Treatment of Pre-GloBE losses in a jurisdiction with a newly introduced Corporate (Income) Tax

Facts

A multinational enterprise (MNE) group is headquartered outside the UAE and operates through a wholly owned subsidiary (“UAECo”) in the United Arab Emirates, a jurisdiction that had no federal corporate income tax until June 1, 2023. The UAE enacted a 9% Corporate Tax (CT) law, effective for financial years starting on or after June 1, 2023.

UAECo adopted a Gregorian calendar year (January 1 to December 31) as its fiscal year. Its first corporate tax period is therefore the year ending December 31, 2024.

In the 2022 and 2023 fiscal years, UAECo incurred substantial accounting losses. It continued to be loss-making in 2024. In light of the enactment of Pillar Two legislation globally, the MNE group becomes subject to the Pillar Two GloBE Rules in 2025 and will prepare its first GloBE Information Return in 2025 for the 2025 fiscal year.

The UAE has enacted a Qualifying Domestic Minimum Top-up Tax (QDMTT), applicable to the Fiscal Years beginning on or after 1 January 2025,[1] aligned with the GloBE Model Rules and transitional guidance.[2]

[1] Cabinet Decision No. 142 of 2024 dated 31 December of 2024, Article 2.

[2] Find details here.

Question

Can the losses incurred by UAECo in 2022, 2023, and 2024, prior to the MNE Group becoming subject to the DMTT in 2025, be carried forward and utilized to reduce or eliminate top-up taxes under the Pillar Two framework?

Executive Summary

Under the mechanics of Pillar Two, losses incurred in a jurisdiction do not reduce future GloBE Income, but they can influence the local effective tax rate through deferred tax accounting. A current-year loss gives rise to deferred tax assets (or GloBE Loss DTAs under Article 4.5), which affect the timing of tax expense recognition and thereby modify the numerator (Adjusted Covered Taxes) in the effective tax rate calculation.. However, the ability to carry forward such losses depends not merely on their accounting treatment but on the Pillar Two framework's internal rules, which provide distinct mechanisms under Articles 4.2, 4.4, 4.5, and the transitional Article 9.1.

The GloBE Model Rules and UAE Domestic Top-Up Tax Rules do provide for loss carry-forward through two pathways:

1. Standard Deferred Tax Accounting (Article 4.4), applicable where covered taxes follow regular deferred tax treatment.
2. GloBE Loss Election (Article 4.5), a mechanism simulating deferred tax treatment for jurisdictions where no tax system deferred tax assets framework existed.

However, both mechanisms are forward-looking. The GloBE Loss Election under Article 4.5 is available only from the first GloBE in-scope year and does not extend retroactively to capture losses from prior periods. Meanwhile, Article 9.1 allows recognition of pre-GloBE deferred tax assets but only if such DTAs are recorded in the financial accounts, which is unlikely in jurisdictions like the UAE that lacked CIT before mid-2023.

Consequently, pre-GloBE losses in 2022, 2023, and 2024 cannot be utilized under Article 4.5, and are unlikely to qualify under Article 9.1 either unless supported by financial statement DTAs. The Pillar Two framework effectively precludes the transfer of pre-GloBE losses into the GloBE regime through either election or transition.

The treatment of 2024 losses differs materially from that of 2022–2023. Because the 2024 fiscal year constitutes the first period covered by the domestic Corporate Tax regime, losses from that year arise under an enacted tax system and may therefore generate DTAs in the financial statements. These DTAs can be carried into the Pillar Two framework through the transitional mechanism of Article 9.1 being recast at the 15 % minimum rate.

Accordingly, it would be favorable to transfer the 2024 loss to the first DMTT year (2025) through deferred-tax accounting rather than rely on the GloBE Loss Election. Doing so preserves the tax value of the loss within Adjusted Covered Taxes and improves the ETR position in the jurisdiction once Pillar Two applies.

We reach these conclusions based on the following considerations.

Mechanics of Loss Utilization under Pillar Two

The OECD’s Pillar Two (GloBE) Rules and the UAE DMTT framework introduce a 15% global minimum tax on large multinational groups, calculated on a jurisdictional basis. Among the technical implementation issues that arise in this context is the treatment of losses incurred before a multinational group becomes subject to Pillar Two. In the UAE, this issue is particularly relevant because companies may have accumulated two distinct categories of losses:

• losses from pre-Corporate Tax (CT) periods, which cannot be carried forward under Article 37(3) of the UAE Corporate Tax Law; and
• losses incurred after the introduction of CT, which may be carried forward under domestic tax rules.

This study seeks to examine the DMTT outcome for both categories. Specifically, it analyses whether and to what extent either type of loss can influence the computation of the jurisdictional effective tax rate once Pillar Two applies. While losses recognized under the UAE CT regime can, in principle, create deferred tax assets (DTAs) capable of being carried into Pillar Two through Article 9.1 transitional adjustments, pre-CT losses, occurring when no corporate tax base or deferred tax accounting existed, present a distinct challenge. They may reduce the domestic tax base but lack the deferred tax element required to provide relief under Articles 4.4 of the GloBE framework, thereby potentially leading to higher top-up tax exposure in the first DMTT years.

Conceptual Framework: Losses under Pillar Two

The Pillar Two/DMTT framework introduces a mechanism to avoid unfairly imposing top-up taxes in such cases. Neither GLoBE Model Rules, nor UAE DMTT adopt a traditional loss carryforward regime. Instead, they rely on deferred tax accounting to smooth out timing differences between financial accounting income and taxable income. Under the GLoBE Model and UAE DMTT Rules, deferred tax expenses and benefits are included in covered taxes (with certain adjustments) to prevent timing mismatches from triggering top-up tax in a given year.

In a high-tax jurisdiction, a loss carryforward typically creates a deferred tax asset (DTA) in the financial statements, which is later released (as a deferred tax expense) when the loss offsets future taxable profits. This increases the book tax expense in those profit years, keeping the ETR from dropping too low. However, in a zero-tax jurisdiction, companies often do not record any DTA for losses (since there is no domestic tax to recover), so the standard deferred tax approach would not provide relief. Losses used later could lead to a near-zero current tax with no deferred tax expense to compensate, thus inviting a Pillar Two top-up tax.

Deferred Tax Adjustments and ETR Alignment

The GloBE and UAE DMTT Rules allow Constituent Entities in a jurisdiction to defer top-up taxes when effective tax rates temporarily fall below the minimum threshold due to deferred tax movements.

This deferred tax accounting is facilitated through two main regimes:

• Article 4.4 relies on standard deferred tax accounting. This mechanism permits the recognition of deferred tax expense and benefit, aligned with financial accounting rules, provided such deferred tax is recorded at the lower of the applicable domestic rate or the Minimum Rate (15%).
• Article 4.5 “GloBE Loss Election” is an alternative introduced to simulate deferred tax accounting in jurisdictions that do not impose corporate income tax or do not permit deferred tax recognition. This mechanism allows the group to elect to carry forward a synthetic DTA equal to the Net GloBE Loss × Minimum Rate, to be applied against future GloBE Income.

Deferred Tax Accounting and the ETR under Article 4.4

Covered Tax expense in a jurisdiction is an inherent element of the DMTT calculation, as it determines the difference between the 15 % DMTT threshold and the Effective Tax Rate (ETR). The ETR, in turn, is computed by dividing the tax expense by the Financial Accounting Net Income earned in the same jurisdiction. Accordingly, the relevant tax amount must be included in the numerator of the ETR formula. The amount recognised for this purpose is referred to as the Adjusted Covered Taxes of the Constituent Entities in that jurisdiction.

According to Article 4.1.1 of UAE DMTT Rules, “the Adjusted Covered Taxes of a Constituent Entity for the Fiscal Year shall be equal to the current tax expense accrued in its Financial Accounting Net Income or Loss with respect to Covered Taxes for the Fiscal Year adjusted by: …

• the net amount of its Additions to Covered Taxes for the Fiscal Year (as determined under Article 4.1.2) and Reductions to Covered Taxes for the Fiscal Year (as determined under Article 4.1.3);
• the Total Deferred Tax Adjustment Amount (as determined under Article 4.4)”.

Under IAS 12:5, current tax is defined as “the amount of income taxes payable (recoverable) in respect of the taxable profit (tax loss) for a period”. It therefore reflects the tax consequence of the period’s taxable results, measured at rates enacted by the reporting date, and forms the starting point of Covered Taxes before deferred-tax adjustments are applied under Article 4.4 of the DMTT Rules.

In contrast, deferred tax assets and liabilities reflect the future tax effects of temporary differences between accounting and tax bases of assets and liabilities, recognised as either deferred tax assets or liabilities in the statement of financial position.

Under Article 4.4.1 of UAE DMTT Rules “the Total Deferred Tax Adjustment Amount for a Constituent Entity for the Fiscal Year is equal to the deferred tax expense accrued in its Financial Accounting Net Income or Loss if the applicable tax rate is below the Minimum Rate or, in any other case, such deferred tax expense recast at the Minimum Rate, with respect to Covered Taxes for the Fiscal Year…” with certain exclusions.

According to Article 4.4.3 “If the taxpayer can demonstrate that a deferred tax asset recorded at a rate lower than the Minimum Rate is attributable to a Pillar Two Loss, such deferred tax asset may be recast at the Minimum Rate in the Fiscal Year in which the Pillar Two Loss was incurred. The Total Deferred Tax Adjustment Amount is reduced by the amount that a deferred tax asset is increased due to being recast under this Article”. Pursuant to para 87 of the Consolidated Commentary, this rule “preserves the basic tenet that a GloBE Loss of EUR 1 should offset GloBE Income of EUR 1”. This paragraph provides an example which is being extrapolated to the UAE soil should read as follows.

If a loss deferred tax asset was recorded at a 9% Corporate Tax rate, a Pillar Two Loss of AED 100 would result in a deferred tax asset of AED 9. When AED 100 of income is subsequently earned, the deferred tax asset of AED 9 reverses and is added to Covered Taxes through the Total Deferred Tax Adjustment Amount. Absent a recast at the Minimum Rate, Top-up Tax of AED 6 (100 x (15%-9%)) would be due when AED 100 of income is subsequently earned. However, permitting a recast of the Pillar Two Loss at the Minimum Rate (i.e. increasing the value of the deferred tax asset recorded from AED 9 to AED 15 in respect of the Pillar Two Loss) prevents this outcome and provides that a loss of AED 100 shelters AED 100 of income.

So, the combined effect of Articles 4.4.1 and 4.4.3 is twofold, defining both the scope of tax recognition and the mechanism for aligning deferred-tax treatment with the Minimum Rate:

• Deferred-tax movements recognised through profit or loss become a component of Adjusted Covered Taxes, thereby influencing the jurisdictional ETR for Pillar Two purposes. The inclusion of deferred tax ensures that temporary timing differences, such as those arising from loss carry-forwards, do not distort the ETR in any given year.
• They establish a symmetrical recasting rule:
  • when the domestic corporate-tax rate is below 15 %, deferred-tax expense is uplifted to the Minimum Rate; conversely,
  • where the domestic rate exceeds 15%, the deferred-tax expense is capped at 15%.

This dual adjustment prevents the jurisdictional ETR from falling artificially low in low-tax environments or inflating above the global minimum in high-tax ones.

Together, these mechanisms ensure that deferred-tax effects are treated on a consistent, standardized basis across jurisdictions, eliminating distortions caused by local rate disparities while maintaining the principle that a GloBE Loss of 1 should offset a GloBE Income of 1 on an equal footing.

This harmonized treatment, however, presupposes the existence of a domestic corporate tax system capable of recognizing deferred tax assets and liabilities in the first place. Where such a system is absent, as in jurisdictions that impose no corporate income tax or where deferred-tax recognition is not permitted, the Article 4.4 mechanism cannot operate.

Comparative Example: GermanyCo vs. UAECo

For example, consider two subsidiaries of the same MNE group:

• GermanyCo, operating in a high-tax jurisdiction with a long-established corporate tax regime (30% tax rate), where Pillar Two was enacted and became applicable from 2024; and
• UAECo, operating in the United Arab Emirates, where no federal corporate income tax existed in 2022, and losses from pre-CT periods cannot be carried forward under the UAE Corporate Tax Law enacted in 2023.

In 2022, both companies incurred a EUR 10 million accounting loss.

Both entities incurred accounting losses throughout 2022–2025, reflecting a prolonged start-up or investment phase within their respective jurisdictions. The first taxable profits were generated only in 2026, when both entities came within the effective scope of Pillar Two. The manner in which these cumulative losses were treated under their domestic regimes, however, differs profoundly, and it is this difference that later creates asymmetry in the Pillar Two calculations.

GermanyCo (High-Tax Jurisdiction)

GermanyCo, governed by IFRS and local tax law, recognizes a deferred tax asset (DTA) in its 2022 financial statements, based on the expectation that the loss will offset future taxable income.

Accounting entry in 2022:

• Dr Deferred Tax Asset: EUR 3 million
• Cr Tax Expense (P&L): EUR 3 million

Accounting entry in 2026 (reversal of DTA upon utilization):

• Dr Tax Expense (P&L) (EUR 3 million)
• Cr Deferred Tax Asset (EUR 3 million)

This reversal records the consumption of the tax benefit as the prior-year loss is used to offset taxable income. Economically, the transaction restores tax symmetry across time: the earlier recognition of a tax credit (reducing the 2022 tax expense) is balanced by a later tax charge (increasing the 2026 tax expense).

We will now turn to examine how the GloBE Rules, in particular, the treatment of deferred tax assets under Article 4.4, magnify the asymmetry between these two entities in 2026. In doing so, we will demonstrate how the same economic loss incurred in 2022 results in radically different outcomes under Pillar Two, depending on whether a tax regime and corresponding deferred tax accounting existed at the time the loss occurred.

The concept of deferred tax is not an easy one to grasp if, like myself, one approaches it primarily as a tax lawyer rather than as an accountant. The idea that a company can “book” a tax expense in years when no tax is payable, and later “reverse” that expense when no tax is paid either, can appear counterintuitive from a legal or economic perspective.

For those accustomed to working with statutory tax liabilities rather than accounting constructs, the deferred-tax reversal, which does not correspond to any cash tax payment, can be understood as the mirror image of the deferred benefit recognised in 2022:

• In that earlier year, the entity booked a tax credit in its profit and loss account (Cr tax expense) based on the expected future deduction. This recognition reduced its post-tax loss for the year by the amount of the future tax benefit it had effectively “earned” through the current year’s financial results. In substance, the company recorded the right not to pay that tax when future profits arise.
• The deferred-tax reversal in a later profit year represents the consumption of that earlier benefit. Economically, this ensures temporal symmetry: the reduction in tax expense recognised during the loss year is balanced by a corresponding increase when the loss is used, even though no cash tax is paid in either period.

This accounting logic is precisely what the GloBE Rules adopt: by including deferred-tax movements, recast at the 15 percent minimum rate, within Adjusted Covered Taxes, so that the effective tax rate reflects economic timing alignment rather than the mere presence or absence of current taxation.

Under Article 4.4, this deferred-tax expense flows directly into the numerator of the jurisdiction’s Adjusted Covered Taxes as part of the Total Deferred Tax Adjustment Amount. Consequently, the deferred-tax movement ensures that the jurisdiction’s ETR remains aligned with its domestic tax rate (30%) and no Top-up Tax arises. The deferred-tax mechanism thus operates as intended: it neutralizes timing differences between accounting income and taxable income and preserves the integrity of the Pillar Two ETR computation.

If Pillar Two had been in effect during the 2022–2026 period, GermanyCo’s jurisdictional ETR would be computed only in years with positive Net GloBE Income:

• 2022 is a loss year, meaning that the GloBE Losses of the Constituent Entities exceed GloBE Income of the Constituent Entities located in that jurisdiction. Commentary 11 to Article 5.1.2 provides that in such case “there is no Net GloBE Income under Article 5.1.2, and therefore, there can be no Top-up Tax computed under Article 5.2 with respect to the jurisdiction. Consequently, the ETR for the jurisdiction need not be computed”.

However, the EUR 3 million deferred-tax asset is recognised in the financials and will reverse in a later profit year, entering Adjusted Covered Taxes via Article 4.4.

• In 2026, when the deferred tax asset reverses as the loss is utilized, the company would again record EUR 10 million of GloBE Income and a deferred-tax expense of EUR 3 million. The ratio (EUR 3 million of Adjusted Covered Taxes over EUR 10 million of GloBE Income) again yields a 30 percent ETR.

ETR₂₀₂₆ = +3,000,000 /+10,000,000 = 30%

This demonstrates the economic symmetry of the deferred-tax mechanism: a tax benefit recognized in the loss year is precisely offset by a deferred-tax expense in the recovery year, ensuring that, over time, the aggregate ETR equals the jurisdiction’s nominal rate. In other words, deferred-tax accounting allows the Pillar Two computation to “look through” temporary timing effects so that only true low-tax outcomes are subject to top-up tax.

UAECo (No-Tax Jurisdiction in 2022)

UAECo also incurs a EUR 10 million accounting loss in 2022. However, in contrast to Germany, in a jurisdiction without corporate income taxation, such as the UAE prior to June 2023, the absence of a deferred-tax system breaks this symmetry. A loss incurred before the advent of the Corporate Tax Law generates no deferred-tax asset. Therefore, there is no subsequent deferred-tax expense when profits re-emerge. When such an entity later earns GloBE Income, its domestic tax computation reflects only the income of the period, since pre-Corporate Tax losses cannot be carried forward under Article 37(3) of the UAE Corporate Tax Law. As a result, UAECo’s current tax in 2026 will equal 9 % of its taxable profits (or 0 % if Free Zone relief applies), but this tax will correspond solely to current-year income. The losses accumulated in 2022 and 2023 (before the introduction of the Corporate Tax regime) have no effect under either domestic tax law or the Pillar Two framework.

GloBE Loss Election Alternative (Article 4.5)

To address this gap, the GloBE Model Rules/UAE DMTT Rules introduce an alternative, elective framework under Article 4.5 (GloBE Loss/Pillar Two Election). According to Article 4.5.1, “In lieu of applying the rules set forth in Article 4.4, a Filing Constituent Entity may make a Pillar Two Loss Election for the UAE. When a Pillar Two Loss Election is made for the UAE, a Pillar Two Loss Deferred Tax Asset is established in each Fiscal Year in which there is a Net Pillar Two Loss for the UAE. The Pillar Two Loss Deferred Tax Asset is equal to the Net Pillar Two Loss in a Fiscal Year for the UAE multiplied by the Minimum Rate”.

Commentary 4:113 explains that this Article “provides an elective rule to effectively carry GloBE losses forward with a deemed deferred tax asset”, thereby allowing to utilize losses for the companies which do not recognised differed tax expanses in their financial accounting. Under the Commentary, “Article 4.5 is generally expected to be of greatest utility as a simplification in jurisdictions that do not impose a corporate income tax or impose one at a very low rate given that when the election is made, Article 4.4 no longer applies and temporary differences may result in Top-up Tax. However, the election may be made for any jurisdiction”. The OECD Commentary provides an example: “if a Constituent Entity is located in a country that does not impose a CIT, an election under Article 4.5 would provide a GloBE Loss attribute at the Minimum Rate for economic losses incurred that otherwise would not have a corresponding deferred tax asset due to the lack of a domestic CIT”.

Hence, in essence, the election mimics the effect of a full-value loss carryforward for Pillar Two purposes, even where domestic tax law or accounting standards wouldn’t recognize one. Article 4.5 replicates the smoothing effect of deferred tax accounting in situations where no underlying tax base exists. In lieu of recognising a conventional deferred tax asset through the books, a Constituent Entity may elect to create a notional Pillar Two Loss Deferred Tax Asset equal to the Net GloBE Loss for the jurisdiction multiplied by the Minimum Rate (15 %). This deemed DTA functions as a synthetic tax attribute, carried forward to offset future GloBE Income on a one-to-one basis.

Conceptually, Article 4.5 extends the economic symmetry achieved by Article 4.4 to jurisdictions without deferred-tax systems. It ensures that a loss of one unit reduces the Pillar Two tax base as effectively as it would in a jurisdiction with full deferred-tax accounting, thereby maintaining parity across no-tax environments. The mechanism is particularly relevant for jurisdictions with no corporate income tax, where no deferred-tax accounting framework exists to capture the future tax effects of losses. In such cases, the Pillar Two Loss Election serves as a substitute for deferred-tax recognition, allowing an MNE to create a deemed deferred tax asset at the 15% minimum rate. Absent this election, losses in no-covered-tax jurisdictions provide no relief in the DMTT computation; with the election, however, the MNE effectively generates its own Pillar Two loss DTA at 15%, ensuring that future GloBE income is not overstated for minimum-tax purposes.

According to Article 4.5.2 of the UAE DMTT Rules, “the balance of the Pillar Two Loss Deferred Tax Asset is carried forward to subsequent Fiscal Years, reduced by the amount of Pillar Two Loss Deferred Tax Asset used in a Fiscal Year”. Article 4.5.3 sets forth that “the Pillar Two Loss Deferred Tax Asset must be used in any subsequent Fiscal Year in which there is Net Pillar Two Income in the UAE in an amount equal to the lower of the Net Pillar Two Income multiplied by the Minimum Rate or the amount of available Pillar Two Loss Deferred Tax Asset”.

So, under the Election, the group will effectively carry forward GloBE losses in that jurisdiction by creating a deemed deferred tax asset at the minimum tax rate (15%) for any Net Pillar Two/GloBE Loss incurred there.

In each loss-making year (for Pillar Two purposes) the jurisdiction’s loss amount × 15% is added to a notional Pillar Two/GloBE DTA pool. In subsequent profitable years, the Pillar Two/GloBE Loss DTA is drawn down: an amount up to 15% of Net Pillar Two/GloBE Income (or the remaining DTA, if lower) is added to Covered Taxes – much like utilizing a tax loss carryforward, but at a fixed 15% tax value. This raises the book tax expense in profitable years, keeping the ETR at 15% and thereby shielding the income from top-up tax. Any unused portion of the GloBE Loss DTA carries forward indefinitely to offset future GloBE income (though if the election is later revoked, any remaining DTA is cancelled to avoid double counting).

Procedurally, the Pillar Two Loss Election is made on the MNE’s first Pillar Two return for the jurisdiction (it must be filed in the first year that the group has any Constituent Entity in that jurisdiction subject to Pillar Two).[1] It is generally a one-time, irrevocable choice, i.e. if not made in that first year, the opportunity is lost, and if made, it remains in force unless deliberately revoked in the future.[2] An election cannot be made at all for jurisdictions with a distribution-based tax system, since those have their own special rules.[3]

The election is jurisdiction-wide (covering all entities in that country) and replaces the normal deferred tax adjustments with the loss-DTA mechanism for that jurisdiction.

When an MNE makes this election for a given jurisdiction, the normal deferred tax accounting rules of Article 4.4 are turned off for that jurisdiction. Disabling the Article 4.4 deferred tax rules could make this election in a high-tax jurisdiction counterproductive since temporary differences there would no longer be smoothed out, potentially causing more volatility or top-up tax on other timing differences.

[1] UAE DMTT Rules, Article 4.5.5. Same Article in GLoBE Model Rules.

[2] According to Article 4.5.4, if the Election is subsequently revoked, any remaining Pillar Two Deferred Tax Asset should be written down to zero as of the first day of the first Fiscal Year for which the election no longer applies. Practically, this prevents double-counting: you don’t get to keep the synthetic loss DTA once you leave the synthetic regime.

[3] Article 4.5.5.

Pre-Scope Losses: Deferred Tax vs. Loss Election Treatment

Having established the functional distinction between the standard deferred-tax mechanism (Article 4.4) and the elective GloBE/Pillat Two Loss regime (Article 4.5), it is necessary to determine which provision governs the treatment of deferred-tax assets that originate before a jurisdiction or constituent entity becomes subject to the Pillar Two framework.

Whether the reversal of a DTA in a profit year after the jurisdiction becomes subject to the Domestic Minimum Top-Up Tax (DMTT) is governed by Article 4.4 or Article 4.5 depends on the origin and legal nature of that DTA.

Where a real (financial-statement) DTA was recorded under the applicable corporate-tax regime before the Pillar Two rules became effective, its subsequent reversal continues to fall under Article 4.4, even if the reversal occurs in a fiscal year when the Constituent Entity is in scope of Pillar Two. This follows from Article 9.1.1 of the UAE DMTT Rules, according to which “when determining the Effective Tax Rate for the UAE in a Transition Year, and for each subsequent year, the MNE Group shall take into account all of the deferred tax assets and deferred tax liabilities reflected or disclosed in the financial accounts of all of the Constituent Entities in the UAE for the Transition Year. Such deferred tax assets and liabilities must be taken into account at the lower of the Minimum Rate or the applicable domestic tax rate. A deferred tax asset that has been recorded at a rate lower than the Minimum Rate may be taken into account at the Minimum Rate if the taxpayer can demonstrate that the deferred tax assetis attributable to a Pillar Two Loss. For purposes of applying this Article, the impact of any valuation adjustment, or accounting recognition adjustment with respect to a deferred tax asset is disregarded”.

This Article expressly imports pre-existing deferred-tax balances into the GloBE framework at transition. Accordingly, the deferred-tax expense (or benefit) that flows through the profit and loss account upon reversal is included in Adjusted Covered Taxes under Article 4.4.1, recast where necessary to the 15 percent Minimum Rate. The Pillar Two regime thus treats the reversal of such pre-existing DTAs as part of the ordinary deferred-tax mechanism rather than as a synthetic Pillar Two loss adjustment.

So, in the first year a group or entity comes into Pillar Two (the “transition year”), existing deferred tax assets and liabilities from prior periods are brought into the GLoBE ETR computation, but with certain caps and adjustments. Generally, deferred tax balances are taken into account at the lower of the applicable domestic tax rate or 15%. This prevents a group from using a high-tax DTA to unduly reduce its Pillar Two top-up tax.

Deferred tax assets arising from pre-Pillar Two losses get special treatment: if the DTA for a loss was recorded at a tax rate below 15% (or not recorded at all), it can be “recast” at the 15% minimum rate, provided the loss would have been a GloBE loss if the Pillar Two rules had applied when that loss arose. In other words, qualifying pre-scope losses are credited with a 15% deferred tax for Pillar Two purposes, ensuring that when those losses are utilized against profits, a corresponding 15% tax expense is counted.

This recasting mechanism can eliminate any top-up tax that would otherwise result from using pre-scope losses. This approach spares groups from an unfair GloBE tax charge just because losses were generated pre-regime.

OECD guidance illustrates the effect: if a Constituent Entity incurred a tax loss of 100 in a year before the GloBE rules applied, and then uses that loss to offset taxable income in a post-Pillar Two year, the rules add a deferred tax expense of 15 (i.e. a 15% tax on the 100 loss) to the covered taxes in the year of use. This 15 unit deferred tax adjustment raises the jurisdiction’s ETR toward 15%, preventing a top-up tax that would have arisen from the loss utilization. In practice, the deferred tax expense enters the Pillar Two tax base (the “Total Deferred Tax Adjustment Amount” under Article 4.4) when the loss is used, offsetting the low current tax. The OECD’s Administrative Guidance confirms that “when the associated tax loss is used in a Fiscal Year in which the GloBE applies,” the deferred tax benefit on that pre-GloBE loss (capped at 15%) is included in covered taxes. This ensures the ETR calculation in the first Pillar Two year reflects the tax value of pre-scope losses, aligning with what the Pillar Two rules would have done had they been in place during the loss year.

Limitations on Retroactive Relief: Articles 4.4, 4.5 and 9.1 Interplay

By contrast, where no actual DTA existed in the pre-scope period (whether because the jurisdiction had no corporate-tax system, or because the entity did not recognise deferred-tax balances in its accounts) the group cannot rely on Article 4.4 or Article 9.1.

In that case, only Article 4.5 (GloBE Loss Election) could create a notional DTA for future offset. If the election is made in the first in-scope year, a Pillar Two Loss Deferred Tax Asset equal to the Net GloBE Loss × 15 percent is established. When the jurisdiction later earns GloBE Income, the draw-down of that synthetic DTA enters Adjusted Covered Taxes under Article 4.1.2(b). Thus, post-election reversals are governed not by Article 4.4 but by the elective regime of Article 4.5. Consequently:

• Article 4.4 applies to reversals of real DTAs recognised before or after transition, provided they were recorded in the financial statements under IAS 12 or equivalent domestic tax law; and
• Article 4.5 applies only where a deemed DTA (GloBE Loss DTA) was created under the elective mechanism in the first Pillar Two year and is later reversed against GloBE Income.

Neither Article 4.5, nor other provisions of the UAE DMTT and GLoBE Model Rules provide a transitional provision for Pillar Two/GloBE Loss Election to retroactively create Pillar Two loss attributes for years when the group was out-of-scope. The election only applies “in lieu of” the normal deferred-tax approach going forward. By design, it must be made in the first Pillar Two year and then applies to losses in that year and future years. Losses incurred in pre-Pillar Two periods are not directly accounted for by Article 4.5, because those earlier years had no GloBE calculation of “Net GloBE Loss”. Instead, the Model Rules rely on the transition mechanism described above to handle such losses via deferred tax in the first year. In fact, the rules explicitly provide that if a GloBE loss election is made for a jurisdiction, the transitional deferred tax rules do not apply for that jurisdiction.

In practical terms, this means that the reversal in a profit year of a deferred-tax asset created before the jurisdiction entered Pillar Two, whether arising from pre-existing corporate-tax losses or temporary differences, should be captured under Article 4.4 via Article 9.1 transition, not under the GloBE Loss Election of Article 4.5. The latter governs only notional DTAs generated prospectively within the Pillar Two period itself.

The remaining deferred tax assets from the pre-regime period are essentially disregarded once a Group opts into the loss election method. This means an MNE cannot double-dip by both recasting a pre-existing DTA at 15% and also using the new GloBE loss DTA – it must choose one approach or the other.

Strategic Implications for MNE Groups

Pragmatically, if a group has significant pre-scope loss carryforwards (and especially if those losses were in a country with a normal tax rate, meaning a DTA exists in the accounts), the group may prefer not to use Article 4.5. By sticking with the default rules, the group can take advantage of the transition relief that recognizes those loss DTAs up to 15%, thereby avoiding top-up tax on subsequent profits. If the GloBE Loss Election were made instead, any accounting DTA from pre-scope losses would no longer adjust the Pillar Two tax calculation (since Article 4.4 is turned off), and at the same time no GloBE loss was recorded in prior years (since the rules weren’t in effect).

The result could be that a post-transition profit gets no relief, triggering top-up tax, under the election method, whereas the deferred-tax method would have prevented it.

In summary

In summary, losses incurred before a group falls within Pillar Two are not forgotten. They are addressed through the Pillar Two deferred tax framework and specific transition rules. The Model Rules allow pre-existing loss carryforwards (and other deferred tax attributes) to be factored into the effective tax rate calculation, typically by carrying their tax effects forward at up to the 15% minimum rate.

The Article 4.5 GloBE Loss Election is a parallel mechanism that an MNE may optionally use going forward, especially in jurisdictions where traditional deferred tax accounting is unavailable or insufficient (e.g. no domestic tax to generate DTAs).

However, Article 4.5’s election does not retroactively cover pre-regime losses. It must be enacted in the first in-scope year and then applies prospectively. So, the Article 4.5 mechanism is applicable to addressing losses once a group is in Pillar Two, but it is not a tool for reaching back to pre-regime losses. Pre-scope losses can still be effectively recognized but through the transitional deferred tax adjustments (including the 15% DTA uplift for qualifying losses) rather than through the loss election.

Prior losses should be handled via the transitional deferred tax adjustments, not by any backdated loss election. Once the election is made, the transitional relief is switched off for that jurisdiction. Thus, companies need to weigh carefully:

• whether to rely on the default deferred-tax approach (with its built-in recognition of pre-Pillar Two losses) or
• to use the GloBE loss election for a fresh start.

For the UAE, however, the GloBE Loss Election would rarely be favourable. Where the jurisdiction already has a corporate-tax system in place, such as for financial years beginning on or after 1 June 2023, and domestic law permits the carry-forward of those post-CT losses, the deferred-tax assets arising from such losses should be recognised under Article 4.4 and imported into Pillar Two through Article 9.1. In this situation, electing under Article 4.5 would disable the ordinary deferred-tax mechanism that allows recognition and reversal of DTAs at the domestic rate (recast to 15%where applicable). Consequently, for UAE entities that had accumulated losses in the first CT-effective periods preceding the DMTT (e.g., FY 2023–2024), remaining under the Article 4.4/9.1 framework ensures that those losses continue to yield a deferrable tax benefit, while an Article 4.5 election would forfeit that advantage.

Acknowledgements

I wish to express my sincere appreciation to Zhandos Taukenov for his valuable comments and observations on an earlier draft of this paper. His insightful feedback helped refine the analysis and clarify several technical aspects of the GloBE framework discussed herein.

The disclaimer

Pursuant to the MoF’s press-release issued on 19 May 2023 “a number of posts circulating on social media and other platforms that are issued by private parties, contain inaccurate and unreliable interpretations and analyses of Corporate Tax”.

The Ministry issued a reminder that official sources of information on Federal Taxes in the UAE are the MoF and FTA only. Therefore, analyses that are not based on official publications by the MoF and FTA, or have not been commissioned by them, are unreliable and may contain misleading interpretations of the law. See the full press release here.

You should factor this in when dealing with this article as well. It is not commissioned by the MoF or FTA. The interpretation, conclusions, proposals, surmises, guesswork, etc., it comprises have the status of the author’s opinion only. Furthermore, it is not legal or tax advice. Like any human job, it may contain inaccuracies and mistakes that I have tried my best to avoid. If you find any inaccuracies or errors, please let me know so that I can make corrections.