Palantir and SpaceX are very promising companies, it's hard to say if their true value will be realized by 2020.

Stripe alone is worth $10-15B if you ask me.

Where he might stumble is the 3rd, however finding $3B of worth in what must be 30+ companies shouldn't be hard, depends how badly the global economy slows in the next 4 years.

I'm not sure this really does much to disprove the insular and bubbly nature of technology investment, however.

The skew is the point though. Startup valuations are supposed to work such that a small number of successes make up for a large number of failures. The press loves to harp on the individual cases, but all that matters for valuation is expected portfolio performance.

Hedging is a thing in investing, yes. I do agree the press reports when businesses are perceived to not be doing well. I don't know I agree they usually point heavily to that being a venture thing. I'm still not sure much of this is the indicator of a bubble, that's all (it's a pretty boring bet?).

> 1 company in each will disproportionately skew.

He's a VC, that's what he's betting on.

Winter 15 batch was 100+ companies.

Reply on news.ycombinator.com