Two Names, One Factory: Related-Entity Risk in Chinese Supply Chains

The first time a buyer describes this structure to me, it comes out as confusion. "I signed with one company. The factory that made my goods is a different company — same name almost, different legal entity. And my payment went to a third account, which I'm told is the group's 'settlement center.' Which company is my supplier?"

Legally, the answer is: the one on the signature page. Commercially, the answer is: none of them, individually — they are one operation wearing three sets of registration papers. And that gap is exactly where a supplier group wants you when something goes wrong. The signing company has no assets. The producing company never signed anything. The collecting company received "intercompany settlements." Each entity can point at the others, and the group's productive assets stay one step ahead of your claim.

This article is about the structure, the statute that addresses it, what the courts require before they will treat related entities as one, and — because litigation is the stupid way to solve this — the contract language that makes the whole fight unnecessary.

The A-B-C Structure, and Why Groups Build It

The classic pattern in a mid-sized Chinese manufacturing group:

• Company A — the "trade" entity. Sometimes an older company with an import-export history, sometimes a fresh registration with a clean name. It signs the sale contract, issues the commercial invoice, employs the export sales staff.
• Company B — the production entity. Owns the plant, the machinery, the land-use rights; employs the workers. Often the group's oldest and most valuable entity, sometimes with state or collective roots.
• Company C — the settlement entity. Receives payments, nets intercompany balances, sometimes handles procurement. Invisible to you until you study your own bank slip.

There are benign versions of this. Groups centralize exports for foreign exchange and tax reasons; related-party settlement centers are common; B may be legally separate for genuine legacy reasons. In my article on the Hong Kong shell I explained the same logic operating across a border; here the entities all sit on the mainland, which changes both the risk and the available remedies.

But whatever the group's intent, the structural effect is constant: the entity that owes you performance holds no assets, and the entities that hold assets owe you nothing — on paper. Limited liability, designed to protect honest shareholders, becomes a routing system for commercial risk. The question is when the law will refuse to accept the routing.

What the Law Actually Says: Article 23 of the Company Law (2023 Revision)

The relevant provision is Article 23 of the Company Law of the People's Republic of China (2023 revision). It has two paragraphs that matter to buyers, and they do different jobs.

Paragraph 1 — vertical piercing. Where a company's shareholder abuses the company's independent legal person status and the shareholders' limited liability to evade debts, seriously harming the interests of the company's creditors, the shareholder bears joint and several liability for the company's debts. This is the classic veil-piercing: the company is a shell, the shareholder used it as a shield, and the shareholder answers personally (or with its own assets) for the company's debts.

Paragraph 2 — horizontal piercing. Where a shareholder uses two or more companies it controls to carry out the conduct described in the preceding paragraph, each of those companies bears joint and several liability for the debts of any one of them. This is the provision buyers in A-B-C structures should memorize. It means the law can treat Company A's debt as Company B's and Company C's debt — if the conditions are met.

Two things to understand about paragraph 2 before you get excited. First, it is aimed at abuse — entities used to evade debts and seriously harm creditors — not merely at the existence of a group structure. Groups are legal. Second, the burden is on you: the court does not dissolve the corporate form because it is convenient. You must establish the facts of abuse, and that is where most of these cases are won or lost.

What You Must Prove: Three Kinds of Commingling

Chinese courts applying these rules look for commingling (混同) across three dimensions — and they weigh the three very differently.

• Financial commingling — accounts mixed, funds shuttling between entities without documentation, Company C collecting your payment then "lending" it onward, shared or unclear bookkeeping, assets titled to one entity while used by another. This is the core of the doctrine and by far the hardest to prove, because the evidence lives in the defendant group's own books. Courts typically want more than suspicious flows: the gold standard is an audit or judicial accounting appraisal showing the entities' finances cannot be meaningfully separated. A buyer who cannot obtain financial records must build the financial case indirectly — which is where the other two dimensions do their work.
• Business commingling — the entities operate as one: the same sales team quotes for A and ships from B; your technical drawings go to B although A signed; C collects; the group issues documents under whichever letterhead is convenient for the moment. Overlapping business scope, shared premises signage, one entity performing another's contract.
• Personnel commingling — same legal representative, same actual controller, overlapping managers and staff on multiple payrolls or none, the "finance director" of A signing receipts for C. Courts also watch for formal appointments that mask a single real decision-maker across all entities.

In practice, strong cases braid all three: personnel and business commingling are provable from documents you already hold, and together they support the inference — later confirmed by financial evidence obtained through court-ordered disclosure or appraisal — that the entities' finances are one wallet with three pockets.

The Evidence Courts Actually Credit

From the claimant's side, the working evidence set looks like this:

• Corporate registration records for all entities: shareholders, legal representatives, registered addresses, business scope, incorporation dates. Overlap in any of these fields is where the analysis starts. These records also reveal the control chain needed for paragraph 2.
• The transaction documents themselves: contract signed by A, proforma invoice from A, quality certificates from B, customs declaration in A's or B's name, bank slip to C. A single shipment producing documents in three entities' names is the A-B-C structure documenting itself.
• Correspondence: the same people signing emails with different entity footers; the factory manager discussing "our" contract though his company is not a party; WeChat records where personnel answer interchangeably for all three entities.
• Payment-flow records: your remittance to C, then whatever fragments can be observed of C's onward transfers — invoices, receipts, intercompany loan notes that surface in later documents.
• Court-ordered measures: property preservation against all three entities at filing; applications for the court to order production of financial records; and, in suitable cases, judicial appraisal of the accounts — the step that most often decides financial commingling.
• Enforcement-stage findings: if you have already obtained a judgment against A and execution comes back empty because B holds everything, that enforcement result is itself powerful evidence for the follow-on horizontal-piercing claim.

Note the sequencing implication: this evidence is built during the dispute, not after judgment. A buyer who preserves broadly at filing, and who kept the factory's direct correspondence all along, arrives at the piercing argument with the file a court needs. A buyer who signed, paid, and saved nothing but the contract arrives with a claim against an empty shell and an expensive research project. The verification work that maps the group before you sign — shareholders, cross-holdings, common controllers — is described in my article on enforcing judgments against Chinese companies, and it is the same map you will need later.

What This Does Not Do — the Honest Limits

A few things buyers over-assume about Article 23:

• It is not automatic. Related entities, shared owners, and centralized settlement are lawful. Abuse — evasion of the specific debt, serious harm to creditors — must be established factually.
• Financial commingling is the battleground. Without access to financial records — through appraisal, disclosure orders, or the defendants' own documents — many cases stall at "strong suspicion." Plan the evidence route before filing.
• It adds defendants; it does not conjure assets. Joint and several liability against three empty entities is still three empty entities. The doctrine pays when the related entities hold real, reachable assets — which is precisely the fact pattern where the group thought it was safe.
• It is litigation, with everything that implies: time, preservation strategy, appraisal fees, appeal cycles. Which is why the next section is the one that matters most.

Prevention Beats Piercing: the Drafting Fix

Horizontal piercing exists for buyers who failed to fix the structure at signing. The fix itself is unglamorous and effective:

1. Name the production entity as a co-seller — or at minimum as an express performance obligor: the contract states that Company B warrants production quality and quantity, and accepts direct obligations to you, regardless of who the signing seller is.
2. Name the settlement entity in the payment clause — payments to Company C's account are expressly recognized as performance, with C's receipt discharging your payment obligation and C warranting onward settlement within the group. This turns C from a stranger into a party with defined duties.
3. Add a group representation: the seller warrants the identities of all group entities involved in performance and their authority, and warrants that performance by an affiliate counts as performance by the seller — closing the "that was B, not us" defense in advance.
4. Preserve the paper trail that maps the group from day one: every document bearing every entity's name, every payment slip, the correspondence. If piercing is ever needed, this file is the case.

Suppliers resist these clauses less than buyers expect. A group that intends to perform has nothing to fear from being bound jointly; the resistance you meet is data about which entity is expected to hold the risk — and, in the worst cases, which one is expected to hold nothing.

The Checklist

1. Map the group before signing: all entities behind the deal, their shareholders, legal representatives, addresses, capital.
2. Match contract party / invoice issuer / B/L shipper / account holder. Three different names means three names in the contract.
3. Put every performing entity on the signature page — as co-seller, performance obligor, or guarantor.
4. Regulate the payment account expressly, by name.
5. Keep the cross-entity correspondence from the first enquiry to the last shipment.
6. If a dispute starts: preserve against all related entities at filing; build personnel and business commingling first; drive toward financial records through the court.

Where a related-entity dispute is already live — an empty judgment, an enforcement that came back with nothing, a group shuffling performance between letterheads — the strategy decisions (which entities to add, when to seek appraisal, how to sequence preservation) are exactly the work of the litigation and enforcement practice. The earlier the group is mapped, the more of the structure the claim can reach.

This article is general information, not legal advice, and does not create an attorney–client relationship. Statutory references are to the Company Law of the PRC (2023 revision) as currently in force; confirm application to your case with counsel. Outcomes vary by case; nothing here is a guarantee of results.