Leveraging shareholder data for beneficial ownership transparency

Discussion and conclusion

The analysis has helped answer the following research questions:

1. Data overlap and alignment: To what extent do shareholder and BO registers capture the same ownership information?

Across both case studies, there is significant overlap between shareholder and BO registers. The findings reinforce that many corporate structures are relatively simple, often involving direct ownership by a small number of individuals. In such cases, shareholder registers can already capture a substantial share of interests and relationships that are later reported through BO declarations. This suggests that shareholder data can play a significant role in supporting BOT systems, including by improving accuracy and usability, reducing duplicate reporting, and lowering reporting burdens by leveraging existing information.

2. Interest types in BO data: What proportion of BO declarations relate to shareholding, compared to other forms of control (e.g. voting rights, appointment powers, or other means of control)?

While the vast majority of interests relate to shareholding, information collected through BO declarations remains essential to capture forms of control not visible through shareholding, including indirect cross-border ownership and other non-share-based mechanisms.

3. Domesticity of corporate networks: To what extent are corporate networks derived from shareholder data domestic, and how does this compare with those derived from BO data?

Corporate networks mostly comprise direct ownership by natural person shareholders, meaning domestic shareholder data can play a significant role in increasing compliance – such as in Zambia – or the verification of BO declarations. In lieu of access to foreign shareholder data, BO data helps identify cross-border ownership structures. While complex structures may be a small proportion of corporate structures, these are the ones of primary interest in many policy use cases for BOT, including fighting financial crime. The analysis further suggests that, even when shareholder data from multiple jurisdictions is combined, it may still underestimate cross-border links, as BO data captures additional relationships. This is especially the case for those involving foreign individuals and interests beyond the ownership of shares.

This work highlights the distinct and complementary roles for shareholder and BO data, and shows that shareholder data could be beneficial for improving BO information and better achieving policy goals. These findings suggest that agencies in charge of companies and BOT reforms should explore:

• Where possible, conducting similar analyses for their jurisdictions as those laid out in this paper, in order to help quantify the potential value of shareholder information and identify next steps, or consider making anonymised datasets available for research.
• Investing in existing or creating new central shareholder registers with up-to-date, accurate, and structured information.
• Aligning shareholder and BO regimes, including coverage of legal entities and reporting periods.
• Reducing the duplication of data collection through pre-populated forms, verification, or exemptions, particularly for simple corporate structures.
• In addition to efforts to strengthen cross-border access to BO information, improving this for shareholder information in order to help understand transnational corporate structures in more detail.
• Potentially, in future and as data-sharing improves, gradually reducing data duplication at the international level, e.g. by expanding exemptions to BO declarations for cross-border corporate structures covering jurisdictions where information is readily available.

Further research

Although Estonia and Zambia represent different contexts, further research is needed to see whether patterns of corporate networks are consistent across more jurisdictions, and whether the findings in this paper are indeed broadly applicable.

The analysis has shown a lack of alignment between the policy and legislative frameworks governing the collection and storage of shareholder and BO information. It was not always possible to explain certain differences in the data, for example: why a significant number of individuals appeared only in shareholder data that was not explainable due to a lack of compliance or below-threshold ownership shares. Further comparative legal and policy analysis will help shed light on these questions.

Finally, further research should examine jurisdictions currently exploring or implementing digital central shareholder registers, such as Norway, and with constitutive effect, such as Pakistan. It is also worth learning from jurisdictions currently implementing reforms that leverage shareholder data for BOT reforms, including Switzerland and Zambia.