A 17% two-day blood loss, is the sharp decline of Broadcom Inc. (AVGO.US) a trap or a windfall?

Last Friday, the stock price trend of Broadcom Inc. (AVGO.US) can be described as "tragic": the closing price that day was $359.93, a significant decrease of $46.44 from the previous day, a decrease of 11.43%. By the end of this Monday, its stock price still did not stop falling, dropping nearly 6% and finally closing at $339.81. Is this fierce selling market reasonable? Analyst Michael Fitzsimmons from Seek Alpha delved into this situation. It is reported that Fitzsimmons is not only a retired electrical engineer, but also has many years of experience in the investment field and is a seasoned investor. Over the past few years, Fitzsimmons has always favored Broadcom Inc. for profound reasons. Ultimately, Broadcom Inc. has the industry-leading "AI Triad" advantage: first, an XPU AI-optimized processor designed for ultra-large data centers, which efficiently handles complex computational demands; second, a leading edge network device switch chip in the high-speed network equipment industry, demonstrating excellent network transmission performance; and third, strategic partnership with VMware virtualization and enterprise software, allowing VMware to run full stack on NVIDIA Corporation (NVDA.US) and AMD (AMD.US) GPUs. Additionally, as a retired computer engineer, Fitzsimmons can confidently say that Broadcom Inc. CEO Hock Tan is one of the sharpest and most visionary leaders in the industry; when it comes to "business", he has a long and legendary history of acquisitionsbeing able to make acquired companies more efficient, profitable, and constantly extract free cash flow, VMware being the latest addition to this brilliant acquisition chain. The investment logic behind this is roughly as follows, and next, let us follow Fitzsimmons' perspective and focus directly on the financial results of Broadcom Inc.'s fourth quarter of fiscal year 2025 for analysis. Fourth Quarter Fiscal Year 2025 Earnings In Fitzsimmons' view, Broadcom Inc.'s financial report for the fourth quarter of fiscal year 2025 was another excellent report: quarterly revenue of $18 billion, a 28% year-on-year increase, and earnings per share under the United States Generally Accepted Accounting Principles (GAAP) surged by 93%; semiconductor business revenue was $11 billion, accounting for 61% of total revenue, a 35% year-on-year increase; infrastructure software business revenue was $6.9 billion, accounting for 39%, a 19% year-on-year increase. However, despite these impressive numbers, they were still labeled as "disappointing". In terms of free cash flow and shareholder returns, Broadcom Inc. has shown strong strength. In terms of quarterly data, Broadcom Inc.'s free cash flow reached a high of $7.5 billion, accounting for 41.4% of revenue, a 36% increase compared to the same period last year. Looking at the entire fiscal year 2025, Broadcom Inc.'s free cash flow reached a staggering $26.9 billion, corresponding to total revenue of $63.9 billion, with a high free cash flow profit margin of 42.1%. Compared to competitor AMD, Broadcom Inc. has an overwhelming advantage in cash flow. AMD's free cash flow profit margin for the second quarter was only 15.0%, about one-third of Broadcom Inc.'s 42.7% during the same period; even though AMD's latest third-quarter free cash flow was $1.53 billion, accounting for 16.6% of revenue, showing some progress compared to before, it is still not half of the level demonstrated by Broadcom Inc. this week. It is with this confidence in cash flow that Broadcom Inc. has brought a series of good news to shareholders: a 10% increase in dividends for the first quarter of fiscal year 2026, at $0.65 per share; year-end long-term debt reduced by $4.3 billion, total long-term debt reduced by $8.2 billion; the board extended the stock repurchase plan, with a balance of $7.5 billion, valid until the end of 2026; quarter-end cash and cash equivalents amounted to $16.2 billion, compared to $10.7 billion at the end of the previous fiscal quarter. Once again, despite shining performance, all of this was still described as "disappointing". Broadcom Inc.'s AI Order Boom: Market Debates Behind the $73 Billion Backlog During the financial report conference call, Broadcom Inc. CEO Hock Tan once again made several stunning announcements: he revealed that the company had once again secured orders totaling up to $11 billion from the same important customer, Anthropic, with the order expected to be delivered by the end of 2026; at the same time, he also unveiled cooperation information with the fifth XPU customer, with an order amount of $1 billion, also scheduled for delivery by the end of 2026. It is worth noting that Anthropic's $110 billion order is specifically tailored for Alphabet Inc. Class C's Ironwood TPU, which will be delivered in the form of a complete cabinet, and will come with network equipment and interconnection technology, effectively providing a complete "system-level" solution. This leads to a key fact: Alphabet Inc. Class C (GOOGL.US) is Broadcom Inc.'s largest AI customer, and its latest Ironwood TPU is not only high-performing, but no longer solely for internal use, thereby bringing a continuous stream of orders to Broadcom Inc. Meta (META.US) and Anthropic have already placed orders, and Fitzsimmons believes that more customers will follow. Reading this, do you have any thoughts about selling stocks? Fitzsimmons certainly does not. So, is the culprit causing the market to go wild and sell-off really the backlog of orders? The answer remains: wrong, for the following reasons. In his opening remarks during the Q&A session, Hock Tan said the following: "Fitzsimmons' current AI switch orders backlog has exceeded $10 billion. The recently launched 102 Tb/s Tomahawk 6 switch chip, which is the only product in the industry with this performance, is still setting record speeds in taking orders; and this is just the tip of the iceberg, Fitzsimmons has also secured record orders for optical components such as DSPs, lasers, and PCI Express switches, all used for AI data centers. Coupled with XPU, the total value of orders on hand now exceeds $73 billion, almost half of Broadcom Inc.'s $162 billion backlog, Fitzsimmons expects this $73 billion in AI orders to be delivered within the next 18 months." It still sounds very impressive, so where is the problem? It turns out that the high $73 billion backlog of AI orders at Broadcom Inc. unexpectedly became the target of everyone's criticism. Just because this figure did not meet the expectations of some people, analysts were eager to compare Broadcom Inc. to Oracle Corporation's earlier performance failure and stock price crash, as if the two were in the same predicament. During the conference call, analysts seemed to have a deep obsession with this $73 billion order data, selectively ignoring the $162 billion backlog of orders at Broadcom Inc.; their first reaction was "this $73 billion is too low", as if the story had already been concluded. However, the reality is far from what they assert. Many analysts and investors have made mistakes Firstly, Fitzsimmons has held shares of Broadcom Inc. since mid-2019 and has listened to all of Hock Tan's conference calls without fail. He made it clear to everyone that he would never provide overly optimistic guidance, and would not easily offer a full-year revenue forecast for 2026; he always practices "low guidance, high delivery", and his financial forecasts are extremely conservative. Moreover, in the rapidly changing artificial intelligence trend, which competent analyst would force him to provide an accurate guidance for an entire 12 months? It is completely unrealistic. Therefore, the first mistake is not understanding the CEO's conservative style of guidance, leading to a complete misinterpretation of the AI backlog. Secondly, this $73 billion is only a snapshot of orders at the end of fiscal year 2025 and does not mean that Broadcom Inc. will stop taking new orders; Hock Tan repeatedly emphasized in the call that this $73 billion is just the "minimum" AI revenue for the next 6 quarters "You could say that $73 billion is the backlog you have today, which will be delivered over the next 6 quarters; given Fitzsimmons' delivery schedule, more orders will be included in this 6-quarter delivery window. Therefore, from one perspective, this $73 billion is the minimum revenue for 6 quarters, but with new orders coming in, the actual figure will be much higher." Since Broadcom Inc. will continue to take orders, and a "large number" of orders at that, how does Hock Tan describe the current order flow? "As the first and only product of its kind, its order flow is still setting record speeds." He added: "I fully expect this $73 billion to continue to grow over the next 18 months, with the numbers increasing over time. I find it difficult to paint a full picture for 2026, so I prefer not to provide an annual guidance, only Q1, and then later on Q2; you ask if this is an acceleration trend? My answer is that as we move forward in 2026, this is likely to be an accelerator trend." A striking Q4 financial report, combined with the CEO's prediction of an "accelerated growth" in orders, has become a reason for the sell-off? In Fitzsimmons' view, this is purely emotional and absurd logic. What really surprised Fitzsimmons was hearing Broadcom Inc. being compared to Oracle Corporationit was simply absurd. Firstly, Broadcom Inc. has strong free cash flow, while Oracle Corporation has been in cash deficit for three consecutive quarters, with a negative free cash flow of $13.18 billion in the latest second quarter of fiscal year 2026; secondly, Broadcom Inc. is reducing debts, and its debt is not for the purpose of building data centers like Oracle Corporation, but for acquiring VMwarean acquisition strategy that not only expands its high-margin software division, but also quickly generates synergistic cash flows. Comparing Broadcom Inc. to Oracle Corporation is unwarranted, illogical, and irresponsible. Valuation In terms of valuation, even after being bloodied on Friday last week, Broadcom Inc. still has a significant premium over the S&P 500, with a trailing twelve-month price-to-earnings ratio of 30.8 times, but this is precisely the truth that Charlie Munger warned Warren Buffett about: if you want to buy a high-quality growth company, you have to be willing to pay a premium; furthermore, Broadcom Inc.'s forward price-to-earnings ratio is not as exaggerated, with the high TTM primarily due to one-time integration costs related to the VMware acquisition, which have now passed. The market continues to severely overlook its free cash flowover the past year, Broadcom Inc. has generated approximately $6.10 of free cash flow per share, compared to the new year dividend of $2.60, demonstrating that the strong cash flow is sufficient to sustain shareholder returns for many years. So, with a company that has a 28% annual revenue growth, a free cash flow of $26.9 billion with a year-on-year increase of 39%, a long-term average dividend growth of 10%, a significant reduction in net debt, year-end cash of $16.2 billion, a total backlog of $162 billion (2.5 times fiscal year 2025 revenue of $638.9 billion), and an order flow that is still "accelerating", what valuation should be given to it? In the rapidly changing AI race, this is certainly a highly rhetorical question, but considering the growth indicators mentioned above, as well as the seasoned and insightful leadership of Hock Tan, and Broadcom Inc.'s "AI Triad" product suite, Fitzsimmons believes it belongs to the top tier along with NVIDIA Corporation and Alphabet Inc. Class C, with the current price and valuation seemingly like a bargain. After all, Broadcom Inc. can rise to the top ten of the S&P 500 market cap, surpassing Berkshire and Tesla, Inc., and this is not without reason. Fitzsimmons has just finished reading the latest report from Morningstar analyst William Kerwin, who described the AI revenue growth as "impressive and accelerating", and raised Broadcom Inc.'s fair value from $365 to $480, a significant rare increase for Morningstar. Remember, under unchanged conditions, a company with a 15% annual growth rate can double its stock price in five years, and Broadcom Inc. has just delivered a 28% annual growth, with a high probability of exceeding this in fiscal year 2026. Profit Margin Risk As for profit margin risk, when Hock Tan was asked whether the lower profit margins of AI TPUs and XPUs compared to the semiconductor average would drag down overall profits, CFO Kristen Spears had already provided the answer in her opening remarks: a quarterly gross margin of 77.9%, higher than the initial guidance, benefiting from increased software revenue and optimization of the semiconductor product portfolio; VMware has a higher gross margin than the semiconductor division and is growing rapidly the infrastructure software division gross margin was 93%, a full two percentage points higher than the previous year's 91%. She also said: "Although the rotation of the semiconductor product portfolio led to a 50 basis-point decrease in gross margin, positive operational leverage continued to drive the operating profit margin up by 70 basis points to 66.2%." With debts continuing to decrease and interest expenses declining, operational leverage will continue to strengthen; in fact, the operating profit margin in the semiconductor division is at 59%, an increase of 250 basis points year-on-year. When responding to Bernstein analyst Stacy Rasgon, Hock Tan directly addressed the question: "Of course, the AI revenue gross margin is lower than the software business, but Fitzsimmons expects the fast AI revenue growth to generate operational leverage in operating expenses, leading to a high absolute operating profit margin; even if the gross margin decreases, operational leverage will still benefit the operating profit level." This has been repeatedly emphasized in several previous conference calls and the actual financial results prove that they predicted the profit margin evolution under the rapid growth of AI revenue very accuratelyoperational leverage is indeed improving. Conclusion In summary, Fitzsimmons believes that analysts and investors misinterpreted Broadcom Inc.'s financial report this week in two major ways: firstly, not understanding Hock Tan's consistent conservative guidance style, and thus completely misinterpreting the AI backlog; secondly, comparing Broadcom Inc. to Oracle Corporation is baseless and in Fitzsimmons's view, irresponsible. However, the stock price crash last Friday was a big gift for investors who had not yet entered the market; Fitzsimmons once again bought around $360 and $350 that day, although his original position was already full, buying the dip in Broadcom Inc. has always been very rewarding, and Fitzsimmons seriously doubts that this time will be any different. That being said, the entire market was visibly shocked last Friday, not just Broadcom Inc.: the S&P 500 fell by 1.1%, the Nasdaq 100 fell by 1.7%, the triple-short Nasdaq ETF fell by 1.9%, and NVIDIA Corporation also fell by 3.3%; if this week sees another sell-off due to panic, profit-taking, and year-end tax planning, it would not be surprising. Before the fireworks of this week, Broadcom Inc. had been the top-performing large tech stock of the year, far ahead. Finally, here is a comparison of total returns over the past year: Broadcom Inc. still dominates and is far ahead of second-place Alphabet Inc. Class CAlphabet Inc. Class C is another favorite of Fitzsimmons, partly due to its deep collaboration with Broadcom Inc., among many other reasons.