FinTech Research Highlights How Alert Models Can Protect Older Customers from Fraud

18 January 2021
FinTech Research Highlights How Alert Models Can Protect Older Customers from Fraud

Posted: 19/12/18

Fraud is responsible for huge losses at financial institutions every year.  In general, older clients are wealthier and also more susceptible to financial fraud and exploitation. This makes such individuals natural targets for financial fraud.  A new research paper titled “Can alert models for fraud protect the elderly clients of a financial institution?” addresses this problem.  The paper was recently featured in the European Journal of Finance and is co-authored by Cal B Muckley, researcher in the ADAPT led FinTech Fusion research programme.

Using account-level transaction data at a major financial institution, the research focuses on over 5 million accounts of clients aged 70 years and older.  Alert models for external fraud comprise a first line of defence for a financial institution to protect its elderly clients, minimise any risk of financial regulatory penalties and safeguard the reputation of the financial institution.  The development of a fit-for-purpose alert model to protect a financial institution’s elderly clients is not straightforward.  In this paper, the researchers provide the first such model.

Speaking about the research, Cal B Muckley said: “Using data provided by a proprietary transaction monitoring system at a major financial institution, we can show that adopting a machine learning approach to design an alert model for the protection of elderly clients has considerable promise.”

The research shows that using sophisticated learning techniques together with rich datasets can lead to a marked increase in efficiency and expediency over industry standard rule-based models. By following this model a financial institution can better protect its elderly clients, reduce operational losses due to fraud and exploitation as well as facilitate a saving in time and resources during the investigation process.

Research paper.

Share this article: