[Paper Review] Identification of Key Companies for International Profit Shifting in the Global Ownership Network

Tembo Nakamoto, Abhijit Chakraborty+1 more|arXiv (Cornell University)|Apr 28, 2019

Corporate Taxation and Avoidance39 references16 citations

TL;DR

This paper proposes a network-based hierarchical model to identify key intermediate companies at high risk for international profit shifting within multinational corporations (MNCs). Using ownership data from the Orbis database, it constructs a Global Ownership Network (GON) and identifies firms in the IN component of the bow-tie structure—particularly in treaty-shopping-friendly jurisdictions like the Netherlands and UK—as central to profit shifting. The model reveals that key firms are geographically concentrated and strategically positioned in ownership hierarchies to enable tax arbitrage.

ABSTRACT

In the global economy, the intermediate companies owned by multinational corporations are becoming an important policy issue as they are likely to cause international profit shifting and diversion of foreign direct investments. The purpose of this analysis is to call the intermediate companies with high risk of international profit shifting as key firms and to identifying and clarify them. For this aim, we propose a model that focuses on each affiliate's position on the ownership structure of each multinational corporation. Based on the information contained in the Orbis database, we constructed the Global Ownership Network, reflecting the relationship that can give significant influence to a firm, and analyzed for large multinational corporations listed in Fortune Global 500. In this analysis, first, we confirmed the validity of this model by identifying affiliates playing an important role in international tax avoidance at a certain degree. Secondly, intermediate companies are mainly found in the Netherlands and the United Kingdom, etc., and tended to be located in jurisdictions favorable to treaty shopping. And it was found that such key firms are concentrated on the IN component of the bow-tie structure that the giant weakly connected component of the Global Ownership Network consist of. Therefore, it clarifies that the key firms are geographically located in specific jurisdictions, and concentrates on specific components in the Global Ownership Network. The location of key firms are related with the ease of treaty shopping, and there is a difference in the jurisdiction where key firms are located depending on the location of the multinational corporations.

Motivation & Objective

• To identify intermediate companies within MNCs that are at high risk for international profit shifting.
• To develop a hierarchical network model that evaluates affiliate positions within MNC ownership structures, rather than relying on financial data.
• To analyze the geographic and structural concentration of such key firms in the Global Ownership Network (GON).
• To examine the relationship between key firm locations and treaty shopping opportunities.
• To reveal how MNCs route investments through specific jurisdictions via intermediate entities for tax optimization.

Proposed method

• Constructs the Global Ownership Network (GON) using firm-level ownership data from the Orbis database.
• Applies bow-tie structure analysis to decompose the GON into IN, OUT, and CCR components to identify central nodes.
• Introduces a hierarchical identification model based on network centrality and ownership layering to detect key firms.
• Classifies affiliates into layers (e.g., holding, holding and conduit, conduit) based on their position in the ownership hierarchy.
• Uses community detection and centrality measures (e.g., in-degree centrality) to identify influential nodes in the network.
• Validates the model by analyzing five Fortune Global 500 MNCs and assessing whether identified firms align with known profit-shifting patterns.

Experimental results

Research questions

• RQ1 Which firms in the ownership structure of MNCs are most likely to enable international profit shifting, and how can they be systematically identified?
• RQ2 How are key profit-shifting firms distributed geographically across the Global Ownership Network?
• RQ3 To what extent do key firms cluster in specific network components, such as the IN component of the bow-tie structure?
• RQ4 How do the locations of key firms relate to treaty shopping opportunities in jurisdictions like the Netherlands and the UK?
• RQ5 What patterns emerge in the ownership layering of key firms across MNCs from different home jurisdictions (e.g., US, UK, China)?

Key findings

• Key profit-shifting firms are predominantly located in the Netherlands and the United Kingdom—jurisdictions known for favorable tax treaties and treaty shopping opportunities.
• These key firms are highly concentrated in the IN component of the bow-tie structure, indicating their central role in capital inflows within the global ownership network.
• For US MNCs, 12% of 'holding and conduit' firms and 8% of 'conduit' firms are located in the US, suggesting potential round-tripping behavior.
• UK MNCs have 11% of 'holding and conduit' firms and 17% of 'conduit' firms in the UK, indicating domestic routing of capital through key intermediaries.
• Chinese MNCs show a strong concentration of 'conduit' firms in tax havens such as the Cayman Islands (29%), British Virgin Islands (24%), and Bermuda (6%).
• The model successfully identifies known profit-shifting entities in five Fortune Global 500 MNCs, validating its ability to detect high-risk intermediaries based on network position alone.