AVGO Stock is Down 26% - Here's Everything You Need to Know

In today's video, we are going to be talking about Broadcom with the ticker symbol AVGO. This is a very popular stock, especially amongst the AI trade because it is the company that is designing the custom ships for Google, Meta, Anthropic, and OpenAI. So, this is a very important company that's seen tremendous growth over the past few years. However, Broadcom stock is down roughly 25% from its all-time highs, and it fell by another 4% on the day after reporting its earnings. And I am recording this video on Thursday and I will be posting it tomorrow because I'm getting to this a little bit late because I have spent the entire day learning about Broadcom, its different businesses, its competitive advantages, its moat, and why investors have been selling off this stock so much despite what seems like a fantastic earnings report. And what I want to do is start off by taking a look at their most recent earnings report, the highlight from it, the transcript, and then I put together a little bit of a Daniel Prong slideshow to really dive into Broadcom's business so we can all understand it more. So let's now take a look at the earnings highlights. All right, so in this first screenshot, we can see that revenue came in at $29.6 billion and was up 86% year-over-year. So, Broadcom nearly doubled its revenue on a year-over-year basis despite it doing $30 billion in quarterly revenue. Now, they also produced operating cash flow of $14.2 billion on the quarter with only $500 million of capex. So, they did $13.7 billion of free cash flow with a 46% free cash flow margin. These are extremely impressive numbers. They also gave us their guidance for next quarter and they are expecting revenue to come in at 34.8 8 billion, which is up 93% year-over-year. So, Broadcom is expecting its revenue to continue accelerating next quarter. Then down below, they say that their custom AI chips revenue grew 221% year-over-year, and it is expected to grow 236% on a year-over-year basis next quarter. So, Broadcom's custom AI chips revenue growth rates are also continuing to accelerate, and they're tripling, literally tripling on a year-over-year basis. So, Broadcom is benefiting significantly from the hyperscalers building their own custom chips like Google's TPUs. This next screenshot shows us a quick table of some of their key metrics and growth rates. So again, revenue up 86%, operating income up 92%, earnings up 95%, earnings per share up 96%, operating cash flow up 98%, free cash flow up 95%, semiconductor solutions revenue up 127%. And then their infrastructure software revenue is up 29%. Across the board, these are ridiculous growth rates. And it's not just growth to the top line, but the bottom line is also growing tremendously. Again, free cash flow basically doubled on a year-over-year basis. So, I want to head over to Stock Unlock really quickly now and show you what the growth rates over at Broadcom look like because this is truly ridiculous. On a trailing 12 months basis, Broadcom's revenue is absolutely exploding. And if we go to a quarterly basis, look at how much the revenue growth rates have started to accelerate over the past five quarters. Revenue is growing tremendously over at Broadcom. As more of the hyperscalers and leading frontier models and AI companies want to design and build their own custom chips, they are going to Broadcom to do so. So, Broadcom is seeing significant tailwinds to its business. And this is the exact same story with the company's free cash flow. Just over the past couple of quarters, free cash flow has nearly doubled from 8 billion on a quarterly basis to 14 billion now. And if we take a look on a trailing 12 months basis, you can see that they have now done $39 billion in free cash flow and the free cash flow is starting to accelerate significantly. So now let's head back over to Broadcom's earnings report. And these screenshots come from the conference call transcript. Now, I'm not going to read the entire text here, but if you want to pause the video and see my underlined highlights, then please feel free to do so. And I would actually recommend doing so as well. But the summary of this screenshot is that Broadcom is delivering custom chips for Google, Anthropic, OpenAI, and Meta, and it is expecting growth to continue accelerating from these businesses. And the revenue that Broadcom is seeing is just scaling rapidly. In this next screenshot from the transcript, Broadcom says that it is expecting 10 gawatt of incremental compute in fiscal year 2028. Anthropic is also on track to become Broadcom's largest customer. And then OpenAI is going to deploy 5 gawatt in fiscal year 2028 and become the second largest customer of the company. And in my opinion, this right here is what the market did not like about this earnings report. And I'll get into more in a couple of screenshots here, but I have more that I want to discuss first. So, moving on to the next screenshot here, Broadcom gives us their guidance, and they are expecting AI revenue to double in 2027 and then double again in 2028. They believe they can produce $230 billion in AI revenue in their fiscal year 2028. They're also expecting to exceed $30 in earnings per share by 2028. And again, they think that they can beat this $30 in EPS in the next 2 years. And remember, Broadcom is a $360 stock today. So if Broadcom can achieve this, then its price to earnings ratio today based on 2028 numbers is only 12. For a business that is so high quality, it is the number one custom chip designer globally. And it is seeing its revenues grow by nearly double on a year-over-year basis, trading for about 12 times 2028 earnings. So the market clearly does not have full faith in this guidance because if they did then I do not think the stock would be selling for 360 bucks today. Again this is one of the highest quality companies in the world right at the frontier of the AI revolution and capex cycle as we call it selling for 12 times 2028 earnings and they believe they can actually exceed $30 in EPS. I mean, for example, if Broadcom simply trades for 20 times earnings in 2028 and actually hits this $30 in EPS target, then it would be a $600 stock, which would be a total return of about 66% over the next couple of years, which is a compounded annual growth rate of 29% relative to the share price today. And again, that's simply the stock trading for 20 times earnings and hitting their $30 EPS targets. So, if Broadcom can achieve that, then the stock does look surprisingly cheap here and it could produce pretty solid returns over the next couple of years. So, it's very clear to me that the market doesn't have faith in this guidance. And I'll show you why. So, going back over to the conference call transcript, an analyst added up all of the growth and saw that 60% of 2027's AI revenue would come from OpenAI and Anthropic. And then 75% of 2028's AI revenue would come from them as well. So what this means is that Broadcom's AI revenue growth and a large large amount of its AI revenue is coming from OpenAI and Anthropic and the market does not like this or have faith in this guidance because OpenAI is not a profitable company. So if a significant amount of their revenue and guidance is coming from a business that isn't even producing profits today, then how much faith can you actually have in those numbers? That is what the market is asking. And that is why the market is skeptical of this guidance and not really valuing Broadcom off of what it's putting in the transcript here. It's kind of like what happened with Oracle stock last year when Oracle said it had something like I'm just pulling numbers from the top of my head, but it was something like $400 billion in RPOS from OpenAI and the stock shot up like crazy. But then over the next few months, the market started questioning like how realistic is this? And I think this is kind of what's happening to Broadcom here. So now let's get into the presentation I made on Broadcom. And by taking a look at the moes and different revenue streams of this business, we're going to be able to better identify if this stock is looking like an opportunity right now or if it is looking a little bit more risky and the market is kind of right here. So let's dive into the presentation. So in this first slide, we have Broadcom's different business units and its revenue breakdown as of its most recent quarter. So the number one revenue generator is its AI semiconductor unit and this has two real units within it. The first one is custom accelerators or XPUs, which are the chips that Google, Meta, and OpenAI are all building with Broadcom. The other unit is AI networking. This unit creates AI clusters that have tens of thousands of chips and networking gear within them to help the systems talk to each other extremely quickly. Then they have nonAI semiconductors. This unit is built of many different products. things like wireless connectivity, think like Wi-Fi and radio, broadband, server storage, enterprise networking, and industrial chips like the ones that go into cars and machinery. Then we have infrastructure software, which might surprise some people, but Broadcom is one of the world's largest enterprise software companies, which is entirely separated from its chips businesses. It's a collection of software companies that Broadcom has acquired over the years and are said to have a 94% gross margin. These software businesses are in private cloud, mainframe, cyber security and fiber networking. The theme is that all of these are missionritical softwares tied to cloud networking and again cyber security. Then we can see Broadcom's revenue diversification and AI accelerators were 41% of revenue, AI networking was 15%, nonAI semiconductors were 14% and infrastructure software was 30% of revenue in the most recent quarter. So Broadcom's revenue is actually pretty diverse and doesn't just come from the hyperscalers or building custom chips. That is where a lot of its acceleration is coming from, but Broadcom does actually have a pretty durable and highly profitable business. All right, now let's talk about where Broadcom sits in the chip stack. So right at the bottom here, we have ASML, which builds the machines that make chips. However, ASML does not manufacture or design any chips itself. It supplies the manufacturers with the machines that actually build the chips. Then we have TSMC which is an example of a manufacturer or factory. TSM takes the chip designs that are handed to it and physically makes the chips using ASML's machines and other machines inside of its fabs. It does not design or sell its own chips. For example, AMD, Nvidia, Marvel, and Broadcom all send their blueprints to TSM to actually make the chip. Then, moving up the stack, we have Broadcom and Marvel, which are the chip designers. They work with companies like Google, Meta, OpenAI, and Anthropic to design chip blueprints that can then be sent to TSM to fabricate. Broadcom is the number one leader in chip design by far, with Marvel being a distant second. And then at the top we have the buyers of chips. So it works like this. The buyers of chips go to Broadcom to design them. Then Broadcom goes to TSM to fabricate them and TSM goes to ASML to get the machines to actually do so. That is how the stack works. Then my next natural question is what is the mode of Broadcom and who can actually compete with the business? So Broadcom's main competitors are Marvel, MediaTek, Alt Chip, and even Google and Meta are building their own in-house design teams. And then we have Nvidia, which compete because they are an alternative to custom chips. So my question was, can these competitors replicate what Broadcom does? And the answer is that Marvel and the Taiwanese houses can genuinely design leading edge silicon chips. So yes, some of their competitors can actually replicate what Broadcom does, but what's challenging to replicate is the execution. Broadcom has the fastest time to market without respected world that is a massive advantage. And then another one of my questions was could Google use competitors for the same services and the answer is yes. Google is already using MediaTek in parallel with Broadcom to build TPUs. So in terms of simply designing, there are other companies that can do what Broadcom does. So then my next question was what is causing companies to stick with Broadcom when there are similar companies out there that can do the same stuff? Well, the first one is switching costs. Each XPU is a multi-year deeply co-engineered relationship. Switching midprogram or even at all means losing a lot of goodwill and time which ultimately is money. So, it's not really in the company's best interest to actually switch from Broadcom. Then there's time to market. Broadcom gets chips to market the fastest, which gives customers a competitive edge as well. There's no respins, which means they also have a cost advantage as there's less waste and bad chips that they need to redo. Then we have the fact that Broadcom has an all-in-one solution. Customers can go to other companies, but no company has the full breadth of what Broadcom does or has. It simplifies supply chains to just stick with Broadcom. And then lastly, we have bundling. Broadcom also has networking gear and infrastructure as part of its packages. Other suppliers do not have this bundling capability. So switching from Broadcom would mean replacing it with multiple suppliers, not just one, which obviously is less efficient. So my summary is that Broadcom does not have a monopoly on its chips designs and companies can replicate what it does there. So Broadcom's moat comes from cost efficiencies scale its marketleading position and brand and being a one-stop shop. So now let's move on to the risks that I have been able to identify based on seeing what other investors are saying and why they're avoiding the stock. So the first one is simply high multiples. The business is trading for about 40 times trailing 12 months free cash flow and earnings. Now this is a high multiple for any business and it means that it's priced for continued extremely fast growth. Then we have customer concentration and in my opinion this is the big one. Investors are concerned about the concentration of Broadcom's revenue guidance. Again, it was noted that 60% of its fiscal year 2027 AI revenue and 75% of its fiscal year 2028 AI revenue is going to come from OpenAI and Anthropic. This also works out to a total of about 58% of the company's fiscal year 2028 revenue guidance just coming from the Frontier Labs, which means that more than half of the business's total revenue would be coming from OpenAI and Anthropic just 2 years into the future. Investors really do not like this. Then we have execution risk. The guidance is somewhat out of Broadcom's control since the chip deliveries are constrained by how quickly land and energy can be secured and ultimately how quickly data centers can come online. If there are any material slowdowns, then that could impact the guidance and Broadcom could end up missing it. Then we have the AI capex cycle and this is the one that bears have been talking about for quite a while and that's that eventually the AI capex cycle will come to an end and the hyperscalers will eventually slow down their spending. Once this happens the entire AI trade will see their fundamentals deteriorate, growth will decelerate and multiples will contract way back down and investors do not want to be in these businesses when that eventually happens. Then we have circular financing because in Broadcom's conference call they were asked about the funding concerns from OpenAI and Anthropic. And to put it simply, Broadcom said it believes these companies will IPO. They will get financing from third parties including Broadcom, which means that Broadcom could end up funding its own customers, which a lot of investors don't like. And then lastly, we have longerterm guidance credibility. Basically, everything that you're seeing on your screen right now is making investors cautious about the guide and the long-term profit potential of this business. In other words, the guide does not have a lot of investor confidence behind it, which is why the stock isn't being properly priced relative to the guidance, at least in my opinion. So, that all sounded pretty scary, right? Well, now let's head into the rebuttals as I see them. The first one is the high multiples. Now, in my opinion, looking at trailing 12 months metrics on a business growing 90% year-over-year gives you the wrong picture. You're starting from the wrong point. You want to be looking at forward metrics when a business is growing this quickly. And on a forward basis, Broadcom is trading for around 21 times earnings in free cash flow, which makes it look much more fairly priced relative to its quality and growth. On customer concentration, Broadcom has said that it's still supply constrained, not demand constrained. This means that if something impacts OpenAI's ability to pay, then the air pocket of demand would be filled by another customer like Meta or Google. Demand currently exceeds supply across all of Broadcom's customers, and it seems like it's going to remain that way for quite some time. Then we have execution risk. The rebuttal to this is that the risk is already being discounted in the guide. In the conference call, the CEO said that they judged the number down from the 30 gawatts of raw demand that they quoted because they don't expect it all to be deployed in time. The current revenue guide is after they revised these numbers down and discussed what customers think is realistic in terms of data centers coming online. Moving on to the AI capex cycle ending. Much of the new demand is coming from inference which scales with AI usage and adoption, not from training new models. Companies are now focused on bringing down the cost to use AI. So adoption grows and it is working. As I showed in my last video, AI token consumption is up 25x year-over-year and 2x month over month. And inference is what is driving this. In other words, if AI adoption and usage grows, then the financing should continue. Moving on to circular financing, Broadcom is not the primary lender and vendor financing is actually normal business practice. Usually, there are a lot of protections, guarantees, or even equity associated with doing so too. Then lastly, we have long-term guidance credibility. And the rebuttal here is Broadcom's track record of conservatism. Broadcom has beaten and raised its stone guidance for 10 straight quarters. The guidance is also based on multi-year agreements and not hopes. So then that leaves us to talk about the conclusion and here are my final thoughts. Number one, Broadcom is the clear market leader and beneficiary from custom chip design and I think that it will continue to grow and be in very high demand. Number two, I think the AI capex cycle will last at least until the end of this decade with custom chip design seeing a higher portion of the capex spend too. Broadcom does not look expensive on a forward basis. In fact, I think it looks quite cheap if its guidance can actually be hit. Again, trading for 21 times forward earnings. Number four, the company has a history of being conservative with its guidance and leaving room for additional upside and guidance to be beaten. and the CEO made it sound like they're actually doing this again with their $30 in EPS for 2028 guidance. Number five, I do think there's risks with OpenAI and Anthropic, but I also believe other customers will fill their gaps if the gaps are actually created from these companies not being able to meet their commitments. And then lastly, I do believe that Broadcom will continue to grow rapidly. It's a diversified business and is becoming one of the most important companies in the world. At a 21 forward price to earnings ratio, I think it's looking attractive and I am considering starting a position here. To give you some more of my perspective as to why I think Broadcom could be looking interesting here, let's take a look at Broadcom's historical forward price to earnings ratio. And right now, it's sitting at a forward PE of about 22.17. Now, during Liberation Day in 2025, it bottomed at a forward PE of about 20, which is slightly below where it is currently trading, but still it's around the same price multiple. And then all the way back here in November of 2023, it was trading for about a 22 forward PE, which again is the multiple it's trading for today. So, November of 2023 and April of 2025. Now, let's go and take a look at how Broadcom has performed since November of 2023. So, it was about 94 bucks back then. And the stock is up roughly 276% since that time. And if we take a look at the returns since April of 2025, it is up 144%. And these share price returns are entirely from the business's earnings growing tremendously because again, these were the last times that the stock was trading for about 22 times forward earnings. So, all of Broadcom share price growth has come from its fundamentals growing tremendously. Now, what that tells me as the potential investor is if Broadcom can actually grow its earnings and, you know, pretty much double them two more times over the next couple of years, then the stock could be looking quite cheap right now and like another one of those opportunities to buy in. The thing that is really hanging me up and the number one concern that I have found is that again 58% of its 2028 revenue projections are expected to come from OpenAI and Anthropic. That is just a large amount of customer concentration and it really means that those two companies have to deliver and be able to meet their spending commitments with Broadcom for Broadcom to ultimately see that growth as well. Then I also do have a question about what does Broadcom look like by 2030 if maybe OpenAI's IPO doesn't happen and maybe its funding stops then how durable are the cash flows truly out to 2030? That is another question that I do have on my mind that I don't know a real answer to. So, if I were to take on a position in Broadcom, just for myself, since I don't really have a strong long-term view on what the cash flows look like, I probably would not make it too large of a position in my portfolio, but I would put it in my portfolio for exposure to the semiconductor industry finally. I know that I've been called out for not having any exposure there, but that's kind of how I would view it. I do think that it is a very stellar company, very high quality, but the revenue concentration and the long-term revenue view I think is a little bit uncertain for me. So that is where I have landed with Broadcom. And if I do end up taking a position, I will definitely let everyone know. But so far, that's my analysis on the business, its moes, why investors have sold it off, and also its potential for future growth. If you enjoyed this video, then please let me know and leave a like on it. And if you want to see more content like this, then please consider subscribing to my channel as well. As always, thank you so much for tuning in.