China’sRushtoFinancializeDataRunsAheadoftheMarket

China’s push to turn data into a tradable financial asset is running into an uncomfortable reality: while policymakers have built the accounting framework for data capitalization, the market still struggles to determine what most data is actually worth.

A recent attempt by a local government financing vehicle (LGFV) in Zhejiang province illustrates the problem. The company planned to use operational data generated by its parking lots — including traffic flows, payment records and vehicle management information — to capitalize data assets and pursue financing.

But the effort stalled almost immediately at the compliance stage.

“In current data capitalization practices, compliance is a prerequisite, but we find it extremely difficult when issuing compliance opinions to enterprises,” Li Yongzhuo, director of the AI and Digital Economy Department at Shengting Law Firm, told Caixin.

Many companies accumulated large amounts of operational data during years of digitalization, but few built governance systems capable of treating that information as a financial asset. Regulators now expect companies to establish systems for data classification, desensitization, lifecycle management and risk control, while ensuring compliance processes are embedded into day-to-day operations rather than added retroactively, he said.

Policy push

The difficulties highlight a widening gap between China’s policy ambitions for data assets and the practical realities of turning data into financeable collateral.

China has spent years laying the policy groundwork for data commercialization. Data was formally recognized in 2020 as a factor of production alongside land, labor, capital and technology. The Ministry of Finance’s accounting rules, which took effect in 2024, established how companies could place data resources onto balance sheets.

Since then, the number of companies capitalizing data assets has risen rapidly. More than 130 listed firms had disclosed data asset capitalization by April 2026, according to Caixin’s tally, while hundreds of non-listed firms, many of them state-owned enterprises and LGFVs, have also joined the push.

Yet the actual balance sheet impact remains modest.

Most companies currently value data assets using the cost approach, recording them based on the expenses incurred to generate or obtain the data. As a result, capitalized amounts are generally small, often ranging from several million yuan to tens of millions. Only a handful of large technology and telecom companies report data assets exceeding 100 million yuan.

In other words, data is entering corporate books, but not yet materially reshaping balance sheets.

At the same time, financing activity tied to data assets has surged, particularly among debt-pressured LGFVs searching for new collateral and funding channels as land-sales revenues weaken.

That boom has been most visible in data asset asset-backed securities (ABS). Wind data shows cumulative issuance reached 16.3 billion yuan ($2.4 billion) by mid-May, more than triple the level recorded for all of 2025.

But many industry participants argue the rapid growth masks a more limited reality: most data asset financing still depends overwhelmingly on traditional credit structures rather than the intrinsic value of data itself.

The industry’s first widely recognized data asset ABS illustrates the point. The product carried a AAA rating largely because of a financing guarantee issued by China Zheshang Bank Co. Ltd. The pledged data assets served primarily as auxiliary credit enhancement rather than the core repayment source.

In practice, repayment still relied on borrower creditworthiness, not future cash flows generated independently by the data.

A China Unicom data center in Zhengzhou, Henan province, in June 2024. Photo: VCG

Valuation problem, regulatory halt

Banks remain reluctant to lend directly against data because valuation methods remain deeply problematic.

The cost approach is viewed as overly conservative. The income approach, which estimates future economic benefits generated by data, depends heavily on subjective assumptions about growth and future cash flow. The market approach — using comparable transactions to determine value — suffers from a lack of actual trades.

“The core contradiction lies in the natural paradox between high-value data and public trading,” Li said. Companies are often unwilling to openly trade their most commercially valuable data because of competitive concerns, while the data that is circulated tends to have limited commercial value.

Without a functioning secondary market and reliable price discovery, banks struggle to treat data as standalone collateral.

That helps explain why LGFVs, rather than data-native technology firms, have emerged as some of the most active participants in the market. For many local state-owned enterprises, the attraction lies less in building a sustainable data business than in finding new ways to unlock financing.

Regulators have taken notice. Applications for data asset ABS have been temporarily suspended, several market participants and people close to regulators told Caixin. Two investment bankers said they received window guidance from stock exchanges on Wednesday telling underwriters to pause new filings.

The sources cited regulators’ concerns that some LGFVs are using the new products to repackage nonstandard debt into tradable securities, sidestepping tighter restrictions on local government borrowing.

The remaining obstacles extend beyond valuation and debt arbitrage alone. Financial institutions also lack mechanisms to continuously monitor data assets after loans are issued, creating concerns about risk control, according to Sheng Jing, general manager of the Data Asset Innovation Department at the Beijing International Big Data Exchange. Compared with traditional collateral, banks have limited ability to track whether data assets deteriorate, lose relevance or are transferred elsewhere after financing is completed, she said.

China’s policymakers have moved quickly to create accounting standards, property-rights frameworks and financing pilots for data assets. But market infrastructure — from compliance systems and valuation standards to trading mechanisms and post-loan monitoring — remains underdeveloped.

Contact editors Jonathan Breen (jonathanbreen@caixin.com) and Kelsey Cheng (kelseycheng@caixin.com)