Humanoid robots may redefine economic indicators
Four key investment themes in 2025: Leading the new direction of market development
Morgan Stanley Research relies on deep industry and macro professional capabilities, a huge global network and collaborative culture to help clients accurately identify investment themes that can bring excess returns. Looking back at the global capital market over the past 30 years, stocks that account for less than 3% have dominated the market value growth of the $80 trillion stock market, and these stocks are in a key position in the core investment theme without exception.
Looking forward to 2025, as the market gradually stabilizes, the four major themes of technology diffusion, longevity, future energy and multipolar world investment led by artificial intelligence (AI) will once again become the focus of the market, and the development speed may even accelerate further. Taking the longevity theme as an example, the global aging population is intensifying, and related fields such as healthcare and elderly care services are ushering in unprecedented development opportunities; multipolar world investment is increasingly prominent due to changes in the geopolitical landscape, and the investment value of emerging markets is becoming more and more prominent.
AI infrastructure: unlimited potential, financing model urgently needs to be innovated
Among many investment themes, AI infrastructure investment has attracted much attention. The market generally wonders whether AI infrastructure investment, which has dominated the growth of US stocks in the past two years, can continue its strong momentum after the market recovers? Morgan Stanley gave a positive answer, believing that it is still in the early stages of AI infrastructure construction. According to the law of computing cycles, a new computing cycle will appear every 10 years, and the availability of computing power and the total potential market size (TAM) will increase tenfold. In the last cycle, the CPU computing power of traditional software applications reached $1 trillion. Based on this, the GPU computing power potential for deploying AI applications is as high as $10 trillion.
The current AI penetration rate is only 5%, which means that infrastructure-related stocks such as semiconductors, electricity, cooling systems, and data center equipment and facilities will have great room for value growth in the future. At present, early AI capital expenditures mainly rely on venture capital and retained earnings of cash-rich technology companies, that is, equity capital. However, as the scale of AI infrastructure construction continues to expand, more efficient forms of financing will be needed in the future. Whether in the public market or the private market, multi-form credit such as unsecured credit, secured financing, securitization and asset-backed financing will play an important role in the construction of AI infrastructure.
AI applications and robots:transforming the economy,opportunities and challenges coexist
Whether large-scale AI infrastructure investment is justified depends on whether it can achieve a leap in productivity. In the global $40 trillion labor market, every 10% increase in productivity can save $4 trillion in costs each year, which is expected to be converted into corporate profits. For most AI application pioneers, a 10% productivity increase is just the lower limit of the target, and many companies are showing greater ambitions. Through research on the AI layout of 3,700 companies, Morgan Stanley can help investors identify AI application pioneers in areas such as financial services, utilities and daily consumer goods that have bargaining power and can maintain productivity advantages.
In addition, for AI to fully unleash productivity opportunities, it must complete the transition from the digital world to the physical world. In the next two decades, the popularization of humanoid robots will fundamentally change the nature of "production" and redefine economic activities. Imagine that the labor force in the factory "produces" like machines producing machines autonomously, with almost no human intervention, and traditional economic indicators such as dependency ratio, retirement age and per capita GDP will be systematically re-evaluated.
Currently, China's humanoid robot commercialization process is rapid, attracting the attention of global investors; the United States is trying to achieve sustainable manufacturing repatriation with the help of AI robots, but the uncertainty of tariffs and global trade policies has made the market full of concerns about its prospects. The widespread application of humanoid robots faces many challenges, and there are technical bottlenecks in fine motion control and integration of real-world data.
