Investing.com-- Asia’s technology sector is poised for a promising year in 2025, with valuations now reflecting elevated bond yields, and earnings trends showing early signs of recovery, Bernstein analysts said in a note.
The sector trades at a forward price-to-earnings multiple of 21.5x and price-to-sales of 2.4x, aligning with its 10-year historical averages. With the impact of higher bond yields seemingly priced in, Bernstein analysts argue that the sector is well-positioned for upside unless yields surpass 5%.
Internet, entertainment, and interactive media are attractive sub-sectors due to their relatively cheap valuations and the potential for tactical earnings support, analysts said.
While semiconductors appear expensive on revenue multiples, Bernstein notes an upward trajectory in earnings revisions. Key growth drivers include AI-related demand, which bolsters optimism for companies like Taiwan Semiconductor Manufacturing (NYSE:TSM) (TSMC).
Memory sub-segments, while controversial, could experience an upward inflection by mid-year.TSMC, SK Hynix Inc (KS:000660), and Samsung Electronics Co Ltd (KS:005930) are identified as top picks within semiconductors, supported by data center AI adoption, Bernstein analysts said.
The brokerage was less optimistic about IT services and software, citing peak valuations and earnings that are unlikely to sustain growth.
Similarly, China’s internet sector started 2025 on a weak footing, weighed down by muted macroeconomic signals and geopolitical risks. However, Bernstein sees potential for improvement as pessimistic valuations normalize.
Companies like Tencent Holdings Ltd (HK:0700) and Meituan (HK:3690) are favored for their risk-reward profiles, while NetEase Inc (HK:9999) also remains attractive despite tempered expectations post-rally.
Bernstein also discusses a broader recovery trend for the tech sector, driven by improving sentiment and valuation resets. However, crowding risks remain low compared to financials and staples, suggesting the sector is underexposed relative to historical norms.