Global investors are facing yet another incredibly crucial earnings season for American large-cap technology companies, as these tech giants are set to disclose their earnings reports intensively this week, with stock prices near historical highs and valuations also at record levels. A key distinction this time is that the expected quarterly profit growth of the so-called 'seven major tech giants' in the USA is predicted to reach the slowest pace in nearly two years, and many Wall Street analysts have indicated that the threshold for exceeding expectations has been raised. Since the end of 2022, the Nasdaq 100 Index, led by these seven tech giants, has undergone an epic market cap expansion of up to 15 trillion dollars, which undoubtedly faces severe tests.

Moreover, it is noteworthy that as the 'four giants' of the US stock market release their earnings this week, the 'low-cost computing power storm' led by DeepSeek is sweeping across the globe, and investors are beginning to question whether the US tech giants' seemingly 'irrational' fervor and expensive AI spending plans are justified. The DeepSeek R1, created by a team of AI engineers from China, dominated the American hot search rankings last week, and the DeepSeek app topped the free app download charts on Apple's platforms in both China and the USA, surpassing ChatGPT in the US downloads. The emergence of DeepSeek R1 marks a significant reduction in AI training and inference costs, achieving a competitive model compared to OpenAI's O1 with an extremely low investment cost of less than 6 million dollars and a performance significantly lower than H100 and Blackwell's H800 chips.

The reason for DeepSeek's dominance on social media trends in the USA and globally stems from the official release on January 20 of its inference model, DeepSeek-R1. This model was validated last week by several notable figures in the tech industry for its performance in critical areas such as mathematics, programming, and reasoning, able to 'compete' with OpenAI's so-called 'strongest reasoning model in human history' O1, while its API calling cost is reduced by 90%-95%.

The combined performance of DeepSeek’s low-cost, ultra-efficient, and competitive model against O1, results from applying 'extreme engineering' and 'fine-tuning' to every aspect of the large model training process. This includes high-efficiency training and data compression strategies oriented towards extreme engineering, leveraging techniques such as multi-layer attention (MLA), FP8 mixed precision, DualPipe parallel communication, and expert gating (MoE) load balancing. This maximizes hardware resource utilization during training, reduces 'unnecessary computing power waste,' and implements innovative AI training measures like 'reinforcement learning + distillation + specialized data optimization.'

In simple terms, DeepSeek continuously minimizes the 'ineffective consumption' of general computing power through its core principle of 'extreme engineering + parallel optimization + precise data filtering + accurate post-training,' focusing resources on the core modules (attention heads, key operators, RL/distillation fine-tuning, etc.) that can significantly enhance model performance. This demonstrates how 'extreme engineering + post-training distillation + professional data integration + targeted reinforcement' can approach or even exceed the performance of mainstream industry large models under limited GPU resources, presenting a strong challenge to the traditional 'massive spending' model. Thus, DeepSeek maximizes the potential of hardware and algorithms, contrasting sharply with the 'extensive spending' approach of American tech giants over a long period.

Therefore, global investors are eagerly hoping that the substantial investments by the 'four giants' of the US stock market in AI can yield active revenue and profit scales, ultimately leading to overall earnings and profit performance that significantly exceed expectations; otherwise, they will view this 'irrational' AI expenditure, which fails to generate any significant profit despite large investments, as 'completely harming the profits attributable to the common stock of the company,' potentially triggering a wave of sell-offs.

The performance of the four giants is seen as the 'hope of the entire village' for the US stock market.

The 'four giants' of US stocks—Apple (AAPL.US), Microsoft (MSFT.US), Meta (META.US), and Tesla (TSLA.US)—will announce their earnings this week. Their actual performance is critical for the US stock market and even the global stock market trend, as these tech giants account for nearly 40% of the weight of the S&P 500 Index and the Nasdaq Composite Index. Their performance is significantly tied to the faith and optimistic outlook of global tech stock investors towards AI. If their earnings generally fall short of market expectations, it could lead investors to question the revenue and profit prospects associated with AI, potentially triggering a global tech stock collapse similar to last summer's.

The 'seven major tech giants' of the US stock market, known as the 'Magnificent Seven,' include: Apple, Microsoft, Google, Tesla, NVIDIA, Amazon, and Meta Platforms. They are the core driving force behind the continuous record highs of the S&P 500 Index. Since 2023, the 'seven tech giants' have been the leading force in the entire US stock market, attracting massive global capital due to their strong revenue driven by AI investments, solid fundamentals, sustained robust free cash flow reserves over the years, and expanding share buyback scales. Google (GOOGL.US), NVIDIA (NVDA.US), and Amazon (AMZN.US) will announce their earnings in February.

Stocks of large tech companies in the US such as Google, NVIDIA, Meta, and Microsoft have surged significantly at the beginning of the year, especially given the risk appetite and the high revenue and profit expectations resulting from these companies spending billions of dollars on AI services. Their stock prices have outperformed the broader market.

However, the earnings season beginning this week may awaken the bullish investors: although the so-called 'seven giants' are still seeing profit growth, far exceeding other segments of the US stock market, Wall Street expects their growth rate to slow down significantly compared to previous quarters. Ultimately, under the pressure of base effects and continuous substantial investments in AI with still somewhat ambiguous profit prospects, the performance growth rate of these tech giants and the market threshold for exceeding expectations are facing increasing pressure.

Since the end of 2022, the market's incredibly optimistic outlook on the AI profit prospects of tech giants has propelled the total market cap of the Nasdaq 100 Index, known as the 'global technology stock barometer', to expand dramatically by about $15 trillion. However, the recent upward momentum of the so-called 'seven giants' and the Nasdaq 100 Index has notably slowed down, mainly due to market concerns that the AI profit prospects will significantly underperform expectations, hindering profit expansion.

Dan Taylor, Chief Investment Officer at Man Numeric, stated that the threshold for exceeding expectations has been raised, saying: 'This should be a fairly decent earnings season, but the benchmarks have been raised, and they may not meet the high expectations of the market. These giants may find it difficult to achieve outstanding performance like last year, especially considering that valuations have risen significantly.'

The profit growth of the "seven major technology giants" in the USA is slowing down.

The performance disclosures of the "seven giants" will begin after Post-Market Trading on Wednesday, during which Microsoft, Meta Platforms, and Tesla will release their earnings reports as scheduled. Apple will follow with its earnings release on Thursday Eastern Time, while Amazon and Google will announce their earnings next week, and the AI Chip dominator NVIDIA will release its earnings on February 26 Eastern Time.

In the long-term bull market of US Stocks that began more than two years ago, the outstanding profit growth data of NVIDIA and Microsoft leading the "seven giants" and the prosperous landscape surrounding AI have been key drivers of the bull market in the US stock market. During this period, approximately 70% of the S&P 500 Index's gains came from large technology companies; however, due to expectations of profit declines and doubts about when all AI investments will yield meaningful returns, the pace of growth has slowed, with even a brief collapse last summer.

It could be a smaller leap.

Statistics compiled by Bloomberg Intelligence show that Wall Street Analysts expect the profits of the seven major technology giants in the USA to grow by 22% year-over-year in the fourth quarter, marking the smallest increase since the first quarter of 2023. Although this is still significantly higher than the overall profit growth expectation of roughly 8% for the S&P 500 Index constituents, it is far below the 51% increase in the first quarter and marks four consecutive quarters of contraction.

Michael Kasper, an Analyst from Bloomberg Intelligence, believes there are reasons for concern. Given that the weight of the technology sector in the S&P 500 Index is about 10 percentage points higher than its weight in the overall profits of the index, this stock analyst is worried that either earnings growth must exceed expectations or valuations need to decline.

The "low-cost computing power wave" led by DeepSeek has begun to make investors question the reasonableness of expenditures by AI giants in the USA. If these tech giants' massive investments in AI still fail to generate satisfactory revenue and profits for investors or performance data that exceeds market expectations, a "tech stock sell-off wave" larger than last summer's scale may occur. Kasper stated, "We are all very clear about how the actual performance of US tech giants will react if it fails to meet the expectations of everyone."

From the perspective of stock price trends relative to expected sales, the valuations of the US tech giants appear to be even more unstable. According to Bloomberg Intelligence's statistics, the trading price of the IT sector in the S&P 500 Index is nearly eight times the expected sales for the next 12 months, approaching the highest level in at least a decade.

The valuation of the Technology Sector of the S&P 500 Index has soared relative to sales.

However, Solita Marcelle, Chief Investment Officer for the Americas at UBS Group's Global Wealth Management division, believes that these valuations merit continued buying by investors, as investments in AI are expected to generate larger overall revenues in about a year. "While the era of easy profits in the AI sector may be over, we believe that this wave of rising technology stocks is far from over, supported by strong revenue prospects for AI," she wrote in a letter to clients this month.

Compared to the current insignificant AI revenue figures, the cash burn scale of American tech giants in AI has been significantly expanding. Microsoft, Google, Amazon, and Meta are expected to collectively spend over $200 billion on capital expenditures in the previous fiscal year, and they have all committed to significantly increasing spending in the current fiscal year. Therefore, in addition to the growth in AI-related revenues, investors will closely monitor spending expectations, especially after the release of DeepSeek R1, focusing on whether the huge AI expenditures by American tech giants are justified.

There are almost no signs that investors are ready to face the "huge disappointment earnings season" of American tech giants. Among the "Seven Major Tech Giants in the USA," the demand for Put Options (to hedge against downside risk) has been shrinking compared to the demand for Call Options after a surge in December.

So far, bullish investors for the S&P 500 Index and the Nasdaq 100 have reaped high returns. Streaming giant Netflix (NFLX.US), as one of the few tech companies that have reported earnings, significantly boosted the Nasdaq 100 Index last week after its subscriber count recorded a record growth.

“Valuations may continue to expand due to bullish market sentiment, but people may feel disappointed with the monetization prospects of AI, potentially triggering a new wave of sell-offs and liquidations,” said Tyler from Man Numeric. However, he emphasized that large tech giants "are still excellent companies that can generate substantial cash flow" and, with strong cash flows and share buyback scale, will remain a "safe haven" during panic sell-offs.