Advanced Risk Modelling

Using its in–depth business knowledge and expertise in data mining and predictive statistical modeling, Mercury Processing Services International operates as a technical service provider and consultant through each step of the credit risk lifecycle – from strategic planning, fraud prevention, to acquisition and portfolio management, and debt collection – in order to provide you with a distinct competitive advantage.

There are several types of statistical models we are developing, which help you manage and optimize decisions across your portfolio and identify measurable benefits.


Credit Risk Modeling Solution

In the card business, credit risk refers to the probability that a customer will stop paying the invoices, while fraud risk represents the probability that a card transaction was not made in compliance with the agreement between a customer and a bank, thus causing loss to the bank. Customers and transactions (i.e. facilities) that meet these criteria are classified as bad.

In order to predict which facility will be bad, we are developing statistical models that are basically a set of characteristics related to the customer demographics or customer transaction volume for the card.

There are several Credit Risk Modeling Solutions we are developing, using data stored on our platforms and data provided by our partners:

Probability of Default (PD) models (Application & Behavioral Scorecards)

  • consistent & unbiased credit processing
  • improved portfolio risk selection 
  • risk based portfolio monitoring and management with impact on decreased provisions and enhanced P&L
  • alignment of pricing and limits with real underlying risks
  • risk based credit policies, cross-selling and up-selling strategies

The Loss Given Default (LGD) models estimate the amount to be lost in case a client does not meet contractual obligations

  • support regulatory compliance
  • reducing provisions and consequentialy reducing write-offs
  • reducing economic capital requirement due to lower risk-weighted assets through the adoption of an internal ratings based approach 

Exposure at Default (EAD) models - estimate the exposure in case a client does not meet contractual obligations

  • supporting regulatory compliance
  • reducing provisions and consequentialy reducing write-offs


Value-Added Solutions

Fraud Scorecards - battle fraud effectively by detecting subtle patterns of unusual activity in data using predictive models.

Collection Scorecards - ranking client creditworthiness on occurrence of default

Mercury Processing Services International's risk modeling solutions are based on proven technology created by SAS, a leader in business analytics software and services currently used at 93 of the top 100 companies on the 2014 Fortune Global 500® list.