央行减持与交易对手风险管理的定量方法

A Quantitative Approachto Central Bank Haircutsand Counterparty RiskManagement Yuji Sakurai WP/25/225 IMF Working Papersdescribe research inprogress by the author(s) and are published toelicit comments and to encourage debate.The views expressed in IMF Working Papers arethose of the author(s) and do not necessarilyrepresent the views of the IMF, its Executive Board,or IMF management. 2025OCT IMF Working Paper Monetary and Capital Markets Department AQuantitative Approach to Central Bank Haircuts and Counterparty Risk ManagementPrepared byYuji Sakurai Authorized for distribution byRomain VeyruneOctober2025 IMF Working Papersdescribe research in progress by the author(s) and are published to elicitcomments and to encourage debate.The views expressed in IMF Working Papers are those of theauthor(s) and do not necessarily represent the views of the IMF, its Executive Board, or IMF management. ABSTRACT:This paper presents a comprehensive framework for determining haircuts on collateral used incentral bank operations, quantifying residual uncollateralized exposures, and validating haircut models usingmachine learning. First, it introduces four haircut model types tailored to asset characteristics—marketable ornon-marketable—and data availability.It proposes a novel model for setting haircuts in data-limitedenvironment using a satallite cross-country model.Key principles guiding haircut calibration include non-procyclicality, data-drivenness, conservatism, and the avoidance of arbitrage gaps. The paper details modelinputs such asValue-at-Risk(VaR) percentiles, volatility measures, and timetoliquidation. Second, it proposesa quantitative framework for estimating expected uncollateralized exposures thatremainafter haircutapplication, emphasizing their importance in stress scenarios. Illustrative simulations using dynamic Nelson-Siegel yield curve models demonstrate how volatility impacts exposure. Third, the paper explores the use ofVariational Autoencoders(VAEs) to simulate stress scenarios for bond yields. Trained on U.S. Treasury data,VAEscapture realistic yield curve distributions, offering an altenativetool for validating VaR-based haircuts.Although interpretability and explainability remain concerns, machine learning models enhance riskassessment by uncovering potential model vulnerabilities. Contents Protecting Central Bank’s Balance Sheet with Tail Risk Measures...............................................................7Value-at-Risk vs. Expected Shortfall..............................................................................................................8Mathematical Definition of Haircuts................................................................................................................8Determining VaR Percentile...........................................................................................................................9External Factors Not Accounted for Within the Modeling Framework............................................................9An Illustrative Toy Model................................................................................................................................9Dependance Between Collateral Volatility and Counterparty Default Risk..................................................10Central Bank’s Uncollateralized Exposure...................................................................................................12Classifying the Modeler’s Situation..............................................................................................................14Setting Duration Bucket................................................................................................................................15 Marketable Assets in a Data-Rich Environment.............................................................................................15 Overview of DASV Model.............................................................................................................................15Duration Approximation................................................................................................................................16Stressing Volatility........................................................................................................................................17Time to Liquidation.......................................................................................................................................18Ensuring Consistency in Haircuts Models for Relevant Markets..................................................................18Combining with Credit Rating Data..............................................................................................................19Application to Real-World Data....................................................................................................................20 Marketable Assets in a Data-Limited Environment.......