Accelerating the Distribution of Financial Products Through Classification and Regression Techniques

Financial products mostly consists in instruments used by households to prepare for retirement and/or to transfer wealth across generations. However, their usage remains low. Current technical solutions aimed at boosting financial products’ consumption mostly revolve around robo-advisors automating portfolio management tasks. Nonetheless, there is no tool to automate the distribution of those products, a gap this study aims to bridge. To do so, a private data-set from a French Fintech is leveraged. It describes at a macro level the structure of 1500+ households and their wealth. This information is fed to standard classification algorithms and regression techniques to predict whether or not households are likely to subscribe to a life insurance or a retirement plan or a real estate program over the forthcoming year and to forecast the associated level of investment. Calibrations show that households’ subscription behavior over the next 12 months towards core investments products can be predicted with a high level of performance ( \(A.U.C >90\%\) ). The information was yet inappropriate to predict the level of investments on those products ( \(R^2<30-40\%\) ). Standard classification techniques could thus be used by financial advisors to accelerate client discovery on their existing portfolio. This should thereby result in productivity gains for those professionals and improve the distribution of financial products.