An exit tax is not a period of time. It is an instant. In most systems that impose one, the taxable event occurs at a single moment, either as residence ends or in the instant immediately before it, and everything that determines the size of the charge is fixed at that instant: which assets are inside the net, what they are worth, who owns them, and through what structure. Before that instant almost everything is adjustable. After it, almost nothing is.
That asymmetry is the whole subject. Advisers meet the same conversation repeatedly, and the difference between the two versions of it is not sophistication but timing. In one version, a founder or an investor arrives with a move planned for the following year and a set of holdings that can still be arranged. In the other, the client has already landed in Europe, has already stopped being a resident of the country he left, and wants to know what can be done now. Those two conversations produce very different bills, and the second one produces a shorter list of options every month it is postponed.
A tax with a date attached
Most planning problems are continuous. Transfer pricing can be adjusted, a financing structure can be refinanced, a holding chain can be reorganized over several years as circumstances change. Exit taxation does not behave like that, and treating it as though it does is the underlying error.
The taxing event is discrete. It happens once, on a day that can usually be identified afterwards to the hour, and it produces a computation that is closed. Whatever relief, deferral or instalment mechanism the departure state offers, it operates on a liability that has already crystallized. Deferral changes when the money is paid. It does not change what was measured or when. Norway assesses the charge on departure and then allows twelve years to pay it, which alters the cash flow and nothing else.
This is why the standard sequence people follow, move first and structure afterwards, produces the worst outcome available. The move is the event. Structuring after it means structuring after the measurement, at which point the planning is no longer about the size of the charge but about how to fund it. An American founder who gives up the passport after arriving in Europe meets the same problem in a sharper form, because the charge follows the status rather than the address.
What the crossing makes irreversible
Three things in particular stop being negotiable once tax residence has ended.
The first is the identity of the taxpayer who holds each asset. Whether shares are held personally or through a holding company, whether an interest sits in a trust, a partnership or a foundation, and where that vehicle is resident, all determine whether an asset falls inside the deemed disposal at all. Moving an asset from one holder to another is itself usually a taxable event. Doing it before departure means doing it in a known system with known consequences. Doing it after means doing it while the asset may already be tied to the departure state.
The second is the composition of the estate at the measuring date. A gain that has been realized in cash before departure is a gain the departure state taxes on its ordinary terms, often with reliefs that a departing person no longer qualifies for. An unrealized gain sitting in a private company at the measuring date is taxed on a valuation rather than a price.
The third is anything that depends on a status which lapses with residence. Participation reliefs, rollovers, discount regimes, business asset reliefs and residence based exemptions almost always require the taxpayer to be resident at the time of the event, and several of them are denied entirely rather than proportionately to someone who has already left. Relief that would have been available in full in one month is unavailable in the next, with no partial version in between.
The departure date becomes a valuation date
The single most underestimated consequence of leaving is that the date of departure becomes a valuation date for every unlisted asset a person owns.
For a listed portfolio this is trivial, although Denmark shows that an easy valuation is not the same as an easy bill. For a private company, an intellectual property portfolio, an interest in a fund with an infrequent net asset value, a property, or a shareholding subject to a shareholders agreement with transfer restrictions, it is not trivial at all. Someone has to establish what those assets were worth on a specific day, and that figure will be the basis of a tax assessment, will be reviewed by an authority with the benefit of hindsight, and may be revisited years later when the asset is actually sold at a price that is now known.
A valuation prepared before the date, by a valuer who was instructed properly, with the company’s own information as it stood at the time, is a defensible document. A valuation prepared two years later, from memory, when the company has since raised money at a much higher price, is an invitation to a dispute. The cost difference between the two exercises is small. The difference in outcome is not.
Evidence is assembled before, never reconstructed after
The same logic runs through everything the departure state and the arrival state will later ask for.
Cessation of residence is a question of fact in most systems, and the facts have to be provable: when the home was given up, when the family moved, when the arrival country’s residence began, where the taxpayer actually was on which dates. Deferral and instalment regimes frequently carry continuing obligations, security to be provided, annual confirmations of address and continued entitlement, notification within short deadlines when a triggering event occurs. Failure on any of those points can accelerate a liability that was otherwise sitting quietly.
None of this is difficult. All of it is nearly impossible to assemble retrospectively. The single most valuable deliverable in a departure file is often not a piece of clever structuring but a dated record of what was true and what was worth what, made while it was still easy to prove.
The arrival structure has to exist before it is needed
The other half of the problem is that the arrival country is also measuring something on the day of arrival, and it is not obliged to wait for the departing person to finish reorganizing.
Where an arrival state grants a step up, it grants it on the assets held on the day residence begins, valued as they stand on that day. The Netherlands does this for shareholdings under article 4.25 of the Wet inkomstenbelasting 2001, in force in its current form from 1 January 2026: a person who comes to live in the Netherlands takes an acquisition price equal to market value at that moment, so pre arrival growth is outside the Dutch base. The provision has exceptions, notably for someone who previously lived in the Netherlands or was already a non-resident taxpayer there, which is exactly the sort of detail that changes the answer for a returning national.
The practical consequence is that assets held through the right vehicle on the day of arrival receive one treatment, and assets moved into that vehicle three months later receive another. A holding company that will exist eventually is worth nothing on arrival day. A structure that is incorporated, funded, governed and actually holding the shares before residence begins is a different instrument entirely. This is also why substance cannot be an afterthought: a vehicle that is not genuinely managed where it is said to be managed will not be treated as resident there, and the treaty position built on it will not hold.
What a treaty can do, and what it needs from you
Treaties are the mechanism that prevents the same gain being taxed twice across a departure, and the good ones do it explicitly.
The treaty between the Netherlands and Germany of 12 April 2012 contains a clear example. Where an individual was resident of one state and becomes resident of the other, the first state may tax, under its own law, the increase in value of shares and comparable interests attributable to the period of residence there. In that case the increase in value taxed by the first state is not included in the tax base of the other state when it determines the subsequent increase in value. The split between pre departure and post departure gain is written into the treaty itself, and the arrival state is required to respect it.
This is precisely the kind of provision that works only if the sequence was planned. Clauses of this type turn on facts fixed at the crossing: which state the person was resident of, on what date, holding what, and whether the departure state actually assessed the gain. The treaty between the Netherlands and the United Kingdom of 26 September 2008 makes the point sharply in the other direction, since its equivalent clause applies only to the extent that part of the departure assessment remains outstanding. A relief drafted around the state of an assessment is a relief that depends on what was done, and when, at the moment of departure.
The order of operations decides the bill
Everything above reduces to a single proposition: with the same assets, the same destination and the same eventual sale, two clients can pay materially different amounts of tax purely because of the order in which they did things.
Realize or hold, before or after. Reorganize or leave in place, before or after. Incorporate the arrival vehicle, or wait. Sign the contract, or exchange next month. Establish the valuation, or hope. None of these is a clever technique and none of them requires an aggressive position. They are ordinary decisions whose only variable is sequence, and sequence is the one variable that becomes unavailable on the day residence ends.
The work therefore belongs in the twelve months before the move, and it belongs to both sides of the border at once. Departure planning done without knowing what the arrival state will recognize produces a structure the new country ignores. Arrival planning done without knowing what the departure state will crystallize produces a bill nobody budgeted for. Handled together, in advance, the exit charge stops being an accident and becomes what it should always have been, a known cost of a decision that was made deliberately.