Data Pledge Financing in China: What Is Actually Being Pledged, and Where the Law Gets Stuck

Editor’s Note — DCC.

This brief summarises 《DEXC+专栏|数据质押融资，出质什么？难点在哪里？》 by Chen Yiqian (陈一芊), Transaction Review Supervisor in the Compliance Department of Shenzhen Data Exchange Co., Ltd. The piece is published under the DEXC+ think-tank column — Shenzhen Data Exchange’s practitioner compliance series — and carries the standard caveat that it represents the author’s academic views, not those of the Exchange, and does not constitute legal advice. DCC is running it because it provides the clearest practitioner-level map of why data pledge financing keeps stalling: the question is not whether data has economic value, but whether the existing Civil Code (《民法典》) categories can accommodate data as collateral without supplementary legislation that has not yet arrived.

The brief connects to the data property-rights registration guide and to the Datatang v. Yinmu data-IP registration case, which noted growing momentum in data pledge financing as a driver of registration activity. Chen’s analysis explains why that momentum has not yet translated into a settled legal mechanism.

What data pledge financing is and why it matters

Data pledge financing (数据质押融资) refers to a transaction in which a data holder — typically an enterprise — uses its data assets as collateral to secure a loan from a bank or other financial institution. The pledge (质押/出质) is the legal mechanism that gives the creditor security over the collateral: if the debtor defaults, the creditor has a prior claim to the collateral’s value, realised by auction, sale, or agreed transfer.

Under Chinese property law, a pledge right (质权) is a form of security interest (担保物权) — one of three categories of real rights (物权, alongside ownership rights and usufructuary rights). A pledge is constituted either by the creditor taking possession of the pledged asset (for chattels) or by registration (for rights). When the debtor defaults, the creditor may apply the collateral’s proceeds against the secured debt on a priority basis.

The practical interest in data pledge financing follows directly from the data-as-asset agenda. Once the 2022 policy document on China’s data foundational system — the Data Foundation System Opinions — endorsed a “three-rights” framework of data resource holding rights (数据资源持有权), data processing and use rights (数据加工使用权), and data product operating rights (数据产品经营权), and once enterprises began mapping data onto their balance sheets under accounting guidance, the question of how to mobilise that recognised value as loan collateral became commercially urgent. Chen’s article is the Exchange’s effort to answer it honestly.

The two routes the Civil Code provides

China’s Civil Code (《民法典》) permits two distinct types of pledge: chattel pledge (动产质权) and rights pledge (权利质权). Chen examines each in turn, because which route is used determines what legal requirements must be satisfied, and each route carries its own difficulties for data.

Chattel pledge (动产质权) treats data as a movable thing capable of possession. Its two defining legal properties are: (1) the subject matter must be a chattel owned or controlled by the pledgor and transferable; (2) the pledge is constituted — and kept alive — by the creditor’s physical possession of the pledged asset. In other words, chattel pledge requires delivery.

Rights pledge (权利质权) treats data as an economically valuable right rather than a thing. The Civil Code (Article 440) lists eligible rights using a “enumeration plus catch-all” structure: bills of exchange, bonds, bills of lading, fund shares, equity interests, transferable intellectual property rights over trademarks, patents and copyrights, receivables, and — critically — “other property rights that may be pledged as provided by law and administrative regulations.” The catch-all clause is not open-ended: only rights that statute or administrative regulation explicitly designates as pledgeable qualify. Civil actors cannot create pledge rights over property interests at will.

Three operational problems with chattel pledge of data

Chen identifies three distinct threshold problems that must be resolved before data can be pledged as a chattel.

First, disposability. Constituting a pledge is a disposition act: the pledgor must have the power to dispose of the pledged asset. For data, this is genuinely uncertain. Data’s characteristics — ease of copying, ease of reuse, potential for simultaneous use by multiple parties — make it structurally difficult to determine whether the pledgor actually has unencumbered control over the data it proposes to pledge. If the pledgor lacks disposal authority, a creditor can invoke the good-faith acquisition rules (善意取得, Civil Code Article 311) only if the creditor neither knew nor had reason to know of the defect — and financial institutions are generally held to a higher standard of diligence in due-diligence processes. Chen’s suggested workaround: creditors should prefer data that has already passed compliance review and been listed on a data exchange, or that has been formally brought onto the enterprise’s balance sheet (入表), as those processes provide an external indicator of legitimate control. This is a practical workaround, not a legal solution — it shifts evidentiary risk rather than resolving the underlying ownership question.

Second, transferability. A chattel pledge is realised by selling, auctioning, or transferring the pledged asset when the debtor defaults. If the data cannot legally be transferred to a buyer, the pledge cannot be enforced. Transferability is not guaranteed: the Data Security Law (《数据安全法》, Article 8) imposes general constraints on data activities that could conflict with a forced transfer, and sector-specific rules — Chen uses the Credit Reporting Industry Regulation (《征信业管理条例》, Article 18) as an example — may require the data subject’s written consent before personal information can be disclosed to a third party in the first place. Assessing whether specific data is legally transferable requires a compliance review of its content, its classification, the consents obtained at collection, and any sectoral rules that apply. The problem for creditors is twofold: at the time of signing the pledge agreement, it may be impossible to confirm whether the necessary consents will exist at the time of enforcement; and even if transferability is confirmed, the creditor must find not just a buyer but a buyer that is itself qualified to receive that data (possesses the relevant consents or authorisations from the original data subjects). This turns a standard enforcement process into a specialised compliance matching exercise.

Third, delivery. Chattel pledge under Chinese law is constituted by delivery — the pledged asset must pass into the creditor’s possession, and the pledge subsists only for as long as the creditor retains that possession. Chen reads “delivery” for data purposes as requiring a transfer of management and control, not a physical spatial transfer. But even on that reading, a creditor that takes control of data becomes a data custodian and assumes the legal obligations that come with it: if the data is damaged, lost, or breached through inadequate storage, the creditor is liable. Data storage is not a neutral technical act — the Data Security Law and data classification rules impose specific storage environment and security requirements that vary by data tier. The practical effect is that accepting data as chattel pledge collateral requires the creditor to either build data security infrastructure or outsource it.

Chen’s recommended solution to the delivery problem is dynamic chattel pledge via a third-party data custodian (数据托管机构). The Supreme People’s Court’s Interpretation on the Guarantee System under the Civil Code (Article 55) provides a template: a three-party agreement among creditor, pledgor, and a monitoring party (监管人) can constitute a pledge over a pool of assets defined by type, specification, and quantity rather than specific items, with the monitoring party holding the assets on the creditor’s behalf. This structure allows data to remain in active use while remaining subject to a pledge, transfers custody obligations to the monitoring institution, and shifts liability for custodian non-compliance away from the creditor.

Why rights pledge of data faces deeper structural barriers

The rights-pledge route appears more conceptually fitting — data is an intangible, and pledging an intangible right seems more natural than treating data as a chattel. But Chen’s analysis shows it faces harder legal barriers.

Data property rights (数据财产权) remain a policy concept, not a legal right. The Data Foundation System Opinions (数据二十条, December 2022) introduced the three-rights framework, and subsequent central planning documents have repeated the call to establish a data property rights system. But as of the time of writing, “data property rights” exist only at the policy layer. The concept of “property rights” (产权) is an economic rather than a legal term; translating it into a specific, legally defined civil right — one that satisfies the Civil Code’s numerus clausus principle requiring that real rights be created by statute — requires legislation that has not yet been enacted. Academic debate continues over whether data property rights should be recognised as a new category of property right. Even if they were, the further step of designating data property rights as pledgeable within the Article 440 catch-all would require an additional statutory or regulatory act. Neither step has occurred.

Data intellectual property rights (数据知识产权) have a pilot but not a statute. The route of treating data as a form of intellectual property has moved further, institutionally: in December 2022, the National Intellectual Property Administration (国家知识产权局) designated eight pilot jurisdictions — Beijing, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, and Shenzhen — to run data intellectual property registration trials. Shenzhen Data Exchange, as the institutional author of this column, has direct operational familiarity with that pilot. But the doctrinal problem persists. The Civil Code (Article 123) defines the objects of intellectual property rights using an enumeration that ends with “other objects as provided by law.” Data that can already fit within existing IP categories — original databases qualifying as copyrighted works, patentable data-processing inventions — does not need the data-IP label and can be pledged under existing IP pledge rules. But data that does not fit existing categories cannot be squeezed into the framework without explicit statutory designation, and that designation has not arrived nationally. The pilot registrations exist and have commercial value, but they do not resolve the underlying “legality” (法定主义) question for pledge purposes.

Chen also notes a commercial reality: intellectual property pledge is already a difficult proposition for lenders. IP value is hard to appraise and its enforced sale value is uncertain. Banks typically resist IP collateral for exactly this reason. A “data IP” pledge compounds the uncertainty by adding an unresolved legal question on top of the existing commercial one. By contrast, chattel pledge — however cumbersome — at least operates within a settled doctrinal framework with established enforcement precedents.

Where things stand: chattel pledge via custodian is the working path

Chen’s conclusion is clear. For data pledge financing in China today:

Chattel pledge via a third-party data custodian is the most viable route. It operates within an existing legal framework, the three-party monitoring structure has Supreme Court backing, and custody arrangements can contain the creditor’s data-security exposure. The three threshold problems (disposability, transferability, delivery) are manageable with careful structuring — in particular, using data that has been listed on a compliant data exchange or brought onto the corporate balance sheet.
Rights pledge via data property rights cannot yet be used. The necessary statutory foundation does not exist. Pledging data as a novel category of property right would require legislation that has not passed.
Rights pledge via data intellectual property is legally uncertain and commercially unattractive. The pilot registrations in eight jurisdictions are genuine, and the Datatang v. Yinmu case documented in DCC’s case study illustrates the registration mechanics. But until the national IP law framework explicitly recognises data as an IP object, using that registration as the basis of a pledge remains legally precarious. And even if the legal question were resolved, lenders’ commercial aversion to IP collateral would remain.

The practical upshot is that data pledge financing is not stuck because data lacks value — it is stuck because the legal infrastructure to give that value a form that existing security-interest law can grip has not yet been fully built. The data property-rights registration guide represents one incremental step in that infrastructure. But registration alone does not resolve the numerus clausus problem, and it does not resolve the transferability question at enforcement.

Why overseas counsel should care

Due-diligence framework for China data-secured lending. Any cross-border financing transaction involving a Chinese counterpart that proposes data assets as collateral should work through Chen’s three-question chattel-pledge checklist (disposability, transferability, delivery) before accepting the security package. Each question has a compliance sub-question that standard asset due-diligence processes were not designed to handle.
The data-as-asset accounting step does not create a pledgeable right. Chinese enterprises bringing data onto their balance sheets under the 2023 accounting guidance generates a book entry, not a legally perfected property right. A balance-sheet recognition of data does not by itself resolve the disposability or transferability questions Chen raises — overseas counterparts that treat balance-sheet data valuation as equivalent to legal ownership of pledgeable collateral are reading the Chinese regulatory framework incorrectly.
Watch the legislative pipeline. The gap between policy intent (the Data Foundation System Opinions’ three-rights framework) and operative legal right (a statute designating data rights as pledgeable) is the core structural risk. When — not if — that legislation arrives, it will reshape the entire data-secured financing landscape. Monitoring the National People’s Congress legislative agenda and the NDA’s data-registration rulemaking is how to get ahead of that shift.
Data exchange listing as a diligence proxy. Chen’s suggested workaround — treating exchange-listed or balance-sheet-recognised data as a proxy for adequate disposability — has practical traction. Overseas counsel structuring transactions with a Shenzhen Data Exchange counterparty can reference the DEXC+ compliance review as part of the security package, understanding that it addresses the diligence concern without resolving the underlying legal uncertainty.

DCC sources

Original: 陈一芊 (Chen Yiqian), 《DEXC+专栏|数据质押融资，出质什么？难点在哪里？》, 深圳数据交易所 DEXC+ 专栏 WeChat Official Account (source).
《关于构建数据基础制度更好发挥数据要素作用的意见》(Data Foundation System Opinions / 数据二十条)
数据知识产权登记指引 (Data Property-Rights Registration Guide, draft)
Civil Code (《民法典》): Articles 115 (real rights defined by statute), 123 (intellectual property objects), 311 (good-faith acquisition), 426 (non-transferable chattels excluded from pledge), 440 (rights eligible for rights pledge).
Supreme People’s Court, Interpretation on the Application of Guarantee Provisions of the Civil Code, Article 55 (three-party dynamic chattel pledge).
Credit Reporting Industry Regulation (《征信业管理条例》, State Council Order No. 631), Article 18.
Data Security Law (《数据安全法》), Article 8.

This is an editorial summary, not a translation of Chen Yiqian’s piece. Conceptual framings and operational extrapolations are DCC’s. Any simplification or error of emphasis is DCC’s responsibility, not the author’s. The original carries the DEXC+ disclaimer that it represents the author’s academic views only and does not constitute legal advice. Not legal advice.