Permanent establishment for a foreign company: cost and likely outcome depends on how the Swedish Tax Agency reads the activity actually carried out in Sweden, not on the parent's size abroad. Where a fixed place of business or a dependent agent is found, extra tax and appeal costs follow; weaker facts often mean the assessment is contained before it reaches court.
Who this concerns
The question surfaces most often for foreign groups that run Swedish-facing operations without a registered branch or subsidiary: a construction or installation project staffed from abroad, an e-commerce operation using a Swedish warehouse and local staff, a sales team that negotiates and signs on the group's behalf, or a consulting team that works from a Swedish address for months at a time. None of these arrangements were set up to create a taxable presence, and in most cases nobody asked the question until Skatteverket (the Swedish Tax Agency) did.
It also concerns groups that assumed the question was settled once, at incorporation or at the start of a project, and never revisited it as the arrangement grew: a warehouse lease that was renewed, an agent whose authority expanded informally, a team that stayed longer than planned. The activity that triggers the assessment is rarely the activity that was originally described in the group's internal structuring memo.
Where the controlling parent sits outside Sweden, the mechanics change in three ways: the underlying contracts and internal approvals usually need to be gathered and translated from another jurisdiction, the group's global transfer pricing position has to be reconciled with whatever profit Sweden claims to attribute, and a bilateral tax treaty may cap or redirect Sweden's taxing right depending on where the group is resident for treaty purposes. None of that removes the exposure; it changes what has to be assembled to answer it.
This sits alongside the broader run of tax dispute assessments and appeals we handle for foreign groups, and the same logic applies whether the question arrives as an information request, a proposed decision, or a completed assessment.
What the law says
Under Swedish law as it currently stands, what is assessed is usually described as fast driftställe, the Swedish equivalent of permanent establishment: a fixed place of business through which the foreign company's activity is wholly or partly carried on in Sweden, or a person acting on the company's behalf who habitually concludes contracts that bind it. Neither test turns on formal registration; both turn on what actually happens on the ground.
Once a fixed place of business or a dependent agent is established, Sweden's taxing right extends to the profit that is properly attributable to that presence, calculated as if it were a separate and independent enterprise dealing with the rest of the group at arm's length. Where a double tax treaty applies, the treaty version of the test can narrow or reshape the domestic one, and the treaty's allocation article then governs how much of the profit Sweden may actually tax.
The two tests do not always point the same way. A group can find that domestic law would treat an arrangement as a permanent establishment while an applicable treaty exempts it, or the reverse, and the treaty position has to be checked separately rather than assumed from the domestic result.
There is no published register of court decisions we can point to for how these tests have been applied in individual cases, and none is cited here. What follows is how the assessment is typically built and contested in practice, not a summary of case outcomes.
How it works in practice
The assessment moves through a fairly predictable sequence, and the outcome usually turns on decisions made early, often before the company realises a dispute has started.
How the inquiry usually starts
Most cases begin with an information request rather than a formal decision: Skatteverket asks for organisational charts, contracts, correspondence about who negotiates what, and evidence of where decisions were actually made. How this request is answered sets the frame for everything that follows, because the agency builds its position from the documents and statements it receives at this stage, not from a fresh investigation months later.
How this differs from a routine tax audit
A general audit checks whether the numbers already reported are correct. A permanent establishment inquiry asks a prior question: whether anything should have been reported in Sweden at all. That distinction matters for how the company should respond, because conceding facts that answer the audit question can inadvertently answer the presence question too, in a way that is hard to unwind later.
The fixed place of business test
The agency looks for a place, not necessarily an office: a warehouse, a construction site, a room made available to staff, or even a home used regularly for the company's business can qualify if the activity carried out there is not merely preparatory or auxiliary to the group's main business. Preparatory and auxiliary functions, storage, display, or purchasing for the group, sit outside the test; sales, negotiation, and service delivery generally sit inside it.
The dependent agent test
Separately, a person in Sweden who habitually negotiates the material terms of contracts, or concludes them in the company's name, can create a permanent establishment even without any fixed premises at all. Job titles do not decide this; what the person actually does, and how much discretion they exercise, does.
Profit attribution once a presence is found
Finding a permanent establishment answers only the first question. The harder one is how much profit belongs to it: what functions, assets and risks sit with the Swedish presence, and what would an independent enterprise performing those same functions have earned. This is where transfer pricing analysis becomes directly relevant, because the same documentation that supports a group's transfer pricing position is usually what the agency relies on to attribute profit to a Swedish presence.
What to check before responding
- Whether any Swedish location, however informal, is used for anything beyond storage, display or purchasing.
- Whether any person in Sweden, staff or agent, has actually negotiated or signed contracts on the company's behalf, and how often.
- Whether the group's board minutes or internal approvals describe Swedish activity differently from how it is described to Skatteverket.
- Whether a double tax treaty with the company's country of residence applies, and what its permanent establishment article actually says.
- Whether transfer pricing documentation already exists for the group, and whether it accounts for the Swedish activity at all.
- Whether prior correspondence with Skatteverket has already conceded facts that narrow the company's room to argue later.
- Whether the individuals who will answer the information request understand the legal significance of what they are being asked.
Where the position gets built for appeal
If Skatteverket issues a formal decision, that decision, not the earlier correspondence, is what an appeal to the administrative court has to be built against. The facts assembled at the information request stage rarely change; what changes is how they are characterised, and that characterisation work is what determines whether an appeal is worth pursuing at all.
The same question comes up in a few related contexts:
Does a foreign parent need an investment screening review before a Swedish subsidiary is set up?
Investment screening and permanent establishment are separate questions, but they often arrive together when a foreign group first structures a Swedish presence. Screening looks at ownership and sector sensitivity before a transaction closes; permanent establishment looks at ongoing activity after the structure exists. Groups planning investment screening before closing a foreign transaction should map the operational footprint at the same time, since the same facts often answer both.
How does transfer pricing documentation interact with a permanent establishment assessment?
Transfer pricing documentation describes how a group prices transactions between related parties; a permanent establishment assessment asks whether a Swedish presence exists at all and, if so, how much of the group's profit belongs to it. Where documentation already exists, it usually becomes the starting point for attributing profit once a presence is found, which is why groups should check what changed in transfer pricing documentation requirements against their Swedish activity specifically.
Can Swedish assets be traced and recovered if a permanent establishment assessment turns into an unpaid tax debt?
Yes, and the mechanics do not stop at the border. Once an assessment is final and unpaid, enforcement can extend to assets the company or its officers hold outside Sweden, following the same cross-border logic used in asset tracing and recovery across jurisdictions. This is one reason contesting the underlying assessment early, before it becomes a final debt, is usually cheaper than contesting enforcement afterwards.
The numbers
Swedish law does not set a bright-line day count or turnover threshold for when a fixed place of business becomes a permanent establishment; the assessment is a facts-and-circumstances test, not an arithmetic one. What actually drives cost and duration is the volume of activity under review, the number of tax years Skatteverket reopens, whether a double tax treaty narrows the domestic test, the state of the group's existing transfer pricing documentation, and how much of the underlying record has to be gathered and translated from outside Sweden.
A case built on a single, well-documented year with a treaty in place typically sits at a different scale of cost than one spanning several years with agent activity spread across multiple people and no contemporaneous documentation at all. Scope widens further when the information request answers were inconsistent, or when the same activity touches more than one Swedish entity in the group.
Neither figure is quoted here, because neither is fixed by statute; both are set by the specific record, and that record is what a first review has to establish before either number can be given honestly.
Where it usually goes wrong
The most common mistake is treating the initial information request as routine correspondence rather than the start of the case. Answers given informally, often by someone with no visibility into how the Swedish activity is actually structured, become the record the agency builds on; walking them back later is far harder than getting them right the first time.
The test also has real boundaries, and the assessment can go the other way when the Swedish presence is genuinely limited to preparatory or auxiliary functions: a warehouse used only for storage without any sales activity, a representative office confined to market research, or an agent who negotiates in the ordinary course of an independent business rather than on the foreign company's behalf. Where the facts fit one of these, the correct outcome is that no permanent establishment exists at all, and the work is to demonstrate that clearly rather than to negotiate a smaller assessment.
A separate error is importing a foreign benchmark. A threshold or exemption that applies under the company's home law, or under a different country's tax treaty, does not transfer automatically to how Sweden reads the same facts; each treaty's permanent establishment article has to be checked on its own terms, and the domestic Swedish test applies in full where no treaty exemption is available.
A third error shows up specifically in groups managed from abroad: assuming that because the parent's own tax department signed off on the Swedish arrangement, the arrangement is defensible in Sweden on the same reasoning. Sweden's agency and courts apply the Swedish test to the Swedish facts, and a characterisation that satisfied a foreign tax authority carries no independent weight here.
What to do next
A first review can usually establish, within the existing documentation, whether the activity described to Skatteverket supports a permanent establishment finding, what profit a presence like this would realistically attract, and whether the information already on file helps or hurts the position. That is the point where self-assessment runs out and a structured case assessment becomes the more efficient next step, particularly where a formal decision is imminent or has already been issued.
Where a decision has already been issued and the company is deciding whether to appeal, the relevant next question is addressed separately in what an appeal to the administrative court costs and is likely to achieve.