Kanghui pivots to AI computing power with huge contracts, but debt and weak core pharma weigh
New AI computing power contracts worth up to 2.4 billion yuan Kanghui's subsidiary signed two computing power service contracts: one worth 415-679 million yuan in July and another worth 1.72 billion yuan in September. These deals show real demand for its new AI computing business, which could bring in steady revenue over five years and help offset weak traditional medicine sales.
These large contracts are the main new growth driver and explain why the stock has nearly doubled this year.
Debt to fund 1.14 billion yuan server purchase To fulfill the big computing contract, Kanghui must buy 1.14 billion yuan of servers, mostly with borrowed money. This will push its debt ratio from 69% to about 78%, increasing financial risk. The company also said the deal's impact on 2026 profit is uncertain, so the market may worry about the cost of this expansion.
This is the main counterweight: the growth comes with significantly higher debt and uncertain near-term profit.
First-half profit turns positive, but only thanks to asset sales Kanghui reported a net profit of 9.09 million yuan for the first half, reversing a loss. However, the profit came mainly from selling idle assets, while the core business still lost money on an adjusted basis. Revenue also fell 17%. This shows the underlying medicine business remains weak.
The profit turnaround is a headline positive, but the reliance on one-off gains and falling revenue is a real concern for future earnings.
Divesting troubled subsidiary Chunsheng Pharmaceutical Kanghui is selling its 51% stake in Chunsheng Pharmaceutical for 64 million yuan, after the unit repeatedly missed profit targets and the original owners couldn't pay compensation. The sale price is only 12.7% above the original investment, so it's a loss-making exit. This removes a drag but also signals past acquisition troubles.
This divestiture is a negative event that highlights management challenges and past poor performance, affecting investor confidence.