Effects of the legal and fiscal framework in UK ownership networks

AI-UK presenter: Leonardo Castro Gonzalez

Project description

We study the different shapes that company ownership and control networks take in the UK. By knowing which companies or humans have shares of other companies, we reconstruct the networks using data from Companies House. Observed networks can go from a fairly simple shape (company-owner) to a complex structure with different branches and long chains of companies.

Put in simple terms, why do we see such a variation in structures? Are we looking at an organic structure, or one made by design?

We focus on networks who have as final beneficiary a human. We suppose a network can grow following three principles:

Logistics/diversification - A companies network might find useful to open or buy shares from a new company to improve its logistics or diversify
Minimise liabilities - A company can decrease the liabilities over a second one by setting a chain of companies in between
Minimise tax burden - A human’s wealth pays less taxes if it is distributed among the profits of a network

Impact

Companies and humans take the legal framework of the UK as guide to create an ownership network. We intend to create a sandbox with wich:

We can model the effect of different fiscal policies in ownerhip networks
We can detect which policies affect the finding of fringe behaviours
Help policy-makers and scientific community to better understand the industrial activity in the UK
Help Companies House to pinpoint which minimal information is necessary to increase transparency

Adapt our model to other legal and fiscal frameworks around the world

Examples

A simple ownership network with two humans (blue nodes) and their owned companies (green).
A more complicated ownership network with branches. The blue node refers to a human, purple nodes refer to middle companies and green nodes refer to leaf companies.

CSS Group @ The Alan Turing Institute