Two-Layer Liquidity Framework

• The two-layer liquidity framework is a model that defines liquidity via two distinct, interacting layers, each with unique constraints and metrics used in various financial contexts.
• It employs separate event scales or on-chain/off-chain mechanisms to capture both fine-grained market microstructure and coarse systemic liquidity dynamics.
• The framework’s mathematical coupling, equilibrium properties, and calibration methods provide actionable insights for market monitoring, regulatory stress tests, and robust network design.

@@@@1@@@@ two-layer liquidity framework refers to any formal system that models or quantifies liquidity or funding availability via two structurally distinct but interacting "layers," which may represent pricing regimes, balance sheet resources, network topologies, or event scales. This approach arises naturally across high-frequency market microstructure analysis, payment/credit network architecture, systemic risk modeling, and regulatory stress frameworks. Each application elaborates both the mathematical coupling between the two layers and the economic rationale for the separation.

1. Structural Paradigms of Two-Layer Liquidity

Across financial contexts, the two-layer approach most commonly manifests as either (a) a division between micro and macro event scales (market microstructure), (b) a split between on-chain (base) and off-chain (credit) mechanisms (payment systems), or (c) the distinction between immediate execution values and terminal valuations (contagion/regulatory models). Abstractly, "layer" denotes any capital, network, or process regime with its own feasible set and liquidity metric, whose interaction with the other layer is explicitly modeled.

For example, in the event-based market liquidity framework of Golub et al., the two layers correspond to directional-change events detected at a fine ($\delta_1$) and coarse ($\delta_2$) threshold, leading to a state space structured as a two-level intrinsic network (Golub et al., 2014). In payment networks, the layers refer to on-chain escrow (Layer 1) and off-chain bilateral debt notes or payment channels (Layer 2), each tracked by its own liquidity constraints and backup guarantees (Ramseyer et al., 2019). In regulatory contagion, the two layers are differentiated by pricing conventions: realized average (VWAP) liquidity for asset sales versus mark-to-market pricing for remaining holdings (Banerjee et al., 2019). A similar layered approach appears in fire-sale and central-bank liquidity models, where asset tranches are segregated by marginal fire-sale recoveries and central bank collateral haircuts, parametrized continuously (Bindseil et al., 2020).

2. Mathematical Layer Construction and Interaction Rules

The mathematical specification of the two layers is context-dependent but always involves explicitly distinct constraint sets, state variables, and transfer mappings.

Market Microstructure (Intrinsic Network):

• Define directional-change and overshoot events at two scales $\delta_1<\delta_2$.
• At each event time, assign a state $(b_1,b_2)$ (up/down at each scale).
• The two-layer intrinsic network has $2^2=4$ states, with allowed transitions determined by which threshold fires.
• The layer hierarchy is formalized: contracting over the fine scale yields a two-state (coarse) process, while the full 4-state system encodes microstructure (Golub et al., 2014).

Payment/Credit Networks:

• Layer 1: On-chain base, enforcing escrowed collateral $C$ for aggregate node budgets $K_i$ via smart contracts.
• Layer 2: Directed off-chain credit graph with per-channel limits $c_{uv}$, constrained such that node $i$’s net dues $\sum_{j\ne i}\ell_{ji}\le K_i$.
• Analytically, the presence of on-chain caps collapses the off-chain network to a star geometry; all routing/liquidity questions become functions of node budgets $K_i$ and the total escrow (Ramseyer et al., 2019).

Systemic Pricing and Fire Sales:

• Layer 1: Terminal mark-to-market price $q=F(\Gamma)$ as the marginal price from an inverse demand curve for each asset.
• Layer 2: Volume-weighted average price $\bar F(\Gamma)$ received for actually liquidated blocks.
• Clearing is enforced via coupled fixed-point equations: $(q,\bar q)=\Phi(q,\bar q)$, with price feedback from liquidation size (Banerjee et al., 2019).
• In fire-sale models, continuous ranking of assets by liquidity yields two per-unit liquidity functions: asset $x\in[0,1]$ has fire-sale liquidity $\ell(x)=1-x^{\Theta}$ and collateral liquidity $c(x)=1-x^{\delta}$. The firm optimally allocates assets between layers depending on cost parameters (Bindseil et al., 2020).

3. Layer-Specific Liquidity Metrics

Each layer admits its own relevant liquidity metric, calibrated either by state transitions, flow feasibility, or price functions.

Intrinsic Network (Market Microstructure):

• Layer $i$ is modeled as a Markov chain (of 2 or 4 states). The information-theoretic liquidity score $\mathcal L_i$ is given by

$$\mathcal L = 1 - \Phi\left(\frac{\Gamma^{[0,T]}-KH^{(1)}}{\sqrt{KH^{(2)}}}\right)$$

where $\Gamma^{[0,T]}$ is the observed cumulative Markov surprisal, $H^{(1)}$ and $H^{(2)}$ are entropy and variance, and $\Phi$ the standard normal CDF. Layer 2 (fine) is sensitive to microstructure illiquidity, Layer 1 (coarse) to systemic stress (Golub et al., 2014).

Credit Networks:

• Define

$$L(K) = \Pr_{\text{stationary}}[\text{payment } x\to y \text{ succeeds off-chain}]$$

where probability is over all state/configuration and payer-payee pairs. In the collapsed "star" model, pairwise liquidity is $L_{ij}=1-1/(K_i+K_j+1)$ and overall liquidity is $L=1-1/(2K+1)$ for homogeneous budgets (Ramseyer et al., 2019).

Fire Sales and Funding Models:

• Aggregate liquidity available for deposit withdrawals is $Y(z)=L(z)+C(z)$ where $z$ is the split between fire-sale liquidation and collateralization, directly summing the two per-unit liquidity integrals. Funding stability is evaluated by comparing $Y(z)$ to deposit size, with (SNNR) equilibrium defined as the region where no rational run occurs (Bindseil et al., 2020).
• In multi-asset fire-sale contagion, the realized liquidation cost (difference between book and actual recovery) and market cap sensitivity to stress or policy are calculable in closed form per layer (Banerjee et al., 2019).

4. Existence, Uniqueness, and Hierarchical Properties

Layered frameworks raise specific mathematical questions about the coupled system: existence/uniqueness of equilibrium, tractability of systemic risk, and the possibility of state contraction.

• In the intrinsic network, a strict hierarchy exists; contracting over one layer yields a coarser process, but microstructure signals are retained only at the finer layer (Golub et al., 2014).
• For constrained credit networks with per-node borrowing caps, classical combinatorial properties (route-equivalence, uniform stationarity) are preserved under the two-layer construction, and all liquidity calculations reduce analytically to the simpler star network (Ramseyer et al., 2019).
• In two-tier pricing contagion models, well-posedness of the clearing fixed point follows from monotonicity and continuity; uniqueness is guaranteed under strict increase of the aggregate value function in each coordinate (Banerjee et al., 2019).
• In general multi-currency (multi-layer/asset) contagion, existence of equilibrium clearing portfolios is given under Tarski's and Brouwer's theorems; uniqueness demands further conditions, particularly in presence of price impact and endogenous feedback (Feinstein, 2017).

5. Calibration, Early Warning, and Policy Applications

Two-layer frameworks are applied to quantify real-time market liquidity, engineer robust payment architectures, and formulate policy tools for systemic stability.

Market Microstructure/FX:

• Calibration typically uses high-frequency tick data, selecting thresholds so that sufficient intrinsic events occur per observation period.
• Empirical case studies (e.g., USD/JPY and EUR/CHF shocks) show that the fine-layer liquidity score often anticipates episodes of systemic stress; persistent drops in $\mathcal L_2$ and, subsequently, $\mathcal L_1$, provide early warning for risk containment (e.g., margining, algorithm throttling, circuit-breakers) (Golub et al., 2014).

Credit/Blockchain Payments:

• By enforcing aggregate collateral constraints via smart contracts on-chain (Layer 1), off-chain payment networks (Layer 2) can maintain optimal tradeoffs between throughput (liquidity) and required capital. Explicit formulae allow direct estimation of network-level liquidity as a function of agent budgets, with practical implications for design of decentralized networks and policy for capital allocation (Ramseyer et al., 2019).

Systemic Risk, Funding, Regulation:

• Two-layer pricing models (VWAP vs MTMP) allow regulators and central banks to distinguish impact of incremental liquidations from true market value erosion, and to compare the benefits of direct bailouts (capital injections) vs indirect asset support. Quantitative criteria for optimal intervention are derived analytically (Banerjee et al., 2019).
• Continuous-layer fire-sale models elucidate how blending fire-sale and lender-of-last-resort (collateral) liquidity determines safe deposit/funding structure, exposure to runs, and minimum cost for maturity transformation (Bindseil et al., 2020).

6. Extensions: Multi-Asset, Multi-Layered, and Nonlinear Regimes

While the two-layer paradigm is foundational, extensions consider networks with more than two currencies/assets or continuous spectra of liquidity, and generalize to arbitrary network topologies or heterogeneous agent constraints.

• Multi-asset contagion models permit institutions to reallocate across layers (assets/currencies), adjust trades to utility maximization, and respond dynamically to market feedback and price impact; equilibrium analysis in this context generally requires iterative algorithms or tâtonnement processes for selection of economically relevant fixed points (Feinstein, 2017).
• Continuous-layer fire-sale models recast the two sources of liquidity (fire-sale and collateral) as integrals over asset quality, admitting comparative-statics and policy insight as model parameters (e.g., haircuts vs market liquidity) shift (Bindseil et al., 2020).
• A plausible implication is that the two-layer approach is a flexible blueprint that can accommodate arbitrarily complex real-world liquidity interactions, so long as the layer coupling admits tractable analysis of equilibrium, efficiency, and systemic fragility.

---

Key sources:

• "Multi-scale Representation of High Frequency Market Liquidity" (Golub et al., 2014)
• "Liquidity in Credit Networks with Constrained Agents" (Ramseyer et al., 2019)
• "Price mediated contagion through capital ratio requirements with VWAP liquidation prices" (Banerjee et al., 2019)
• "Fire Sales, the LOLR and Bank Runs with Continuous Asset Liquidity" (Bindseil et al., 2020)
• "Obligations with Physical Delivery in a Multi-Layered Financial Network" (Feinstein, 2017)