Permanent Establishment Risk for Remote Teams

The short answer most employers want is the wrong one. There is no universal day-count that turns a remote worker into a taxable presence, and permanent establishment risk for remote teams does not switch on at 183 days, 90 days, or any other magic number (Thomson Reuters). A permanent establishment (PE) is what creates a corporate tax filing obligation for the employer in the worker's country — and whether one exists turns on treaty language, facts, and conduct, not a calendar.

That picture became sharper on 18 November 2025, when the OECD Council approved the 2025 Update to the Model Tax Convention. The update rewrote the Commentary on Article 5 to spell out when an employee's home can count as a "place of business" for the company that employs them (OECD). The new guidance introduces a 50%-of-working-time benchmark and a "commercial reason" test that, between them, settle a question employers had been guessing at since 2020.

This guide translates that primary text into operating rules. It separates the two PE triggers employers actually face — the fixed-place-of-business PE created by a home office, and the dependent-agent PE created by remote salespeople — and shows where national rules in the UK, US, and India diverge from the OECD baseline.

Key takeaway: Under the OECD 2025 update, an employee's home office generally does not create a permanent establishment unless they work there more than 50% of their total working time over a 12-month period AND the company has a commercial reason for the work being done in that country — convenience, cost savings, and talent retention do not count.

What is a permanent establishment, and why does it matter for remote teams?

A permanent establishment is the threshold at which a country can tax a foreign company's business profits earned inside its borders. Under the OECD model, two routes create one: a fixed place of business through which the enterprise carries on its activity, or a dependent agent who habitually concludes contracts on its behalf (OECD).

The stakes are concrete. A PE forces the employer to register for corporate tax in the worker's country, file returns there, and attribute a share of group profit to that location. It can also trigger payroll, social security, and transfer-pricing obligations. For a company that hired one engineer in Lisbon thinking it was a simple payroll question, the discovery of a PE converts a single headcount into a foreign tax footprint.

Remote work raised the question because the old Commentary assumed places of business were offices the company controlled. When millions of employees moved to kitchens and spare bedrooms, tax authorities had to decide whether a home an employer never sees, rents, or controls could still be "at the disposal of" the enterprise. The 2025 update is the OECD's answer.

[UNIQUE INSIGHT] The framing matters more than most coverage admits. PE is an employer problem, not an employee problem. The remote worker rarely owes extra tax personally from creating a PE; it is the company that inherits a filing obligation and a profit-attribution argument it never planned for. That asymmetry is why HR-led "work from anywhere" policies so often outrun the tax function.

Does a home office create a permanent establishment under the OECD 2025 update?

Usually not — and the 2025 update finally puts a number on "usually." If an individual works from home for less than 50% of their total working time over any 12-month period commencing or ending in the fiscal year, that home is generally not a fixed place of business, so no PE arises (EY). The test looks at what actually happens, not at what the contract says.

The conduct-based design is deliberate. A clause in an employment agreement stating that the home is "not a place of business" does not control the outcome; the regular pattern of where the work is genuinely performed does (EY). An employer cannot paper over a full-time home-based role with a recital.

Crossing 50% does not automatically create a PE, though. It only moves you to the second gate — the commercial-reason test below. A worker who spends most of their time at home still escapes PE if the company has no commercial reason for the activity being performed in that country.

The commercial reason test

Once the 50% threshold is met, the home office risks PE only if the business has a commercial reason for the work being done in that country — for example, the employee meets local customers or suppliers there (Deloitte). The test asks why the work happens in that location, not merely how often.

What does not count is the part employers most need to absorb. Offering remote work for employee convenience, to retain talent, to save on office rent, or to cut costs is not a valid commercial reason (Deloitte). If your German engineer works from Munich because she wants to live near family, that is convenience. If she works from Munich because that is where your German customers are and she visits them, that is a commercial reason — and the PE risk is real.

How do you calculate the 50% working-time threshold?

You measure actual working time, across a rolling 12-month window, against where the work is performed. The OECD frames the benchmark as less than 50% of total working time over any 12-month period commencing or ending in the fiscal year concerned (EY). The rolling window is the trap most planners miss — it is not a clean tax-year reset.

Two design choices make the test workable. First, it is conduct-based, so it follows real attendance records, not job descriptions. Second, it is proportional rather than absolute, so a part-time employee and a full-timer are measured against their own total working time, not a fixed hours figure.

The table below shows how the two gates combine. Both columns matter; a PE generally needs both a "yes" on time and a "yes" on commercial reason.

Scenario	>50% of working time at home?	Commercial reason in that country?	Home office PE likely?
Engineer works from home for personal convenience, no local clients	Yes	No	No
Salesperson works from home and meets local customers there	Yes	Yes	Yes
Employee occasionally works from home, mostly at HQ abroad	No	Either	No
Consultant based at home serving the local market the firm targets	Yes	Yes	Yes

Source: thresholds and commercial-reason test per EY and Deloitte. Outcomes are general illustrations; treaty-specific facts can change the result.

[PERSONAL EXPERIENCE] In reviewing distributed-team policies, the cleanest cases are the boring ones: a back-office or product role, performed at home for lifestyle reasons, with no customers or suppliers in the country. Those sit comfortably inside the safe zone. The files that keep tax directors awake are sales, account management, and "founder-adjacent" roles where the person both lives somewhere and sells there.

When does a remote salesperson create a dependent-agent PE?

This is the trigger that catches employers who think they have no office abroad at all. A dependent-agent PE arises when a person acting for the enterprise habitually concludes contracts, or habitually plays the principal role leading to contracts that the enterprise then concludes without material modification (Osler). No premises are required — the agent's activity is the PE.

That second limb is the post-BEPS expansion. Before BEPS Action 7, a salesperson who negotiated every term but routed final signature through head office could often avoid PE through a commissionaire arrangement. The widened test closed that gap: habitually playing the principal role in routine, rubber-stamped contracts now counts (Osler).

For remote teams, the practical line is between selling and supporting. A remote employee who answers product questions, demos software, or relays leads to a home-country sales desk is on safer ground. A remote employee who negotiates pricing, agrees terms, and effectively closes deals — even if a manager abroad signs — is squarely in dependent-agent territory.

Selling vs. supporting: where the line sits

The activity that matters is who drives the contract, not who holds the pen. If your remote salesperson habitually leads negotiations that head office routinely approves unchanged, that pattern can create a PE regardless of where the signature lands. Compare jurisdictions and their treaty positions on our jurisdiction directory before placing client-facing staff abroad.

Independent agents are different. A genuinely independent intermediary acting in the ordinary course of its own business — a third-party distributor, not your employee — generally does not create a PE for the principal. The distinction turns on real independence, both legal and economic.

How does the UK treat permanent establishment for remote workers?

The UK codifies the same two routes but adds firm registration mechanics. A company has a UK permanent establishment if it has a fixed place of business through which its business is wholly or partly carried on, or if an agent acting on its behalf habitually exercises authority to do business there (GOV.UK). Both limbs mirror the OECD model.

The compliance deadline is where the UK gets specific. A non-UK company trading through a dependent-agent permanent establishment must register for UK Corporation Tax within three months of becoming chargeable to it (GOV.UK). That is a hard clock, and it starts when liability begins — not when the company notices.

For a foreign employer with a UK-based remote closer, the sequence is unforgiving. The salesperson's habitual contract activity can create the PE; the three-month registration window then runs; and a missed deadline brings the usual late-filing and late-payment consequences. Treaty relief may limit how much profit the UK can actually tax, but the registration obligation is procedural and applies regardless.

How do US ECI rules change the analysis?

The US runs a two-layer system that does not map cleanly onto "PE." When a foreign person operates a US trade or business, US-source income connected to it becomes Effectively Connected Income (ECI), taxed at graduated rates (IRS). The trade-or-business threshold is the first layer, and it can be crossed without a treaty PE.

The treaty layer sits on top. Under US tax treaties, the US may tax business profits only to the extent they are attributable to a US permanent establishment, and that PE threshold is generally higher than the trade-or-business threshold (IRS). So a treaty-country employer might have ECI exposure on paper yet owe no US tax because no PE exists under the treaty.

[UNIQUE INSIGHT] The two-layer structure produces a counterintuitive result employers should plan around. A company from a non-treaty country gets no PE shield, so ECI rules can bite at the lower trade-or-business level. A company from a treaty country gets the higher PE threshold as protection. For US-inbound remote hiring, your home country's treaty network with the US matters as much as the worker's day count.

Why is there no universal day-count that creates a PE?

Because PE depends on the type of activity and the treaty, not on time alone. There is no universal magic number of days that automatically creates a permanent establishment; thresholds vary by treaty, depend on facts and circumstances, and turn on whether the worker's presence serves a commercial reason in that country (Thomson Reuters). The popular "183-day rule" answers a different question entirely.

The 183-day figure governs individual tax residency and the dependent-personal-services article — whether the employee owes income tax in the host country. PE governs whether the employer owes corporate tax there. Conflating the two is the single most common error in remote-work tax planning, and it produces both false comfort and false alarm.

Some treaties do use time-based tests for specific cases, most visibly construction and building sites, which often carry a 12-month presence rule. But that is a narrow, activity-specific provision, not a general PE clock. For an office worker at home, time is only one input into the 50% test, and even crossing it does nothing without a commercial reason.

Where does national implementation diverge from the OECD guidance?

The OECD Commentary is influential, not binding, and several countries decline to follow it. India does not accept the new home-office tests, and Israel and other countries maintain differing interpretations, so local rules can diverge from the OECD position (EY). A clean answer under the 2025 update can still produce a PE under a reserving country's domestic practice.

This divergence is the operational risk that survives the update. The OECD harmonized the model, but each treaty partner applies its own version, and reservations carry through to how its tax authority audits. A distributed team spread across compliant and reserving countries faces two different rulebooks at once.

The planning implication is jurisdiction selection. Placing remote staff in countries that follow the OECD baseline narrows the gap between the rule and the audit. Several low-friction hubs that distributed teams favor — Estonia, Cyprus, Ireland, and Portugal — sit inside well-developed treaty networks; weigh their positions side by side on our comparison tool and model the corporate-tax impact with the calculator.

Frequently asked questions

Does one remote employee abroad create a permanent establishment?

Rarely on its own. A single back-office employee working from home for personal convenience, with no local customers or suppliers, generally fails both the 50% and commercial-reason tests, so no PE arises (Deloitte). Risk rises sharply when that employee sells to or serves the local market. Read more on our blog.

Is 183 days the trigger for a permanent establishment?

No. The 183-day rule concerns an individual's tax residency and personal-services taxation, not the employer's PE. PE has no universal day-count; it depends on treaty language, facts, and commercial reason (Thomson Reuters). Treating 183 days as a PE switch is the most common planning mistake.

Does the OECD 2025 update apply automatically to my treaties?

Not instantly. The OECD Council approved the update on 18 November 2025, but it revises a model and its Commentary (OECD). How quickly it shapes a given treaty depends on each country's adoption and any reservations, such as India's non-acceptance of the home-office tests (EY).

What happens if my company creates a PE in the UK?

A non-UK company trading through a dependent-agent PE must register for UK Corporation Tax within three months of becoming chargeable (GOV.UK). Profits attributable to the PE are then taxable in the UK, subject to any treaty relief. Missing the deadline brings standard late-registration consequences.

Are remote salespeople riskier than remote engineers?

Generally, yes. A salesperson who habitually concludes contracts, or habitually leads negotiations that head office approves without material change, can create a dependent-agent PE with no premises at all (Osler). An engineer doing back-office work for convenience is far less exposed under the 50% and commercial-reason tests.

Bottom line for distributed teams

The 2025 update is good news handled carefully and a liability ignored. For most remote roles — back-office, product, engineering performed at home for lifestyle reasons — the 50% threshold and commercial-reason test leave the employer outside PE, because convenience and cost savings are explicitly not commercial reasons (Deloitte). That is a workable safe harbor for genuine "work from anywhere" policies.

The exposure concentrates in two places: client-facing roles that sell into the local market, and countries that decline the OECD tests, such as India (EY). Map your roles against both gates, treat salespeople and reserving countries as the high-risk quadrant, and stop relying on day-counts that were never the rule. Compare treaty-friendly bases on our jurisdiction directory and read related analysis on the blog.

Disclaimer: This article is for informational purposes only and does not constitute tax, legal, or financial advice. Tax laws change frequently and vary by jurisdiction. Consult a qualified professional before acting.