Most structuring memoranda written for Egyptian groups begin with tax and reach currency in an annex. In practice the order is reversed. A shareholder outside Egypt is not primarily asking what rate applies to a distribution. The shareholder is asking whether the distribution can leave, in what currency, on what evidence, and how long it will take. Those questions are answered by banking and exchange practice rather than by the tax code, and they determine which structures are worth building at all.
This is not an argument against planning the tax position, it is an argument about sequence. A structure that is efficient on paper and cannot move cash is worse than a plainer one that can, because the first has a cost today and a benefit that never arrives. The groups that get this right decide the route the money will take before they decide who owns what.
The order in which the questions actually arrive
Ask a European shareholder in an Egyptian business what keeps the file open and the answer is rarely a rate. It is the interval between a board approving a dividend and the money appearing in a European account, and the documentation the Egyptian bank will ask for in between.
That interval is not a tax variable and it cannot be modeled the way a tax charge can. It responds to the availability of foreign currency, to the rules and practice the Central Bank of Egypt applies at the time, and to the order in which banks meet demands on that currency. What the Central Bank publishes is narrower than the assumptions built on it. Its foreign exchange page shows a single market exchange rate as a weighted average, 52.1460 pounds to the dollar on 17 September 2026, and it describes its monetary policy framework as transitioning to a flexible inflation targeting framework in accordance with the liberalization of the foreign exchange market.
Its index of circulars on foreign currency activities runs to thirty three items, on export proceeds, cash deposit and withdrawal limits, card use abroad, licensing of exchange companies and the allocation of foreign currency, and none of them is addressed to dividends or profits as such. The closest is a circular of 14 June 2017, which removed the ceiling of one hundred thousand US dollars or the equivalent, once a year per client, on customer requests to transfer abroad, and let banks execute them with no maximum subject to the customer identification rules. Two circulars from 2013 tell banks to give priority in providing foreign currency, one to a list of import categories and the other to clients who had surrendered their own foreign currency proceeds, and a dividend is in neither list. So the published material gives a rate, a ceiling lifted in 2017 and a priority order dividends do not appear in. None of it gives a distribution a date, and what a group meets is the bank’s own currency position and its documentary requirements on the day.
The practical consequence is that a group should hold, in writing, an answer to three questions before the structure is drawn. What documentation does the Egyptian bank require to execute a remittance of this type. What has the group’s own experience been on timing over the last four quarters. And what happens to the group’s covenants and working capital if that timing doubles. A structure designed without those answers is a diagram, not a plan.
What the treaty actually gives
The Netherlands and Egypt have a double taxation treaty, signed at Cairo on 21 April 1999 and in force since 20 May 2000, and it appears with those dates and the status in force on the Dutch Ministry of Finance’s overview of treaties on direct taxes as at 1 July 2026. Its terms are better than most groups assume, and the operative line is not the headline rate.
Article 10 caps source taxation of dividends at 15 per cent of the gross amount where the beneficial owner is a resident of the other State. It then goes further. Where the beneficial owner is a company whose capital is wholly or partly divided into shares, resident in the other State, and holding directly at least 25 per cent of the capital of the paying company, the State of the paying company shall not levy tax on the dividends at all.
That is a zero rate rather than a reduction, and it carries two conditions the rate itself does not show. Paragraph 4 of the same article withdraws it where the relation between the two companies has been arranged or is maintained primarily with the intention of securing the reduction. And the multilateral instrument adds a holding period the treaty text does not contain: the Egyptian position deposited in 2020 and the Dutch position deposited in 2019 both list this treaty and both notify its article 10(3) under article 8 of the convention, neither reserving on that article, so the ownership condition has to be met throughout a 365 day period that includes the day of the payment of the dividends.
Article 11 caps interest at 12 per cent of the gross amount, and leaves it taxable only in the recipient’s State of residence where a Contracting State, a statutory body, a political subdivision, a local authority or a central bank is payer or recipient, or where the loan was granted, guaranteed or insured by a government, a central bank or a government owned or controlled agency, a financial institution included. Article 12 caps royalties at 12 per cent. Both figures are high by modern European standards and low relative to what an unrelieved position can cost.
The twenty-five per cent line is the reason a European holding above an Egyptian operating company is usually built as a majority position rather than as a minority stake alongside local partners. Groups that end up at twenty per cent for commercial reasons should know they have crossed a tax line as well as a governance one. The 365 day period is the reason the holding has to be in place well before the distribution rather than arranged around it.
Why the Dutch company has to exist before the cash does
The Dutch parent is not a rate instrument. Dutch corporate income tax under article 22 of the Wet op de vennootschapsbelasting 1969 is 19 per cent on the first 200,000 euro of taxable profit and 25.8 per cent above that, in the text in force from 1 January 2026. Nobody chooses the Netherlands for those numbers.
The Dutch company is chosen because the participation exemption of article 13 lets it receive dividends and realize gains on a qualifying shareholding of at least 5 per cent of nominally paid up capital without a second layer of tax, because it gives access to the treaty position described above, and because it is a counterparty European banks, auditors and buyers can assess without a translation exercise. The exemption is not unconditional. Paragraph 9 of that article withdraws it for a participation held as a portfolio investment unless the participation qualifies, and qualifying turns on the subsidiary being subject to a profit tax that is a real levy by Dutch standards, or on its assets, directly or indirectly, generally consisting for less than half of low taxed free investments. A real operating business normally meets that, but it is a test rather than a presumption.
Timing matters more here than in most jurisdictions. Article 46 bis 3 of the Egyptian income tax law, in the text substituted by Law 151 of 2026, taxes capital gains on the disposal of quotas in companies or of securities not listed on the Egyptian exchange, whether the gain is realised in Egypt or abroad, and where the person disposing is not resident it puts on him the duty to compute and remit the tax within sixty days of the transaction. The next article measures that gain against a sale, an exchange or any other form of disposal. Inserting a holding company above an Egyptian subsidiary once the shares have appreciated is one of those disposals, and it is taxed as one. Inserting it once distributions have started means restructuring a payment stream that the Egyptian bank has already documented under a particular set of instructions. Both are avoidable by putting the holding in place early, when the shares are worth little and no flow exists yet.
Pricing, when currency is the scarce input
A group that cannot move dividends freely will be tempted to move value some other way, and the usual instrument is a service charge or a management fee from the European parent to the Egyptian company. This works, up to a point, and the point is transfer pricing.
The Egyptian administration examines related party charges, and article 9 of the treaty lets either State adjust profits where conditions made or imposed between associated enterprises differ from those which would be made between independent enterprises. So the question is the ordinary one: what was actually provided, by whom, and what would an independent party have paid for it. The same article adds that having concluded a cost sharing arrangement or a general services agreement is not in itself such a condition, so the agreement is not the weakness; the missing service is. A charge that rises whenever a dividend is delayed and falls when it is not answers the question badly. So does a charge that has no supporting evidence of the work performed.
The disciplined version is to price the services on their own merits, document them as they are delivered rather than at year end, and accept that the resulting figure will not be large enough to substitute for a dividend. Treating a service agreement as a repatriation mechanism converts a currency problem into a transfer pricing assessment, which is a worse problem because it comes with penalties and it is permanent.
Debt, and why a shareholder loan is not the workaround either
Lending into Egypt from the European parent has an obvious attraction. Interest is deductible where it is paid, and the treaty caps the withholding at 12 per cent, which compares favorably with an unrelieved position.
It has three limits worth stating plainly. The interest still has to leave the country, so a loan does not solve the constraint that motivated it. The rate charged has to be defensible on both sides, and a rate set to move as much value as the deduction will bear is not. And the Egyptian company’s capital structure has to survive a debt to equity test rewritten in 2023 and again in 2026. Item 1 of the first paragraph of article 52 of Law 91 of 2005 disallows the debit interest that a juridical person within article 47 of that law pays on loans and advances above twice the average of equity, measured on financial statements prepared under Egyptian accounting standards. Article Five of Law 30 of 2023 phases that baseline in, at four to one for tax year 2023, three to one from tax year 2024 to tax year 2027, and two to one for tax year 2028. Law 151 of 2026 replaced the text of the item without disturbing that schedule, and added one exception at four times average equity, available only where the borrower is carrying out national infrastructure projects designated by decision of the Prime Minister on the Minister’s proposal, the loans and advances come from unrelated parties, and the borrower’s participation in those projects is not less than 25 per cent of the total value of its own investments. Banks, insurance companies and financing companies designated by the Minister sit outside the item entirely. An ordinary shareholder loan is measured against three to one for now, a figure worth confirming locally each year rather than carried forward from the last file.
On the Dutch side there is a further item to check before designing any lending flow through a third vehicle. The Wet bronbelasting 2021 taxes interest and royalties, and since 1 January 2024 dividends as well, paid to an affiliated entity in a low taxed jurisdiction at the highest percentage of article 22, which is 25.8 per cent for 2026. A low taxed jurisdiction there is a state designated by ministerial regulation, either because it does not subject companies to a profit tax or does so at a rate below 9 per cent, or because it is on the EU list of non-cooperative jurisdictions. Egypt is on neither of the two lists in force for 2026. An offshore company inherited from an earlier structure may well be. Being off the lists is not the end of the enquiry either. Article 2.1 makes the recipient taxable in five alternative cases, and only the first of them is residence in a listed jurisdiction. The second catches benefits attributed to a permanent establishment in one, and the fourth and fifth catch a recipient that its own state does not treat as entitled to them. The third is the one a third vehicle runs into: a recipient resident under the tax rules of a state that is neither the Netherlands nor a listed jurisdiction, entitled to the benefits with as its main purpose or one of its main purposes the avoidance of tax for someone else, and an artificial arrangement or transaction, or series of them, artificial to the extent it is not set up on valid business reasons that reflect economic reality. Paragraph 6 gives a recipient that meets conditions set by ministerial regulation a presumption against both of those tests, which the inspector can rebut. Paragraph 7 adds a sixth case, for a reverse hybrid.
Substance, on both sides of the Mediterranean
The Egyptian question is whether the operating company is a real business with a real capital structure and related party dealings priced as an independent party would price them. That is usually satisfied, because the operating company is genuinely operating.
The Dutch question is harder, because the holding company is the entity nobody staffs. It has to make its own decisions, in the Netherlands, with the means to bear the risks it reports. A board that ratifies decisions taken in Cairo will be characterized as one, and the treaty benefit described above depends on the Dutch company being what it says it is. It depends too on the company being resident in one place only. Both states notified the treaty’s article 4(3), which deemed a person other than an individual resident in both States to be resident only where its place of effective management is situated, under article 4 of the multilateral instrument, and neither reserved, so that tiebreaker is replaced. The competent authorities are now to endeavour to settle residence by mutual agreement, having regard to the place of effective management, the place where the person is incorporated or otherwise constituted and any other relevant factors, and in the absence of that agreement the company gets no relief or exemption under the treaty except to the extent and in the manner they agree. Article 7 of the multilateral instrument withholds a treaty benefit where it is reasonable to conclude that obtaining it was one of the principal purposes of an arrangement or transaction that resulted directly or indirectly in that benefit, unless it is established that granting it would accord with the object and purpose of the relevant provisions. That test reaches these treaty reliefs, though not the Dutch domestic exemption, which carries its own conditions. Both positions are argued at the worst possible time, during a sale.
Exit, which is where the structure is finally graded
Every one of these decisions is read again when the group is sold, by a buyer’s adviser with no attachment to the history. A Dutch holding with proper minutes, priced intercompany agreements and a documented treaty position is a structure a buyer can underwrite. The same holding with a nominal board and a stale service agreement is a discount on the price, usually an indemnity rather than a stated criticism.
Currency practice is read at exit too, in a different way. A buyer will ask how much of the group’s cash has actually left Egypt over the last three years, and the answer is either a number or an excuse. Groups that built a workable repatriation route early can point to a record. Groups that planned around the tax rate and postponed the currency question have nothing to show.
What we would do for an Egyptian group today
Start with the route the money will take, and document what the group knows about it. Then decide the shareholding, with the twenty-five per cent line and the 365 day period in view rather than discovered afterwards. Put the Dutch holding in place before value has accumulated and before flows have started, because both make insertion expensive.
Price the intercompany relationships as if they will be examined, since they will be, and keep them stable regardless of what is happening to the currency. Keep the Dutch company genuinely resident and genuinely deciding. And treat the Egyptian operating company as the asset it is, rather than as a source of cash to be extracted by whatever channel is least obstructed this quarter. The channel changes. The record of how the group behaved does not.
This article is informational and does not constitute tax or legal advice. The treatment of any structure depends on its facts and on the law of each jurisdiction involved. Each engagement is subject to scope and applicable regulation.