Dynamic Funding Gap Risk Measurement
Static gap schedules once sufficed. Then markets turned volatile, liabilities repriced faster than assets, and balance-sheet mismatches became visible only after the damage was done. We built a different approach.
Continuous simulation
Track your funding position as it evolves, not as it was last reported.
Explainable models
Every output traces back to documented assumptions and transparent methodology.
Near real-time alerts
Surface deterioration before it crystallizes into a liquidity event.
The DFGRM methodology in five stages
Exposure identification
We begin by cataloguing all contractual cash flows, their timing, and their triggers. This includes drawn facilities, undrawn commitments, derivatives margin calls, and off-balance-sheet obligations. The goal is completeness: a gap that is not mapped cannot be measured, and a commitment that is omitted will eventually surprise you.
Cash flow mapping
Normalize disparate data sources into a unified temporal structure, reconciling timing conventions and behavioral assumptions so that assets and liabilities speak the same language.
Treasury systems, loan servicing platforms, and liability ledgers rarely share a common format. We build normalization layers that reconcile day-count conventions, currency mismatches, and counterparty hierarchies without forcing a rip-and-replace of existing infrastructure.
Stress scenario construction
Design distributions of market and behavioral shocks rather than single-point scenarios, capturing the fat-tailed nature of funding stress and the correlation between liability outflows.
Gap quantification
Run thousands of simulation paths to produce a dynamic funding gap distribution, revealing not just a single number but the shape of your vulnerability across time horizons.
Dynamic rebalancing signals
Alerts are calibrated to the cost of false signals versus the cost of missed warnings. Each threshold is documented, defensible to auditors, and linked to a specific governance action: review at amber, hedge at red, restructure at critical.
Mathematical foundations
Stochastic cash flows
Cash flows are modeled as stochastic processes rather than deterministic schedules. Each liability stream incorporates a behavioral component: deposit outflow sensitivity to rate differentials, wholesale rollover probability as a function of counterparty confidence, and contingent drawdown intensity linked to borrower credit stress. The result is a set of correlated random variables whose joint distribution captures the interaction between funding sources under stress.
Correlation structures
Funding stress rarely arrives through a single channel. Our framework models the correlation between deposit flight, wholesale rollover failure, and contingent commitment drawdowns using copula methods calibrated to historical episodes of systemic funding pressure. This preserves the tail dependence that linear correlation matrices miss: the tendency for multiple outflows to spike simultaneously when confidence collapses.
Dynamic rebalancing
The gap is not a static scalar but a time-dependent surface. We apply optimal control techniques to identify rebalancing triggers that minimize expected liquidity cost subject to a constraint on tail-risk probability. The solution is a set of dynamic thresholds that adapt as the balance sheet evolves, market conditions shift, and new information arrives.
Model validation
Every engine is validated against historical out-of-sample episodes where funding stress actually occurred. We require that the model would have flagged deterioration before the event, without generating excessive false positives during tranquil periods. Validation metrics include the Brier score for probability calibration, the expected shortfall backtest, and the binary accuracy of alert thresholds at multiple horizons.
How DFGRM compares
Fleplionciuvu
Continuous simulation engine
Static gap tools
Quarterly schedule-based
Temporal resolution
Measures gap at a single reporting date versus tracking how it evolves continuously between dates, capturing rollover cliffs and repricing acceleration as they occur.
Behavioral modeling
Incorporates dynamic outflow sensitivity, deposit beta shifts, and contingent drawdown probability rather than assuming contractual cash flows will arrive as scheduled.
Scenario depth
Generates distributions of correlated stress paths instead of running one or two deterministic shocks that may not reflect the joint nature of funding crises.
Alert calibration
Thresholds are statistically calibrated to balance false positives against missed warnings, with full documentation for audit and governance review.
Guidelines for liquidity resilience
Liquidity planning benchmarks
-
Short-term gap coverage 100%
-
Stable funding ratio target 85%
-
Liquid asset buffer 30%
-
Wholesale funding limit 25%
-
Stress scenario frequency 12/yr
-
Deposit beta review cycle 90d