AI trading enters "shift period"! Wedbush and Goldman Sachs both speak up: the chip sector is too crowded, and tech giants see a good opportunity for layout.

Including Microsoft Corporation (MSFT.US), NVIDIA Corporation (NVDA.US), Meta Platforms (META.US), and Palantir (PLTR.US), many tech stocks have recently experienced significant declines. In response to this, investment bank Wedbush has described the current market as entering the "Twilight Zone" - a confusing market environment that continues to play out, but also highlights emerging opportunities. The Goldman Sachs Group, Inc. also stated that amid the recent intense volatility in chip stocks, large-cap tech stocks may be becoming more attractive as investment themes in artificial intelligence (AI). Wedbush: Tech stocks in a "Twilight Zone," but opportunities are emerging Although concerns about rising capital expenditure and the potential threshold of computing power and storage costs are driving the current pullback, Wedbush believes that the market will provide new opportunities for patient investors. Upon closer examination, Wedbush stated that the current market is in a "temporary air pocket stage." This year, tech giants have spent up to $700 billion on AI infrastructure development, but investors in large tech companies - especially shareholders of Microsoft Corporation and Meta - are becoming increasingly frustrated. In a report to clients, Wedbush wrote, "Essentially, we are in a 6 to 12-month window where data center and computing infrastructure development is rapidly advancing. However, large tech giants such as Microsoft Corporation, Meta, and Amazon.com, Inc. (AMZN.US) and Alphabet Inc. Class C (GOOGL.US) are still in a waiting period, hoping to see a real wave of commercialization." Further analysis by Wedbush indicated that this issue has affected all hyperscale cloud service providers, including Alphabet Inc. Class C. Just a few weeks ago, Alphabet Inc. Class C was considered the "golden child" in this group. However, recent departures of several AI researchers from the company have further exacerbated investor concerns. Meanwhile, Meta may be in a key period of business transformation, but the company led by Mark Zuckerberg is investing significant funds, and investors seem unwilling to continue waiting for the ultimate realization of this transformation. Apple Inc. announced a significant increase in product prices this week, which may have caused a "negative impact" on the market. Investors are beginning to question whether computing power and storage costs have become too high to sustain and whether AI infrastructure development needs to slow down. However, Wedbush explained, "We believe that these costs will start to gradually decline over the next year. As a massive wave of consumer-grade AI hardware, physical AI deployment, and enterprise AI applications erupt in the coming years, these concerns will become distant memories, much like the construction of the Las Vegas Strip in the 1950s, which seemed costly at the time but ultimately created long-term value." Wedbush concluded, "Overall, many trading days may make the current tech market seem filled with various inexplicable trends. However, in our view, this creates mismatches and opportunities that allow investors to position themselves as true winners in this ongoing tech and AI bull market with long-term growth potential." Goldman Sachs Group, Inc.: Investors should position themselves in tech giants amid chip stock volatility Christian Mueler-Grissman, strategist and head of asset allocation research at Goldman Sachs Group, Inc., also shares a similar view. He stated that amid the recent intense volatility in chip stocks, large-cap tech stocks may be becoming more attractive as investment themes in AI. Mueler-Grissman pointed out that while chip manufacturers and beneficiaries of AI capital expenditure have been the primary drivers of market growth, rather than hyperscale cloud computing companies, chip stocks are actually the most volatile part of the entire AI industry chain. This sector not only has highly concentrated investor positions but also has accumulated significant leverage through tools such as exchange-traded funds (ETFs) and options. Mueler-Grissman said, "If you still believe that the trend of AI development will continue to be positive, then you should allocate investments appropriately to hyperscale cloud computing companies and perhaps reduce the proportion of chip sectors in your portfolio, as the chip sector is the most volatile part of the entire AI capital expenditure industry chain." Over the past few months, chip stocks have been the most popular AI investment direction in the market, with the Philadelphia Semiconductor Index (SOX) gaining about 150% in the past year. In contrast, large-cap cloud computing companies including Amazon.com, Inc., Oracle Corporation (ORCL.US), Microsoft Corporation, Alphabet Inc. Class C, and Meta have lagged behind, as investors have questioned the massive capital expenditure in data centers by these companies. Reflecting on the significant rise in chip stocks, Mueler-Grissman said, "To some extent, reducing allocations to this sector, diversifying investments, or rebalancing may be a reasonable choice." From a broader market perspective, Mueler-Grissman stated that Goldman Sachs Group, Inc.'s risk appetite indicator continues to rise recently, initially driven by the improvement in corporate profits driven by AI capital expenditure and further boosted by the reopening of the Strait of Hormuz. He said this creates what is called a "golden-haired girl range" - where inflation expectations decrease, while corporate profit growth remains relatively strong. However, he emphasized that even if investor sentiment remains bullish, it does not mean that investors should automatically turn bearish on the market at this point, despite indicators showing a clear increase in market risk appetite. Mueler-Grissman added, "When everyone is bullish, it doesn't mean you should automatically turn bearish. But if the market starts to question the factors driving the current uptrend, there is indeed a risk of correction and adjustment in the market." Star analysts during the Internet bubble period advised: to divest in cloud giants and focus on the semiconductor sector However, Dan Niles, a well-known semiconductor analyst and founder of Niles Investment Management, warned investors that AI trading is about to hit a "speed bump." He believes that smart money should flow into areas where actual AI capital expenditures are made - the chip industry, instead of the hyperscale cloud computing companies that are pouring in huge amounts of money. Niles expressed his concerns and pointed out that there is a significant shift happening in enterprise AI strategies. Just a few months ago, companies were focused on maximizing "Tokens," encouraging employees to generate as many AI Tokens as possible. Now, the focus has shifted to "Token minimization," as companies are realizing that they cannot exhaust the entire year's AI budget in just a few months. He said, "Just like Uber Technologies, Inc., you can't spend the entire AI budget in four months and not run into issues when reporting earnings and issuing guidance." Looking ahead, Niles warned that due to the previous push for "Token maximization," quarter performance estimates for June are expected to be quite strong. However, as companies begin to route more requests to lower-cost open-source models - some of which cost only one-eighth of high-end models - performance guidance for the September quarter may face pressure. He said, "My question is, if you are shifting tasks to cheaper models, what will your September performance guidance look like?" "This is where I believe these companies may hit a 'speed bump' when reporting earnings and issuing future guidance." When discussing his investment portfolio strategy, Niles said that he is gradually reducing his exposure to hyperscale cloud computing companies while also beginning to trim some semiconductor holdings - even though the semiconductor sector has been performing well. He noted that "the semiconductor index has doubled," but also emphasized that as the logic behind AI-related investments continues to evolve, investors need to become "more selective." Niles's views reflect a reality where as more investors begin to question the rationality of the massive capital expenditure by hyperscale cloud service providers in AI infrastructure expansion, the threshold for AI trading is rising. The market is demanding visible orders, cash flow, and profit realization, as mere long-term potential is no longer enough. This means that investors will turn to areas such as the semiconductor industry that benefit from the wave of AI capital expenditures. Investment firm Columbia Threadneedle Investments also holds a similar view. The firm believes that with the continued acceleration of investment in AI infrastructure, the uptrend in tech stocks is likely to continue for at least the next few quarters, with the revenue and profit expectations of related companies continuously being revised upward. However, as the AI race enters a new stage, changes in internal fund flows and competitive landscape are occurring, with investors starting to focus more on which companies can truly benefit from the wave of AI capital expenditure. Tiffany Wade, Senior Portfolio Manager at Columbia Threadneedle Investments, stated in an interview on Monday that companies related to AI infrastructure are still in a strong growth cycle, and market expectations for their revenue and profit prospects continue to improve. "The prospects of technology companies related to AI infrastructure spending are very promising, and their revenue and profit forecasts are constantly being revised upward."