How Tax Treatment Creates a Fifteen to Twenty Percent Cost Gap Between Network Models
“For the Chinese law firms, the tax is really calculated based on their total fees without the deduction of the cost of their foreign counsel. So that the total cost that we will charge of our clients would be normally would be like 15% or 20% higher than to use the kind of alliances structure.”
Inside Two Collaboration Models
Penghu brings a perspective rare in discussions of international legal networks: direct experience working within two fundamentally different structures. “I also worked in Dentons before,” Penghu explains, “and I think in my, if I’m correct, I think Dentons used this kind of Swiss alliances and they have a kind of internal cost and profit sharing mechanism.”
The Swiss alliance structure at Dentons employed systematic profit distribution: “to guarantee that everyone involved, all the lawyer involved, relatively fair profits.” This internal mechanism represented a formal approach to managing cross-border work, with explicit rules about how fees flow between different parts of the network.
The contrast with best friends networks is structural. “But for the best friend network, for example, for my law firm, we will usually be engaged as a lead counsel. Then we contract the local work to a local foreign law firm,” Penghu describes. The lead counsel model places one firm as the primary client contact, with other firms operating as subcontractors rather than alliance partners.
This distinction matters because it determines how costs flow and how taxes apply. In an alliance structure, each entity bills for its own work within an internal framework. In a best friends network, the lead counsel often bears the full revenue and then pays out costs—a difference that creates the tax gap Penghu identifies as fifteen to twenty percent.
The Tax Calculation That Changes Everything
Penghu quantifies what other discussants describe only generally: the actual cost differential between collaboration models. “For the best friend network… I think the biggest difference is their cost,” she explains, “because I use the best friend network quite often, and I think their cost is kind of higher than the strategic alliances.”
The mechanism is straightforward but consequential: “for the Chinese law firms, the tax is really calculated based on their total fees without the deduction of the cost of their foreign counsel.” This means if a Chinese firm bills a client 1 million RMB and pays 400,000 RMB to a foreign law firm, Chinese tax authorities calculate tax on the full 1 million rather than on the 600,000 net revenue.
The impact compounds: “So that the total cost that we will charge of our clients would be normally would be like 15% or 20% higher than to use the kind of alliances structure.” This isn’t a small efficiency loss—it’s a substantial price premium that makes best friends networks systematically more expensive for clients.
Penghu’s precision here—”15% or 20%”—reflects someone who has calculated these costs in real transactions. For a firm deciding between network models, this differential represents a significant competitive disadvantage. For clients, it’s a hidden cost that may not be apparent when comparing quotes but that ultimately flows through to the final bill.
The tax treatment also explains why larger firms might pursue alliance structures despite their complexity and overhead. If the alternative is a twenty percent cost penalty, the investment in formal alliance infrastructure becomes economically rational.
AI Platforms and the Accountability Gap
Penghu describes a shift in how Chinese companies approach routine legal work. “One of my friends who’s the in-house counsel of a listing company… told me that they have stopped, they are kind of, they have stopped hiring an external law firm for its day-to-day legal work,” she explains. The replacement: “they have recently engaged an AI legal platform company to help them to review their daily contracts and to help them to do their legal research.”
The appeal is primarily economic: “they think it’s a very kind of cost-efficient because it’s much, much more cheaper than engage external company.” The cost advantage is substantial enough to drive rapid experimentation: “I think recently many Chinese companies, like many Chinese clients, they tend to try and they tend to try with some kind of AI legal platform.”
But Penghu identifies a critical limitation: “AI, as we always said, AI cannot be held liable, right?” This accountability gap creates a secondary cost: “So they must, alternatively, kind of engage internal in-house counsel to help them to review the work that the AI does. So to some extent, it’s also a cost burden for the company who uses legal tech platforms.”
The contrast with external lawyers is clear: “if they engage a lawyer, like us, engage an external lawyer, then a lawyer will help to review all the work and I will be responsible and liable for the work they have done.” Penghu suggests this liability assumption is particularly important for certain clients: “I think these are kind of a big difference, especially for the state-owned Chinese companies. Sometimes they should have a person to be responsible for the work that they have done.”
The analysis reveals that AI adoption in legal services isn’t a simple cost-cutting story. Companies save on external fees but must invest in internal oversight. The net savings depend on the volume of work and the risk tolerance of the organization.
The International Limitation of Legal Technology
Where AI platforms show promise for domestic work, Penghu sees clear boundaries when work crosses borders. “For domestic, I think from the cost perspective, I think the legal tech platform, especially domestically, I think it’s a kind of a lowest method,” she argues.
But the calculus changes completely for international matters: “for the international ones, because I don’t, in my experience, I don’t see many that very good or many very good AI tech platforms that can handle very good international legal issues because many Chinese companies ask about, ask questions relating to a lot of jurisdictions.”
The multi-jurisdictional challenge proves insurmountable for current platforms: “So I don’t see, I don’t see there’s one legal tech platform that help the Chinese clients to solve all their questions, to answer all their questions.” The result is a bifurcated approach: “So I think they are either the best friends network that I use most often for the transactions or the strategic alliances is the kind of mass for international transactions or international issues.”
Penghu’s assessment suggests that while technology may disrupt routine domestic legal work, complex cross-border matters still require human networks. The knowledge synthesis required to navigate multiple legal systems, the judgment needed to reconcile conflicting jurisdictional advice, and the relationship management essential to coordinating across borders—these capabilities remain beyond current AI systems.
This creates a bifurcated market where technology and human networks serve different segments rather than competing directly. Firms that position themselves primarily for cross-border work may face less immediate AI disruption than those focused on routine domestic services.
Understanding What Chinese Clients Actually Need
Penghu identifies multiple factors that drive Chinese client preferences in cross-border work, starting with language. “Chinese companies, they also prefer someone who can speak Chinese,” she explains. This isn’t just convenience—it shapes the entire service delivery model.
A regulatory constraint proves equally important: “there’s also a very practical issue, which is, it’s a kind of the foreign exchange control. So sometimes the Chinese companies, they prefer to engage a Chinese law firm or a law firm with a Chinese branches or offices located here in China, so that they won’t face their kind of foreign exchange control issues.”
This foreign exchange consideration means client “preferences” are often regulatory necessities disguised as choices. A Chinese company can’t easily pay a purely foreign law firm without navigating currency controls. Engaging a firm with Chinese presence solves this problem administratively, even if it’s not the cheapest or most direct solution.
Penghu also notes client behavioral patterns: “many Chinese clients, they’re quite pushy and demanding.” This creates a need for cultural buffering: “I think if we let the Chinese client facing the California councils, there might be a very much kind of cultural difference.”
The result is significant unpaid coordination work: “In that case, so that I will receive a lot of complaints, which I cannot charge because it’s just a kind of like coordination or just like to comfort both sides.” This coordination burden—managing expectations, translating not just language but business culture, mediating between different working styles—becomes part of the service whether formally recognized or not.
The One-Stop Solution Imperative
When asked about client preferences, Penghu is unequivocal: “I agree with Bernardo. I think Chinese companies, they also prefer one-stop solutions.” She outlines multiple reasons for this preference.
Language remains central: “for many reasons, first of all, as Pengfei said, the language barriers, they prefer someone who can speak Chinese.” But the foreign exchange control issue proves equally decisive: “And secondly, I think if we just like, I say secondly, there’s also a very practical issue, which is, it’s a kind of the foreign exchange control.”
The cultural mediation requirement reinforces the one-stop preference: “And certainly I think many Chinese clients, they’re quite pushy and demanding. I think if we let the Chinese client facing the California councils, there might be a very much kind of cultural difference in between… so that I will receive a lot of complaints, which I cannot charge because it’s just a kind of like coordination or just like to comfort the both sides.”
The cumulative effect of these factors—language, regulatory compliance, cultural translation, and relationship management—makes distributed service models unattractive to clients even when theoretically possible. “In that case, personally, as a Chinese lawyer, I would also prefer a one-stop solution,” Penghu concludes.
This client preference has significant implications for network design. Firms can’t simply connect clients to a network of specialists and expect satisfaction. Someone must take responsibility for the complete experience, absorbing the language, regulatory, and cultural friction that clients won’t or can’t manage themselves.
About the Speaker: Penghu (Penny) is a Partner at Winteam Law Firm based in Shanghai, where she focuses on cross-border investment and international trade matters. Penghu has worked with hundreds of companies including foreign investment entities in China and Chinese companies expanding abroad. Her experience includes working at Dentons, providing her with firsthand knowledge of both formal alliance structures and best friends network models. Penghu brings practical insights on how tax treatment, regulatory constraints, and client behavioral patterns shape collaboration choices in the Chinese legal market. As a contributor to LexChina Forum discussions, she offers perspective on the rapid evolution of legal technology adoption in China and its limitations in cross-border work.
