Pillar 2 Algorithm

Introduction

Context

Welcome to the OpenSource Pillar 2 Algorithm wiki page! This platform serves as a comprehensive resource for understanding and implementing a pioneering open-source algorithm specifically designed to perform Pillar 2 computations and prepare the GloBE Information Return.

BEPS Pillar 2 represents a significant step in global tax reform, introducing the concept of a Global Minimum Tax to prevent aggressive tax planning and profit shifting by multinational enterprises. Given the intricate nature of international taxation, coupled with the unique structures of individual corporations, computing the tax liabilities under the Global Minimum Tax framework is inherently complex. The scope of the regulation and the various election options available to entities further complicate the computations, requiring in-depth analysis and consideration.

Objectives

Recognizing this complexity, Algonomia has developed a sophisticated algorithm to address these challenges. By opting to release it as open-source, Algonomia aims to foster collaboration among tax specialists, economists, and administrations. This approach not only ensures the refinement and enhancement of the algorithm but also promotes transparency, demonstrating a commitment to a fair and unified international tax landscape.

Feature Overview

The algorithm will present the following features:

• Scope and Groupe perimeter analysis
• Safe Harbour computation and analysis
• GloBE Income and Adjusted covered tax computation
• Election treatments
• Effective Tax Rate and Top-Up Tax computation
• Aggregated tax impact on each tax payer including different features of tax systems (losses carry forward, tax credits on WHT...)

Data structure

Algonomia API Framework

The Transfer Pricing Algorithm uses Algonomia API Framework to receive input data and send output data for the computations. In the following sections, it is assumed that data has been transferred to the Transfer Pricing Algorithm using this framework.

Input

The input data follows the usual division required by the Algonomia API Framework : Data, Rules, Other.

Data

The following tables presents the data points allowed as "Data" :

Input Data Points

Role	Label	Definition	Typing	Multiplicity behavior	Mandatory
Matching	ATP Declaring Entity segmentation	Allows to define the segmentation of the data input that corresponds to the base aggregation level for the Transfer Pricing Computation (PLIs will be aggregated at this level of segmentation)	Couple (String; Integer)	Several Values can be defined as "ATP Declaring Entity segmentation", but the Integer has to be unique among the set	Yes
Matching	ATP Tax Payer	Allows to define the segmentation of the data input that corresponds to the aggregation level for the Tax computation, it is necessarily a subset of the ATP Declaring Entity segmentation defined	Boolean	Completes the definition of the ATP Declaring Entity segmentation with a boolean (String; Integer; Boolean)	No, if absent no tax computation are performed
Matching	ATP Counterpart Entity segmenfation	Allows to define the segmentation of the data input that corresponds to the base aggregation level for the Transfer Pricing Computation (PLIs will be aggregated at this level of segmentation), if it is not defined or null for certain data entries the Counterpart entities will be sourced from the associated Rules	Couple (String; Integer)	Several Values can be defined as "ATP Counterpart Entity segmentation", but the Integer has to be unique among the set	No, if absent the Counterpart entity may be automatically defined using the Rules file

Rules

The following tables presents the data points allowed as "Rules" :

Input Rules data points

Role	Label	Definition	Typing	Multiplicity behavior	Mandatory
Matching	ATP Rule segmentation	Allows to define the segmentation level of the Rule component to be retrieved	Couple (String; Integer)	Several Values can be defined as "ATP Rule segmentation", but the Integer has to be unique among the set	Yes
Matching	ATP Counterpart Entity segmentation	Allows to define the segmentation that defines the Counterpart Entity, it is recommended to use the same definition as the Declaring Entity segmentation to allow for smooth computations for the counterpart entities and their tax impacts	Couple (String; Integer)	Several Values can be defined as "ATP Counterpart entity segmentation", but the Integer has to be unique among the set	No, if absent from the data as well, no TP computations are possible

Serie of files

A series of structured files corresponds to the selection of a sequence of structured files whose data and rules can be processed in a coordinated manner in the dashboard or the calculations of the algorithm.

The selected structured files must form a sequence where their temporal referencing perimeters follow each other in a disjointed and adjacent manner (e.g., a succession of months, quarters, years).

Impact of series on Financial data

Initially, these relative temporal reference values are sought within the same structured file, and then in the preceding structured file that corresponds to the correct relative reference in the matching file within the sequence.

Only one single value works this way for TP computation : ATP Tax losses carry-forward N-1.

Impact of rules

The optimization calculation will be further impacted as transfer pricing rules can cover a reference period exceeding a single structured file. The optimization calculation will then treat the rule file variable as a shared variable.

Computational steps

Data received by the algorithm are processed using the standardized Rule matching methodology. At this stage every single entry contains both the financial data and the rules to be applied for both Transfer Pricing Computation and Tax impact computation.