Good day, I'm Teacher Liu from Jiaxi Tax & Finance Company. Over the past 12 years serving foreign-invested enterprises and 14 years handling registration procedures, I've seen more than a few managers scratch their heads over this one decision: "Should we set up a branch or a subsidiary in China?" It's a question that often leads to late-night debates and frantic calls to your tax advisor. The difference isn't just about a name on a door; it's a fundamental fork in the road that can dramatically alter your tax burden, your operational flexibility, and even your exit strategy. Inbound investment rules are changing all the time, and getting this wrong can cost you a fortune. This article will cut through the noise, sharing real cases from my own practice and highlighting the specific, sometimes surprising, tax differences between these two structures. Our goal is to arm you with the knowledge to make a decision that aligns perfectly with your business strategy, not just your immediate tax savings.
税收居民身份差异
Let's start with the most fundamental concept: tax residency. In China, a subsidiary, being a separate legal entity incorporated under Chinese law, is typically considered a Chinese tax resident. This means it's liable to Corporate Income Tax (CIT) on its worldwide income, just like any local Chinese company. However, the practical application is often more nuanced. For many foreign investors, the subsidiary's business is primarily focused within China, so the "worldwide income" concern is minimal. But it's a critical point when you consider the tax implications of passive income, like dividends from an overseas parent or interest from foreign bank accounts.
A branch, on the other hand, is not a separate legal entity. It's an extension of the foreign parent company. In China, a branch is generally treated as a non-resident enterprise. This distinction is massive. The branch is only liable for CIT on income that is "effectively connected" with its permanent establishment (the branch itself) in China. This "effectively connected" principle is where the rubber meets the road, and frankly, where I've seen many companies get tripped up.
I recall a case from about five years ago. A German manufacturing firm set up a branch in Shanghai simply to do market research and coordinate with suppliers. They had no sales in China, just costs. The parent company thought, "No revenue, no tax, no problem." But they also had a German engineer who came to Shanghai for six months to oversee a supplier's quality control. The engineer signed a few minor contracts for the parent while at the branch office. The tax bureau in Pudong argued that this activity was "effectively connected" to the permanent establishment, and they re-characterized a portion of the parent's global profit as taxable in China. The branch ended up with a significant tax bill plus penalties. It was a classic case of underestimating the "nexus" rules. So, when you choose a branch, you're not just choosing a structure; you're choosing a complex relationship with the Chinese tax authority about what constitutes your taxable footprint.
利润汇回税务处理
This is the aspect that usually gets the CFO's attention right away. When a subsidiary makes a profit and wants to send that cash back to its foreign parent as a dividend, there's a clear tax process. The subsidiary first pays its 25% CIT (or preferential rate if it's a high-tech or small low-profit enterprise). Then, when it distributes the after-tax profit as a dividend, it must withhold a 10% dividend withholding tax (WHT) for the parent, unless a tax treaty reduces that rate. For example, the treaty with Hong Kong often reduces it to 5%, but only if the parent is the "beneficial owner" and meets substance requirements. This is a classic, predictable tax cost.
For a branch, the calculation is completely different. A branch's profit, after paying the 25% CIT in China, is generally considered repatriated to the head office automatically. There's no additional dividend WHT at the moment of remittance. This sounds great on paper, doesn't it? You save that 10% or 5% WHT. But hold on. The benefit is often illusory. While you avoid the second layer of tax in China, the profit remitted to the head office becomes taxable in its home country. If the head office's home country has a territorial tax system (like Hong Kong or Singapore), the remitted profit might indeed be tax-free. But if the parent is in a country with a worldwide tax system (like the US), the profit is subject to tax there, with a foreign tax credit for the CIT paid in China.
Let me give you a personal insight from my experience. A large US tech company came to me, thrilled that their branch structure in Beijing could save the 10% dividend WHT. They were all set to go. I told them, "Let's look at the big picture." We modeled their 3-year plan. After accounting for the US corporate tax rate (which was higher than China's at the time) and the foreign tax credit limitations, the branch actually resulted in a higher global effective tax rate than a subsidiary would have, because the dividends from a Chinese subsidiary are often eligible for a 100% dividends-received deduction in the US under certain conditions. The branch's "savings" in China were completely eaten up by the higher US tax on the remitted profits. So, never look at the China tax cost in isolation. You must model the entire global tax impact. The branch's repatriation advantage is a "saving" that can quickly turn into a "cost" depending on your parent's tax regime.
增值税处理差异
Many people focus solely on Corporate Income Tax and forget about Value-Added Tax (VAT). In my 14 years of doing this, VAT is often the bigger operational headache. A subsidiary is a standard VAT taxpayer, just like any Chinese company. It issues "Special VAT Invoices" (Fapiao), which allow its customers to claim input VAT credits. This is crucial in a B2B environment where your Chinese clients need to offset their own VAT liability. If your subsidiary is a "general taxpayer" (which it almost always will be), you can fully participate in the VAT deduction chain.
A branch can also register for VAT and become a general taxpayer. However, the complication arises from its lack of independent legal personality. For cross-province operations, this is a nightmare. A branch in Shanghai and a branch in Guangzhou are not separate entities. They cannot issue VAT invoices to each other. If your head office in Shanghai procures goods that are stored in a warehouse in Guangzhou, you are dealing with a "cross-province transportation" or "branch-to-branch" movement of assets. Under Chinese VAT law, this can be deemed a taxable sale if the goods are ultimately sold by the Guangzhou branch. This means you have to issue a VAT invoice from the Shanghai branch to the Guangzhou branch, which is effectively a transaction between two parts of the same legal entity. This creates a huge administrative burden.
I remember a logistics client, a Japanese company, who set up a branch in Qingdao to handle exports. Their main operation was in Tianjin. They initially thought, "One company, one VAT registration, easy." But when goods moved from Tianjin to Qingdao for export, the Tianjin tax bureau wanted a VAT output tax on that internal transfer, treating it as a sale to the Qingdao branch. The Qingdao branch then had to claim the input tax credit. It was a circular mess of reconciliation, paperwork, and massive "Advance Payment" tax adjustments that tied up cash flow for months. A subsidiary structure would have completely avoided this "internal transaction" trap. If you have a multi-location operation, the VAT implications of a branch can be a silent profit killer. You are not just dealing with tax law; you are dealing with a very rigid invoicing system.
亏损弥补与税务优惠
Let's talk about the silver linings and the financial pitfalls. A subsidiary is its own taxpayer. If it incurs losses in its first year, it can carry those losses forward for up to 10 years to offset future profits (for qualifying high-tech enterprises and small low-profit enterprises, the carryforward period is even longer). This is a huge benefit for startups or companies entering a new market. The subsidiary's losses are independent and belong only to it.
A branch, however, is a different animal. Because it's not a separate legal entity, its profits and losses are generally consolidated with the foreign parent's global results for its home country tax return. For China-specific CIT, the branch's losses can only be carried forward and offset against its *own* future China-source taxable profits. But here's the catch: If the branch closes down, those unrelieved losses are generally gone. They cannot be transferred to the parent or any other entity. This is a significant risk.
Furthermore, when it comes to tax incentives, the subsidiary wins hands down. China offers a wide array of tax holidays, reduced rates (e.g., 15% for High and New Technology Enterprises or HNTE), and super deductions (e.g., 100% additional deduction for R&D expenses). A subsidiary can apply for these benefits independently. For instance, if your subsidiary develops software, it might qualify for a "two-year exemption, three-year half-rate" on CIT. A branch, however, often struggles to qualify for these incentives. The tax authorities are extremely strict about granting HNTE status to a branch because it's not a "resident enterprise" in the same legal sense. The branch's R&D activities are seen as being conducted by the foreign parent, not an independent entity. If your business model relies on local R&D or using preferential tax policies, a subsidiary is almost always the right choice. The branch's inability to access these benefits can be a deal-breaker, even if the initial tax setup is simpler.
常设机构认定风险
This is a topic I discuss with every client who leans towards a branch. The concept of a "Permanent Establishment" (PE) is the bedrock of international tax law, and choosing a branch means you are explicitly creating one. But the danger is that you might accidentally create a PE even if you don't want one. If you set up a branch formally, the tax authorities know exactly where to look. Every piece of revenue generated by that branch is subject to Chinese tax. The risk is low in terms of misinterpretation—you've declared your presence.
The risk is paradoxically higher if you try to operate informally for a while before setting up a legal entity. I've seen foreign companies send a team of salespeople to China for 183 days. They rent an apartment, sign a few contracts, and think they are not taxable. The Chinese tax law, consistent with the OECD Model Tax Convention, says that if a non-resident enterprise has a "place of business" or a "dependent agent" habitually concluding contracts, they have a PE. If that PE is not registered, all the profit attributed to it is taxable, plus penalties and interest. This is what we call a "shadow PE."
So, my advice is always black and white: if you choose a branch, be ready for the full compliance burden. You cannot hide. The administrative cost is higher because you must maintain Chinese GAAP accounting for the branch, file tax returns, and deal with the local tax bureau's scrutiny. I recall a British engineering firm that had a "service project" office in Shenzhen. They thought it was a "short-term project" and didn't register. After 12 months, the local tax bureau discovered them through a tip-off. The tax bureau audited them and determined that the project had created a PE from day one. The penalty was 0.5% of the total project value per month of non-registration. That's a brutal penalty. So, when you think about a branch, you are not just choosing a tax structure; you are choosing a level of tax audit exposure that a subsidiary, with its clear corporate veil, does not carry to the same degree.
转让定价与合规要求
Transfer pricing (TP) is another arena where the branch and subsidiary differ greatly. For a subsidiary, transfer pricing is a major compliance focus. You must prepare a Local File, Master File, and Country-by-Country Report if you exceed certain thresholds. The subsidiary must ensure that all related-party transactions (with the parent or other related companies) are at arm's length. The tax authorities scrutinize this aggressively to prevent profit shifting.
A branch, however, presents a unique TP challenge. Since the branch and the head office are the same legal entity, there are no "related-party transactions" in the traditional sense. There are only "internal dealings" or "head office allocations." However, China's tax law still requires the branch to use an arm's length profit attribution. How do you determine the profit of the branch? You must attribute assets, risks, and capital to it. This is called the "Authorized OECD Approach" (AOA). It is incredibly complex. You have to create a "functional analysis" of the branch as if it were a separate enterprise.
In my practice, I've seen many companies with branches drastically understate their taxable profit by allocating too many costs to the head office. The tax bureau is wise to this. They will often use a "deemed profit method" or apply a minimum profit margin to the branch's turnover, especially for service industries. For example, if a branch provides consulting services, the tax bureau might deem a 30% profit margin on its costs, regardless of the books. This creates a huge tax liability that is hard to dispute. The subsidiary is almost always better from a TP perspective because the rules are clearer and you have more control over the documentation. The branch is a "black box" that the tax authorities love to open.
Let me give you a piece of advice from a practical standpoint: if you must use a branch, hire a professional to do a proper "profit attribution" analysis upfront. Don't just use a simple cost-plus method. The cost-plus method often fails because it ignores the value of the intangibles and the functions the branch performs. I had a Swiss client whose branch in Nanjing was doing high-level R&D. They allocated only a 5% markup on costs to the branch. The tax bureau audited them and said, "The R&D you are doing is creating value for the entire global group. The branch must be attributed a share of the global profit from those intangibles." The result was a significant adjustment and a lengthy negotiation. So, the message is clear: the compliance burden for a branch's internal accounting and tax filing is not simpler; it's often more complex and more judgmental.
结论与前瞻性思考
To summarize, the decision between a branch and a subsidiary is not a simple tax calculation. It's a strategic business choice. The subsidiary offers legal isolation, access to local tax incentives, clear transfer pricing rules, and a predictable dividend WHT. It's generally the right choice for long-term, substantive operations in China. The branch offers the illusion of simplicity and a potential saving on dividend WHT, but it comes with higher audit risk, complex profit attribution rules, inability to use tax holidays, and messy cross-region VAT issues.
My personal view, after 14 years of doing this, is that the subsidiary is the default winner for 90% of foreign companies. The branch is only suitable for very specific situations, such as a pure representative office (which is now heavily restricted), a short-term project with a clear exit, or a structure where the home country's tax system makes the branch more favorable (e.g., a country with a very high territorial exemption). Looking forward, I see the Chinese tax authorities getting even more sophisticated in auditing branches. The digitalization of tax administration ("Golden Tax System Phase IV") means that every transaction, every Fapiao, and every internal journal entry is monitored in real-time. The opacity of a branch's accounting is becoming a liability, not an asset. The future trend is towards more transparency. Therefore, when you are planning your China entry, think beyond year one. Think about your exit strategy, your future financing, and your long-term IP protection. The subsidiary, with its clear corporate structure, provides a much better platform for all of these future needs.
Jiaxi Tax & Finance's Insights:
Based on our extensive experience guiding foreign companies through the "branch vs. subsidiary" maze, we believe the core insight is often overlooked: the choice is less about the current tax rate and more about the operational friction and exit complexity. A branch may save a few percentage points on dividend WHT in the short term, but it creates significant friction in daily operations—especially around VAT, cross-region logistics, and internal cost allocations. Moreover, closing a branch in China is notoriously bureaucratic and can take years, while dissolving a subsidiary, though also complex, follows a clearer legal path. We strongly advise clients to build a "10-year tax model" that includes potential exit costs, inflation, and regulatory changes before making the final call. The administrative burden of a branch is often a "hidden tax" that many businesses fail to account for.
