Zhongji Innolight, a leading Chinese producer of optical transceivers used in artificial intelligence data centres, saw its shares fall on their Hong Kong debut Thursday, as a global downturn in AI-related investor sentiment weighed on what had been one of the market’s most anticipated listings.
The stock dropped as much as 3% in early trading, falling 2.86% to HK$953 (approximately $121) by mid-morning, undercutting expectations for a strong opening. The muted reception came despite the company completing Hong Kong’s largest initial public offering of the year.
The listing raised HK$53.4 billion (approximately $6.8 billion), underscoring continued appetite for AI-linked hardware suppliers even as sentiment toward the broader sector cooled.
Zhongji Innolight manufactures optical transceivers, components that transmit data at high speeds between servers within the sprawling data centres that power AI models. Demand for such hardware has surged alongside the global build-out of AI computing infrastructure, positioning the company as a key supplier in the ecosystem.
The soft debut mirrored a wider retreat across AI-exposed equities, as investors reassessed valuations following a prolonged rally. Concerns over the pace of AI spending and stretched valuations have recently rippled through chip and semiconductor stocks, dampening enthusiasm for new offerings in the space.
The performance stands in contrast to some recent Hong Kong flotations that surged on their first day, highlighting how quickly market conditions can shift for firms tied to fast-moving technology trends.
Hong Kong has emerged as a preferred venue for large Chinese share sales, with several major companies choosing the city to raise capital and broaden their investor base. Zhongji Innolight’s offering ranks among the standout listings of the year by size.
The company’s early trading suggests investors remain cautious about pricing near-term momentum into AI supply-chain names, even those with strong order books. How the stock stabilises in the coming sessions will offer a gauge of underlying demand for the sector’s hardware backbone.