A revenue of 150 billion marks a "watershed" for the AI business! Morgan Stanley maintains an "overweight" rating on Broadcom Inc. (AVGO.US) and warns that success or failure depends entirely on expectations.

It has been noted that Broadcom Inc. (AVGO.US) will announce its fiscal year 2026 Q3 earnings after the U.S. stock market closes on September 2 (early morning September 3 Beijing time). The performance data is almost certain to be impressive. Morgan Stanley has explicitly pointed out that the real suspense lies in whether "impressive" is enough. This summarizes the situation before entering the third quarter. The key question is not Can Broadcom Inc. deliver good results? but whether it can surpass the high threshold set privately by some investorsraising the fiscal year 2027 AI revenue expectations to exceed $150 billion. In contrast, Morgan Stanley's forecast is $120 billion. The risk of short-term volatility lies precisely in the gap between these two numbers. Remember, the experience of the second quarter serves as a caution: when market expectations are excessively high, even stellar performance may not satisfy investors. Morgan Stanley maintains an "Overweight" rating on Broadcom Inc. However, the title of this research report is particularly noteworthy: "The main risk before the earnings release lies in market expectations, not fundamentals." The report indicates that Morgan Stanley's forecast for the quarter ending in July is very detailed and broadly aligned with Wall Street's expectations: Revenue is projected at $29.4 billion, a year-over-year increase of 84.3% and a quarter-over-quarter increase of 32.5%. AI-specific revenue is projected to be $16 billion (quarter-over-quarter growth of 48%), including $10.8 billion from custom ASIC business and $5.2 billion from AI networking business. Gross margin is projected at 74.0%, slightly above Wall Street's average expectation of 73.5%. Earnings per share are projected at $3.24, slightly above the consensus expectation of $3.22. By any historical standard, this is a remarkable quarter. But Morgan Stanley believes investors need some more time to look at the following set of data. Broadcom Inc. recorded $10.8 billion in AI chip revenue in Q2 of fiscal year 2026, a year-over-year increase of 143%. According to Broadcom Inc.'s Q2 earnings report, CEO Hock Tan provided guidance for Q3 indicating that AI revenue will "grow more than 200% year-over-year, reaching $16 billion." Therefore, the $16 billion in AI revenue for Q3 is not a surprise; it is merely the performance baseline set by management. The report indicates that for the guidance of the fourth quarter ending in October, Morgan Stanley forecasts revenue to reach $34.8 billion (a year-over-year increase of 93.4%), with AI revenue accelerating further with a quarter-over-quarter growth of 32% to reach $21.2 billion. This number is where potential surprises beyond expectations may actually materialize. The dispute regarding Broadcom Inc.'s valuation in 2027 There is another more important point of discussion that is completely unrelated to this quarter's performance. Broadcom Inc. had previously set the guidance for fiscal year 2027 AI revenue at "far exceeding" $100 billion. In the last quarter, management demonstrated increasing confidence and signaled that this growth momentum would continue "deep into 2028." The report shows that Morgan Stanley forecasts AI revenue for fiscal year 2027 to be about $120 billion. However, some investors' expectations have gradually climbed to $150 billion or even higher. The divergence between $120 billion and $150 billion is not just a debate over forecasting methods, but also a divergence in valuation logic. If it is $120 billion, Broadcom Inc. appears reasonably valued at its current stock price level; but if it is $150 billion, then the stock appears very cheap. If the guidance provided on September 2 implies an expectation of $120 billion, no matter how good the actual Q3 performance is, investors expecting $150 billion will feel disappointed. Even Morgan Stanley mentioned this dynamic relationship in the report: "The underlying potential business may continue to perform exceptionally well, but it may not necessarily exceed the most aggressive expectations." This is a politely worded warning, reminding investors to be aware of the downside risks in this situation. The TPU supplier rivalry Another pending question worth addressing is whether Alphabet Inc. Class C is diversifying its custom chip suppliers away from Broadcom Inc. Recent reports indicate that MediaTek is involved in the Tensor Processing Unit (TPU) project, AMD is participating in TPU v10, and Marvell Technology, Inc. has entered into a warrant agreement with Alphabet Inc. Class C. These are seen as signals that this hyperscale cloud computing giant is expanding its ecosystem. According to the report, Morgan Stanley's position remains unchanged. There is indeed supplier diversification, but Broadcom Inc.'s first-mover advantage and existing advantages are extremely solid. The firm estimates that even if Alphabet Inc. Class C introduces other potential suppliers, Broadcom Inc. will still retain about 80% of its long-term TPU market opportunity. The framework provided by the analysts is: "If anything, the fact that multiple semiconductor companies are competing around TPU highlights the vast scale of this market opportunity." This is a very important shift in perspective. When multiple leading chip companies are vying for the share of the same custom chip project for a client, it does not mean this market space is shrinking; rather, it indicates that this market is large enough that everyone wants a piece of the pie. The report shows that in Morgan Stanleys ranking of AI computing stocks, Broadcom Inc. (AVGO.US) ranks second, only behind its top pick NVIDIA Corporation (NVDA.US). The significance of Q3 earnings goes beyond Broadcom Inc.'s industry background There is another background worth noting. Based on FactSet data as of August 28, 2026, the semiconductor industry has just reported a 142% year-over-year profit growth in Q2 of 2026. In fact, this sector is the largest contributor to the overall profit growth of the IT industry. If the semiconductor sector is removed from the computation of the IT industry, the overall profit growth rate of the IT industry would plummet from 75.3% to 38.3%. The dominant position of the chip industry cycle is thus evident. Broadcom Inc. is at the core of this cycle. According to data up to the end of August, Broadcom Inc.s stock price has increased by 7.40% so far this year, while the Standard & Poors 500 index has increased by 12.28% during the same period. Over the past year, the stock has generated a return of 25.44%, and a return of 316.02% over the past three years. Its performance this year has lagged behind the market, reflecting the sell-off that occurred following the release of the second quarters surprisingly strong earnings report. This perfectly exemplifies the "expectation risk" that Morgan Stanley warned about again before the Q3 earnings report. The business is exceptionally strong, and the thresholds are extremely high. Both points can hold true simultaneously. On September 2, investors will know which side prevails.