<![CDATA[This study assesses the systemic stability of the fintech lendingecosystem by linking three analytical pillars: (i) a multilayer network oflinkages (platform–investor–custodian bank–payment rails–dataproviders), (ii) liquidity risk through the Liquidity-at-Risk (LaR)framework, and (iii) a flow-based macroprudential policy evaluation (e.g.,dynamic cash buffers and circuit breakers). We construct a network mapof bipartite investor–platform exposures, platform–custodian banklinkages, and dependencies on payment rails, then calculateconcentration and centrality metrics, as well as investor overlap acrossplatforms. Next, we estimate daily LaR (14-day horizon, α=99%) fromcash-in/out, settlement, and disbursement flows, and develop thePlatform Run Index (PRI)—a nowcasting indicator that combinesredemption pressure, settlement queue length, pricing spread deviation,and operational stress. Contagion dynamics are measured by losspropagation from platforms to banks/rails based on an exposure matrix,while policy effectiveness is identified using stepwise Difference-in-Differences and event studies on staggered rollouts of liquidity rules. Themain results show that funding concentration (high HHI) and reliance ona few banks/rails increase loss amplification and potential spillover tobanks. LaR peaks with a surge in cash-outs and settlement queues,marking a run-prone zone even without a significant increase in defaults.PRI exceeding the p90 threshold predicts a spike in withdrawals thefollowing day, making it a suitable trigger for adaptive policy. Agentbasedsimulations show that funding shocks and operational outagesincrease run probabilities and lengthen queues, and—when combined—result in material loss amplification. Causal evaluations show that thecombination of dynamic cash buffers and flow-based circuit breakerssignificantly lowers PRI, reduces LaR violations, shortens queues, andmitigates early contagion. The implication is that systemic resilience infintech lending requires diversified escrow and rail systems, real-timePRI-based monitoring, multilayer stress testing (LaR + ABM) withperiodic backtesting, and operational resilience standards(SLA/latency/failover). These findings support the design of dynamic,flow-based macroprudential policies to balance innovation, inclusion,and stability.]]>
