원문 보기
Welcome to The Explainer. Today we will analyze Broadcom's incredibly rapid transformation. We are talking about a company that has transformed from a manufacturer of conventional chips into an undisputed AI giant with a market capitalization of $1.79 trillion. We will use the latest verified market data for October 2026 to find out how exactly this company became Nvidia's most important competitor in the field of AI.
We will look at fundamentals , competitive advantage, and, of course, the real risks that investors have to consider now. So, let's begin. 221%. Just realize that number for a moment. This is a staggering annual revenue growth from AI semiconductors. And, frankly , it was this number that completely changed the investment thesis regarding Broadcom.
We see a business that generated $16.7 billion in revenue from AI semiconductors in the third quarter of fiscal 2026 alone . For the first time in the company's history, AI revenue accounted for 56% of the total. The whole narrative has changed. The company is no longer valued on its outdated portfolio.
The market perceives it as a key driver of the specialized chip revolution. So, here's the plan for our episode today. First: a large-scale transformation in the field of AI. Second: analysis of the financial leap. Third: dominance in the specialized chip market. Fourth: positive and negative signals.
And finally, fifth: scenarios, goals, and verdict. Well, let's start with the first section—large-scale AI transformation. From a manufacturer of multi-profile microcircuits to an AI infrastructure center. Let's go back a little to early October 2026, when the company's shares made a real splash, jumping 9.36% in just three days.
This increased market capitalization by $153 billion almost overnight. So what was the catalyst? On October 1, Anthropic's IPO filing was published, revealing the scope of their partnership with Broadcom. The very next day, news emerged that Broadcom's banking syndicate had begun forming a debt package for a whopping $60 billion.
This includes a $42 billion senior secured tranche from major banks and an $18 billion subordinated tranche led by Blackstone. All of this is aimed at financing the purchase of AI chips and the construction of data centers for Anthropic and other clients. By October 6, the market had reacted. The volume of institutional purchases soared to 30.8 million shares.
For comparison, that's 137% above their 20- day moving average, which pushed the stock up 3.67% in one session to $375.81. But you know, context is the most important thing when you analyze such market movements. While that three-day jump was incredible, let's look at the results for the entire year of 2026.
The year-to-date return of just 9.18% actually turned out to be lower than the S&P 500's 14.22% over the same period. And, frankly, that somewhat masks the extreme volatility beneath the surface. In early 2026, Broadcom shares reached a high of $495 before falling to $289.96. This is a brutal 41% drop from peak to trough.
However, if you step back and look at the 5-year return, it perfectly demonstrates the phenomenal power of multi-year growth thanks to Broadcom's move into AI. They delivered a massive return of 739.94%, absolutely outperforming the broader market over this longer period. Let's move on to the second section.
Analysis of the financial leap, detailed review of results and cost estimates for the third quarter of fiscal year 2026. Let's take a look under the hood of these third quarter figures. AI semiconductors are now the absolute foundation of business. Looking at the third quarter, total revenue reached a record $29.59 billion , up 86% year-over- year.
But just look at the AI segment. It brought in $16.7 billion , which is the same staggering 221% we were talking about. Meanwhile, the infrastructure software segment, which is based on the huge $61 billion acquisition of VMware, brought in $ 8.75 billion. So this software division is vital. Why? Because it generates disproportionately high profits.
We are talking about an operating margin of 84%. And this provides incredibly stable regular cash flows. These funds help fund Broadcom's dividends and reduce debt, effectively stabilizing the highly cyclical nature of the chip business. And that's the thing about this AI revenue. It's not just growing , it's accelerating.
It rose from $8.4 billion in the first quarter to $10.8 billion in the second and to $16.7 billion in the third quarter. It just accumulates. And management is forecasting an astonishing $21.7 billion in the fourth quarter. If they achieve this, it would mean an annual growth rate of 236%. It is this acceleration on an increasingly larger revenue base that is why the market is currently valuing Broadcom at such a premium.
Okay, but here's where the real financial strain is, and we really need to pay attention to it. As they sell more of these AI chips, their gross margin actually shrinks. It fell from 78% to 75% in the third quarter, and the forecast indicates 73% in the fourth quarter. So, why is this happening? That's because these specialized AI chips require increasingly large amounts of incredibly expensive, high- bandwidth memory.
This structurally reduces profitability compared to their software or legacy products. However, and this is critically important, management is ingenious in its use of operational leverage. This means that their total revenue is growing much, much faster than their operating expenses. Because they manage costs so well, they increased their non-GAAP operating margin to a record 67.9%.
This allowed them to generate a whopping $13.67 billion in free cash flow in the third quarter alone. This brings us to the third section: the dominance of the specialized chip market, a protective moat that ensures transparency in AI revenues. We hear about Nvidia GPUs all the time these days. They're grabbing all the headlines, but Broadcom is waging a very different war for dominance in AI data centers.
They absolutely dominate the application-specific integrated circuits (ASIC) market. They occupy approximately 70% of the market share. Their closest competitor is Marvell Technology with a market share of around 22.5%, leaving only a tiny market share for others. Morgan Stanley even calls Broadcom the clear number two behind Nvidia in the broader AI computing landscape.
But, unlike Nvidia, which sells general- purpose chips to everyone, Broadcom completely owns the niche of specialized solutions. If you want to know what a specialized AI ASIC is and why it's important, let's take a look. These are chips designed from the ground up for very specific hyperscaler workloads.
Remember Google's TPU. Hyperscalers prefer these specialized chips because they provide much better performance for their algorithms. Additionally, they are significantly cheaper to operate when scaled, significantly reducing the total cost of ownership. And, perhaps most importantly, having their own chips reduces these tech giants' dependence on expensive general- purpose Nvidia hardware.
This dominance translates into what Broadcom internally calls its “six leading customers .” We are talking about multi-year contracts for the development of several gigawatts of capacity. Google, for example, has signed a deal through 2031 for next- generation TPUs, including the 2-nanometer TPU V9, expected in 2028.
Meta has extended its partnership with MTIA chips until 2029. OpenAI recently deployed its first- generation accelerator, Jalapeno, which reportedly outperforms Nvidia's Grace Blackwell on inference tasks. And then there's Anthropic, which is on track to become Broadcom's largest customer by 2027. Therefore, such concentration creates huge switching costs.
These companies are “ locked in” to this ecosystem. But it also means that 50–55 % of Broadcom's total revenue is heavily dependent on just a few of the largest AI players. Let's move on to the fourth section. Green and red flags: comparing multi-year visibility and risks. Every great investment thesis has two sides.
So let's objectively compare Broadcom's enormous advantages with its most acute fundamental, operational, and regulatory risks. On the "green " side, you have incredible visibility of prospects. Management clearly sees the opportunity to generate $115 billion in revenue from AI in fiscal year 2027 and is targeting $230 billion in 2028.
Additionally, with a forward P/E ratio of around 20.9 on expected EPS growth of 66%, Broadcom's PEG ratio is only 0.38. This indicator strongly suggests that stocks are significantly undervalued relative to their tremendous growth. Add to that 15 consecutive years of dividend growth , and the situation looks pretty attractive.
But on the downside , we have the same extreme concentration on six clients that we just talked about. We have a downward trajectory in gross margin due to expensive memory components. And importantly, management noted that capacity constraints at TSMC, their primary manufacturer, have held back the supply chain in 2026, and this may delay the recognition of some of this revenue.
We should also stop and discuss this interesting dynamic in funding news in October. Broadcom has agreed to provide Anthropic with a loan of up to $42 billion to fund their infrastructure costs. Think about that for a second. These are, in essence, mutual costs . Broadcom funds the very company that buys its products.
This creates incredible short-term revenue and completely cements Anthropic as a key customer. But this creates huge contingent liabilities. Analysts at Simply Wall Street noted that the entire sector could accumulate up to $370 billion in senior debt by mid-2029. So, if the AI spending cycle slows even slightly, this cyclical funding could significantly increase downside risk for Broadcom.
Objectively speaking, we must also consider the geopolitical and regulatory landscape as the situation escalates. In Europe, the EU is stepping up antitrust oversight over Broadcom's complete overhaul of VMware's licensing system . After Broadcom canceled perpetual licenses , some European cloud providers reported staggering price increases.
We are talking about 800–1500%. If regulators step in and force Broadcom to stop bundling these products, it could threaten the high-margin software revenues that are the company's financial foundation . Meanwhile, in China, which accounts for about 17% of Broadcom's total revenue, the state-owned asset regulator is reportedly scrutinizing the use of Broadcom switches in government data centers, where their share currently reaches 90%.
Any serious move towards import substitution there could deal a serious blow to their revenues in the APAC region. So, let's summarize everything in the fifth chapter: scenarios, goals, and verdict. Modeling the future for AVGO. So, looking at this data objectively, we can identify three different paths for these stocks.
The optimistic scenario, to which we give a 30% probability, assumes almost flawless implementation of plans. This is where they hit that huge goal of $230 billion by 2028, which will push earnings per share to $35–$40. This gives us a target price in the $650–$800 range. Our baseline scenario, to which we give a 50% probability, assumes stable and confident implementation of plans.
In this scenario, they achieve about 80% of their ambitious goals. Gross margin stabilizes at around 73%, and no major regulatory shocks disrupt the business. This gives us a target of $500–550. But we must definitely consider a pessimistic scenario with a probability of 20%. If AI capex starts to slow, if this cyclical funding model collapses, or if the EU increases pressure on VMware, the stock could fall significantly to the target level of around $215–$290.
When we combine these probability-weighted scenarios, we get an estimated average 12- month price target of $494. This is about 31% higher than the closing price as of October 6, which was $375.81. Based solely on financials, extremely high multi-year contract predictability, and a low PEG ratio , the data supports a Buy recommendation for Broadcom.
However, this is provided, of course, that investors correctly assess their positions, taking into account the very real risks of client concentration and regulation that we have discussed. So, in concluding this review, the main question you as an investor should ask yourself is this: does the broader market really understand the magnitude of this multi-year shift to specialized chips?
And how reliable are these hyperscaler contracts really? Or perhaps the short-term hype, debt costs, and this staggering 221% growth have already outpaced realistic calculations? Be sure to pay close attention to gross margin in the next earnings report, because honestly, that will be the truest indicator of Broadcom's continued ability in this area. Thank you for being with us.
Comments 0