The Challenge
Securitization is a credit transfer process involving Banks, Rating Agencies (e.g. Standard & Poor’s, Moody’s, …), SPVs that convey the financial assets sold by banks and Investors who purchase the debt generated by these non-performing loans (NPLs) through portfolios of non-performing loans.
In this context, an important Italian SPV has requested to modernize its analysis on the portfolio of receivables sold in order to identify new portfolio and predictive analyses to be provided to its Investor Clients, useful both for the composition of the portfolio that investors will purchase and for the forecast of collection of non-performing loans from original debtors.
The Solution
The proposed Machine Learning service was defined starting from a Pilot Project that allowed the identification, on a subset of Customers, of different machine learning models, both Unsupervised and Supervised and Predictive.
In particular, the Models have allowed the identification of a Classification of the Portfolio with different degrees of “certainty of collection” of the portfolio, the behavioral profiling of the components of the portfolio between similar portfolios useful for making decisions during the establishment of a portfolio and finally the Monitoring of the credit to be recovered with temporal and quantitative prediction.
The Benefits
The benefit of such a Machine Learning solution lies in the enrichment of Non-Performing Credit (NPL) analysis as it constitutes the possibility of combining the results of Machine Learning algorithms with already present Business Intelligence analysis and results of classic statistical algorithms already present.
Humanativa’s Role
Implementation of the Machine Learning Pilot service, from the creation of Machine Learning Models to Advanced Analytics.