We investigate how financial intermediation by rural banks (BPRs), commercial banks, and fintech P2P lending influences per-capita household consumption, MSME production, and Gross Regional Domestic Product (GRDP) across Indonesia’s 34 provinces during 2022–2024 (N = 102). Using Partial Least Squares Structural Equation Modeling (PLS-SEM) via SmartPLS with bootstrapped inference, we estimate a system of simultaneous equations that captures both direct and indirect (mediated) pathways from each intermediation channel to regional output. Results show that BPR intermediation (X₁) and commercial-bank intermediation (X₂) significantly and positively affect MSME production (H4 and H5 supported; p < 0.05), whereas fintech P2P lending (X₃) does not (H6 rejected; p = 0.161). Fintech P2P lending positively influences per-capita consumption (H3 supported at the 90% level; p ≈ 0.078), while BPR and commercial-bank intermediation have no significant effect (H1 and H2 rejected). MSME production strongly enhances GRDP (H7 supported; p < 0.01), and consumption is positively associated with GRDP at the 90% level (H8 supported; p ≈ 0.085). Direct paths from intermediation to GRDP (H9–H11) are not significant, supporting a mediated mechanism: BPRs and commercial banks primarily operate through MSME production, while fintech channels influence GRDP via consumption. Model fit is acceptable (SRMR = 0.091), although some covariance-fit indices are weak (discussed in the manuscript), so inference relies on bootstrapped path estimates and mediation diagnostics. Policy implications suggest regulators should reinforce BPR and commercial-bank support for MSME financing, while guiding fintech to expand safe consumer access that enhances household purchasing power without introducing systemic risks.