The rapid expansion of direct-to-consumer (DTC) markets has transformed brand community participation and user-generated content (UGC) into important mechanisms influencing customer-based brand equity (CBBE), repurchase intention, and electronic word-of-mouth (eWOM). However, conventional analytical approaches frequently examine these relationships through linear structural models, potentially overlooking nonlinear interactions, heterogeneous consumer responses, and complex mediation pathways. This study proposes a novel Brand Community and User-Generated Content Equity Prediction Algorithm (BCUGC-EPA), an explainable hybrid analytical framework designed to quantify and predict the interconnected effects of brand community engagement and UGC characteristics on brand equity formation and post-purchase behavioural intentions. The proposed algorithm integrates Partial Least Squares Structural Equation Modelling (PLS-SEM), Extreme Gradient Boosting (XGBoost), and Shapley Additive Explanations (SHAP) within a unified predictive architecture. PLS-SEM estimates direct, indirect, and mediating relationships among community participation, UGC credibility, perceived authenticity, customer-based brand equity, repurchase intention, and eWOM, while XGBoost captures nonlinear dependencies and higher-order interactions. SHAP provides interpretable estimates of variable contributions to consumer behaviour predictions. The study adopts a quantitative research design using structured questionnaires administered to customers of DTC brands, supplemented, where available and consented to, by anonymised community engagement indicators and UGC interaction records. Model evaluation will compare BCUGC-EPA against conventional PLS-SEM prediction, Multiple Linear Regression, Random Forest, Support Vector Regression, and standalone XGBoost under identical data partitions and validation procedures. Predictive performance will be evaluated using the coefficient of determination (R²), root mean squared error (RMSE), mean absolute error (MAE), and cross-validated predictive stability. Statistical significance will be examined through bootstrapped path coefficients, mediation effects, and confidence intervals. Comparative results will be presented through predictive performance bar charts, observed-versus-predicted scatterplots, SHAP importance diagrams, mediation path diagrams, and interaction response surfaces. The central hypothesis is that integrating theory-driven structural relationships with nonlinear machine learning will improve predictive accuracy while preserving meaningful interpretation of how community participation and UGC influence consumer outcomes. The study is designed to determine whether customer-based brand equity mediates the relationships between brand community participation, UGC characteristics, repurchase intention, and eWOM, and whether these pathways differ across consumer engagement levels. Its anticipated contribution is an empirically testable framework for explaining and predicting customer retention, brand advocacy, and equity development in DTC markets, offering actionable evidence for optimising community-driven marketing strategies and consumer relationship management.