Predictive Savings Service

This paper highlights the persistent concern of low personal and household savings across various demographics globally. The factors leading to this deficit include stagnant or low wages, high levels of personal debt, cost of living increases, economic instability, and lack of financial literacy. The paper proposes a machine learning driven solution to address prevalent lack of financial education and indebtedness by implementing a two-strategy approach. The first strategy involves analyzing past cash flow and determining earning power, liabilities, and outflows. The second strategy entails implementing a segmentation engine that takes into account various individual characteristics to determine the future likelihood of certain cash-intensive events occurring. By combining both data analytics streams, the proposed solution provides a predictive savings service that can anticipate upcoming events with a degree of likelihood and allocate in advance the finances required to meet those future obligations. The report outlines various algorithms that can be used for cash flow forecasting and recommends the use of Open Banking as a reliable source of customer financial data. The proposed solution is expected to provide short and long-term debt reduction and reduce personal financial risk of bankruptcy. Continuous optimization and feedback will be essential in keeping the cash flow forecasting models up to date.