The Role of Big Data Analytics in a

Globalised World
Maqbool Lalljee
Director: Tax Risk Solutions
Introduction
• Globalization and the increasing loss of tax from avoidance, evasion-illicit activities

• Drive to make data accessible across tax administrations

• Developing tools, capacities and capabilities to recoup tax loss

• Increasing use of Big Data Analytics ( advanced analytics)

• Use case examples – potential tax avoidance and tax evasion structures

• Benefits realisation

• Key takeaways
1 Globalization
Increasing loss of tax
Lose of tax from avoidance, evasion-illicit activities

Globalisation over the last 10 years has seen considerable tax losses through:

• Tax avoidance over $240 billion are lost annually due to tax avoidance by multinational companies
https://www.oecd.org/tax/beps/

• About $190 billion of global tax revenue is lost through offshore tax evasion. Gabriel Zucman, 2015

• From money laundering amounts to around $2.8 trillion, or 5% of global GDP annually Source: European Parliament,

Offshore activities and money laundering: recent findings and challenges, 2017

• Estimated untaxed wealth of individuals in tax havens is around 10% of global GDP Source: Business Insider article,
2017
2
Access to Global Data

Across Tax & Customs

Administrations
Drive to make data accessible to Administrations
• Base erosion and profit shifting (BEPS) Action 13 – Country by Country Reports - 135+ countries
collaborate on the implementation of the BEPS package of sharing data across administrations

• Common Reporting Standards (CRS) - requirement for financial institutions in participating

countries/jurisdictions, to help fight against tax evasion – 100+ jurisdictions committed to the CRS

• Data and intelligence made available by CITES (Convention on International Trade in Endangered
Species of Wild Fauna and Flora)

• Each administration now collects vast amounts of data annually through returns, electronic
lodgements, tax amnesty – voluntary disclosure campaigns

• Number of public disclosures are made available to combat tax mischief – for example work of
International Consortium of Investigative Journalists and availability of their data and information

• More and more data is collected commercially and made easy to blend with the administrations'
data ( see examples )
 Examples of growing source of Bureau van Dijk data

holding

Note: Figures collected January-June 2020. Figures are rounded.

 … and its coverage is constantly increasing

Note: Figures are rounded.

 Growing beneficial ownership and sanctions data

Note: January 2020 figures. Percentage increases are year-on-year. Figures are rounded.
 Growing ownership links…

Note: January 2020 figures. Percentage increases are year-on-year. Figures are rounded.
 Easier to blend data with a wide range of identifiers…

Note: January 2020 figures. Figures are rounded.

3
Developing Tools

Capacities and Capabilities

 Developing tools, capacities and capabilities to recoup tax
loss
• Around mid-2000s, some of the larger Tax & Customs administration started investing into:

a) large data warehouses, with relational databases, creating data marts – risk marts for tax risk
data analytics

b) triggered reviewing the IT estate, rationalising the mainframe systems that didn’t speak to each
other

• Large IT and data programmes were initiated with private sector consultants employed

• New recruitment of specialist commenced not just for IT, data but in the areas of international
taxation
Increasing use of Big Data Analytics

• Around 2015, with the growing phenomenon of “Big- Data” (larger volumes of data – digital
data)
• Larger investments being seen in:
a) data architectures – creating data-hubs, using new toolkits, including Artificial Intelligence
b) recruiting new brand of data analysts/data scientists

• 10 years ago less than 10 Tax and Customs administrations employed “Big Data” and
advanced analytics

• 2020 there are more than 40 administrations employing advanced analytics and transforming
their IT Infrastructures
2010: typical data Architecture of a large tax administration
 Your own tax data Bureau van Dijk’s External information
efiled and elodged global ownership • electronic data from financial
• returns for individuals, data feed from institutions, third parties (for
partnerships and companies with example, CITES), and
VAT and tax information government exchanges
• customs transactional data • tax jurisdictions reports
• customs duties • public disclosures; for example,
• corporate financials ingested into the lake by linking key the Panama Papers
• commercial data and intelligence variables and gluing data • open source information

Hadoop data lake

Upstream risk management Downstream risk management

• enabled by Bureau van Dijk’s API, advanced
• enabled by business intelligence analytics, neural networks, text and data mining
• executed by junior analysts • output used by tax and policy professionals, data
• automated with artificial intelligence scientists, senior leadership
Note: CITES – the Convention on International Trade in Endangered Species of Wild Fauna and Flora

16
Use of advanced data analytics

Global Map 2010 - under 10 Tax Global map 2020 - more than 40 Tax
Administrations Administrations

Author’s research
4
Use case examples
Potential tax avoidance and
tax evasion structures
Complex large multinational structures – input for risk model
Risk Model outputs: Example 1: possible South East Asian
Tax Avoidance Structure( Source Bureau van Dijk , Orbis Data Explorer )

• Foreign MNE ( US Global Ultimate/beneficial

Owner)
• Use of Dutch of BV ( Holding and
Finance/Investment) entities identified
• Swiss Intellectual Property Holding identified
• Review of any inter-company loans, interest
payments, royalty payments up-wards from SEA
entity
• Treaties applications – withholding taxes (none ?)
on Interest payments, Royalty Fees and
Dividends – interrelated application between SEA
State, Netherlands, Switzerland, USA
• Permanent Establishment ( possible challenges
on foreign entities - substance tests )
• Review potential Transfer Pricing issues – arms
length principle – intra-group transactions
Example 2 neural network model for detecting illicit
sources of finance
Non-transparent ownership structure, including
1 sanctioned individuals and/or those showing
non-compliant or criminal behaviour

2
High increase in income or profit year over year, 1
or a new very large influx of capital

3 Cost and sales ratios of the business not in line 8 2

with the industry benchmarks

4 Sale or purchase of companies with undervalued

or overvalued stocks

5 Unclear or unverifiable origin of money 7 9 10 3

Financial transactions with subsidiaries based in
6
tax havens

7 Transactions in goods and services outside

the company’s normal trading areas
6 4
8 Large transactions with fictitious companies

9 Nominee officers of related offshore companies 5

10 Use of special purpose vehicles for large

financial transactions

21
Example 2 complex secretive structure

Source: Bureau van Dijk’s Orbis. Note: AHC/IHC – Asset/Investment Holding Company; Nominee – IHC’s nominee officer and shareholder

22
5 Benefit Realisation
Benefits realisation

• Centralising analytics
• Saving technicians’ time through automation
• Accurate output for tax challenges
• Senior team given the access through web-interface
• Support tax and analytics teams empowered
• Data and predictive data modelling capacity enhanced
• Real investigation/litigation successes through data led investigations and analytics
• Able to react to urgent analytical output for policy or rapid campaigns
• Contribution to tax yield achievement in $ hundred of millions
• Better analytical capability to review/risk assess Country by Country Reports and other
BEPS recommendations
• Setting the scene for digital development – make use of cutting edge technology and
Artificial Intelligence
6 Key takeaways
Takeaways

• Big data and advanced analytics play an important role in closing the tax gap

• Investment in cutting edge technology and sophisticated algorithms produce powerful,

efficient neural networks, that reveal income from fraudulent and illicit activities.

• Investment into global company data and a 21st century IT infrastructure will reduce
the considerable financial losses to revenue and customs administrations.

• This should be an informed investment decision.

26
Thank you
maqbool.Lalljee@ | bvdinfo.com