Search

CN-121981813-A - Asset loss rate estimation method

CN121981813ACN 121981813 ACN121981813 ACN 121981813ACN-121981813-A

Abstract

The invention discloses an asset loss rate estimation method, which relates to the technical field of financial risk management and comprises the following steps of S1 data acquisition, S2 default probability and main body rating determination, wherein the S1 data acquisition comprises the steps of acquiring internal data and external market data of a target client, acquiring a guarantee mode, a mortgage countermand detail and present value data of the target debt, S2 default probability and main body rating determination, namely, calculating future annual default Probability (PD) of the target client by utilizing a Merton model based on the internal data and the external market data of the target client in the step S1, mapping the default probability to a standard Prime main body rating, and outputting an internal main body rating according to an internal and external rating mapping rule, wherein the default probability prediction based on the Merton model and the default loss rate calculation fused in a multimode mode can be organically combined through an integrated automatic evaluation flow, and full-link automatic metering from data input to risk output is realized by means of a standardized rating mapping mechanism, and the accuracy and the efficiency of risk estimation are remarkably improved.

Inventors

  • ZHANG TONGYU
  • CHEN YIYING
  • FENG LIYANG

Assignees

  • 中富数字科技有限公司

Dates

Publication Date
20260505
Application Date
20251224

Claims (10)

  1. 1. The asset loss rate estimation method is characterized by comprising the following steps of: s1, acquiring internal data and external market data of a target client, and acquiring a guarantee mode, a mortgage resisting object detail and present value data of a target debt; S2, determining the default probability and the main body rating, namely calculating the future annual default Probability (PD) of the target client by utilizing a Merton model based on the internal data and the external market data of the target client in the step S1, mapping the default probability to the standard Primer main body rating, and outputting the internal main body rating according to an internal and external rating mapping rule; S3, determining the default loss rate and the debt rating, namely parallelly calculating the default loss rate (LGD) of the target debt based on the data of the target debt in the step S1 through a WATERFALL theoretical model, a market parameter model and a mortgage present value model, and synthesizing the LGD values calculated by the three models to obtain a final LGD estimated value; And S4, multiplying the default probability PD obtained in the step S2 by the final LGD estimated value obtained in the step S3, and calculating and outputting the expected asset loss rate of the target debt.
  2. 2. The method of claim 1, wherein the internal data in step S1 includes bank-to-credit ledger details and customer financial data, and the external market data includes industry index data and market interest rate data.
  3. 3. The asset loss rate estimation method according to claim 1, wherein the calculation of the default probability by the Merton model in step S2 is based on the Black-Scholes-Merton option pricing model by the following relation, wherein the default probability pd=1-N (d 2): Wherein E represents the equity value, V represents the asset value, D represents the liability value, r is the risk-free interest rate, T is the liability expiration period, and other parameters are defined according to model criteria.
  4. 4. The method of claim 1, wherein calculating the LGD by WATERFALL theoretical model in step S3 includes defining legal repayment priority for each debt of the same debtor, calculating repayment amount for each debt according to repayment priority based on the assignable value of the mortgage, and further calculating recovery rate and LGD, and determining repayment priority by referring to standard Prer WATERFALL method.
  5. 5. The method of claim 1, wherein the step S3 of calculating the LGD by a market parameter model is implemented according to a formula LGD=MAX (Floor, 1-C% -BG%), floor is a preset LGD base line value, C% is a mortgage recovery rate, BG% is an enterprise clearing reference recovery rate, floor value is determined by a main guarantee mode of a target debt and a main mortgage type, C% is determined by a mortgage type, an evaluation value and a conversion rate, and BG% is determined by a ratio of an enterprise clearing value to a total debt.
  6. 6. The method of claim 1, wherein mapping the probability of breach to the standard Prayer rating in step S2 is performed by querying a standard Prayer credit rating migration matrix.
  7. 7. The method for estimating a loss rate of an asset according to claim 1, wherein in step S3, the calculation of LGD by the present value model of the mortgage is performed according to the formula LGD=1- (present value of mortgage/remaining principal of debt).
  8. 8. The method of estimating a loss rate of an asset according to claim 1, wherein synthesizing LGD values calculated by the three models in step S3 means calculating an average value thereof.
  9. 9. The asset loss rate estimation method according to claim 1, wherein the correspondence between the recovery interval defined by the recovery rate rating map preset in step S3 and the rating adjustment sub-term is +3 when the recovery rate is 100%, +2 when the recovery rate is 90% or more and less than 100%, +1 when the recovery rate is 70% or more and less than 90%, 0 when the recovery rate is 50% or more and less than 70%, 0 when the recovery rate is 30% or more and less than 50%, 0 when the recovery rate is 10% or more and less than 30%, 1 when the recovery rate is 0% or more and less than 10%, 2 when the recovery rate is 0% or more and less than 10%, and generating the debt rating after adjustment is to add or subtract the rating of the internal body.
  10. 10. The method of estimating a loss rate of an asset according to claim 1, wherein the processing of step S2 and step S3 are performed independently of each other and in parallel.

Description

Asset loss rate estimation method Technical Field The invention relates to the technical field of financial risk management, in particular to an asset loss rate estimation method. Background In the management of modern commercial banks, the accurate estimation of potential loss of credit assets is a core link of credit risk management, and is also a basis for meeting supervision compliance and fine configuration of internal capital, the expected loss rate of assets is used as a key index for quantifying credit risk, the calculation of the expected loss rate of assets depends on the accurate measurement of risk parameters such as default Probability (PD), default loss rate (LGD) and the like, the bank can not only realize the accurate classification of customer and debt risk by scientifically evaluating the loss rate of assets, but also provide decision basis for credit approval, risk pricing and post-credit management, and can further provide sufficient reserve and capital according to the calculation, thereby improving the overall risk resistance capability and the operational robustness of an organization; However, the existing asset loss rate estimation technology has significant shortcomings, the traditional method usually breaks the metering processes of PD and LGD, an integrated automatic estimation system cannot be formed, a large amount of manual intervention exists in a chain from data input to final expected loss rate output, the efficiency is low, inconsistency is easy to generate, the PD prediction often depends excessively on historical financial data in a model layer, prospective capture of market dynamics is lacking, the LGD estimation adopts a single method, the strict nature of law clearing sequence, the robustness of market parameters and the requirement of estimation efficiency cannot be considered, in addition, the mapping relation between internal risk rating and external standard (such as standard rating) is not systematic, so that the risk result lacks comparability, and the overall accuracy, efficiency and application value of the asset loss rate estimation in risk management practice are finally influenced. Disclosure of Invention The invention aims to overcome the existing defects, provides an asset loss rate estimation method, can organically combine the default probability prediction based on the Merton model with the default loss rate calculation based on multi-model fusion through an integrated automatic estimation flow, realizes full-link automatic metering from data input to risk output by means of a standardized rating mapping mechanism, remarkably improves the accuracy and efficiency of risk estimation, and can effectively solve the problems in the background art. In order to achieve the above purpose, the invention provides the following technical scheme that the asset loss rate estimation method comprises the following steps: s1, acquiring internal data and external market data of a target client, and acquiring a guarantee mode, a mortgage resisting object detail and present value data of a target debt; S2, determining the default probability and the main body rating, namely calculating the future annual default Probability (PD) of the target client by utilizing a Merton model based on the internal data and the external market data of the target client in the step S1, mapping the default probability to the standard Primer main body rating, and outputting the internal main body rating according to an internal and external rating mapping rule; S3, determining the default loss rate and the debt rating, namely parallelly calculating the default loss rate (LGD) of the target debt based on the data of the target debt in the step S1 through a WATERFALL theoretical model, a market parameter model and a mortgage present value model, and synthesizing the LGD values calculated by the three models to obtain a final LGD estimated value; s4, multiplying the default probability PD obtained in the step S2 by the final LGD estimated value obtained in the step S3, and calculating and outputting the expected asset loss rate of the target debt; The problems that PD, LGD and the final loss rate calculation flow in the traditional method are cracked and a unified automation system cannot be formed are solved through the steps, full-link automatic metering from original data to final risk quantization indexes (expected loss rates) is achieved through organic series connection of the four steps, and the efficiency and accuracy of risk metering are improved. Further, in the step S1, the internal data comprise details of a public credit standing book and customer financial data, the external market data comprise industry index data and market interest rate data, and the quality of a model result is ensured from the source by ensuring that a risk metering model can fully utilize microscopic information lend financial data accumulated in the bank and combining external data reflecting macroscopic ec