The relationship between mental health and finance (and/or socioeconomic factors) has been explored with evidence for bidirectional relationship between both. Previous studies relied on subjective surveys, which were conducted after mental symptoms had occurred and thus prone to recall bias. Moreover, varying subjective diagnostic scales across studies make standard comparisons challenging. This research focuses on assessing the mental health of individuals with bipolar disorder using objective financial data, statistical analysis, and privacy-preserving machine learning. It also includes data instrumentation, collection, and participant informed feature identification for efficient modelling.