China’s tech companies are looking to expand internationally, both to take advantage of growth opportunities as well as to mitigate slowing growth in their domestic market.
There are plenty of examples of Chinese tech companies with vibrant offshore businesses: Shein, Bytedance, and smaller fintechs like LianLian and Airwallex.
But it’s slow going for the two fintech giants, the duopoly of Alibaba and Tencent. They are the country’s two greatest technology conglomerates, and they’ve amassed a string of investments in tech businesses beyond China. But directly growing their own fintech footprints outside of the mainland has been difficult.
They are prisoners of their own success.
China’s fintech industry is arguably the most sophisticated in the world. China leads the world in e-wallet adoption. The payment parts of Alibaba and Tencent (Alipay and WeChat Pay), leveraging the humble QR code to do away with point-of-sale machinery, turned everyone’s mobile into a scanner that could enable the movement of money.
The two companies leveraged a huge population that was, on the one hand, largely banked, but on the other hand, mostly ignored by banks. China’s state-controlled financial system has never been robust. Alipay, grounded in e-commerce, and WeChat, built on games and messaging, pushed against an open door.
Today most Chinese people, particularly urbanites, conduct all aspects of their lives on just two clusters of apps: entertainment, transportation, dining out, dining in, shopping, running a business, government services, financial services – you name it, it’s probably facilitated on either an Alibaba or a Tencent server.
Alibaba’s fintech spinoff, Ant Financial, is still considered the most valuable fintech startup in the world, even after a government crackdown that scuttled its planned IPO in 2020: it’s now valued at $78.5 billion, a far cry from the $315 billion touted in the run-up to its planned listing, but still bigger than Stripe ($70 billion).
While the two conglomerates remain dominant in their home market, that market power isn’t translating into winning business models overseas – which means the innovations they continue to deliver aren’t traveling either.
Take two recent examples of how these companies are presenting themselves to international audiences.
Tenpay Global
Tenpay Global is the cross-border payments unit of Tencent Financial Technology, and its international ambassador is Yang Wenhui, Singapore-based general manager of TFT.
At Money 20/20 Asia in April, Yang extolled the company’s groundbreaking work to enable payments via QR code. There are 1.3 billion users on the WeChat platform (mostly on Weixin, the version designed for mainland-based users). He said the company wanted to bring that experience to cross-border payments.
“We’re proud of what we’ve achieved…the change is global,” Yang said, citing QR-based e-wallets run by central banks in Southeast Asia. The trend is now gaining steam in the Middle East, Africa, and Europe.
But what does that cross-border opportunity look like? Wang says the company’s mission is to serve cross-border consumption and remittances for users at home and abroad. “A more open network and interconnectivity is a blessing for the whole industry.”
For example, foreign workers in Hong Kong can use WeChat to remit money back home to the Philippines or Indonesia through local partnerships. Overseas businesses sending parcels to China can use WeChat to ensure payments reach people like local delivery people.
Wang notes that as a first-mover, there is now a gap between what WeChat facilitates in China and what people can achieve in the rest of the world. Can WeChat expand into that space? “China moved first in digital payments, leaving a gap versus connectivity in the rest of the world, so we can play a role.”
A concrete example of this gap is the fact that visitors to China can’t pay for anything unless they have a local bank account (to which they could link a WeChat account).
This fact is partly because no foreign payments rival operates onshore. There’s no Revolut in China, or any other platform that would enable money to move in and out of the country. That would contravene the government’s capital controls.
Tenpay’s solution is…download WeChat. The company last year enabled local acceptances for foreigners who rely on a credit or debit card.
This isn’t the same thing as how Southeast Asian institutions are connecting their domestic payments infrastructure. If a Singaporean using PayNow visits Bangkok, she can pay locally with PayNow, and the transaction shows up in her Singaporean account. She doesn’t need to download PromptPay, the Thai counterpart. Similarly, a Thai person doesn’t need to download PayNow if he’s visiting Singapore and wants to pay with his PromptPay wallet.
But Tencent’s solutions still orbit around the need for third parties to download WeChat to do anything, whether the starting point is an overseas wallet or credit card. Tenpay’s cross-border relationships are designed to facilitate onboarding new WeChat users, rather than to ease the way for others to use their own payment tools.
Tenpay also signed a deal last year with Western Union to allow people remitting money to send the money to a mainland-based WeChat account.
“We need to find the balance between security and convenience,” Wang said by way of explanation, citing local KYC rules.
The company is aware of the problem, and is working on ways to let visitors use their own mobile payment tools within China. Before Covid, China received over 100 million visitors, so this is not a trivial problem to solve.
But the company’s take on why it’s hard for these cross-border initiatives to gain traction is “awareness”, Wang said. “We need to overcome the mindset that people fear downloading an app. This requires trust. This is a top priority, enabling convenience on top of compliance.”
Similarly, Wang says the company is eager to join in mutual-recognition schemes of QR codes within Southeast Asia. The company touts its infrastructure, its userbase, and its merchant network. Along with stakes in leading Southeast Asian tech platforms, Tencent certainly brings a lot to the table. Bank of Thailand officials have told DigFin they’d also like to find a way to include China and India in these schemes.
But it’s hard to see how payment switches such as PayNow or PromptPay would include WeChat Pay without reciprocity: without users having to download WeChat. Is that purely due to Chinese regulations? Tencent wouldn’t be the first Big Tech company to insist others play by its rules (hello, Microsoft!). But there’s lots of competitive offers in Southeast Asia and other regions, whereas WeChat evolved in a world where it didn’t really have a competitor other than Alibaba – and as fierce as that competition is, the two companies have their lanes, even in China.
One area where Tenpay is looking for overseas opportunities is e-commerce. Tencent has stakes in platforms such as Shopee (although Tencent has recently said it would reduce its stake in the platform’s parent, Sea). Sellers on these platforms may use WeChat when dealing with mainland customers (or WeChat users overseas), and Tenpay is working to provide value-added services.
This is an attractive business: “We get end-to-end information of the users, which eases the transaction,” Wang said. But it’s early days, and the landscape is very competitive.
Wang says Tenpay has begun serving sellers on one regional e-commerce platform, but declined to say which one. When it comes to shipping goods to Chinese buyers, Tenpay is encouraging use of QR-based payments, citing its domestic infrastructure and network.
But it appears to be slow going. “We need help from the [e-commerce] platforms,” Wang said. “Industry KYC and KYB [know your business] rules require every participant to exchange information properly. Then we can build a robust ecosystem.”
One barrier is China’s local data-governance rules. Wang compares this to GDPR in the European Union: “Each country has its own need.” He says government-level agreements will be necessary to allow remittance companies to satisfy privacy-protection rules.
Tenpay now has a Major Payments Institution license in Singapore and it’s open for business, says Wang: “We’re establishing cross-border payment platforms and we hope we are welcomed by the industry.”
No doubt many fintechs and other businesses will want access to the WeChat network. It’s striking, however, that this Big Tech legend seems to be finding it challenging. The network, the user base, and the wide-open spaces that made Tencent a global tech giant don’t exist outside of China. The regulations that buttressed its duopoly position at home, throwing sand in the gears of would-be foreign competitors and enabling WeChat to develop its app around the intricacies of local privacy rules, now look like barriers to partnerships outside of China.
Ant Insurance
Ant Insurance recently briefed the international media on the technological capabilities of its insurance brokerage platform. They are indeed impressive.
So are the company’s results. In the first half of 2024, insurance companies selling products via Ant recorded claims payouts totaling Rmb14 billion ($1.9 billion), representing 30 percent year-on-year growth.
While China’s consumers are struggling in the face of a crushing real-estate collapse, digital platforms that can deliver good services at affordable prices can flourish.
Ant caters to younger people in their 20s and 30s, often in smaller cities, who collectively represent 600 million policyholders. On the other side of its platform are 90 insurance companies providing more than 1,000 products with coverage across life, health, motor, medical, pet and e-commerce.
Fang Yong, head of claims technology in Hangzhou, outlined the way the company is using artificial intelligence on top of its trove of user data, instead of relying on a (mostly) unprofessional salesforce of agents. Technology lets Ant broke products more cheaply. Its health insurance service, Haoyibao, began in 2018 with critical illness; now it offers a range of health-insurance products across a diverse customer base.
Its pet insurance offering, provided by China Continent Insurance and Zhongan Insurance, uses nose-print recognition AI to identify cats and dogs. In the West, vets typically microchip pets, but that’s not common in China. Instead, owners upload photos of their pets’ noses via the Alipay app. The accuracy rate is above 99 percent, says Fang.
More recently, Ant Insurance has bundled a quartet of AI-enabled services. These let users compare various products and choose the one that is most competitive for their needs. Another service can personalize a user’s portfolio of insurance cover, to identify gaps or areas of over-coverage. A third service uses generative-AI-powered chatbots to answer questions about policies or coverage, underwriting and claims.
Fourth, the company uses AI to speed up settlements, guiding users in real time as they upload photos or other information. Fang says most claims filed via Haoyibao are now processed within two days.
The claims settlement piece itself is the most complex part of Ant Insurance’s platform. It requires a blend of supporting AI tools, including document scanning via mobile (including AI to decipher blurry photos or sniff out fraud), extraction of structured data, a tool to investigate claims, and a decision tool to determine most outcomes.
This last aspect in turn required the company to build a knowledge graph of more than 20,000 diseases, nearly 8,000 types of surgery, and symptoms, and match these to each user’s symptoms.
Fang says today, the system makes fully automated decisions for 80 percent of outpatient claims and 40 percent of in-patient cases. By placing large-language model chatbots on top of the analytics, the system can also provide explanations in case someone asks why they’re being refused compensation.
“The inexplainability and hallucination problems of large-language models are not acceptable,” Fang said. The company breaks the decision-making process into separate links that LLMs can interrogate separately. Fusing LLMs with traditional analytics can satisfy more and more claim questions. “This creates accuracy and explainability,” Fang added.
This is a huge time-saver for the claims team. Even if the platform fails to make a decision, it can narrow the context that humans need to investigage. For example, an auditor’s investigation of a claim might run over 100 pages, but the AI will guide claims officers to the dozen or so relevant pages that need to be read.
In short, this is bleeding-edge stuff for insurance companies. Imagine what the response would be from global carriers if Ant Insurance showed up in their domestic market with this kind of capability.
But it won’t happen. First of all, AI decisioning relies on big data to be reliable. That means Ant Insurance can only deploy it if it has a big customer base. It doesn’t outside of China.
Could it set up overseas and build one? Not any time soon.
Fang noted that insurance is heavily regulated everywhere, including in China. Obtaining licenses is not easy.
Secondly, Fang noted Ant Insurance’s platform works thanks to the vast ecosystem built by Alipay, which generates valuable customer insights.
Third, the company’s AI prowess is built upon reading Chinese-language documents. Overseas it would have to start learning new languages, and the jargon that goes with different medical rules.
“Our expertise in terms of technology is not so easily replicated,” Fang said.
A fourth challenge is different market needs. Ant Insurance evolved to meet the needs of users and insurers in China; those needs are different overseas. For example, pet owners in many countries microchip their animals, so there might not be any demand for Ant’s nose-print recognition. Another example is that insurance companies elsewhere readily share information with hospitals, so there wouldn’t be a need for the document scanning that Ant has mastered.
In short, there’s little prospect that Ant Insurance could replicate its success, either technologically or commercially, in overseas markets, even big ones. That also means its innovations are not going to be adopted outside of China quickly or comprehensively.
Local heroes
While it’s true that ‘every market is different’, the China tech ecosystem is truly unique. One feature of that uniqueness is the Tencent-Alibaba duopoly.
That duopoly is less potent today. New tech companies like Shein and Bytedance have emerged outside of the old duopoly. The government’s push for an e-RMB is also eating away at the dominance of the Alipay and WeChat Pay closed-payments networks.
But these remain the biggest tech companies, with stakes in other companies worldwide that suggest global ambitions. Both have backed virtual banks in Hong Kong and Singapore, for example, and both have fully digital banks in China (MyBank and WeBank). But these offshore banks all have to act within the constraints of local market size, regulation, partnerships, and demand. They’re not spinning up a MyBank International.
Other companies have had success exporting their technology but not the business: Zhong An and Ping An, for example, sell their capabilities as software vendors. Tencent’s WeBank has recently set up a similar sales effort in Hong Kong.
Yet as other Chinese tech companies find their way into international opportunities, the Alipay-WeChat duopoly looks ever more like a domestic institution.
Although they own stakes in other leading tech platforms and wallets, these haven’t cohered into cross-border versions of Alipay or WeChat Pay. The closest is Ant’s Alipay+, a cross-border service for overseas acquirers, merchants and mobile payments providers. It is more open than WeChat, helping customers pay with their own preferred method. Other Ant companies also provide merchant payment processing and SME money transfers.
Alibaba and Tencent became massive thanks to embedding payment apps into everything else (hallmarks of their competitiveness and culture of innovation), in the specific context of when Chinese financial services were lousy and regulators welcomed insurgents who could support consumers and small businesses. And bigness begat bigness.
Now they’re so big they can’t gain a similar data advantage anywhere else, and there’s only so much localization that makes sense to them, so the model can’t export.
