Discussions surrounding US interest rates have recently shifted from "when will rates be cut" to "whether rates will be hiked."
The firm's baseline scenario currently assumes the US Federal Reserve will not adjust interest rates before the end of 2026. However, extending the outlook to June 2027, DWS still anticipates a roughly 0.5% reduction space for US rates.
She forecasts the Federal Reserve will implement two rate cuts before May of next year, with the federal funds rate target range expected to fall to between 3% and 3.25%. She explained that US inflation is anticipated to remain around 3.2% in 2026 but is expected to decline to approximately 2.3% by 2027. In other words, if inflation subsides gradually after a brief rebound in the third and fourth quarters as projected, the Fed will regain room for policy easing.
However, she also cautioned that if inflation data fails to improve as expected, the timing of rate cuts could be pushed back further.
The recent sharp decline in the South Korean stock market triggered a minor rout in global AI-related stocks. She believes the increased market volatility is not surprising. The reason is that Asian tech markets, such as South Korea and Taiwan, have been supported this year by AI, memory, and semiconductor exports, becoming a hot trade pursued by global capital. When valuations rise to historically elevated levels, any negative news, policy risks, or individual company events can easily trigger a chain reaction of liquidations in leveraged products and retail positions.
She pointed out that the South Korean market is particularly influenced by leveraged products and retail investor participation, which is why it often experiences larger declines than other Asian markets when sentiment reverses. Over the past few months, whenever risk events emerged in Asian markets, South Korean stocks were often the first to come under pressure, followed by Taiwan and Japan. This reflects not necessarily a rapid deterioration in corporate fundamentals, but rather the natural consequence of excessive concentration in popular trades.
Nevertheless, DWS has not altered its positive view on emerging markets due to these developments. She stated that DWS remains overweight on emerging markets in its CIO investment portfolio, with a key reason being continued optimism towards tech leaders in Taiwan and South Korea.
Fundamentally, AI demand, capital expenditure from hyperscale cloud companies, and orders for memory and high-end chips continue to support the earnings of related companies. She emphasized that what truly warrants attention is the "rate of change" in corporate earnings guidance and capital expenditure from cloud giants, not just short-term stock price volatility.
Regarding US stocks, the market's primary concern remains whether AI has already formed a bubble. She noted that DWS is not inclined to define the next 12 months as a "bubble burst" scenario but will closely monitor several indicators, including PE valuations, earnings performance, index targets, and corporate balance sheets.
She acknowledged that US stock valuations are already at relatively high levels compared to historical averages. DWS's current target for the S&P 500 index is 8200 points, corresponding to a forward PE of 23.5 times, which is in the high end of the historical range. However, the key support for these high valuations lies in the continued strength of earnings.
US tech companies continued to deliver market-beating earnings performances in the first-quarter results season. DWS also expects US corporate earnings to potentially maintain double-digit growth until the end of this year.
She also mentioned that although some tech companies have begun entering the bond market for financing since the second half of 2025, overall balance sheets and leverage levels have not shown significant tightening. Therefore, judging that the AI bubble is about to burst based solely on high valuations is an oversimplification.
On the Asian tech supply chain, she believes AI hardware demand is gradually shifting from a traditional cyclical pattern to structural growth. Taking memory, chips, and related high-end components as examples, some companies have secured revenue visibility through 3 to 5-year long-term contracts, making their profit models no longer entirely dependent on short-cycle demand. This is also why the market is willing to re-evaluate the valuations of some tech companies in South Korea, Taiwan, and Japan.
However, she also cautioned that the focus of the AI supply chain can change rapidly as technology stages evolve. The market's focus has shifted from GPUs and training demand to inference, CPUs, and NAND, among other segments. This signifies that AI investment is not a one-time purchase of a single theme but requires ongoing observation of which part of the entire supply chain truly possesses a comparative advantage.
Comments