Impact of digitalization and the COVID-19 pandemic on the AML scenario: Data mining analysis for good governance

Abstract

. The article deals with the impact of digitalization and COVID-19 on the choice of AML scenarios for reforming the system of tactical and strategic monitoring of transactions carried out by economic entities based on providing good governance. The study period is 2011-2020; the objects of the study are 140 countries. Calculations are performed using Data-Mining methods, such as AML scenarios based on the classification tree method (one-dimensional CART branching method) and clustering of countries according to relevant AML scenarios based on agglomerative methods. There are three stages of research. The first builds a comprehensive system of indicators which involves financial inclusion indicators of the population, the ranking of countries on the Basel AML Index, and effectiveness of the AML policy implementation at the country level. The se cond stage considers countries’ clustering according to the AML scenarios and formalizes the portraits of countries’ clusters. The third stage examines the impact of digitalization and COVID-19 on the choice of AML scenarios. According to the empirical results, adaptability to external factors are formalized in the possible scenarios as a result of such effects. Moreover, the countries’ clustering prove s that the money laundering risks relevant to the country lower and the implementation of the AML measures by the state grows more effective with higher financial inclusion for the population in the country. The study results can be helpful for authorized bodies in providing good governance while conducting financial monitoring and analysis of information on transactions carried out by economic entities. The study considers three stages of its implementation. The first stage involves the construction of a comprehensive set of indicators for the further calculations, included determining the priority financial inclusion indicators with the PCA-method. During the study the general set of financial inclusion indicators are analyzed. They characterize the following forms of financial services access to population: opening (availability) an account, in particular in a financial institution; the amount of savings; the amount of outstanding housing loan; loan to start a business - in terms of such criteria as income level, employment, living rural. The second stage of the study considers clustering the countries according to the AML scenarios based on analysis of financial inclusion indicators, indicators of AML policy effectiveness and directly Basel AML Index Scores within countries; formalizing the portraits of countries’ clusters taking into account both simultaneous and separate consideration of relevant financial inclusion indicators, ranking of countries on the Basel AML Index Score, and indicators of AML policy effectiveness at the country level. While clustering the countries empirical research was implemented on the basis of agglomerative methods of minimum variance. Within that method the countries were divided into four groups (clusters) under the conditions of their similar characteristics within three groups of indicators that characterize: financial inclusion, efficiency of AML policy and Basel AML Index. The number of clusters was developed with variance analysis. The affiliation of countries to each cluster in the study was determined by k-means method (within using the agglomerative methods). conducted clustering (12 classified st cluster, 17 - 2 nd 12 to 3 rd 15 - to 4 th cluster) shows that the higher the for the effective the TF