SETTLEMENT DESK NOTE // REF: DN-2026-04 // CLASSIFICATION: INTERNAL-DISTRIBUTION // DATE: 18 APR 2026

TARGET2 Weekend Settlement Frictions: Empirical Telemetry

Observation: Telemetry captured during weekend clearing cycles reveals a operational gap between continuous, 24/7 tokenized asset settlement channels and traditional TARGET2 liquidity accounts. The bottleneck occurs at the retail sweep interface during high-volume periods.

SCT Inst / RT1 Telemetry: Peak latency reached 412ms on Saturday at 14:00 CET, compared to a baseline average of 15ms on weekdays. TARGET2 liquidity exposure gap measured at 0.14 bps.

Analysis: Direct funding via TARGET2 remains closed over the weekend. Automated sweep algorithms designed for Tier-1 LCR buffering faced operational dry-spells of up to 45 minutes before secondary clearing routes activated. This suggests that wholesale DvP settlement utilizing off-ledger cash legs introduces an asynchronous queuing risk that cannot be resolved solely via DLT performance upgrades.

// UNRECONCILED DESK SCRIBBLER NOTES (PRE-RECONCILIATION):

"There is a severe telemetry data gap on Saturday morning between 02:00 and 04:00 CET. Direct participant node maintenance at [redacted participant] went offline, leaving our telemetry logs for this interval unverified. Desk remains divided on whether the 412ms spike is a structural queuing bottleneck or simply localized ISP routing failure. Under stress conditions, we cannot confirm stable convergence parameters."

- NOTE (J.L.): Need to check if EBA Clearing actually processed the 03:15 batch or if it rolled over. The private feeds conflict directly with the official Eurosystem operational timeline. Still unreconciled.

Methodology & Data Limitations: This telemetry is limited to public EBA RT1 batch settlement flows and does not capture bespoke bilateral liquidity sweeps or private OTC funding windows. Latency spikes may be influenced by systemic ISP routing issues or localized node failures. The sample reflects a single isolated weekend cycle and should not be used as a structural trend model without historical calibration.