According to Woofun AI, JPMorgan Chase (JPM.US) revealed a significant split in the return on investment in the field of artificial intelligence: infrastructure giants Google (GOOGL.US) and Amazon (AMZN.US) received an optimistic revaluation, while application-layer and platform companies Meta (META.US) and Fiverr (FVRRR.US) faced a decline in profits or a traffic crisis.
This structural difference shows that the commercialization process of AI technology is not evenly distributed, but rather profoundly affects revenue structure, cost efficiency, and cash flow performance under different business models.
The growth logic of Google Cloud (GOOGL.US) is undergoing restructuring, and the core driving force stems from the inclusion of self-developed chip TPU sales. J.P. Morgan Chase (JPM.US) raised Google Cloud's (GOOGL.US) revenue growth forecast for the third quarter of 2026 sharply from 88% to 100%, and from 90% to 105% in the fourth quarter. More significant adjustments are reflected in medium- to long-term forecasts: full-year growth expectations jumped from 70% to 107% in 2027, and increased from 55% to 63% in 2028. These figures are all agency predictions, and the company has yet to announce results. JPMorgan Chase (JPM.US) pointed out that the main reason for the increase was the inclusion of more Tensor Processing Unit (TPU) sales in the model, and believes that the adjusted forecast may still be conservative. TPU is a special chip designed by Google (GOOGL.US) for machine learning calculations, and its sales growth is becoming an important variable for the bank to evaluate Google Cloud (GOOGL.US) revenue.
This means that understanding the growth of Google Cloud (GOOGL.US) requires further observation of the revenue composition. The 107% growth forecast includes more TPU sales contributions, and cannot all be attributed to the acceleration of traditional cloud service consumption. The profit margins, delivery pace, and cash recovery characteristics of different businesses also influence how much profit revenue growth can ultimately be converted into. At the group level, JPMorgan Chase (JPM.US) raised Alphabet's (GOOGL.US) 2027 GAAP EPS (GAAP EPS) forecast by about 3% to $14.05, and by about 4% to $17.29 in 2028; raised its target price from $420 to $425 in December 2027, maintaining an 'incremental' rating, corresponding about 25 times the 2028 expected earnings per share.
Cloud business growth expectations have increased dramatically, while the Group's profit forecast has been revised relatively moderately. How much profit elasticity can be brought about by the growth of a single business still depends on its revenue scale, cost structure, and other business performance of the group. Financing was also included in this adjustment. JPMorgan Chase (JPM.US) included the $5 billion debt financing completed by Waymo (GOOGL.US) in early October as described in the report, and assumed an additional $80 billion in debt financing in 2027. The latter is a predictive assumption and cannot be considered an announced or completed financing.
The ability of AWS (AMZN.US) to monetize computing power is also promising. Benefiting from the expansion of OpenAI cooperation, large-scale GPU deployment, and continuous price increases in reserved computing power, JPMorgan Chase (JPM.US) raised AWS (AMZN.US) growth expectations, but emphasized the need to pay attention to delivery implementation and profit margin risks.
According to data compiled by Woofun AI, JPMorgan Chase (JPM.US) expects AWS (AMZN.US) revenue to accelerate to 40% in 2027, and believes this forecast may still be conservative. The judgment has two main supports: first, the expansion of the scale of cooperation with OpenAI and the deployment of an additional 2 million GPUs from 2027 to 2028; the second is the ability to price AI computing power. The additional computing power determines the potential scale of delivery, and the price affects the revenue that can be achieved with the same resources. According to reports, AWS (AMZN.US) has increased the price of GPU reserved computing power by 15% since October 7. This is the fourth consecutive quarter in which the price of reserved computing power for related AI has increased.
The report also indicates that H100/H200 and B200/B300 related prices have increased cumulatively by 40% to 75% since the beginning of the year. Reserved computing power refers to a service arrangement where customers reserve a certain amount of computing capacity in advance. The price increase in the report focuses on specific GPUs and reserved AI computing power, and it cannot be judged from this that all AWS (AMZN.US) products are increasing by the same margin. J.P. Morgan Chase (JPM.US) also emphasized that customer needs have a certain breadth: the computing power distribution of AWS (AMZN.US) is scattered, and the largest customer share is only a single-digit percentage. As a result, its proof of growth includes both the increase brought about by large-scale AI partnerships and the support of a wider range of customer needs.
Together, these factors drove the bank to raise Amazon's (AMZN.US) 2027 and 2028 forecast: Group sales will increase by about 1% to 2%, operating profit will increase by about 3% to 5%, and GAAP earnings per share will increase by about 2% to 3%. J.P. Morgan Chase (JPM.US) maintains a target price of $365 for December 2027 and an “increase in holdings” rating, based on an estimated 2028 earnings of $13.34 per share, which is approximately 27.5 times higher. Revenue and profit improvements support its positive stance, but AWS (AMZN.US) still needs to translate cooperative expansion and deployment plans into actual delivery. The report also listed risks that generative AI monetization is early, visibility is limited, and continued investment could impact profit margins. The increase in prices provided revenue support, and its final profit contribution also depended on equipment, operating and delivery costs.
Compared with the revised forecasts of the two cloud businesses, Meta (META.US)'s current adjustment mainly reflects an increase in expenses. Despite improvements in advertising and AI monetization, huge legal expenses mentioned high capital expenses leading to negative free cash flow, and JPMorgan Chase (JPM.US) lowered its earnings per share forecast, suggesting short-term spending pressure. J.P. Morgan Chase (JPM.US) added a $10 billion legal fee accrual to the 2026 third quarter model. The report attributed this to the settlement reached by Meta (META.US) with 52 state attorneys general, and indicated that the fee was not included in the $165 billion to $169 billion annual cost guidance given in the company's second-quarter earnings report. Affected by this, the bank lowered Meta (META.US)'s 2026 GAAP earnings per share forecast by about 10% and by about 2% in 2027 to $27.99 and $31.08, respectively.
Meanwhile, JPMorgan Chase (JPM.US) raised the total GAAP cost forecast for 2027 to about US$210 billion, an increase of about 18% year over year; if the above $10 billion legal fees were excluded from the 2026 base, the 2027 fee increase would be about 25%. The lower apparent growth rate was partly affected by the increase in one-time costs in the previous year. There was a slight improvement on the revenue side. The bank raised Meta (META.US)'s 2027 revenue forecast by about 0.9% to $307.9 billion, but this increase was insufficient to fully offset the impact of the increase in consumer spending forecasts. Capital expenditure and cash flow provide another layer of observation. According to the reporting model, Meta (META.US)'s capital expenditure in 2027 was approximately $242.7 billion, and free cash flow (FCF) was about negative $72.8 billion.
Free cash flow reflects the balance of cash generated from operating activities after deducting capital expenses; negative values in the model indicate that operating cash is not yet sufficient to cover capital expenses for the same period. Meta (META.US)'s core contradiction is that improvements in advertising and AI monetization need to keep up with the huge scale of investment. J.P. Morgan Chase (JPM.US) believes that AI ad ranking and recommendation capabilities are improving user engagement and advertising returns, and is optimistic about the monetization path of consumer, enterprise intelligence, and model interface services, so it still maintains a “increase in holdings” rating and a target price of $920. However, the bank also clearly indicated that if AI monetization lags behind, continued investment will suppress operating profit, earnings per share, and free cash flow. Maintaining ratings reflects confidence in longer-term returns, and recent spending pressure remains.
The Fiverr (FVRRR.US) traffic crisis and transformation difficulties have further highlighted the impact of AI on traditional service platforms. Changing AI search habits led to a sharp drop in visits, the platform's transformation to high value was slow, and JPMorgan Chase (JPM.US) lowered its revenue and profit forecasts to maintain a neutral rating. Citing SimilarWeb data, J.P. Morgan Chase (JPM.US) indicated that the year-on-year decline in visits to the Fiverr (FVRRR.US) US website increased from 10% in the second quarter to 22% in the reporting period for which data had been obtained in the third quarter. The monthly data presented in the report was a 19% decrease in July and a 24% drop in August, so this statistic cannot be directly taken as a complete third quarter result.
According to JPMorgan Chase (JPM.US), changes such as AI summaries in Google (GOOGL.US) searches are reducing users' chances of paying per click or searching for organic links. Even though Fiverr (FVRRR.US) invests in generative engine optimization (GEO) to try to increase exposure in AI answers, the click rate from big language models is still lower than traditional searches, putting pressure on the platform to acquire potential customers. The service supply side is also facing adjustments. Fiverr (FVRRR.US) is transforming from a trading market focusing on low-value projects to a higher-end, higher-value work platform.
However, according to the report, high-value businesses currently account for only 15% of the platform's total commodity transaction volume (GMV), and management expects the transformation to take at least six quarters. According to this, J.P. Morgan Chase (JPM.US) predicts that Fiverr (FVRRR.US)'s 2027 revenue will drop by about 15%, as in 2026, and the 2027 and 2028 revenue forecasts will be lowered by about 6%, and adjusted profit before interest, tax, depreciation and amortization (EBITDA) forecasts by 21% and 14%, respectively. The bank maintained a 'neutral' rating and the target price dropped from $11 to $10.
Other travel platform adjustments and summaries show that judging AI returns requires testing revenue, cost, and cash item by item, and each subject's logic is different. The Uber (UBER.US) restructuring involves about 3,300 people. JPMorgan Chase (JPM.US) estimates that it can save more than $600 million a year, but considering the company's plans to reinvest capital into price cuts, service choices, and business growth, it has not yet been counted as net cost savings. Lyft (LYFT.US) faces additional fees and cash expenses due to a driver identity classification dispute settlement; the bank maintains a neutral rating and a target price of $19. Together, these adjustments show that judging the returns of AI and platform businesses requires individual testing along revenue sources, cost changes, and cash flow. Google Cloud (GOOGL.US) TPU sales, AWS (AMZN.US) computing power delivery and prices, Meta (META.US) fees and capital expenses, and Fiverr (FVRRR.US) traffic and transformation progress will each determine whether their respective growth logic can be implemented. This is an in-depth stress test of the market's ability to monetize the AI application layer following the infrastructure arms race. The trend of differentiation may continue throughout the cycle.