The death of holding companies

Friday 22 April 2022

Isabella Denninger

McDermott Will & Emery, Munich


Report on a session of the IBA Taxes Committee at the 11th Annual IBA Finance & Capital Markets Tax Virtual Conference

Session Chair

Rhiannon Kinghall Were Macfarlanes, London


Ramona Azzopardi WH Partners, Ta' Xbiex

Maarten de Bruin Stibbe, Amsterdam

Rachel Fox William Fry, Dublin

Ayzo van Eysinga AKD, Luxembourg


The panel – represented by common jurisdictions for holding companies, such as Luxembourg, Ireland, the Netherlands and Malta – discussed the death of holding companies.

At the beginning, the speakers discussed what has been driving the attack on the use of holding companies, and then focused on the following three areas: (1) Denmark cases; (2) Anti-Tax Avoidance Directive 3 (ATAD 3); and (3) Pillar 2.

Rhiannon Kinghall introduced the topic with a flashback stating that, in the past, it was not particularly controversial that holding companies were entitled to the benefits of tax treaties and European Union directives on inbound flows. However, the viability of holding companies has been under scrutiny following a litany of changes. The position has shifted materially in recent years, with an accumulation of measures that can restrict access to the benefits of tax treaties and EU directives. This shift has a long history, going back to changes to the Organisation for Economic Cooperation and Development (OECD) model treaty in the 1970s, but the challenge to the treaty status of holding companies began in earnest in 2015, when the OECD base erosion and profit shifting (BEPS) project recommended measures to restrict the inappropriate use of tax treaties under Action 6, with the result that higher and higher bars are set for holding companies to benefit from relief from withholding taxes and exemptions to non-resident capital gains tax.

Setting the scene: providing insights into holding companies

The speakers first provided insights into the typical holding companies used in their jurisdictions to explain what is being 'attacked'.


Ayzo van Eysinga explained that Luxembourg has always been known for its holding companies. Lux Holdcos were used by everyone, for example, large multinational enterprises (MNEs), small and medium-sized enterprises (SMEs) with cross-border elements, and funds. However, if one compares today's practice with that of 20 years ago, a lot has changed: in the early 2000s, consultants met clients abroad; then, clients began to travel to Luxembourg (2010 onwards) and later set up businesses there (creating substance from 2015 onwards). Thus, Luxembourg no longer comprises mere letterbox companies, and consultants do not really talk about special purpose vehicles (SPVs) anymore. At present, the focus is on investment structures/funds (so-called Master LuxCos), that is, platform entities covering investment and treasury activities and advice.


Rachel Fox first listed the types of holding companies that are set up in Ireland, such as collective investment vehicles, multinational groups to hold shares in subsidiaries on a regional basis and private equity vehicles for investments in portfolio companies. She then explained that there has been a move away from low-substance intellectual property (IP) holding companies because passively holding IP was not the best way to benefit from the Irish onshore IP regime, with the result that the trend was to put a lot more substance into companies that exploit IP in Ireland.

Fox further explained that, in the past, the focus for holding companies was on: (1) making sure they were doing enough to avoid becoming tax resident elsewhere; (2) directors being where key decisions were made; and (3) where the board meetings were held. Nowadays, the focus is on putting more substance into investment holding companies to avoid issues in source countries that require a certain level of suitably qualified employees, particularly for active investment companies.

Regardless of this, Ireland does not currently have anything codified in terms of substance requirements. Substance is most relevant for active trading companies that want to get the benefit of the 12.5 per cent tax rate rather than the higher 25 per cent rate. Furthermore, substance is something that the Irish Revenue focuses on in terms of VAT registration (eg, a mere holding company typically does not get a VAT number due to the lack of sufficient substance).


Maarten de Bruin provided a retrospective on the taxation regime 30 years ago, inter alia, the then existing well-known Dutch 'crown jewels', such as: (1) the participation exemption rule; (2) extensive treaty network; and (3) no withholding tax on outgoing passive payments. Additionally, neutral joint venture standing, flexible corporate law and investment protection treaties also had great attraction power. However, at the beginning of 2000, these 'benefits' were under scrutiny (cf the Primarolo report). The political climate in the Netherlands started to tilt due to the financial crisis (2008) and the OECD BEPS project (2013 onwards). Thus, the Netherlands introduced new ruling policies (2004 and 2019), for example, withholding tax on dividends for holding coops.

De Bruin explained that the Netherlands also introduced various principles with reference to substance, for example:

  • relevant substance criteria, which focus on: (1) place of residence and knowledge of directors; (2) place of board meetings; (3) bookkeeping and bank accounts; and (4) more recently, the requirement to have an annual payroll of €100k and office space for at least 24 months; or
  • sufficient relevant 'economic nexus' in the Netherlands with regards to rulings that require that the requesting entities have a real business and real people in the Netherlands (for the risk and account, and sufficient personnel in the Netherlands). Thus, no ruling is allowed if the main purpose of a transaction is tax avoidance in the Netherlands, or if a low-tax jurisdiction is directly involved.

De Bruin then also stated that if the relevant substance criteria are not met at the level of the Dutch company, no ruling and/or automatic exchange of information is possible, and the source country may deny (domestic/treaty) benefits in the absence of a beneficial owner/principal purpose test. Thus, the development of more substance must be taken seriously in the Netherlands.


Ramona Azzopardi outlined that, ten to 15 years ago, many Maltese companies were incorporated with little substance because, pursuant to domestic corporate and tax laws, substance was not required.

She then made a few remarks about the economic situation and the set-up of holding companies: Malta is home to various companies from different industries, such as pharmaceuticals, shipping, gaming, blockchain and crypto. Foreign investors typically set up a two-tier Maltese structure: a holding company and one or more operating companies that share the same offices and, at times, the same directors, premises and staff. Usually, the holding company acts as a mere holding vehicle that benefits from the participating holding regime and full imputation system. However, some holding companies are also set up to ring fence their assets from their operations, for example, tech companies segregate their technology in separate companies and, in turn, charge the intra-group for the use of technology.

Azzopardi then explained that, over the years, there have been substantial changes in the terms of substance requirements. Today, some companies are headquartered in Malta, and a huge influx of foreign workers and C-suiters have made this country their home. However, substance requirements have been rather indirectly imposed through: (1) banks, which refuse to open accounts for companies with no substance in Malta; or (2) regulated businesses, which require key people to be present in Malta.

It has become very difficult to obtain a tax residence certificate in recent years, unless a company has real substance in Malta, for example, the company is required to have its management and control based in Malta. Furthermore, substance is also required in terms of receiving a VAT number.

Denmark cases: reactions in the Netherlands?

Kinghall moved to the Denmark cases and asked de Bruin whether there had been any reactions in the Netherlands in light of the Denmark cases.

De Bruin explained that, until 2020, the relevant substance criteria operated as a safe harbour, for instance, for the exemption of dividend withholding tax. Thus, if a taxpayer met these criteria, the structure was deemed non-abusive. As a result of the Danish cases, the relevant substance criteria, however, no longer function as a safe harbour, but split the burden of proof between the taxpayer and tax administration as follows: if the taxpayer meets the relevant substance criteria, the tax administration must prove that the structure is abusive. Conversely, if the taxpayer does not meet the relevant substance requirements, its structure is presumed to be abusive; that is, the taxpayer must prove that its structure is not abusive.

Furthermore, a debate arose regarding whether the Danish cases also had an impact on the participation exemption; that is, whether a Dutch Holdco must have substance to benefit from the EU Parent-Subsidiary Directive for incoming payments. Scholars argue that because the Dutch participation exemption predates the EU Parent-Subsidiary Directive, the participation exemption should not be affected by the Danish cases. However, opposite views are also held. The Dutch Government takes the position that further case law is required.

Holding companies under ATAD 3

Fox initially stated that the expected directive involves two-year look-back rules, which means that if the directive is introduced in 2024, the tests will look back to what is happening now (2022). Thus, holding companies must examine whether they will fall into these rules and take steps to avoid the unpleasant effects of the expected directive. Such steps may potentially be: (1) amending existing structures by either putting enough substance in holding company jurisdictions; or (2) winding up structures to align the holding company activity more closely with where the key decision makers are.

Ideally, holding companies will make sure that they fall out of the scope of the directive by not passing the so-called 'gateways'. This can be done by: (1) making sure that they have not outsourced both the administration of day-to-day operations and decision making on significant functions; or (2) by qualifying for one of the exemptions.

For entities that are within scope of the expected directive, an annual report must be filed, stating whether the following indicators of minimum substance are met:

  • having local premises for exclusive use;
  • holding an active bank account somewhere in the EU; and
  • at least one local director (who acts solely for that group) or a majority of local full-time employees who are suitably qualified.

If the holding company does not meet all the indicators, it needs to rebut the presumption by showing the commercial, non-tax reason for being established in an EU Member State, and that it has not been interposed to reduce the tax liability of the shareholders or group.

Additionally, if the directive is enacted in its current form, it will have far-reaching consequences for EU-based holding companies classified as shell companies under the directive, particularly because (further) access to tax treaties determines whether holding companies minimise withholding taxes or taxation of capital gains so that interposing a holding company in the structure does not lead to tax leakage.

Also worth mentioning is that one of the elements highlighted in the draft directive is the fact that outsourcing arrangements, particularly to professional third-party service providers, is seen as a risk, and the use of professional third-party directors is not going to be enough to meet the local director test. Thus, given how common such outsourcing arrangements are, many entities will be affected.

Finally, Fox drew attention to the fact that the EU Commission plans to follow-up with proposals that aim to target non-EU shell companies; thus, any advantage of using non-EU holding companies to circumvent the rules of the expected directive is expected to be short-lived.

Kinghall then asked the speakers about the implications for holding companies in their respective jurisdictions.


According to van Eysinga, the draft directive claims that existing anti-avoidance legislation does not include any measures that target low-substance factors, and therefore, it does not have to evaluate existing measures. Van Eysinga did not agree with this. It was his strong belief that BEPS/ATAD-driven measures, such as the principal purpose test and multilateral instrument; the general anti-avoidance rule; and EU Council Directive 2011/16 (as amended) ('DAC 6') with the main benefit test, already triggered most of the holding companies to meet the minimum substance indicators, such as having their own premises, holding a bank account and having a resident director. Thus, Luxembourg holding companies will not pass the gateway into ATAD 3 and will have sufficient substance to meet the substance indicators.

He pointed out that ATAD 3 may be an opportunity for Luxembourg to consider the carve-out rules in the draft directive for holding companies with the same Member State parent. Thus, he imagined that non-EU funds (eg, the British Virgin Islands) with a Master LuxCo that is under attack by ATAD 3 may move to Luxembourg to rely on the carve out to fall out of ATAD 3.

In a nutshell, it was van Eysinga's opinion that Luxembourg holding companies are still the gateway to Europe, but not to ATAD 3.


De Bruin explained that, in autumn 2020 (ie, before the EU started its ATAD 3 initiative), a government commission was established to publish a report on conduit companies. In the report, the commission advised the Dutch Government to support the ATAD 3 initiative, but also proposed some unilateral measures such as: (1) eliminating the safe harbour for interest/royalty conduits; (2) expanding the exchange of information; and (3) limiting the access of shell companies to investment protection treaties.

Additionally, further scrutiny of trust and corporate service providers was proposed. In this respect, various initiatives are ongoing. The Dutch Government, for instance, announced a legislative proposal that will make certain trust services, such as the offering of conduit companies by trust and corporate service providers, illegal and announced that it will also consider a total ban on corporate service providers. In terms of the ongoing Ukraine crisis, this far-reaching idea raised further support from the public because approximately 400 Russian conduit companies are based in the Netherlands and managed by corporate service providers.


Azzopardi stated that there are several holding companies that will not be affected by the proposed directive, such as companies that: (1) are listed internationally; (2) are regulated by financial services laws; and (3) offer crypto services. Additionally, there are many holding companies that may be out of scope of the directive because the ultimate beneficial owners are resident in Malta.

However, other Maltese companies that meet the gateway criteria (constituting the majority of holding companies in Malta) would need to satisfy the minimum substance requirements. In particular, in terms of the implementation and interpretation of the directive with respect to national law (minimum standard), one must wait and see whether, for instance, shared offices/shared directors within the group will tick the box.


Fox concluded that the Unshell proposal (if passed) will reduce the number of substance-related disputes. This is because of the bright-line test and a certain level of clarity as to what sufficient substance means. Thus, companies that are not falling through the gateways and that are clearly meeting the minimum substance indicators should be relatively comfortable that they will not have a substance-related challenge.

However, for companies that fall through the gateways and have to rebut the presumption, it raises the risk of a substance challenge because, even if they rebut the presumption in their home country, because the details of their level of substance is automatically exchanged with other EU countries, they are effectively flagged as low-substance companies. Tax authorities can then ask for audits to be opened on these companies in cases of doubt, which is likely to lead to even further substance-related disputes.

In terms of defending an attack, the substance must be strengthened in accordance with the indicators in the directive, that is, having a bank account within the EU; having suitable premises; and having full-time employees or a local director who is dedicated to the group, not a professional director. It will also require the company to show the commercial, non-tax-related rationale for interposing the holding company.

Pillar 2

Kinghall asked about the implications for low-taxed holding companies and addressed the specific question of whether it is opportune to consider a jurisdiction that has implemented the income inclusion rule (IIR) rather than consider a low-tax jurisdiction that does not implement IIR and thus will be subject to the undertaxed payments rule (UTPR).


Azzopardi explained that Malta is most likely to qualify as a low-tax jurisdiction in light of Pillar 2. Although companies that are tax resident in Malta are generally subject to income tax at a flat rate of 35 per cent, a lower effective income tax rate may result due to a full imputation system whereby shareholders may benefit from tax refunds.

At present, Malta is most likely to adopt the IIR. Currently, discussions are still ongoing as to whether Malta will include the qualified domestic top-up tax. If Malta opts for the qualified domestic top-up tax, changes to tax law would be necessary. Such amendments would certainly take time (uncertain implementation date) due to the change from a full imputation system to a conventional tax system. If necessary, Malta may also reconsider its currently very stringent deductions to bring them in line with those of other countries. In Azzopardi's view, the UTPR would be much more complicated to manage and administer. Referring back to the Unshell Directive, Azzopardi thought that the EU would have to make sure there are adequate and realistic substance requirements, that is, there should be no one-size-fits-all directive, otherwise the EU will be less attractive for businesses than other countries.

Concluding remarks

Kinghall summarised that there has been a consistent attack on holding companies, for example, by the OECD, due to the Denmark cases and through developments within the EU. However, in her opinion, there will be no escape and it is likely that conversations will continue. In reflecting on the conversation, her final question was: what's dead and what's alive, that is, what structures are likely to remain intact, and are mere SPVs no longer viable?


Van Eysinga's answer was that he was on an IBA private equity panel in Washington in 2005, which was introduced with the question of whether holding companies would still exist in five years, and the current session's title was 'The death of holding companies', which is much more aggressive, but rather catchy. Thus, he was curious about what the title would be 15 years from now.

However, in his view, holding companies will not die and ATAD 3 will not 'kill' them – although, the now rare, cheap, low-substance and ad hoc SPVs may face some threats, but well-established, properly staffed platform investment entities will be fine.


Fox did generally not see the death of holding companies. First, holding companies with no substance that were set up purely as a conduit to benefit from tax treaties have been dying for some time; thus, the Unshell Directive is likely to be the final blow. Second, according to Fox, legitimate securitisation vehicles, such as the s110 company, will continue to be viable and fall outside the scope of the Unshell Directive, although further work is needed on the wording of the directive to clarify this.

In terms of the future of holding companies, it must be noted that: (1) holding companies now have more people-based substance; and (2) groups are accepting that if they want to get the benefit of the regimes in traditional holding company jurisdictions, they need to locate suitably qualified people there. If one remembers the early stages of BEPS, the thinking was that companies would move their assets to where the substance was rather than looking at what might be the most tax-efficient location. However, in actual fact, companies moved the substance to tax-efficient locations instead. A similar trend in terms of holding companies is likely.


De Bruin noted that clients must understand that setting up a company and visiting twice a year is not sufficient anymore because the rules have changed. He observed a client leave the Netherlands by way of liquidating Dutch holding companies due to the various substance requirements.


Azzopardi drew attention to the fact that the draft directive should not make Europe less attractive than the rest of the world for holding companies