The AI bull market has entered the "realization era"! Morgan Stanley and JP Morgan both predict the S&P 8000 points, while the semiconductor sector and the South Korean stock market are violently rebounding, confirming the "main trend of profit growth."

Since August, against the backdrop of a significant rebound in technology stocks led by the semiconductor sector and the broader theme of AI computing infrastructure, the extreme volatility in the global stock market is rapidly dissipating. Recently, two major financial giants on Wall StreetMorgan Stanley and JPMorganhave released research reports stating that the primary driver for the continued upward movement of the S&P 500 is shifting from valuation expansion to earnings upgrades coupled with the commercialization of AI. JPMorgan raised its year-end 2026 target from 7,800 points to 8,000 points last week and adjusted upward its earnings per share (EPS) trajectory for this year and next. Likewise, Morgan Stanley also raised its 2026 target to 8,000 points and its 12-month target to 8,300 points, clearly stating that the upward adjustments are mainly due to earnings rather than valuation. At least seven Wall Street firms now expect the S&P 500 to reach 8,000 points by the end of 2026. As the earnings expectation-driven S&P 500 index approaches 8,000 points, it is becoming a new bullish consensus, resonating strongly with the price behavior following the de-leveraging of the AI computing infrastructure theme in July: the Philadelphia Semiconductor Index experienced a nearly 29% drop from its June 22 high to its July 29 low, before quickly rebounding around 20% from the low; the South Korean benchmark index KOSPI rebounded nearly 22% just about two weeks after its July 30 low, entering a technical bull market. The July drop in South Korea included significant leverage from ETFs and forced liquidation factors, with the scale of leveraged products declining from about $50 billion to $17 billion; however, the fundamentals of AI Memory (the super bull market in AI-driven memory chips) did not collapse simultaneously, with the industry still discussing tight supply of DRAM/HBM and a demand gap projected for 2027. Thus, this round of AI-driven bull market seems increasingly aligned with the positive feedback loop of de-leveragingmarket position rebalancingrisk re-assumptionFOMO sentiment heating up, rather than a dead cat bounce after the peak of an AI profit cycle. Another Wall Street financial giant, Citadel, provided evidence of capital flow indicating that this rebound has further transitioned from fundamental repair to a self-reinforcing buying phase. Its official August report indicated that the S&P 500's second quarter EPS growth rate was about 33%, and the earnings upgrade path is among the steepest since at least 2000; meanwhile, the index hit record highs, but the 12-month forward P/E ratio fell from about 23.1 times in October last year to 20.1 times, showing that it is primarily earnings expansion rather than multiple expansion that is driving the index higher. From buying shovels to who uses shovels to dig out profits: the ultimate metrics for AI investment have turned into ROIC and free cash flow. JPMorgan has raised its 2026 S&P 500 index target from 7,800 points to 8,000 points, citing an unusually strong performance during the second quarter earnings season, with increasing evidence that large-scale AI investments are translating into stronger business performance. The bank also raised its earnings forecasts for 2026 and 2027; however, persistently high interest rates, geopolitical risks, and the large new supply in capital markets continue to limit its assumptions on valuation multiples. With 87% of S&P 500 constituents having reported earnings, JPMorgan's strategists led by Dubravko Lakos-Bujas noted that the earnings outlook remains robust, and this strength is widely distributed across multiple sectors. This strong performance has prompted the bank to raise its 2026 EPS forecast to $365, suggesting a 35% increase compared to the previous year. This forecast also exceeds the current market consensus expectation of $358. For 2027, JPMorgan raised its EPS forecast to $420, which indicates a further 15% increase. A portion of the exceptionally strong earnings growth comes from the rising private investment value held by listed companies. JPMorgan's strategist team projects that according to valuation adjustments recorded in the first half of 2026, these valuation changes contributed about $18 to the S&P 500 index EPS. Excluding this effect, the normalized EPS for 2026 is approximately $347. Even based on this adjusted figure, the year-on-year earnings growth still reaches about 28%, underscoring the underlying strength of AI-driven corporate profitability. One of the most significant changes in this quarters earnings season is that the market discourse around AI investment in hyperscale cloud computing companies is undergoing a shift. Investors are increasingly shifting their focus from the scale of AI capital expenditures to whether these expenditures can generate attractive returns on invested capital (i.e., ROIC). JPMorgan believes that the latest results have provided encouraging evidence indicating that commercial monetization is beginning to take shape. According to JPMorgan, some of the strongest examples come from Google, Amazon, and Microsoft, whose stronger cloud business growth, backlog expansion, and improved visibility of operating cash flow successfully surpassed the high expectations set by investors. JPMorgan stressed that the growth rates of cloud businesses among North American tech giants provide some of the clearest evidence that AI investments are translating into stronger customer demand. Amazon's cloud computing division, AWS, has accelerated its year-on-year revenue growth to 37%, Microsoft's Azure cloud computing business grew by 43%, and Google's cloud computing division, Google Cloud, showed the strongest performance with a record revenue increase of 82%. Backlogs also expanded significantly. Google Cloud's backlog increased by $52 billion from the previous quarter to $514 billion. AWS's backlog reached $496 billion, growing 36% quarter-on-quarter and nearly 2.5 times the level a year ago. As these hyperscale cloud giants continue to invest heavily in new capacity, these figures provide substantial visibility for their future revenues and profits. This is also why JPMorgan can still raise its index profit forecasts even as AI capital expenditures are headed toward approximately $900 billion in 2026 and over $1.2 trillion in 2027; it is essentially betting on the business loop of AI CapExcloud revenuebacklogoperating profit, rather than merely betting on GPU shipments. Morgan Stanley took this logic further: the next stage of Alpha is spreading from AI infrastructure suppliers to AI adopterscompanies that can truly leverage AI to enhance labor productivity, reduce costs, expand profit margins, and generate cash flow may receive higher valuations than those merely possessing the AI concept. The strategist team at Morgan Stanley, led by Michael Wilson, believes that investors are also becoming more selective, increasingly willing to reward companies that can combine earnings growth with strong free cash flow and operational efficiency. This is also why Morgan Stanley favors hyperscalers (i.e., cloud computing super giants) over merely AI semiconductor or computing-related stocks in the medium to long term. Chip stocks may still return to a tactical upward trajectory after the momentum liquidation in July, but cloud computing giants possess a combination of existing cash cow businesses, AI infrastructure, models/platforms, customer distribution channels, and future AI monetization options, leading to a more favorable risk-reward profile. Morgan Stanley has recently emphasized that the market is transitioning from the early cycle of a bull market into the mid-cycle, with the leadership weight of the bull market shifting from purely high beta to earnings quality + cash flow. The bull market has not departed from technology; instead, it is upgrading from technology solo dance to profit diffusion: high quality is the truly scarce asset in the next phase. JPMorgan provides the most crucial bullish evidence for the current AI bull market: the high AI-related revenue growth from AWS, Azure, and Google Cloud, along with hundreds of billions in backlog, starts to prove that AI capital expenditures are not merely a cost black hole but are transitioning toward visible revenues. This is also why the bank can raise index profit forecasts even as AI capex approaches approximately $900 billion by 2026 and exceeds $1.2 trillion by 2027. The two Wall Street giants are not actually presenting contradictory strategiesJPMorgan continues to confirm that the AI super cycle is the engine of index profits, while Morgan Stanley informs investors how the next stage of the bull market should diffuse. With 87% of S&P 500 companies exceeding earnings expectations, the median earnings growth rate for the Russell 3000 increasing to 15%, and the earnings revisions improving in the financial and consumer sectors, it suggests that the market is beginning to see conditions for diffusion from Mega-cap AI (the tech giants closely associated with AI) to higher-quality financials, consumption, AI applications, and more broadly, profit growth companies. The most critical selection criteria are no longer just growth but whether growth can turn into cash: companies that see simultaneous upgrades in EPS and free cash flow are receiving significantly higher returns, while firms with EPS growth but deteriorating cash flow are beginning to be penalized by the market. From a multi-month investment horizon, Morgan Stanley believes that hyperscale cloud giants possess more attractive risk-return characteristics compared to AI semiconductor forces. The institution's strategists emphasized that these companies have resilient core businesses, attractive relative valuations, and market awareness of the investment returns related to AI computing infrastructure and the options brought by AI adoption that is still insufficient. This combination means that hyperscale cloud companies can not only continue to benefit from ongoing growth in cloud business but may also reap ongoing improvements in returns related to their massive AI investments. As JPMorgan and Morgan Stanley both focus on quality amid the diffusion of market leadership and upward momentum, this expansion of the earnings recovery range increasingly assures Morgan Stanley's strategists that market opportunities are no longer confined to a few mega-cap stocks. With 87% of S&P 500 constituents exceeding earnings expectations, the median earnings growth rate for Russell 3000 reaching 15%, the strongest growth since 2021, and earnings forecasts improving across various sectors, the overall fundamental environment has significantly strengthened. Meanwhile, Morgan Stanley's statistical data shows that the breadth of earnings expectation revisions for the S&P 500 has returned to 23%, and 76% of industry groups are recording positive earnings revisions, both of which are nearing their respective cyclical highs. However, investors are becoming increasingly discerning about the specific quality of growth. Hence, Morgan Stanley prefers high-quality companies with strong free cash flow, AI adopters, large financial stocks, and discretionary consumer goods companies. Within the tech sector, hyperscale cloud giants still outperform semiconductor stocks from a longer-term investment horizon. The most significant downside risks are also very clearlong-term U.S. Treasury yields, oil prices, and the cost of capital market financing; JPMorgan is reluctant to raise the valuation multiple above 20 times, and Morgan Stanley continues to warn about long-term yields, essentially indicating that if this bull market continues to soar, it will be primarily driven by earnings expansion rather than re-relying on multiple expansion.