Equity markets have continued to reach new highs, driven largely by the outperformance of semiconductor stocks. We believe that such concentrated market movements warrants some caution as it implies that investors do not have confidence in the broad economy.
The US Federal Reserve has also largely withheld any conviction on interest rate cuts in the near-term, citing robust employment and sticky inflation. Market analysts thus far have also displaced any possibility of a recession as businesses and consumers appear to stay resilient. On the other hand, there are signs showing that US unemployment is gradually rising with the unemployment rate reaching 4%, from a low of 3.4% in 2023. As the economy continues to weaken gradually, we are aligned with expectations that one interest rate cut is likely on the table this year.
China’s market performance has started to lag after the government’s policy announcements to stabilize their property market. Some profit taking is expected, given the sharp upward momentum, in our view. Market participants are likely waiting for further updates to be announced on the property industry before having conviction that the slowdown is under control. We uphold the view that the property industry will not result in a systemic risk, given the government’s focus on the issue.
In the technology sector, we are optimistic on the long-term productivity gains from AI applications. One study found that software developers using Microsoft’s GitHub CoPilot, their AI assistant, completed coding tasks up to 56% faster than those not using the tool. Expectations on AI applications are highly optimistic, with investment in hyperscalers expected to grow at least 20% per year up to 2030. While automating tasks such as summarising emails, customer service and image generation would see cost savings, we believe that AI will have to evolve to artificial general intelligence for it to see major contributions to the economy. Generative AI has made significant improvements in recent times, but it lacks the ability to adapt when queries falls outside of its training data. Advancements in general intelligence should mitigate this limitation, paving the way to a more versatile AI solution. Moreover, end-consumer use still seems lacking as current mobile hardware is too slow to process AI tasks efficiently. However, with massive AI investment persisting, there are already significant improvements occurring and we shall continue to look for opportunities as companies innovate and monetise AI features.