The Intrinsic Value Podcast 对 $GOOGL 的判断变化

Want Google's cloud revenue? It's there.

yet that said he just invested tens of billions of dollars in Alphabet and we know that it was Buffett now and not Abel who was calling that decision which is interesting and basically he's called Is negligence of tech companies a mistake?

And although I would say investing in Google was probably one of the safest bets on tech you can make. Prices are high, but boy, I mean, what a quarter we just saw from Google again.

And so anyways, I'm glad that we will be able to talk about that one in our winner episode because it's definitely more fun to revel in the successes and we bought that one pretty close to the bottom and it made it our biggest position by far and it's worked out well.

But looking at Adobe again, I mean, it's clearly not at the same quality caliber of Alphabet, of course, and yet it still is the market leader in its industry by a very wide margin.

And so we talked about the Alphabet comparison and I actually don't think it's a totally off comparison to make because just like how I thought Alphabet as the dominant player in search was best position to integrate AI into search and even though that wasn't a guarantor of success, it did give them a pretty competitive advantage.

And probably the problem is that Adobe feels more threatened by AI, at least I would argue, than Alphabet ever has. Again, last year we bought Google and there was certainly a lot of people who told us that stupid search will die and they have no clue what they're doing in terms of AI.

So it might just be the hind bias of me thinking that Google has not been threatened to at least the same extent as Adobe has.

I was expecting for it to be easier for me to wrap my head around the trade desk after already having gone deep on Alphabet and Roku,

But on the other hand, you can analyze a Google investment which was [clears throat] also done during peak negativity and it turned out different.

Yeah. For me, Uber, Google, and Adobe are all very, very significant positions in my portfolio,

I'm very proud of Alphabet. I think there's a lot to say about this one and and fortunately mostly positive things to say. It's somewhat funny that Google has been our largest holing from pretty much the beginning of this journey and yet it's probably also the company we spent least amount of time talking about and it might be because you know everybody talks about Google.

There's so much to say about it, but there's also so many people talking about it, we don't really feel the need to say that much about it.

Back when you pitched Google, the big question was whether search would be disrupted and whether the DOJ antirust investigations could potentially even split up the company, And again, like those concerns pushed the stock down to as low as $150 per share about one and a half years ago.

And fortunately, you made the wise decision of buying a lot of stock for the portfolio back then between about $150 and $190. It went up a bit after averaging up. So, we sit on a gain of almost 100%.

It is funny because when I first covered Alphabet, if you go back and listen to that episode, I definitely was not pounding the table and and calling it a screaming buy.

Uh I did this very basic some of the parts valuation model and it sort of suggested that it was fairly valued at between 180 and and $200 per share. And I really just felt like there was more potential upside though that you couldn't necessarily cover through a valuation model.

At the time, OpenAI's Chat GBT dominated the LLM race and barely anyone had ever heard of Claude and and Gemini was certainly seen as being way behind.

And a lot of people were convinced Google search was dead. And I saw that all over Finwit and Reddit and it did feel like for the first time in my life, Google search was facing a real credible threat.

And the way people were searching for information was potentially changing dramatically. And half a year later though, Google's Gemini was then seen as the leading LLM. And Google used its dominance over search to incorporate AI into Google search.

And so one of the big concerns was that even if people came to Google still for queries, queries that were answered by Gemini or AI overviews powered by Gemini, that might not be worth as much as traditional Google search in terms of advertising monetization.

So the pendulum did swing pretty dramatically, which for anyone who wants to talk about efficient markets, it is hard for me to fathom how a company can have its value swing by the order of magnitude of trillions within a calendar year and then say Mr. Market isn't susceptible to overexaggerating to the upside or to the downside based on the narrative of the day.

Whereas Chad GBT got tons of users quickly, but serving answers to hundreds of millions of people using LLMs is a really tough business to be in because the compute is just so expensive, right?

I mean, the cost to answer the types of basic questions that everyday people ask is significantly less for traditional Google search to be able to answer in a satisfactory way than for LLMs.

I wouldn't say because the market has found a new appreciation not only for Gemini but also for Google's cloud business which has benefited hugely from this entire AI race.

And I certainly didn't capture all these different variables in my model last year. And directionally, I felt like it was their game to lose. I'm happy to make that bet all day.

And that's how I I thought about it at the time. And I don't know, maybe in hindsight, it it was probably more than a fair price.

It was a pretty generous price from the market. Although, I would also say that back then the narrative was quite brutal. So, it was not an easy thing as you did to just say, "Hey, I have the guts. I invest in Google." And I go against pretty much the entire market back then.

And of course, you know, you could argue and there are some people doing that that the same happened to Google where it basically went from trading at a price to operating cash flow in the mid- teens to, you know, the 30s.

And yet, you can already see that that alone cannot explain the doubling of the stock. And it basically grew in pretty much a straight line, 10% a year. You know, 0% 10, 20, 30, and today we're at 40%.

I should add that operating cash flows in this quarter have been abnormally high just due to the impact of the SpaceX IPO because Google had an equity stake. We all know SpaceX went public went quite well in the beginning not so much by now but still obviously was a huge impact on Google's earnings and it says that the current multiple is 20 again only for the sake and SpaceX.

If you adjust for that you know it's quite a lot higher than 20. And obviously, we use Google and Alphabet interchangeably. And I think it's perfectly possible that we might see a pullback here at some point, but it could just as well be that Google keeps chugging along, compounding its intrinsic value for a long, long time to come. That that's certainly the hope if you zoom out long enough.

Although to be fair, I would say that with the benefit of hindsight, which is, you know, something that you always have to account for in today's episode, getting Google at a forward PE of 18 times was probably more than just a fair price.

but how about instead of doing that, we just talk a bit more about Google's latest results because I think there's a lot to discuss. I mean, the stock went down 8% the following day, although I personally think the earnings were quite spectacular.

>> In a word, they were breathtaking. Alphabet just defies all laws of economic gravity. The the law of large numbers just seems not to apply to them. Apparently, the market though seems to be more focused on the potential downside of this capex spending, overbuilding data center capacity.

And and I do see that as being a pretty valid concern, right? If we're being honest, the Alphabet that investors have the chance to buy today is a very different company profile than the company that existed 2 years ago.

And so I I don't mean that simply because of all the changes with AI, but from a financial lens, the capital intensity of Alphabet has changed profoundly as they commit spending tens of billions of dollars for years to come.

and all that will eventually show up as depreciation costs on the income statement to say nothing of the interest costs that will come from the debt that they're taking on to finance all this spending. Yeah. The question is what the ROI, the return on investment will be on all of those investments.

And I guess you can probably say that's a trillion dollar question here. argue that Google probably has a unique advantage in terms of LLMs, for example, because we talked about Adobe's AI distribution advantage just a couple of days ago and Google obviously has one of the biggest distribution advantages in the world. I mean, just AI overviews is a fantastic example of that.

Absolutely. So, and you know, you have this kind of perfect storm of caution around AI generally, caution around the capex spending, and then also some caution from the market with Google trading at a relatively rich valuation, right?

It's certainly not as cheap as it was when we first bought it a year and a half ago. And it just looking at the numbers, Google is now trading at a price to free cash flow per share of over 70 times.

And that's mainly because capex investments of over $200 billion annualized are eating away a significant chunk of cash flows. And so free cash flow is just operating cash flow minus capex.

So as they blow up this spending on data centers, that means that their free cash flow, the money that's basically left over for them to potentially distribute to shareholders is significantly reduced.

So maybe that multiple is artificially inflated at 70 times free cash flow. Or to the point a moment ago about the capital intensity of Alphabet changing, maybe this is the new normal where Alphabet is nothing close to the asset light software search business it once was.

And for them to continue growing, they have to invest massive amounts in physical projects in the real world, right? They now own a whole lot of data centers directly and indirectly. And that again changes the company's profile from being an internet company, a software company to a company that does things in the real world.

And so that is a big change for them. and for context you know management is guiding for capex of about $200 billion this year and estimates are actually as high as $300 billion for next

I don't know if we'll actually see that. I think it's realistic that we might be in the middle. So talking 250, 260 billion. But still, I mean, those are incredible numbers.

And if you just do some quick math here, Alphabet is running at roughly $475 million in revenue and an operating margin of 34 35% in 2026.

So just the current depreciation and amatization is a little less than $60 billion, which would be about 12% of revenue. If we now assume that the capex numbers that I mentioned above are true, so we will actually talk about 200 billion this year and then 300 billion for next year and the capex split of servers and chips is about 2/3 servers and 1/3 chips which is what you usually tend

Obviously that also depends on the revenue growth that accompanies the capex spend and the assumptions that we make here are quite bearish. I mean 300 billion in capex is quite a lot of money. But whatever that number will actually be investors want to see returns sooner rather than later. I think that's what the market is telling us especially in the last couple of days.

And in the last few years Google spent tens of billions of dollars each quarter to buy back shares which led to a buyback yield of about 4%. And just last year, we sort of look for these companies that turn into share cannibals. And it looked like Google might become that. Today, I would say that's probably not what we're going to see from Google. I mean, that part of the shareholder return is something that we'll no longer see with Google.

Instead, Google has already raised more than $50 billion in long-term debt and announced it will issue about $85 billion in new shares this year. So a lot of the stock that was bought back in the last few years is now issued again to fund capbacks, to fund data centers, to fund AI research.

So yeah, I mean getting to the more positive news behind the earnings release, I would say that Google's cloud growth is just absolutely off the charts. Growth went from low30s% year-over-year to more than 80% in the last quarter.

And so the most insane number is without a doubt though the backlog which is now standing at half a trillion dollar growing $375% year-over-year almost four times year-over-year. And according to Google about 50% of that backlog is expected to be recognized within the next 24 months. So that's a lot of pending unearned revenue.

And so even with about 40% of that backlog coming from anthropic that is obviously a concentration risk but the point is there is a lot of real demand supporting a huge amount of business for Google cloud and these are very much not just numbers on paper

But now I think I'm pretty confident that will be the case. I just don't see an end and especially if you just look at a 24-month time frame. Um I'm pretty confident Google will get that money and then it just absolutely insane looking at the size of that backlog.

I would say that the only downside and probably also why the market sold off the stock on these earnings is that Google for the first time since IPO reported a quarter of negative cash flows which is why Google needs to raise all that debt and that equity to still fund capex investments because for the first time in a long time even the cash flows of Google are not enough to fund their new investment cycles which is insane to say and and think about

but you know if this happened a year ago again I think I would have been more cautious about that than today because the numbers does clearly show that the demand is actually there and that these investments are paying off and if I can bet on arguably the best company in the world and you know incredible tailwinds that we certainly see I think I can continue to sleep well with you know Google being the largest position in our portfolio even though it's not a value play at today's valuations

I have to say though and this is important and this might also sound counterintuitive at first I wouldn't buy Google shares at today's prices considering the other opport opportunities.

And I know some would probably argue that if you wouldn't buy a stock today, you probably should sell it if it's already a position in your portfolio. I would personally disagree with that. I think when you have these highquality compounders in your portfolio, I would just let them run and sort of, you know, give them the benefit of a doubt for all the returns that they generated from you.

And this kind of goes back to our discussion at the beginning of the episode. You know, great companies tend to surprise you to the upside. So, I'm personally okay with holding a stock that is valued, let's say, optimistically. If the stock declines in the short term, you know, that's fine with me given that I intend to own it for the long term anyway, and we're sitting on 100% returns.

And if there's a positive surprise that makes the stock fairly valued again, which could argue maybe it's a bit over that today. Say they have significantly faster growth or maybe less margin pressure than I thought and the market thought. You're still invested to benefit from all of that.

Want Google's cloud revenue? It's there.

We just compared you know Google which is at 475 billion which is also insane and still Amazon is significantly bigger

And all of this is a long-winded way of saying I like both Amazon and Google a lot. And it's mainly because both of them feel like they would benefit massively from AI. But if AI should be a disappointment in whatever way that would be, they still would be fine and they still would be two of the best businesses in the world.

Google is up 75%. Amazon is up 0.5% at the time of recording. So it does feel like one has a lot more room to catch up.

I with Alphabet, even if the valuation is stretched, it's such a good business. So diversified, I'm I'm pretty happy to let it run. Wouldn't have the same confidence in Reddit, even if it is still a good business.

You can do that for Google because Google is a much more mature business. It's way more diversified. Things will not change as quickly.

And ironically, Google is part of the reason for that. So, Reddit does have some dependency risk on Google. You hear about it on every Reddit earnings call. The analysts love to ask about it and it's because a lot of their traffic does come from Google search. So, a bit more than half of daily Reddit users visit the website via Google search. So, they start on Google search, they search something, the best uh brownies in town, and then they go and find a local subreddit or whatever it is that gives them recommendations. And they're not necessarily logged into the app, though. They don't necessarily have accounts. So, that's why we were talking before about the importance of logged in users, people who have accounts and are recurringly coming back versus people who just Googled something. Google suggested a Reddit thread for them to check out and then that's how they ended up on Reddit. So really the entire debate of how LLMs will change the legacy Google search model, how people will have their queries answered, that's all fundamentally very very relevant to Reddit. And so maybe some listeners are surprised to hear that, but Reddit is truly one of the best places on the internet to get answers from real people. and Google is of course in the business of delivering the best answers to people who are searching for queries. So there's a natural relationship between them and it kind of es and flows.

Well, another thing that Reddit has just announced or at least debated in order to reduce the dependency is not to renew their data licensing deal with Google because Google currently pays Reddit $60 million a year to use their data for their LLM results. And I already found that shockingly low back when you pitch Reddit to me about a year ago. And I think they should get multiples of that, which for a company of their size would be a whole lot of money. So, I'm actually glad that they play their cards right here. But I got to say too on this earnings call, you didn't get any update on when we can actually consider money coming in through those licensing deals or if that will even happen in the future. And now that companies like Alphabet in particular are starting to make money from AI and LLMs, I would think that the price that they're willing to pay for good data would increase. So yeah, you could probably argue that Reddit's corpus of human interaction from over the last 20 years. Actually, just last night, I was reading through a Reddit thread from 16 years ago that was still relevant and interesting to me today. So that that's the cool thing about Reddit. And uh yeah, that deal was signed in 2024. And back then, nobody knew how AI overviews on Google would affect web traffic, right? It could have been that they convert at similar rates to a classic Google search, but now two years in, Google has a billion Gemini users and clickthrough rates for original sources plunged by roughly 60% because users can find exactly what they're looking for answered in the AI overview, as you kind of alluded to, Daniel, and instead of going to the underlying website for additional information, they just do a follow-up question to the AI and get a customtailored answer. So, Google is almost in a way siphoning traffic from websites and kind of maybe cannibalizing itself in some way and then funneling people into Gemini chats.

>> Even for Google, this has been a debate for quite some time whether they are cannibalizing themselves because these AI overviews can perhaps not be monetized in the same way than you know classical search traffic. But now we know that the company is saying that's not the case. They can monetize it in the same way. So now the focus is sort of shifting to companies like Reddit and saying well Google is not affected but maybe you are. And I don't know if AI and LLM weren't already as far as they are in terms of adoption. I think this could have even turned into a more fundamental question. But I think we are past that point now where you know even though Reddit is responsible for about 5 to 10% of

And then just the second piece of that to follow up on Ron's question, do you see any world where you're not licensing data to Google and OpenAI next year?

And and a good example for why that might not end up being the case is that they just went through some major and unexpected business changes. And while I think they were the right thing to do and the right decision, those things just wouldn't happen at, you know, Google, Amazon, or Netflix, those sorts of companies.

But Chinese competition fears has not yet shown up in Google stock.