LLM-Based Examination of Eligibility Criteria from Securities Prospectuses at the German Central Bank
2026-06-25 • Computation and Language
Computation and Language
AI summaryⓘ
The authors address the challenge of checking if securities qualify as collateral for the German Central Bank, a process that is usually slow because it involves reading complicated, mixed-language documents. They use Large Language Models (LLMs) instead of traditional methods to better handle messy text and bilingual content. Their method breaks the task into parts like extracting and understanding information, improving accuracy. They also created a new way to evaluate results based on meaning rather than just exact text matches. Their approach shows high precision, meaning it rarely accepts ineligible documents by mistake.
Collateral eligibilityGerman Central BankLarge Language Models (LLMs)Named Entity Recognition (NER)Information ExtractionOptical Character Recognition (OCR) noiseBilingual text processingEvaluation metricsDocument-level classification
Authors
Serhii Hamotskyi, Akash Kumar Gautam, Christian Hänig
Abstract
Verifying the eligibility of securities as collateral is a key responsibility of the German Central Bank. However, manually verifying these assets against legal and financial criteria within lengthy, semi-structured, and often bilingual prospectuses is a resource-intensive task. While previous efforts utilized traditional Named Entity Recognition (NER) for information extraction, these methods can struggle with OCR noise, linguistic variance, and rigid span-based constraints, and the need for manually annotated training data for each relevant annotation type. In this paper, we present the first case study applying Large Language Models (LLMs) to the eligibility examination process, shifting the paradigm toward a generative Information Extraction pipeline. Our approach decomposes the task into extraction, normalization, and interpretation, allowing for greater flexibility in handling noisy text and interleaved German-English content. We further introduce a value-based evaluation methodology using LLM-as-a-judge, which offers a more semantic assessment than location-based metrics. Our results demonstrate that LLM-based systems achieve high precision (up to 91%) in document-level eligibility, exhibiting a conservative operating profile that minimizes false acceptance.