百融雲-W: 業績點評：2024H1階段性承壓，全年營收有望向上

24H1收入增長穩健，淨利潤階段性承壓

2024H1，公司收入13.21億元，同比+6%；毛利潤為9.67億元，同比+8%。收入與毛利潤的增長主要得益於對人工智能(AI)技術的成功應用和商業化。2024H1，公司淨利潤為人民幣1.43億元，同比-31%，主要原因是上半年銷售與營銷開支的大幅上升，由去年同期的4.61億元上升至5.07億元，用於增加品牌及業務宣傳以提升公司品牌知名度以及持續獲取優質流量以提高轉化效率。我們認為，公司淨利潤受銷售與營銷開支而下降是階段性的，未來受益於公司在宣傳與優質流量獲取上的投入，下半年收入與淨利潤有望環比增長。

Steady revenue growth in 24H1E; we see net profit pressures as temporary

Bairong’s unaudited 24H1E results show both revenue and gross profit yoy growth, with revenue up 6% to RMB1.32bn and gross profit up 8% to RMB967m, while net profit fell 31% yoy to RMB143m. The company attributes revenue and gross profit growth to the successful commercialization of artificial intelligence (AI) applications. The weak net profit was mainly due to substantially heavier sales and marketing expenses at RMB507m (vs RMB461m in 23H1), as the company stepped up brand and business promotions to raise brand awareness and increase quality-user traffic to drive better sales conversions. The temporary expenditure being publicity and promotional investments to acquire better-quality user traffic, we are expecting both revenue and net profit to grow in 24H2E.

BaaS金融行業雲數據亮眼，成為業績增長驅動力

2024H1，BaaS金融行業雲收入為5.89億元，同比+20%，主要得益於促成的交易規模快速上升：今年上半年資產交易規模同比+23%至261.50億元。2024H1，公司和客戶的合作加深，聯合開拓各種營銷活動，促成資產交易規模的快速拓展；此外，隨著更多自有用戶貢獻收入，結合公司在生成式AI上的獲客優勢，持續提升獲客效率，並帶來更高的ROI水平。我們認為，公司在核心業務BaaS金融行業雲上的亮眼數據或將在未來帶動公司業績持續向上。

The financial industry segment of Bairong’s business-as-a-service (BaaS) generated RMB589m in cloud data revenue in 24H1E, up 20% yoy, mainly due to facilitated transactions as asset transactions scaled up 23% yoy to RMB26.15bn. Bairong deepened collaborations with customers and jointly developed various marketing activities to quickly raise the scale of asset transactions. And as more of its own users contributed revenue, combined with the Bairong’s edge in customer acquisition on generative AI, its customer acquisition efficiency improved, which brought higher ROIs. In light of the boom in financial industry cloud data, the company’s core BaaS driver, we believe profitability could continue to improve in the future.

技術持續更新，助力公司持續盈利

2024H1，公司成功將智能語音應用與大模型相結合，將完全基於大模型的交互時延壓縮在500毫秒之內，低於同業1秒以上；同時，通過純自研，公司的智能語音大模型進一步迭代出支持情感識別和情感語音輸出的功能。對於未來的長期佈局，公司的大模型進一步豐富模態支持，結合3D數字人的智能語音交互硬件已經開發出第一個商業化版本，並已與客戶進行意向溝通，可廣泛適配於銀行大廳接待、零售百貨導購等多元應用場景。我們認為，公司在技術上的持續更新有望使公司技術領先同業，從而獲得穩健的收入，並在未來助力業績增長。

Tech upgrades pushing Bairong toward industry leadership and profitability

The company integrated large language model (LLM) capabilities into its AI VoiceGPT application in 24H1, while keeping LLM interaction latency within 500ms, faster than the industry average of over 1 second. Its proprietary smart voice LLM also iteratively supports emotion recognition and emotional voice output functionalities. Bairong’s long-term LLM buildout plans includes multimodal support and it has developed the first commercial version of its VoiceGPT with 3D avatars to interface with potential clients. This is adaptable to multiple application scenarios, including a bank’s reception and a department store’s shopping guide. We believe the company’s innovations will help advance its pacing toward industry tech leadership, which would bring stable revenues to fuel earnings growth in the future.