Morgan Stanley Global Macro Forum: Reading the Signals: Moonshot, Macro Data, and Bank Earnings

研报英文原文证据摘录

Morgan Stanley Global Macro Forum: Reading the Signals: Moonshot, Macro Data, and Bank Earnings

Key Takeaways

• Early thoughts on Moonshot: Last week, Chinese AI startup Moonshot released Kimi K3, an open-weight large

language model that the company claims delivers frontier-level performance at 40–70% lower cost than comparable

US models. While these claims need to be validated and geopolitical, regulatory, and strategic considerations

surrounding adoption of Chinese-developed models need to be kept in mind, this underscores the rapid pace of

innovation outside the US and raises the possibility that leading-edge AI capabilities may become increasingly

accessible at lower cost. The broader read-through is that continued improvements in model efficiency have the

potential to reduce the cost of using AI to perform a wide range of economic activity, leading to the classic "Jevons

Paradox" that can drive even greater demand for compute (and associated AI infrastructure).

• Macro takeaways from the strong Q2 bank earnings: Economic and macro sentiment remains constructive. US

consumer continues to be resilient – spending and credit are running better than expected. Commercial loan growth is

broadening; consumer loan growth is positive but increasingly selective and overall credit trends remain benign with

problem assets declining rather than migrating into losses. The AI investment cycle is still accelerating and the

opportunity is spreading across the economy

• Macro data takeaways – not a “Messi” print: June inflation is a clear signal about disinflation. We maintain our out-

of-consensus call that the Fed will remain on hold this year and cutting twice next year. The Fed will see inflation

running closer to 2.0% on an annualized basis in 2H26, allowing them to stand pat.