Bond University
Research Repository

Global money laundering appeal index: application of principal component analysis
Tiwari, Milind; Gepp, Adrian; Kumar, Kuldeep
Published in:
Journal of Money Laundering Control
DOI:
10.1108/JMLC-10-2021-0108
Licence:
Other
Link to output in Bond University research repository.

Recommended citation(APA):
Tiwari, M., Gepp, A., & Kumar, K. (2023). Global money laundering appeal index: application of principal
component analysis. Journal of Money Laundering Control, 26(1), 205-211. https://doi.org/10.1108/JMLC-102021-0108

General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
For more information, or if you believe that this document breaches copyright, please contact the Bond University research repository
coordinator.

Download date: 02 Nov 2026

Global Money Laundering Appeal Index: An Application of Principal Component
Analysis
Milind Tiwaria1, Adrian Geppa, Kuldeep Kumara
a

Bond University, Australia

Abstract
The paper aims at developing a global ranking system determining a country's appeal as a
destination for money laundering. Using Principal Component Analysis (PCA), four
components relating to economic feasibility, financial liberty, government spending, and tax
regime were critical in influencing a country's money laundering appeal. It is the first attempt
to use a statistical technique to understand the underlying components of a country's money
laundering appeal. It could be used to develop more effective preventative strategies.
Key Words: Economic growth; Forensic accounting; Money laundering; Principal Component
Analysis; Walker Gravity Model
JEL Classification: C38; G32; E39; K42

Corresponding author: Milind Tiwari, Bond University, 14, University Drive, Robina, Queensland – 4226,
Australia, e-mail address: mtiwari@bond.edu.au
1

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

1. Introduction
Money laundering presents social, economic, and political harm to a country and the global
economy by providing an opportunity to launder and reinvest criminal proceeds. Such
opportunities lead to economic distortions, erosion of financial sectors, reduced government
revenues, and other socioeconomic effects (Barone et al., 2018; Degryse et al., 2019; Walker
and Unger, 2009). The extent of economic and financial disruptions can be assessed based on
Barone and Masciandaro (2011) estimates. They found money laundering operations to
produce legal assets worth USD 108.72 billion in Eastern Europe in 2009. As a result, efforts
have been made to measure money laundering flows, construct money laundering risk
indicators and uncover other such instances (Basel Institute on Governance, 2014; Collin,
2019; Ferwerda and Kleemans, 2019). Walker (1999) can be regarded as a pioneer in the
domain. He postulated a range of factors contributing to the attractiveness of a destination for
money laundering. The factors contributing to attractiveness have often been used to estimate
the magnitude of money laundering, and there is often a debate around the accuracy of the
generated estimates (Collin, 2019).
Efforts have been made to improve the proposed Walker Gravity model, and one such effort in
this regard was by Unger et al. (2006). The attractiveness component has been improved only
through the addition or removal of variables. No attempts in the literature have been made to
use quantitative techniques to develop an index of money laundering appeal. This paper
attempts to use a standardized and replicable methodology to condense into a single measure
the complex and multifaceted phenomenon of a country's appeal as a destination for money
laundering, thus avoiding the difficulty of precisely calculating illicit financial flows.
This paper uses principal component analysis (PCA hereafter), with a mix of standardized and
unstandardized components relating to attractiveness, economic freedom, and money

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

laundering risk to come up with an index of money laundering appeal (Basel Institute on
Governance, 2014; Khan et al., 2018; Kim and Holmes, 2016). Such an index would act as an
additional tool to support policymakers and investigators in allocating relevant resources
effectively and developing suitable preventative strategies; for banks and professionals, the
indicator would improve their due-diligence operations. Such efforts would combat the
phenomenon of money laundering responsible for the stagnation of economic growth through
tax evasion, corruption, and the creation of non-competitive markets.
2. Data and Methodology
The data on 30 variables incorporated for analysis in the paper have been collected for 150
countries. The year chosen for analysis is 2014 to facilitate comparison with the results of an
earlier study that used the traditional gravity model to develop the attractiveness of money
laundering for countries. In the case of missing variables, the variable information available
for the most recent year, past or present, is considered for analysis. Finally, not to exclude
countries from analysis, the variables for which no information is available, the variable value
is recorded as 0 for that country. Table I below provides the list of variables used for analysis:
Variables

Source

Description

Per Capital Gross Domestic Product

The World Bank

Assessment of country's macroeconomy

(GDP), Government Expenditure, Tariff

(2020)

Rate, Income Tax Rate, Corporate Tax
Rate, Tax Burden (% GDP), Public Debt (%
GDP),

FDI

Inflows

(in

millions),

Unemployment, Inflation (%), Financial
Deposits
Basel AML Score

Basel Institute on

Ranking countries based on their

Governance

assessed risk of money laundering or

(2014)

terrorist activity to come up with Basel

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

Anti-money laundering (AML) Risk
Index
Economic

Freedom

Score,

Business

Freedom,

Financial

Freedom,

Fiscal

Kim and Holmes

Assessment

of

economic

freedom

(2016)

available to a country's population to

Freedom, Government Spending, Labor

come up with the Index of Economic

Freedom,

Freedom

Monetary

Freedom,

Trade

Freedom
Attractiveness Score, Banking Secrecy

Khan et al. (2018);

Estimate the magnitude of money

(BS), Conflict (C), Corruption (COR),

Walker and Unger

laundering for countries subject to their

Member

(2009)

attractiveness for it

of

Egmont

Group

(EG),

Government Attitude (AG), Swift Member

Table I. List of variables
The use of variables for accessing countries' attractiveness has been limited to estimating the
magnitude of money laundering. It presents an opportunity to develop an index of a country's
appeal as a destination for it. The rationale to incorporate additional variables stems from works
directing attention towards the link between money laundering and economic cycles, economic
growth, and taxes (Barone et al., 2018; Johannesen et al., 2016). Such an index would differ
from the Basel AML Risk Index and other similar indices that do not measure the country's
actual money laundering activity. These indices' focus has been to assess the risk level based
on adherence to AML/CTF (Countering Terrorism Financing) standards and other such risk
categories. The rankings of such global indices are more suitable in a probabilistic rather than
economic sense.
This paper uses PCA to reduce a large number of variables into fewer components representing
a specific phenomenon; in this case, money laundering appeal. The data are assessed to
determine how the variables covariate. The number of components is selected based on

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

accepted standards – the Kaiser-Harris criterion and parallel analysis. The unrotated
components are subject to rotation to ensure the pattern of loadings is easier to interpret. The
components are then extracted to be normalized and combined in a composite indicator using
the proportion of variance explained by each component as weights (Ferwerda and Kleemans,
2019; Watkins, 2006).
3. Results
The suitability of the data for PCA is assessed by using a correlation matrix, Kaiser-MeyerOlkin Measure of Sampling Adequacy (KMO), and Barlett's Test of Sphericity (Kaiser, 1960).
The high correlation between variables (as seen in Figure 1), a KMO value of 0.768, and a
significant sphericity test, as presented in Table II suggest the data are suitable for PCA.

Figure I. Correlogram of variables

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

Kaiser-Meyer-Olkin Measure of Sampling Adequacy.
Bartlett's Test of Sphericity

Approx. Chi-Square

0.768
4931.351

df

435

Sig.

0.000

Table II. KMO and Bartlett's Test of Sphericity

The decision of the number of components to be extracted is based on a mix of Kaiser-criterion
(components with eigenvalues > 1) and parallel analysis (Watkins, 2006). The comparison of
eigenvalues obtained from Kaiser-criterion and parallel analysis (Table III) are presented below
and suggests that four components should be extracted:

Initial Eigenvalues
Component

Total

Random Data Eigenvalues (Parallel Analysis)

% of variance

Total

Decision

1

10.922

36.406

2.067

Accept

2

3.063

10.210

1.916

Accept

3

1.996

6.654

1.779

Accept

4

1.878

6.260

1.679

Accept

5

1.390

4.633

1.595

Reject

Table III. Comparison of eigenvalues from Kaiser-criterion and Parallel Analysis
The extracted components were subject to varimax rotation to ensure suitable factorization.
The output is presented in Table IV and Table V below:

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

1
Attractiveness

0.817

FD

0.791

Freedom from Corruption

0.789

Property Rights

0.780

COR

-0.755

Economic Freedom Score

0.723

Business Freedom

0.678

GDP

0.670

FDI Inflows in millions

0.561

Component
2
3

4

0.608

Labor Freedom
Public Debt as percent of
GDP
C
AG
Unemployment
Monetary Freedom

0.808

Inflation in percentage

-0.725

Investment Freedom

0.699

Financial Freedom

0.553

0.657

Trade Freedom

0.606

Tariff Rate

-0.504

BS
Basel AML Score
EG
Government Expenditure

0.916

Government Spending

-0.913

Tax Burden as percent of

0.665

GDP
Swift
Income Tax Rate

0.903

Fiscal Freedom

-0.888

Corporate Tax Rate

0.774

Table IV. Rotated component matrix

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

Variance
Component

% of variance

Cumulative %

1

22.949

22.949

2

14.859

37.809

3

12.228

50.037

4

9.494

59.530

Table V. Variance explained by components
The first component (1) comprised the set of variables related to a country's economic
feasibility and explained 22.9 % of the overall variance. The second component (2) explained
14.85% of the overall variance and included proxies for financial liberty. The expenditure
incurred and income source to support it by a country's government are grouped in the third
component (3) and explain 12.22% of the overall variance. Finally, the fourth component (4)
relates to a country's tax regime and explains 9.49% of the overall variance. Consequently, a
country's appeal for money laundering can be explained as a follows:
𝑀𝑜𝑛𝑒𝑦 𝐿𝑎𝑢𝑛𝑑𝑒𝑟𝑖𝑛𝑔 𝐴𝑝𝑝𝑒𝑎𝑙 = 𝑓(𝐸, 𝐹, 𝐺, 𝑇)
Where,
E = Economic feasibility of a country to invest in
F = Financial liberty available to people to transact and do business
G = Government spending
T = Tax regime
Each principal component's variance is used as a fraction of the model's overall variance as a
weight to obtain the index. The index is normalized using the Min-Max criterion and is
expressed as follows:

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

𝐺𝑙𝑜𝑏𝑎𝑙 𝑀𝑜𝑛𝑒𝑦 𝐿𝑎𝑢𝑛𝑑𝑒𝑟𝑖𝑛𝑔 𝐴𝑝𝑝𝑒𝑎𝑙 𝐼𝑛𝑑𝑒𝑥
4

= ∑

(Sij ∗ wj ) = (𝑆1𝑖 ∗ 𝑤1 ) + (𝑆2𝑖 ∗ 𝑤2 ) + (𝑆3𝑖 ∗ 𝑤3 ) + (𝑆4𝑖 ∗ 𝑤4 )

𝑗=1

Where the subscript i indicates the country (150 in total), j is the component and wj is the
proportion of variance explained by each component, that is, 𝑤1 = 22.9%, 𝑤2 = 14.85%,
𝑤3 = 12.22%, and w4 = 9.49%, respectively. Sij is the relevant value from the PCA for
each component. According to the index, for 2014, the most appealing countries for money
laundering were Denmark, Sweden, and the United States, and the least appealing countries
were Sudan, Iran, and Iraq (Ranking results are in Table A of the appendix).
4. Conclusion
The use of PCA instead of a more theoretical approach is validated by comparing rankings with
Khan et al. (2018), and the results were found to be similar. Additionally, Denmark's being
found an appealing destination for money laundering in 2014 aligns with a Danish bank's
involvement in one of the largest such scandals around the same time (Bjerregaard and
Kirchmaier, 2019). The robustness of results can be considered suitable for future work along
with the incorporation of more variables. At present, the paper proposes an approach where a
complex phenomenon of money laundering appeal can be placed into a single composite
indicator, allowing for ranking of geographical locations. Such an index might not only inform
national strategies to prevent money laundering but provides an opportunity to use a similar
approach to develop more localized hot spot maps that could move analysis at the sub-national
level. Finally, concerns about whether a low Basel AML Risk score (Manning et al., 2020)
results in lower chances of money laundering are addressed through this paper.

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

Acknowledgment
The authors are grateful to Basel Institute on Governance, The Heritage Foundation, and Wahaj
Ahmed Khan for their assistance in providing the data.
Funding
This work was supported by the Australian Government Research Training Program.
References
Barone, R., D. Delle Side, D. Masciandaro, 2018. Drug trafficking, money laundering and the
business cycle: Does secular stagnation include crime? Metroeconomica 69, 409-426.
Basel Institute on Governance, 2014. Basel AML Index 2014 Report. (Basel Institute on
Governance).
Bjerregaard, E., T. Kirchmaier, 2019. The Danske Bank Money Laundering Scandal: A Case
Study. Available at SSRN 3446636.
Collin, M., 2019. Illicit Financial Flows: Concepts, Measurement, and Evidence. World Bank
Research Observer 35, 44-86.
Degryse, H., A. Karas, K. Schoors, 2019. Relationship lending during a trust crisis on the
interbank market: A friend in need is a friend indeed. Economics Letters 182, 1-4.
Ferwerda, J., E.R. Kleemans, 2019. Estimating Money Laundering Risks: An Application to
Business Sectors in the Netherlands. European Journal on Criminal Policy and Research 25,
45-62.
Johannesen, N., T. Tørsløv, L. Wier, 2016. Are less developed countries more exposed to
multinational tax avoidance? Method and evidence from micro-data. The World Bank
Economic Review.
Kaiser, H.F., 1960. The application of electronic computers to factor analysis. Educational and
psychological measurement 20, 141-151.
Khan, W.A., S.T. Jawaid, I. Arif, 2018. Where does a nation's wealth go? Evidence from a
third world country. Journal of Money Laundering Control 21, 426-476.
Kim, A.B., K.R. Holmes, 2016. 2014 Index of economic freedom. The Heritage Foundation in
Partnership with Wall Street Journal.
Manning, M., G.T. Wong, N. Jevtovic, 2020. Investigating the relationships between FATF
recommendation compliance, regulatory affiliations and the Basel Anti‑Money Laundering
Index. Security Journal.
The World Bank, 2020. World Bank Open Data.
Unger, B., M. Siegel, J. Ferwerda, W. de Kruijf, M. Busuioic, K. Wokke, G. Rawlings, 2006.
The amounts and the effects of money laundering. Report for the Ministry of Finance 16.
Walker, J., 1999. How Big is Global Money Laundering? Journal of Money Laundering
Control 3, 25-37.
Walker, J., B. Unger, 2009. Measuring global money laundering. Review of law and economics
5, 821-853.
Watkins, M.W., 2006. Determining parallel analysis criteria. Journal of modern applied
statistical methods 5, 344-346.

© Emerald Publishing Limited.
This AAM is provided for your own personal use only.
It may not be used for resale, reprinting, systematic distribution, emailing, or for any other commercial purpose without the permission of the publisher.

