Profit-driven churn prediction for the mutual fund industry: A multisegment approach

This paper proposes a novel approach for profit-based classification for churn prediction in the mutual fund industry. The maximum profit measure is redefined to address multiple segments that differ strongly in the average customer lifetime values (CLVs). The proposed multithreshold framework for churn prediction aims to maximize the profit of retention campaigns in binary classification settings. The multithreshold framework is empirically tested on data from a Chilean mutual fund company with varying and heterogeneous individual CLVs. Our results demonstrate the virtues of the proposed approach in achieving the best profit when compared to other metrics. Although presented in the context of investment companies, our framework can be implemented in any churn prediction task, representing an important contribution for decision-making in business analytics. (C) 2020 Elsevier Ltd. All rights reserved.

Automated summary

This study introduces a profit-based classification approach for churn prediction in the mutual fund industry, redefining the maximum profit measure to address varying customer lifetime values. Empirical testing demonstrates the method's superiority in maximizing profit, offering an important contribution to decision-making in business analytics.