Step-by-step computation for Macro-step 1

Purpose

The purpose of the Scope and Perimeter analysis is to construct the GloBE perimeter by identifiying all the Constituent Entities and fixing their legal characteristic, thus allowing further treatment. It also allows to define the different sub-groups and sub-perimeters (JVs, MOMNEs, Investment entities...) in each jurisdiction that will be the basis of the next computational steps.

General Logic

In order to perform the Perimeter analysis a serie of computations on the shareholding graph needs to be perform. This will allow to finalize the direct and indirect ownership interests as well as the controlling interest necessary to draw the limits of the different sub-perimeters. Moreover a series of graph transversal will be necessary to qualify certain entities as member of Joint Ventures, Excluded Entities and or Investment entities.

Data Input

The Data input for Macro-step 1 consists in :

• The list of all potential entities and their characteristics;
• The list of all direct shareholdings and their characteristics or the ownership interests and Control rights held links
• If possible the Controlling Interest links

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Macro-step 1 consists in :

• The final characterization of each Constituent Entity and out of group entity with regard to GloBE rules, ready for input in the GIR;
• The final ownership and controlling interest matrices for use in Macro-step 3.'s re-allocations and various elections, as well as the Charging Mechanisms of Macro-step 5
• The final list of JVs and Minority-Owned sub-groups in each Jurisdictions as well as groups of Investment entities

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed Step by step approach

Step 1.1. - Sanity checks on Ownership and Controlling interests

Purpose

The purpose of the Sanity checks is identify and report all the data structure issues in the shareholding structure of the group, before computing the indirect ownership interests. It serves as an important reminder that the group's perimeter is at the basis of the Pillar 2 computation and that no accurate result can be achieved without ensuring its quality.

General Logic

This Step 1.1. will consist in a sequence of computations, checks and controls aiming at ensuring that depending on the level of input, the data provided is coherent and usable.

Data Input

The Data input for Step 1.1. consists in :

• The list of Entities for which legal characteristics are provided (within or outside the future GloBE Perimeter);
• The list of pre-calculated shareholding interests both ownership interests and control rights held if links are available for each shareholding relationships as percentages;
• The list of shares, characteristics of these shares with regards to financial and control rights, for each Entity in the perimeter as well as outside;
• The list of Fall-back elections for the MNE group : assumptions for undocumented shares, ownership or controlling interests (re-normalization; auto-filling with either UPE or generic out-of-group entity or UPE)

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.1. consists in :

• The list of subsidiary entities absent from the list of Entities for which legal characteristics have been defined;
• The list of parent entities absent from the list of Entities for which legal characteristics have been defined;
• The list entities with Legal characteristics completely absent from the shareholding structure;
• The list of subsidiaries with incoherent shareholding structures : incoherent single values (negative, not numerical; higher than 100 for percentages...); incoherent aggregates values (totals not equal to 100%; computational errors...)
• The list of auto-cleaned elements of the structure using the fall-back mechanisms
• The list of errors for each faulty elements
  • unknown subsidiaries
  • unknown parents
  • unknown shareholding relations
  • typing errors
  • constraints errors
  • coherence controls

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The list of Micro-Step is a follows :

• Micro-Step 1.1.1. Unknown subsidiaries
• Micro-Step 1.1.2. Unknown parents
• Micro-Step 1.1.3. Entity absent from the shareholding structure
• Micro-Step 1.1.4. Typing and constraints controls
• Micro-Step 1.1.5. Shareholding coherence controls

Micro-Step 1.1.1. Unknown subsidiaries

This step is pretty simple as it basically corresponds to a VLOOKUP between the matrix of shareholdings and the list of entities with characteristics to verify that any entity present as the target of a shareholding relationship from at least one parent entity is also present in the list of legal entities.

Micro-Step 1.1.2. Unknown parents

This step is pretty simple as it basically corresponds to a VLOOKUP between the matrix of shareholdings and the list of entities with characteristics to verify that any parent entity present as the owner of a shareholding relationship to any subsidiary entity is also present in the list of legal entities.

Micro-Step 1.1.3. Entity absent from the shareholding structure

This step is pretty simple as it basically corresponds to 2 VLOOKUPs between the matrix of shareholdings and the list of entities with characteristics to verify that any entity present in the list of entities is also present as the owner or the target of at least one shareholding link.

Micro-Step 1.1.4. Typing and constraints controls

This step is pretty simple as it basically corresponds to a verification of the different values according to their definition, typing and constraints.

The current list of these controls are as :

• percentages need to be doubles between 0 and 1;
• integers and doubles need to be positive or null;

Micro-Step 1.1.5. Shareholding coherence controls

This step requires to add the shareholding relationships to verify that the aggregates add-up to the given totals or percentages for each subsidiary entity and apply the fall-back mechanisms.

The current list of these controls and corrections are as follows :

• In any case, an error is returned containing the type of error and the correction made
• do the sum of percentages for both ownership interests and/or control rights held sum to 100%, for each subsidiary entity;
  • If the fall-back election has been taken, the percentages are corrected to the choice made :
    • re-normalization (for total values strictly positive and lower or higher than 1): values are put back to 100%, by dividing to the total percentage reached;
    • automated allocation of the missing shareholding interests (for total values positive and lower than 1) to either :
      • a generic independent third party owner;
      • the UPE;
• do the sum of each Share type for the subsidiary entity and compare to the provided total (if not provided at all, or equal to 0, this test is skipped);
  • If the fall-back election has been taken, the numbers of shares are corrected according to the choice made :
    • re-normalization (for a total of different from the total provided): the new total is considered as true, unless equal to 0;
    • automated allocation of the missing shares (for positive totals lower than the total provided) to either :
      • a generic independent third party owner;
      • the UPE;

Step 1.2. - Generation of the Ownership Interests and Control rights matrices

Purpose

The purpose of this step is to build the necessary input for the various computations that will follow and may require matrix computations.

General Logic

This Step 1.2. will consist in the application of a few transformation of the data cleaned or checked at the Step 1.1. and form the initial ownership interest matrix and the control rights matrix if the data is available.

Data Input

The Data input for Step 1.2. consists in :

• The same input as the Step 1.1.
• The output of the Step 1.1.

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.2. consists in :

• The final direct ownership interest matrix;
• The final direct control rights matrix;

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The list of Micro-Step is a follows :

• Micro-Step 1.2.1. Transformation of Shares data into percentages
• Micro-Step 1.2.2. Initial construction of the matrices
• Micro-Step 1.2.3. Removal of direct auto-circular rights

Micro-Step 1.2.1. Transformation of Shares data into percentages

This step applies to both financial and control rights, and consists in the computation of the ownership interests and control rights as percentages for each subsidiary-parent relationship present in the output data of Step 1.1. when the information provided is not directly provided as ownership interests and control rights but with detailed shares and share types.

Logic

The computation is quite straight forward :

• for each subsidiary entity, the total of financial rights (and control rights respectively) are computed as the sum for all share types of the produce of the number of shares of a given type across all of its shareholders multiplied by the financial (or control rights respectively) of the given type;
• for each shareholder of this entity, the financial rights (and control rights respectively) are computed as the sum for all share types of the produce of the number of shares of a given type owned by this specific shareholder multiplied by the financial (or control rights respectively) of the given type;
• for each shareholder of a given we can then generate a unique ownership interest (and control right) value that corresponds to the ratio of the values determined in the previous two bullet points.

Example

• Input data

Entity: XYZ Corporation

Shareholders: Entity 1, Entity 2, Entity 3

Share Types: Type A, Type B

Share Types and Rights:

Share Type	Financial Rights	Control Rights
Type A	1.5	1.2
Type B	0.8	1.8

Share Distribution:

Shareholder	Type A Shares	Type B Shares
Entity 1	10	20
Entity 2	15	5
Entity 3	5	15

• Computation of Total Financial and Control Rights for XYZ Corporation:

Formula: TotalRights=∑(Number of shares of each type×Rights of that type)

Total Financial Rights	Total Control Rights
77.0	108.0

• Computation of Financial and Control Rights for Each Shareholder:

Formula: RightsforShareholder=∑(Number of shares of each type owned by shareholder×Rights of that type)

Shareholder	Financial Rights	Controlling Rights
Entity 1	31.0	48.0
Entity 2	26.5	27.0
Entity 3	19.5	33.0

• Computation of % interests

Formula: Interest=Shareholder’s Rights​/Total Rights

Shareholder	Financial Rights	Control Rights	Ownership Interest (%)	Control Interest (%)
Entity 1	31.0	48.0	40.26	44.44
Entity 2	26.5	27.0	34.42	25.00
Entity 3	19.5	33.0	25.32	30.56
Total	77.0	108.0	100.00	100.00

Micro-Step 1.2.2. Initial construction of the matrices

This step applies to both ownership interests and control rights, when already described as unique percentage values qualifying the rights of a given shareholder over one of its subsidiaries. At this stage there is a single value for a given couple entity-shareholder.

This may require to aggregate ownership interests and control rights that may have been expressed as multiple links even under their percentage form.

Once this preparation step is over, it consists in the construction of the Adjacency Matrix of the Shareholding Graph.

Micro-Step 1.2.3. Removal of direct auto-circular rights

Some Entities may hold ownership interests and/or control rights in themselves. It may make sense in the context of the management of shares in a legal entity, this makes the computation of indirect links impossible.

Therefore Micro-Step 1.2.3. empties the diagonal of the adjacency matrices of the shareholding graphs to remove of such links.

Once this is done, the matrices need to be re-normalized to reallocate the rights based on a total of rights that exclude the rights that were "self-owned" by the various entities. This normalization does not treat entities that have 0 parents in the shareholding structures to avoid division by 0.

Step 1.3. - Indirect ownership/control rights computation

Purpose

The purpose of this stage is to compute the Indirect ownership matrix, to allow all the computation that require indirect ownership values between entities.

General Logic

This Step 1.3. will be based on the Indirect ownership computation using Baldone ownership.

Data Input

The Data input for Step 1.3. consists in :

• The Ownership Interest Matrix computed in Step 1.2.
• The Control Rights Matrix computed in Step 1.2.

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.3. consists in :

• The Indirect Ownership Interest Matrix
• The Indirect Control Right Matrix

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The detailed computation method is detailed in the Indirect ownership computation using Baldone ownership and is applied similarly for the Control rights matrix.

Step 1.4. - Controlling interest matrix computation

Purpose

The purpose of the the Step 1.4. is to control the coherence of the Controlling interest matrix. The ability to identify the shareholder having a direct or indirect Controlling interest in a given Entity is key at several stages of the Pillar 2 computation.

However in most expected cases this matrix will not be available and will need to be supplemented by an Approximate Controlling interest matrix.

General Logic

This Step 1.4. will consist alternatively in the application of quality controls using Control rights and/or ownership interest matrices, or in the computation of the computation of the Approximate controlling interests matrix using majority rules along the shareholding chain.

Data Input

The Data input for Step 1.4. consists in :

• The precise Controlling Interest matrix;
• or alternatively the Indirect Control Rights matrix from Step 1.3;
• or as a minimal requirement the Indirect ownership Matrix from Step 1.3.

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.4. consists in :

• The list of errors in the Controlling Interest Matrix;
• The Approximate Controlling Interest;

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The list of Micro-Step is a follows :

• Micro-Step 1.4.1. Verification of the unicity of the shareholder having a controlling interest at each level of the shareholding chain
• Micro-Step 1.4.2. Computation of the Controlling interest matrix using majority rules

Micro-Step 1.4.1. Verification of the unicity of the shareholder having a controlling interest at each level of the shareholding chain

Logic

Using a Graph Transversal Algorithm to go up the direct Ownership interests (or Control rights if available) and check that at each stage only a single Parent entity is deemed to have a Controlling Interest in the Controlling Interest Matrix.

At each stage only one Parent has a Controlling Interest, and only this shareholding chain can have shareholders with Controlling Interests.

If it is not the case, this Matrix is deemed incoherent, will generate an error and will need to be supplemented by the Approximate Controlling interest matrix at the next step.

Example

TechCorp Group Structure (based either on Control rights preferably or ownership interests)

• UPE (Ultimate Parent Entity): TechCorp
• Subsidiaries: Subsidiary A, Subsidiary B, Subsidiary C
• Sub-Subsidiaries:Sub-Subsidiary A1 (owned by both Subsidiary A and B), Sub-Subsidiary B1 (owned by Subsidiary B), Sub-Subsidiary C1 (owned by Subsidiary C and A)

Potential Conflict in Control

Sub-Subsidiary A1 is co-owned by both Subsidiary A and B, while Sub-Subsidiary C1 is co-owned by Subsidiary C and A. These arrangements create potential conflicts in controlling interests, as there are multiple parent entities for some subsidiaries.

Verification Process for Unicity of Controlling Interest

The verification process involves analyzing each path from the lower-tier subsidiaries to the UPE, ensuring that only one controlling interest is present at each level:

1. Starting from Sub-Subsidiaries:
   • Examination begins with Sub-Subsidiaries A1, B1, and C1.
2. Traversing Upwards:
   • For Sub-Subsidiary A1, the shareholding paths for both Subsidiary A and B are assessed.
   • For Sub-Subsidiary C1, the assessment involves paths from both Subsidiary C and A.
3. Identifying Conflicts:
   • The process checks for overlapping controlling interests using the value stored in the Controlling Interest matrix. For example, the dual control of Sub-Subsidiary A1 by both Subsidiary A and B would be a conflict.
4. Error generation :
   • An error is returned by the control if for example bot A and B are deemed to have a Controlling Interest in A1 in the Controlling Interest Matrix

This approach ensures a clear and coherent control structure within the MNE, crucial for the coherent application of major components of the GloBE computation.

Micro-Step 1.4.2. Computation of the Approximate Controlling interest matrix using majority rules

Logic

To build the Approximate Controlling interest matrix, and given an adjacent matrix of indirect ownership interests (or control right if available), we have developed a simple majority rule to select among the in-group parents of a given entity the entities for which we should register a controlling interest. It however requires to have already identified the UPE, as well as the entities that are part of the MNE Group, UPE needs to be "In group".

Principle : For each entity B deemed "In group"

1. Short-list the "in-group" parent entities of B in which the control rights of A in B is higher than the control rights of the UPE in B reduced by the share of the control rights of the UPE in B obtained through A,
2. If the set is reduced to the UPE, only the UPE has a Controlling Interest in the Entity
3. Apply a simple majority rule for each possible Degree of "Parentality" in the graph, each entity being tested only once with its smallest degree (important when there are loops)
4. In case of equality for a given degree, all the equal values receive the Controlling Interest

This echoes the computation of the impact of the offset mechanism of §2.3. and preserves this logic.

Example

To illustrate the computation of the Approximate Controlling Interest Matrix using the majority rules for the TechCorp Group, we'll start by laying out the initial data of control rights in a table. Then, we'll apply the specified principle to compute the final matrix.

Initial Data: Control rights (could be based on ownership interests if not available)

Entity	Controlled by	Control rights (%)	In group
Subsidiary A	TechCorp (UPE)	60%	True
Subsidiary B	TechCorp (UPE)	50%	True
Subsidiary C	TechCorp (UPE)	60%	True
Sub-Subsidiary A1	Subsidiary A	30%	True
Sub-Subsidiary A1	Subsidiary B	20%	True
Sub-Subsidiary B1	Subsidiary B	80%	True
Sub-Subsidiary C1	Subsidiary C	15%	True
Sub-Subsidiary C1	Subsidiary A	15%	True

Computation Steps for the Approximate Controlling Interest Matrix

1. Calculate TechCorp's Indirect Control rights in Sub-Subsidiaries Through its Subsidiaries:
   • For each sub-subsidiary, compute TechCorp's partial Indirect Control rights through each subsidiary.
   • Sum the indirect interests to get TechCorp's total Indirect Control rights in each sub-subsidiary.
   • (please note that this methodology works in cases without loops)
2. Apply the Majority Rule:
   • Compare the subsidiary's indirect control rights in the sub-subsidiary with TechCorp's total indirect control rights.
   • Register a controlling interest for the subsidiary if its interest is higher than TechCorp's total indirect control rights minus the share obtained through that subsidiary.

Computation Results

1. For Sub-Subsidiary A1:
   • TechCorp's total control rights = (60% * 30%) + (50% * 20%) = 18% + 10% = 28%
   • For A
     • Subsidiary A's interest = 30%
     • TechCorp's total control rights outside of A = 10% =< 30%
   • For B
     • Subsidiary B's interest = 20%
     • TechCorp's total control rights outside of B = 18% =< 20%
   • A and B qualify for degree 1, but 30% > 20% and therefore A has the Controlling Interest
   • TechCorp's is alone for degree 2 and thus also has the Controlling Interests
2. For Sub-Subsidiary B1:
   • TechCorp's total indirect control rights through Subsidiary B = 50% * 80% = 40%
   • Subsidiary B's interest = 80%
   • Subsidiary B has a controlling interest in B1.
   • TechCorp's is alone for degree 2 and thus also has the Controlling Interests
3. For Sub-Subsidiary C1:
   • TechCorp's total indirect control rights = (60% * 15%) + (60% * 15%) = 9% + 9% = 18%
   • For A
     • Subsidiary A's interest = 15%
     • TechCorp's total control rights outside of A = 9% =< 15%
   • For C
     • Subsidiary C's interest = 25%
     • TechCorp's total control rights outside of C = 9% < 15%
   • A and C qualify for degree 1, and 15% = 15% and therefore both A and C have the Controlling Interest
   • TechCorp's is alone for degree 2 and thus also has the Controlling Interests

Final Approximate Controlling Interest Matrix

The resulting matrix will reflect the above computations, indicating a controlling interest where the majority rule is satisfied:

Controlling / Controlled	TechCorp (UPE)	Subsidiary A	Subsidiary B	Subsidiary C	Sub-Subsidiary A1	Sub-Subsidiary B1	Sub-Subsidiary C1
TechCorp (UPE)	True	True	True	True	True	True	True
Subsidiary A	False	False	False	False	True	False	True
Subsidiary B	False	False	False	False	False	True	False
Subsidiary C	False	False	False	False	False	False	True
Sub-Subsidiary A1	False	False	False	False	False	False	False
Sub-Subsidiary B1	False	False	False	False	False	False	False
Sub-Subsidiary C1	False	False	False	False	False	False	False

This matrix provides a clear visualization of the controlling interests within the TechCorp Group based on the majority rule computation.

Limitations

The test required by the GloBE model legislation is not a computational test but a legal test, and therefore it is complicated to come-up with a perfect computational solution. In most cases, the Approximate Controlling Interest Matrix should allow to generate a proxy for the Controlling interest matrix, with limited numbers of ambiguous situations.

In the next Step it is important to understand that the Approximate Controlling Matrix will replace the Controlling Interest Matrix if it is not available. When possible issues can arise from the remaining possible ambiguous situation, specific mention will be made.

Step 1.5. - Definition of the MNE Group

Purpose

The purpose of the Step 1.5. is to draw the perimeter of the MNE Group at the basis of the GloBE Perimeter, using the Controlling Interest Matrix and either the UPE or the Consolidation method.

General Logic

This Step 1.5. will consist in the use of the Controlling Interest Matrix to compute and/or verify the various elements that are used to compute and generate the GloBE Perimeter.

Data Input

The Data input for Step 1.5. consists in :

• The list of Entity Characteristics
• The Controlling Interest Matrix
• The Approximate Controlling Interest Matrix (verification only)

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.5. consists in :

• The updated UPE list
• The updated list of "In group" entities
• The list of errors generated by the step

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The list of Micro-Step is a follows :

• Micro-Step 1.5.1. UPE identification or control
• Micro-Step 1.5.2. Control of the consolidation methods along shareholding chains
• Micro-Step 1.5.3. Identification or control of the MNE group's perimeter

Micro-Step 1.5.1. UPE identification or control

This step is used to identify the UPE as either :

• the only entities with:
  • either :
    • no Controlling interests or,
    • which shareholder with Controlling Interest is not consolidated with no reasons listed in article 1.2.2.b) "size or materiality grounds"/, or "is held for sale",
  • and that has full consolidation as consolidation method;
• the only entities
  • marked "In group",
  • with no Controlling Interests of shareholders also marked "In group",
  • with consolidation method = full

If one or several entities are marked as UPE, errors should be returned in the following cases :

• Consolidation method for UPE is not full,
• UPE has Controlling Interests, either
  • among the entities marked "in group"
  • that have a consolidation method different from "not consolidated" or that have selected one of the reasons listed in article 1.2.2.b) "size or materiality grounds"/, or "is held for sale"

Micro-Step 1.5.2. Control of the consolidation methods along shareholding chains

Explanation of the requested data structure

The method of consolidation plays a key role in the construction of the GloBE perimeter, as it is the main criterion used to define the Group. It is also combined with the ownership interests to identify the Joint Ventures.

It is also important to note that "JVs of JVs" are not included in the GloBE perimeter, and only subsidiaries of Heads of JVs that are consolidated on a line by line basis by the Head of JV are included.

When consolidation methods are stored as an entity characteristic, the interpretation becomes ambiguous. For instance, consider a scenario where 'full consolidation' is used as an entity characteristic. The immediate question that arises is whether this refers to the consolidation method employed by the Ultimate Parent Entity (UPE) or is it specific to the consolidating entity in the context of Joint Ventures. This distinction is crucial. If a subsidiary is fully consolidated by the UPE, it implies a different level of control and financial integration compared to a Joint Venture where full consolidation might denote a shared control scenario.

Moreover, if the consolidation method is recorded in the context of shareholding relationships, another layer of complexity is added. It becomes challenging to discern who exactly is consolidating. For example, in a multinational group with a web of cross-shareholdings, it might be unclear whether a particular entity is being consolidated proportionally by one member of the group and fully by another. This ambiguity can lead to significant issues in financial reporting and governance.

This explains the requested data structure, requiring a consolidation method both by the UPE and for JVs only by the first consolidating entity, to allow for the eliminations of "JVs of JVs" from the GloBE perimeter at Step 1.7..

Controls on consolidation methods

If entities sandwiched between fully or proportionally consolidated can be "not consolidated", the general tendency along the shareholding chain needs to be that a parent is supposed to have a "stronger" consolidation method as a subsidiary.

Therefore the following controls can be applied to check for the coherence of the consolidation methods either by the UPE or by a given first consolidating entity:

an entity cannot have consolidation method = full/proportional when its direct parent with Controlling Interest has consolidation method = equity OR (consolidation method = not consolidated AND NOT one of the reasons listed in article 1.2.2.b) "size or materiality grounds"/, or "is held for sale") 

This can be checked using a Graph Transversal Algorithm on the direct shareholding structure, starting at either the UPE or any first consolidating entity and by testing for the existence of a Controlling Interest using the Controlling Interest Matrix.

The error is returned to indicate an incoherence in the consolidation method.

Furthermore an error should be returned for each Entity that is reported as first consolidating entity if :

• its consolidation method to UPE is not equity or;
• it is not the UPE;

Micro-Step 1.5.3. Identification or control of the MNE group's perimeter

Entity is part of the MNE Group if :

• UPE has Controlling interest in the entity
• and consolidation method
  • full/proportional/equity
  • or not consolidated for reasons in article 1.2.2.b) "size or materiality grounds"/, or "is held for sale"

Control of the "In group" indicator : an error is returned if a entity is predicted as "in group" but marked out of group in the forced input

Step 1.6. - Excluded entity identification

Purpose

The purpose of the this Step is to identify the Excluded entities based on the different mechanisms described in the model legislation, especially when they involve the use of the shareholding structure.

General Logic

This Step 1.6. will consist in identifying all the entities excluded from the GloBE perimeter based on their characteristics alone (article 1.5.1.) to then use a graph transversa algorithm to test for the exclusion of their subsidiaries not covered by article 1.5.1. but potentially by article 1.5.2. depending on the elections made, their characteristics and the shareholding structure.

Data Input

The Data input for Step 1.6. consists in :

• The Ownership Interest Matrix
• The list of characteristics of the entities
• The list of elections made for the entities

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.2. consists in :

• The list of Excluded entities for the 2 articles as well as their exclusion mechanism

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The list of Micro-Step is a follows :

• Micro-Step 1.6.1. Exclusion based on article 1.5.1.
• Micro-Step 1.6.2. Exclusion based on article 1.5.2.
• Micro-Step 1.6.3. Application of election from 1.5.3.

Micro-Step 1.6.1. Exclusion based on article 1.5.1.

This step simply verifies for each entity in the group if they match one of the criterion of the article 1.5.1., in the relevant characteristic.

Micro-Step 1.6.2. Exclusion based on article 1.5.2.

This step utilizes the methodology for Indirect ownership computation using Baldone ownership. It run as follows:

At the beginning 1.5.2.a) and b) are treated separately

• Identify all the entities that match the different criterion on the characteristics are activities of the entity listed for the 1.5.2. a) and b)
• Test to identify the candidates for the 1.5.2.a)
  • Build a new version of the Ownership Interest Matrix that only maps the shareholding structure for the entities also identified in the previous bullet point.
  • Create a generic Excluded Entity to which are attributed all the Ownership interests of any Excluded Entity according to article 1.5.1.
  • To maintain the integrity a generic not excluded entity is created to which are allocated all the Ownership Interests necessary to sum to 100%
  • Apply the Indirect ownership computation using Baldone ownership to this specific matrix
  • Define the set of entity that remain for the "chain test" by excluding the entities that do not respect the test of article 1.5.2.a) : 95% Indirect Ownership Interest by the Generic excluded entity AND the criterion of the 1.5.2.a).i or ii.
  • These form the 1.5.2.a) candidates
• Test to identify the candidates for the 1.5.2.b)
  • Build a new version of the Ownership Interest Matrix that only maps the shareholding structure for the entities also identified in the first bullet point.
  • Create a generic Excluded Entity - Excluding Pension Services Entities to which are attributed all the Ownership interests of any Excluded Entity according to article 1.5.1., but for those corresponding to "Pension Services Entity"
  • To maintain the integrity a generic not excluded entity is created to which are allocated all the Ownership Interests necessary to sum to 100%
  • Apply the Indirect ownership computation using Baldone ownership to this specific matrix
  • Define the set of entity that remain for the "chain test" by excluding the entities that do not respect the test of article 1.5.2.b) : 85% Indirect Ownership Interest by the generic Excluded Entity - Excluding Pension Services Entities AND the criterion of the 1.5.2.b)
  • These form the 1.5.2.b) candidates
• For the set of all candidates, run the following Recursive Algorithm
  • i=0
  • Count the number of Direct Parents
  • While(i<number of Direct Parents)
    • If Test(direct Parent) returns Excluded Entity 1.5.1. => Return Excluded Entity 1.5.2.
    • If Test(direct Parent) returns Excluded Entity 1.5.2. => Return Excluded Entity 1.5.2.
    • If Test(direct Parent) returns not Excluded => Test next parent, i++
  • Return not Excluded
  • This algorithm needs to be managed to ensure convergence in case of loops

Micro-Step 1.6.3. Application of election from 1.5.3.

This is a lot simple than the previous one : it sets a specific status to the elected entities so that they are not treated as excluded.

Here we make the strong hypothesis that the election does not break the chain that resulted from the step 1.2.2. analysis by disqualifying an entity as excluded with the election. If it is shown to be incorrect, we will need to add the election at the identification steps of the step 1.2.2..

Step 1.7. - Joint-Venture identification

Purpose

The purpose of the step 1.7. is to identify all the Joint Ventures and their JV Heads to carve them out from the general GloBE Perimeter.

General Logic

This Step 1.7. will consist in the identification of the Entities matching with the criterion of the Joint Venture, before forming the different a JV Sub-group by grouping the JV subsidiaries under the JV Heads that have a Controlling Interests in them.

Data Input

The Data input for Step 1.7. consists in :

• The Indirect Ownership Interest Matrix;
• The Controlling Interest Matrix;
• The list of characteristics of the entities with the consolidation method;

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.7. consists in :

• The list of JV Sub-groups identified by the JV Heads;
• The list of JV Subsidiaries in each JV Sub-groups;

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The list of Micro-Step is a follows :

• Micro-Step 1.7.1. Identification of all JV entities
• Micro-Step 1.7.2. Identification of all JV Heads
• Micro-Step 1.7.3. Iterative identification of the JV subsidiaries to the JV Sub-group

Micro-Step 1.7.1. Identification of all JV entities

This step identifies the set of entities that qualify as JV entities:

• marked as "In group"
• consolidation method = "equity";
• Indirect Ownership Interest by UPE >= 50%

Micro-Step 1.7.2. Identification of all JV Heads

To identify the JV Heads, this steps selects all the entities identified at the step 1.7.1. that do not have any shareholders with a Controlling Interest (or Approximate) in them.

If additional information on first consolidating entity is available only keep as JV Heads the entities that have the UPE as first consolidating entity.

Micro-Step 1.7.3. Iterative identification of the JV subsidiaries to the JV Sub-group

The following iterative process is applied to the set defined at the step 1.7.1. reduced by the JV Heads identified at the step 1.7.2.

Case 1 : only the consolidation method by UPE is available

• N= CARD(set defined at 1.7.1.)
• SET = set defined at 1.7.3.
• While (N>0 and card(SET)>0)
  • Affect each JV entity of the SET to the JV Sub-group where
    • at least one entity has a Controlling interest in it;
    • the direct shareholder with the highest Indirect Control right is located, if multiple JV sub-group match the first point;
  • Subtracts the affected JV entities from the SET
  • i--
• If at i = 0, the SET is not empty, all the remaining JV entities become independent JVs in their own JV sub-group

Case 2 : additional information is available for first consolidating entity

• N= CARD(set defined at 1.7.1.)
• SET = set defined at 1.7.3.
• All entities in SET that have Consolidation method by first consolidating entity = "equity" are disregarded, from SET (that is subsequently updated)
• While (N>0 and card(SET)>0)
  • Affect each JV entity of the SET to the JV Sub-group where the first consolidating entity is affected
  • Subtracts the affected JV entities from the SET
  • i--
• If at i = 0, the SET is not empty (impossible), all the remaining entities are returned as ERRORS (as

Step 1.8. - MOCE and MOMNE identification

Purpose

The purpose of the step 1.8. is to identify all the Joint Ventures and their JV Heads to carve them out from the general GloBE Perimeter.

General Logic

This Step 1.8. will consist in the identification of the Entities matching with the criterion of the Minority-Owned Constituent Entity, before forming the different a Minority-Owned Sub-group by grouping the MOCEs under the Minitority-Owned Parent Entities that have a Controlling Interests in them.

Data Input

The Data input for Step 1.8. consists in :

• The Indirect Ownership Interest Matrix;
• The Controlling Interest Matrix;
• The list of characteristics of the entities with the consolidation method;

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.8. consists in :

• The list of MOMNEs identified by their MOPEs;
• The list of MOCEs in each MOMNE;

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The list of Micro-Step is a follows :

• Micro-Step 1.8.1. Identification of all MOCEs
• Micro-Step 1.8.2. Identification of all MOPEs
• Micro-Step 1.8.3. Iterative identification of the MOCEs to the MOMNEs

Micro-Step 1.8.1. Identification of all MOCEs

This step identifies the set of entities that qualify as JV entities:

• marked as "In group"
• consolidation method different from "equity";
• Indirect Ownership Interest by UPE < 30%

Micro-Step 1.8.2. Identification of all MOPEs

To identify the MOPEs, this steps selects all the entities identified at the step 1.8.1. that do not have any shareholders with a Controlling Interest (or Approximate) in them.

Micro-Step 1.8.3. Iterative identification of the JV subsidiaries

The following iterative process is applied to the set defined at the step 1.8.1. reduced by the MOPEs identified at the step 1.8.2.

• N= CARD(set defined at 1.8.1.)
• SET = set defined at 1.8.3.
• While (N>0 and card(SET)>0)
  • Affect each MOCEs of the SET to the MOMNEs where
    • at least one entity has a Controlling interest in it;
    • the direct shareholder with the highest Indirect Control right is located, if multiple MOMNEs match the first point;
  • Subtracts the affected MOCEs from the SET
  • i--
• If at i = 0, the SET is not empty, all the remaining MOCEs become independent MOCEs on their own.

Step 1.9. - Specific CE treatments

Purpose

The purpose of this step is to complete the treatment of the GloBE perimeter for various entity types : Permanent establishments, Investment entities, MOMNEs within JVs.

General Logic

This Step 1.9. will consist in the completion of the perimeter analysis to finalize the GloBE perimeter, the different Micro-steps will follow a mechanical order.

Data Input

The Data input for Step 1.2. consists in :

• The input from the previous Steps
• The data generated from the previous Steps

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Data Output

The Data output for Step 1.2. consists in :

• The update list of MOMNEs, MOCEs and their compositions
• The list of Investment Entities and the sub-groups they form in each sub-group
• The affectation of Permanent establishment to the sub-groups of their Parent Entity

The detailed matrix is available in the relevant section of the Pillar 2 Data Structure.

Detailed computations

The list of Micro-Step is a follows :

• Micro-Step 1.9.1. Identification of MOMNEs and MOCEs within JV sub-groups
• Micro-Step 1.9.2. Affectation of the Permanent Establishment to the sub-groups of their Main Entities
• Micro-Step 1.9.3. Isolation of Investment Entities within each subgroups

Micro-Step 1.9.1. Identification of MOMNEs and MOCEs within JV sub-groups

As they are supposed to be treated as separate MNE Groups for the purpose of most of the GloBE Rules, and as they are composed solely of entities consolidated on a line by line basis by their JV Heads, it is necessary to isolate the MOMNEs and MOCEs from the point of view of the JV Sub-Group.

In order to achieve this purpose, the step launches the Micro-Step 1.8. on each JV Sub-Groups, and utilizes the JV Head and its Indirect Ownership Interests in the JV Subsidiaries to apply the criterion.

Micro-Step 1.9.2. Affectation of the Permanent Establishment to the sub-groups of their Main Entities

This stage is quite simple has it only requires to affect each Permanent establishment to the Sub-group of their Parent Entities.

Micro-Step 1.9.3. Isolation of Investment Entities within each sub-groups

Identification of investment entities

The following entities are identified as Investment entities

• The Investment entities that match the activity criterion of the a) : "Investment Fund"; "Real Estate Investment Vehicle"
• The Investment entities that match the activity criterion of the b) or c) and that are owned directly or indirectly through a chain of entities defined at the first bullet point at least at 95% or 85% respectively => the same algorithm as Step 1.6.2. for Excluding Entities, with the notable difference that the sets are treated completely separately

Isolation in their respective sub-groups

At this stage all investment entities have been affected to specific sub-groups or are still in the Main perimeter under the UPE.

The Investment Entities of each sub-group will then be grouped within their respective sub-groups to form isolated sub-groups identified with an Investment Entity marker.