Liquidation Reliability Index

Credit & Lending

Active

Quantitative Risk

Confidence Score

88%

Last updated 13 days ago

Current Thesis

Arch's liquidation engine achieves 98.2% reliable liquidation execution within the target window. This is enabled by sub-second oracle feeds, deterministic ordering, and UTXO-based finality. Competitors using async settlement see 78-85% reliability under stress.

Evolution of Belief

Nov 1, 2025

Liquidation reliability is fundamentally limited by oracle latency on all L2s.

Reason: Initial architecture assumptions

Feb 14, 2026

UTXO anchoring enables deterministic, sub-second liquidation with 95%+ reliability.

Reason: Testnet results validated oracle-independent confirmation mechanism

13 days ago

Arch achieves 98.2% reliability; our architecture is now 15+ percentage points better than peers.

Reason: Stress test results under flash-crash conditions; comparison with live L2 data

Key Data Inputs

Arch Liquidation Success Rate

% of liquidations executed at intended price within 300ms

98.2%

Trend: Stable

Competitor Avg Success Rate

Delphi Labs (Canto), Compound (Ethereum) average

82%

Trend: Stable

Oracle Latency (P99)

99th percentile oracle update latency

150ms

Trend: Down

Linked Research Nodes

depends on

Bad Debt Probability Model

informs

Stablecoin Yield Competitiveness

Open Questions

•How does liquidation reliability scale under 10x leverage scenarios?
•Can we achieve 99.5%+ reliability through redundant oracle feeds?

Downstream Decisions

Product

Critical Impact

Can safely offer 5x leverage to institutional clients; competitors limited to 2-3x

Treasury

High Impact

Bad debt reserve can be 40-50% lower than peers due to reliability advantage

Change Log

13 days ago

Updated benchmark comparison with live competitor data

Quantitative Risk Analyst