Alphabet Waymo valuation cut 40% by Morgan Stanley to $105B | Hacker News Reader

Alphabet Waymo valuation cut 40% by Morgan Stanley to $105B

cnbc.com

Why does this read as “mythology specialists cut valuation from an arbitrarily high made-up number to another arbitrarily high made-up number“

I mean, people do discuss the latest episode of game of thrones ... this is not much different. At this point the market has gone well beyond its utility as a tool. The overinflation with fleeing capital makes it seem more like a spectacle.

The astonishing bit of this otherwise vacuous article is that Morgan Stanley thinks Alphabet is undervalued by ~20%:

In terms of Alphabet’s current value, Morgan Stanley has a price target of $1,450

I don't understand how those public company people come with such estimates.

I asked some, tried to wrap my mind around of what they sad, but I still don't understand them.

Basically it's wildly convoluted guess based on looking what companies already doing a similar thing is valued. In the end, it usually ends up by people doing the this line of work calling their peers on an anonymously bought sim card, and asking for how much they will buy it for.

I think, it is stupid

What you describe is likely one of multiple approaches, another of which (and arguably the most important) being fundamental analysis based on company financials.

How else do you find the value of something, other than to ask others what they’d pay for it?

That's what you ultimately have to do, but you can take users, engagement, retention, growth and monetization over time compared to competition.

Discounted cash flow models are popular. Morgan Stanley probably did one to get the 20% undervalued number. Though they are still basically spread sheets with a lot of numbers you have to kind of guess in them.

This approach works for established companies, but breaks down when considering services like waymo, whose value is purely a matter of potential. Waymo is currently worth nothing (if not less) from a cash flow perspective, but potentially worth hundreds of billions in the future, with their effective market value lying at some (entirely subjective) point in between.

WeWork was valued at >60 billion by the same bankers who valued it at 20 billion a couple days later.

> That assessment values Waymo at about $20 billion

Wait so is it $105bn or $20bn? $105bn is still an insane valuation given the serious unsolved problems they face. $20bn seems more reasonable.

"eventual projected" valuation probably means after the product is available on the market. Looks like right now it's all R&D and Testing.

With cars on the road, it's not exactly vapor ware though. The main concern seems to be that the testing is taking longer than they initially hoped.

Full level five autonomy without safety drivers is going to require extended R&D investments. However, with Google rolling around in cash the eventual outcome seems to be not that uncertain. Hence the insane valuation.

I'd say the main risk is increasingly other companies getting there first. E.g. Tesla seems to be quite stubbornly moving forward with its approach and unlike Waymo they have many vehicles on many roads. Also, the German car manufacturers are making progress.

Hard problems; but with feasible paths to a solution and a huge potential market once technology gets there. IMHO this won't take more than a decade and possibly a lot shorter. What will take longer is regulation catching up with reality. Safety drivers will likely be mandatory long after they stop adding any value whatsoever.

The reality is that most car accidents are caused by humans and increasingly in vehicles that don't have assisted & autonomous capable technology on board. Human drivers are by far the biggest risk factor on the road and autonomous driving is made harder by their presence. So, ironically, the job gets easier as that becomes rarer.

>Hard problems; but with feasible paths to a solution and a huge potential market once technology gets there. IMHO this won't take more than a decade and possibly a lot shorter.

I’d consider that an extremely rosy outlook, and would argue nobody has yet demonstrated a “feasible path to solution” for level 5 autonomy (or in fact, lower levels). Self-driving seems as far away as it ever has been, much like fusion power.

What's a valuation meant to be? If it is to represent the value that the company will create for shareholders over its entire lifetime, all these numbers seem very low for a company that is near lead position in vehicle autonomy.

A valuation is what the market (i.e. the average trader) would pay for it. Pretty simple.

In this context a valuation is then impossible, though. The market does not have access to Waymo except through Alphabet.

The market value for waymo now is $20b based on google price. The analyst target is 105b. If they succeed according to the cash flow model the analyst came up with. Which is, of course, horse shit as all analyst models are, and you don't need to be a rocket scientist to notice that they have no monetizable technology at present and are unlikely to in the time frames they 've boasted about in the past (aka we were supposed to be at level-5 two years ago).

FWIIW you can look at the market price as a probabilistic prediction as to whether waymo will succeed to the analysts expectations; about 1 in 5 chances, discounted by long term rates. Seems about right to me.

How do they get $20bn and acknowledge that Tesla will have the winners market share (while Tesla has a market cap of $45bn and does all the other things.)

Oh, the horror, this side project from Google harsly ten years old is only worth $100 billion...

Yeah, whatever. Full self driving will happen. Waymo will be the first to it. And then Morgan Stanley will backtrack.

You think there is a 100% that both are true? That there is no world in which they are not? Because I see a lot of worlds where at least one is not true. Moreover even if they are true, full self driving might take a really really long time.

All of this uncertainty needs to be taken into account in the DCF analysis. Time value of money, yada yada.

Self-driving for some value of most conditions seems inevitable but you’re probably looking at different use cases maturing over decades.

The main wild card IMO is that general purpose self-driving turns out to be really hard and it’s cheaper for those with money to just hire drivers even after self-driving becomes technically viable. That’s dystopian though and there’s not much economic history to support the idea.

Fusion seems iffier in the sense that other energy sources/techs just turn out to be better in one or more ways.

Full self driving will happen.

Yes, and so will nuclear fusion. But it will take much longer than anyone thought.

Waymo should be the first in the western world, but in China I imagine that any barely working self-driving car service would be able to launch, provided it was Chinese-made.

I agree full self driving will happen. It's one of those things like computer will beat humans at chess one day was. But as to Waymo being first that seems a bit up in the air. They may be in the lead at the moment but there's a fair bit to go and it only takes one competitor to come up with better algorithms.

>It's one of those things like computer will beat humans at chess one day was.

No, those are completely different things. The rules of chess are set in stone. The rules of driving are based on communication - driving is an inherently social activity. Having to cover all the weird corner case situations that might occur on the road, fully autonomous driving is essentially equivalent to strong AI, and no one's remotely close to that yet.

It's not just that you need "strong AI" that can function in social situations, it's that you are frequently centimeters and milliseconds from disaster if that AI makes the wrong decision, unlike, say, a flying drone.

There really is no competition. With the weight of Google’s CAPTCHA - billions of images trained on every week - and the fact that they are already FSD in one city makes this a no-brainer.

“Based on discounted cash flow analysis”... bwhahahaha

Hey, could you please not post unsubstantive comments to Hacker News? Especially not shallow dismissals, as described at https://news.ycombinator.com/newsguidelines.html.

If you know more, a great thing to do is to share what you know, so the rest of us can learn.

OK fair point. I’ll elaborate. I think that a DCF model of anything as speculative with such a wide range of outcomes as this company is an exercise in futility. Yes you can build such a model but it is so dependent on your assumptions that it’s pure GIGO. Having worked a long time in finance, I am confident that any sell side model of this company starts with the valuation as the input and then dcf assumptions are built to arrive at that number.

Reply on news.ycombinator.com