Data processing method, storage medium and electronic equipment

Abstract

The embodiment of the invention provides a data processing method, a storage medium and electronic equipment, and the method comprises the steps: extracting similar asset historical data corresponding to a target asset pool of a to-be-predicted object, the to-be-predicted object being an asset securitization product initiated by a target initiator, and the similar asset historical data being historical data of a similar asset pool; analyzing the historical data of the same kind of assets according to preset pressure parameters; calibrating a pressure parameter reference sequence corresponding to the target asset obtained through analysis based on a parameter true value, corresponding to the preset pressure parameter, of the target asset, and obtaining a predicted cash flow collection table for cashing by applying a pressure parameter calibration sequence obtained through calibration to a target cash flow collection table of a target asset pool; and distributing the cash flow of the target asset pool according to the cash flow payment mechanism of the to-be-predicted object based on the predicted cash flow collection table and a credit level increasing measure adopted by the target initiator to obtain an expected cash flow of the to-be-predicted object.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of the Internet, and more specifically, to a data processing method, a storage medium, and an electronic device. Background Art

[0002] When making investment decisions, users often need to predict the cash flow of asset-backed securities products as a reference for decision-making. However, since the cash flow of asset-backed securities products needs to be processed from the cash flow of the underlying asset pool assets, there are many data sources, a large amount of data, and a high proportion of unstructured data, which makes data collection and processing difficult. In addition, asset-backed securities products have complex cash flow payment mechanisms and credit enhancement measures, and the calculation process is long and the amount of calculation is large. It is difficult for users to directly calculate and process data to obtain accurate predicted cash flows. It can be seen that the data processing methods in the relevant technologies have the problem of low accuracy of data processing results due to the difficulty of data acquisition and the complexity of data processing. Summary of the invention

[0003] The embodiments of the present application provide a data processing method, a storage medium and an electronic device to at least solve the technical problem that the data processing method in the related art has low accuracy of data processing results due to the difficulty in data acquisition and the complexity of data processing.

[0004] According to one aspect of an embodiment of the present application, a data processing method is provided, comprising: extracting historical data of similar assets corresponding to a target asset pool of an object to be predicted, wherein the data of similar assets is stored in a designated database, the object to be predicted is an asset securitization product initiated by a target initiator, the historical data of similar assets is historical data of a similar asset pool, the similar asset pool is a collection of similar assets of a target asset in the target asset pool, the asset type of the similar assets is the same as the asset type of the corresponding target asset and does not belong to the target asset pool; parsing the historical data of similar assets according to preset pressure parameters to obtain pressure parameters corresponding to the target asset. A benchmark sequence, wherein the preset pressure parameter is a parameter that affects the cash flow of the target asset; based on the true value of the parameter corresponding to the target asset and the preset pressure parameter, the pressure parameter benchmark sequence corresponding to the target asset is calibrated to obtain a pressure parameter calibration sequence corresponding to the target asset, and the pressure parameter calibration sequence is applied to the target cash flow aggregation table of the target asset pool to obtain a forecast cash flow aggregation table for redemption; based on the forecast cash flow aggregation table and the credit enhancement measures adopted by the target initiator, the cash flow of the target asset pool is allocated according to the cash flow payment mechanism of the object to be predicted to obtain the expected cash flow of the object to be predicted.

[0005] In an exemplary embodiment, the designated database also stores at least one of the following information: asset pool data of the target asset pool; cash flow payment mechanism data indicating the cash flow payment mechanism; credit enhancement data indicating the credit enhancement measures; historical principal and interest repayment data indicating the historical principal and interest repayment of the object to be predicted; the data used in the designated database for cash flow prediction of the object to be predicted is underlying data, and the underlying data includes historical data of similar assets; the method also includes: streaming query in a multi-threaded concurrent manner to see whether the underlying data stored in the designated database is updated; if the underlying data is updated, synchronizing the updated underlying data to the cache of each node among the multiple nodes, wherein the underlying data used for cash flow prediction of the object to be predicted is the underlying data cached on one of the multiple nodes.

[0006] In an exemplary embodiment, the data used by the designated database for cash flow prediction of the object to be predicted is underlying data, and the underlying data is synchronized to the cache of each node among the multiple nodes; the underlying data includes the historical data of the same type of assets; the amount of memory data on each node is less than or equal to the threshold of the amount of memory data corresponding to each node; the method also includes: in response to the received target prediction request, according to the specified load balancing strategy, the target computing task corresponding to the object to be predicted is allocated to the target node among the multiple nodes, so that the target computing task is executed on the target node based on the underlying data cached on the target node, wherein the target prediction request is used to request cash flow prediction for the object to be predicted, the specified load balancing strategy is a strategy for balancing the computing tasks allocated to the multiple nodes, the target computing task is a computing task for cash flow prediction of the object to be predicted, and the target computing task is executed on the target node using multi-threaded computing.

[0007] In an exemplary embodiment, the method further includes: in the case where there is an asset set of the same type of assets initiated by the target initiator, determining the asset set of the same type of assets initiated by the target initiator as the same type of asset pool; in the case where there is no asset set of the same type initiated by the target initiator, determining the asset set of the same type of assets initiated by other initiators except the target initiator as the same type of asset pool; after extracting the historical data of the same type of assets corresponding to the target asset pool of the object to be predicted, the method further includes: based on the cross-reference relationship between fields, standardizing the historical data of the same type of assets according to a set of specified fields to obtain the standardized historical data of the same type of assets.

[0008] In an exemplary embodiment, the parsing of the historical data of the same type of assets according to preset stress parameters to obtain a stress parameter benchmark sequence corresponding to the target asset includes one of the following: when the same type of asset pool includes a static structure asset pool and the historical data of the same type of assets includes the historical data of the static structure asset pool, parsing the historical data of the static structure asset pool according to a first stress parameter to obtain a first parameter benchmark sequence corresponding to the target asset, wherein the first stress parameter includes early repayment rate, default rate and recovery rate, and the stress parameter benchmark sequence corresponding to the target asset includes the first parameter benchmark sequence; when the same type of asset pool includes a cyclical structure asset pool and the historical data of the same type of assets includes the historical data of the cyclical structure asset pool, parsing the historical data of the cyclical structure asset pool according to a second stress parameter to obtain a second parameter benchmark sequence corresponding to the target asset, wherein the second stress parameter includes the repayment rate, default rate, recovery rate, yield rate and purchase rate, and the stress parameter benchmark sequence corresponding to the target asset includes the second parameter benchmark sequence.

[0009] In an exemplary embodiment, the pressure parameter baseline sequence corresponding to the target asset is calibrated based on the true value of the parameter corresponding to the target asset and the preset pressure parameter to obtain the pressure parameter calibration sequence corresponding to the target asset, including: based on the difference between the true value of the parameter corresponding to the target asset and the preset pressure parameter and the parameter value of the same pressure parameter in the pressure parameter baseline sequence corresponding to the target asset, the pressure parameter baseline sequence corresponding to the target asset is shifted and adjusted to obtain the pressure parameter calibration sequence corresponding to the target asset.

[0010] In an exemplary embodiment, before obtaining the forecasted cash flow aggregation table for redemption by applying the pressure parameter calibration sequence to the cash flow aggregation table of the target asset pool, the method further includes: calculating the expected cash flow recovery information of each target asset according to the asset pool data of the target asset pool, wherein the expected cash flow recovery information of each target asset is used to indicate the expected future cash flow recovery of each target asset; aggregating each target asset in the target asset pool according to each aggregation date of the object to be predicted, and generating a first candidate cash flow aggregation table based on the cash flow recovery information obtained on each aggregation date, wherein the cash flow recovery information on each aggregation date is used to indicate the cash flow recovery on each aggregation date; determining the cash flow aggregation table that best matches the latest balance of the target asset pool among the first candidate cash flow aggregation table and the second candidate cash flow aggregation table as the target cash flow aggregation table, wherein the second candidate cash flow aggregation table is a cash flow aggregation table integrated by the cash flow aggregation table of the target asset pool configured for the object to be predicted and the actual performance of the target asset pool calculated by statistics.

[0011] In an exemplary embodiment, the predicted cash flow aggregation table is cached according to a specified data reading mode, and the specified data reading mode is used to balance the total amount of table data and the cache load; the method also includes: calling the cash flow distribution sequence calculation method corresponding to the object to be predicted in the preset method module library under the account type corresponding to the target initiator, to obtain the cash flow payment mechanism of the object to be predicted.

[0012] In an exemplary embodiment, the method further includes: determining the information to be displayed of the object to be predicted based on the historical principal and interest payment data of the object to be predicted and the expected cash flow of the object to be predicted, and distributing the information to be displayed to a display interface for display, wherein the information to be displayed is used to describe the principal and interest payment cash flow during the complete life cycle of the object to be predicted.

[0013] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when run.

[0014] According to another aspect of the embodiments of the present application, there is further provided an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the steps in any one of the above method embodiments through the computer program.

[0015] Through the present application, since the historical data of similar assets corresponding to the target asset pool of the object to be predicted can be extracted, the collection standards, processing principles and usage methods for large quantities of data can be standardized, which can effectively reduce the time-consuming long process calculations and reduce the errors caused by manual calculations; the historical data of similar assets are parsed according to the preset pressure parameters to obtain the pressure parameter benchmark sequence corresponding to the target assets, wherein the preset pressure parameters are the parameters that affect the cash flow of the target assets, and the pressure parameter benchmark sequence corresponding to the target assets is calibrated based on the actual values ​​of the parameters corresponding to the target assets and the preset pressure parameters to obtain the pressure parameter calibration sequence corresponding to the target assets, and the pressure parameter calibration sequence is applied to the target cash flow collection table of the target asset pool to obtain the predicted cash flow collection for redemption. The table can be used to set different scenarios such as early repayment and default recovery for the underlying asset pool assets, and the risk level of asset securitization products can be tested through stress testing. In addition, based on the predicted cash flow aggregation table and the credit enhancement measures adopted by the target initiator, the cash flow of the target asset pool can be allocated according to the cash flow payment mechanism of the object to be predicted, and the expected cash flow of the object to be predicted can be obtained, thereby achieving a transparent, repeatable, and well-interpretable modeling process and result display. By adopting a distributed computing framework, the data can be processed comprehensively and accurately. Therefore, the technical problem of low accuracy of data processing results due to the difficulty of data acquisition and complex data processing in the data processing methods in related technologies can be solved, thereby achieving the effect of improving the accuracy and reliability of data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of an application scenario of a data processing method according to an embodiment of the present application.

[0017] Figure 2 It is a flowchart of an optional data processing method according to an embodiment of the present application.

[0018] Figure 3 It is a schematic diagram of an optional asset securitization process according to an embodiment of the present application.

[0019] Figure 4 It is a schematic diagram of an optional cash flow payment mechanism according to an embodiment of the present application.

[0020] Figure 5 It is a flowchart of another optional data processing method according to an embodiment of the present application.

[0021] Figure 6 It is a flowchart of another optional data processing method according to an embodiment of the present application.

[0022] Figure 7It is a flowchart of another optional data processing method according to an embodiment of the present application.

[0023] Figure 8 It is a flowchart of another optional data processing method according to an embodiment of the present application.

[0024] Fig. 9 It is a technical framework diagram of an optional data processing method according to an embodiment of the present application.

[0025] Fig.10 It is a module logic diagram of an optional data processing method according to an embodiment of the present application.

[0026] Fig.11 It is a structural block diagram of an optional data processing device according to an embodiment of the present application.

[0027] Fig.12 It is a block diagram of a computer system structure of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0029] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] According to one aspect of an embodiment of the present application, a data processing method is provided. Optionally, in this embodiment, the data processing method may be applied to, but is not limited to, Figure 1The hardware environment shown includes a terminal device 102 and a server 104. The server 104 can be connected to the terminal device 102 via a network, and can be used to provide services (e.g., application services, etc.) for the terminal device 102 or a client installed on the terminal device 102. A database can be set on the server 104 or independently of the server 104 to provide data storage services for the server 104.

[0031] The above network may include but is not limited to at least one of the following: wired network, wireless network. The above wired network may include but is not limited to at least one of the following: wide area network, metropolitan area network, local area network, and the above wireless network may include but is not limited to at least one of the following: WIFI (Wireless Fidelity), Bluetooth. The terminal device 102 may be but is not limited to a PC (Personal Computer), a mobile phone, a tablet computer, etc. The server 104 may be but is not limited to a cloud server, a server cluster or other server types.

[0032] The data processing method of the embodiment of the present application can be executed by the server 104, or by the terminal device 102, or by both the server 104 and the terminal device 102. The data processing method of the embodiment of the present application can be executed by the terminal device 102 or by a client installed thereon.

[0033] Taking the data processing method in this embodiment executed by the terminal device 102 as an example, Figure 2 is a flow chart of an optional data processing method according to an embodiment of the present application, such as Figure 2 As shown, the process of the method may include the following steps:

[0034] Step S202, extracting historical data of similar assets corresponding to the target asset pool of the object to be predicted, wherein the data of similar assets is stored in a designated database, the object to be predicted is an asset securitization product initiated by the target initiator, the historical data of similar assets is historical data of the similar asset pool, the similar asset pool is a collection of similar assets of the target asset in the target asset pool, and the asset type of the similar assets is the same as the asset type of the corresponding target asset and does not belong to the target asset pool;

[0035] Step S204, parsing historical data of similar assets according to preset pressure parameters to obtain a pressure parameter benchmark sequence corresponding to the target asset, wherein the preset pressure parameter is a parameter that affects the cash flow of the target asset;

[0036] Step S206, calibrating the pressure parameter reference sequence corresponding to the target asset based on the actual value of the parameter corresponding to the preset pressure parameter of the target asset to obtain the pressure parameter calibration sequence corresponding to the target asset, and applying the pressure parameter calibration sequence to the target cash flow collection table of the target asset pool to obtain the forecast cash flow collection table for redemption;

[0037] Step S208, based on the forecast cash flow aggregation table and the credit enhancement measures adopted by the target initiator, the cash flow of the target asset pool is allocated according to the cash flow payment mechanism of the object to be forecasted to obtain the expected cash flow of the object to be forecasted.

[0038] The data processing method in this embodiment can be applied to the scenario of cash flow forecasting for asset securitization products. Here, asset securitization refers to the process in which an enterprise or financial institution combines its assets that can generate cash income, and then issues securities products to investors with the support of its cash flow. This behavior converts illiquid assets into securities that can be freely bought and sold in the financial market, making them liquid. Asset securitization products are ostensibly supported by "assets", but in fact they are supported by the "cash flow" generated by the assets. It is a technology for redistributing and reorganizing the cash flow of the asset pool.

[0039] Taking debt assets as an example, the asset securitization process is as follows: Figure 3 As shown, it may include the following steps:

[0040] (1) The original equity holder issues a loan and forms the underlying creditor's rights asset;

[0041] (2) The original beneficiary, as the initiator, entrusts the trustee to establish a special purpose trust with its creditor's assets as trust property;

[0042] (3) The trustee issues the current asset securitization products to investors and acquires the underlying assets of the original equity holders within the limit of the investors’ subscription amount;

[0043] (4) During the term of the subsequent products, the principal and income of the current securities shall be paid within the limit of the cash flow generated by the trust property;

[0044] (5) The intermediary institution provides underwriting, financial consulting, fund custody, registration and custody services and charges a certain fee. The relevant fees are also paid within the limit of the cash flow generated by the trust property.

[0045] For the above-mentioned asset securitization products, the asset pool they own is the combination of the assets therein. Since the cash flow of asset securitization products needs to be processed from the cash flow of the underlying asset pool assets, there are many data sources, a huge amount of data, and a high proportion of unstructured data, which makes data collection and processing difficult; coupled with the complex cash flow payment mechanism and credit enhancement measures, the calculation process is long and the amount of calculation is large, making it difficult for investors to obtain the expected cash flow of asset securitization products, resulting in difficulties in investment decisions, resulting in a situation where the primary market issuance volume is large but still lacks liquidity, and the primary and secondary prices deviate.

[0046] If investors calculate expected cash flows manually, they will face several challenges: first, the amount of data collected is large, the data is difficult to obtain, and the data processing and cleaning are complicated. Investors need to prepare hundreds of data fields such as basic information data of asset securitization products, asset pool data, cash flow payment mechanism data, credit enhancement data, historical principal and interest payment data, etc. These data come from multiple documents such as the prospectus, trust contract, trustee report, rating report, etc., and a large amount of unstructured data needs to be further processed. In addition, the amount of asset pool data is usually in tens of thousands, and there is generally no public access channel. The possibility of manually calculating the cash flow of the asset pool is zero; second, there are many calculation steps and it takes a long time to calculate by yourself In order to obtain the expected cash flow, it is necessary to perform multiple calculation steps, such as statistics of asset pool stress scenario parameters, calculation of basic scenario asset pool cash flow, calculation of asset pool cash flow considering stress scenario, and allocation of asset pool cash flow to asset securitization products. The calculation process is long and there are many calculation details. In addition, the cash flow payment mechanism of each asset securitization product is unique. Manual calculation is difficult to apply manual calculation templates of other asset securitization products and must be calculated from scratch. Therefore, manual calculation takes a lot of time. Third, the calculation accuracy is low. Due to the rich data, large amount of unstructured data, long calculation process and large amount of calculation, manual processing is bound to have errors, making it difficult to guarantee the accuracy of cash flow forecasts.

[0047] In order to at least partially solve the above technical problems, in this embodiment, a data processing method is adopted to standardize data collection standards and processing principles, effectively reduce calculation time, and reduce errors caused by manual calculations, so as to provide more accurate expected cash flows of asset securitization products, fill market gaps, and provide investors with more convenient related services, thereby promoting price discovery of asset securitization products, improving market pricing efficiency and securities liquidity, and providing more convenient financial service support.

[0048] In this embodiment, based on the cash flow of the underlying asset pool of the asset securitization product and taking into account credit enhancement measures, the cash flow of the asset pool can be split into the asset securitization product in accordance with its cash flow payment mechanism to obtain the expected cash flow of the asset securitization product. The systematic calculation reduces the cost for investors to obtain the expected cash flow, and facilitates investors' further valuation, investment decision-making, cash flow management, and position analysis. In addition, in this embodiment, different scenarios such as early repayment and default recovery can be supported for the underlying asset pool assets, and the degree of risk borne by the asset securitization product can be tested through stress testing so that investors can conduct risk management, investment decisions, credit rating, etc.

[0049] Here, credit enhancement measures refer to measures that issuers can take through various means and methods to ensure timely and full payment of investors' interest and principal in order to strengthen the guarantee of capital repayment of securitization products and protect investors when the asset quality of the asset pool deteriorates. It can be an important feature that distinguishes asset securitization products from other bond products. Among them, credit enhancement measures usually include internal measures and external measures. External measures refer to increasing third-party guarantees or insurance, etc. Internal measures do not introduce external institutions. Credit enhancement is usually achieved through dozens of methods such as the securitization product's own structure and trigger mechanism design, asset pool excess spread, etc.

[0050] Optionally, credit enhancement measures may include structural design, trigger mechanism and asset pool excess spread, where structural design refers to the division of the issued securitized products into products of different structures such as priority / intermediate / subordinate, and the asset pool receipts are paid in accordance with the agreed sequence of cash flow repayment mechanism, and the subordinated securities provide credit loss protection for the priority securities; the trigger mechanism refers to the setting of accelerated repayment and default events. Once the relevant events are triggered, the cash flow payment mechanism will be rearranged. The arrangement of the trigger mechanism alleviates the impact of time risk to a certain extent and provides a certain degree of credit support; the asset pool excess spread refers to the weighted average interest rate of the asset pool being greater than the service fee rate of the relevant participating institutions and the face value interest rate of the priority securities of the asset securitization products, which provides certain support for the repayment of principal and interest of the asset securitization products.

[0051] Here, the cash flow payment mechanism refers to the order of cash repayment arrangements for using the asset pool to recover cash flow to pay related expenses and redeem the principal and interest of asset securitization products. Taking the principal and income sub-account as an example (that is, the asset pool income and principal are used to pay the interest and principal of asset securitization products respectively. If a single account is insufficient to pay the corresponding amount, it can be supplemented by transferring between accounts), the cash flow payment mechanism of asset securitization products is as follows: Figure 4 As shown, it may include the following steps:

[0052] (1) In terms of the income account, in the absence of a default, taxes and trust fees are actually the highest priority, followed by the priority payment of service agency fees (some products stipulate that the loan service agency can only obtain the remaining service fees after paying interest at all levels and replenishing the principal account), and then the highest priority and second priority interest respectively;

[0053] (2) Subsequently, if there is a loan default or a transfer from the principal account to the income account, the income account will first transfer the above amount of funds to the principal account. After that, the remaining service agency remuneration and the income within the secondary tier limit can be obtained in turn;

[0054] (3) If there is any balance in the income account, it will be transferred to the principal account;

[0055] (4) With respect to the principal account, after the income account is replenished, the principal of each series at each level will be repaid in turn. Any surplus will become the excess income of the secondary tier.

[0056] Exemplarily, in this embodiment, historical data of similar assets corresponding to the target asset pool of the object to be predicted can be extracted. The similar asset data is stored in a designated database, the object to be predicted is an asset securitization product initiated by the target initiator, the similar asset historical data is the historical data of the similar asset pool, and the similar asset pool is a collection of similar assets of the target asset in the target asset pool.

[0057] Optionally, in this embodiment, the designated database may be directly connected, for example, the designated database may be directly connected through the data access and processing module to obtain the required data from the designated database.

[0058] Optionally, historical data of similar assets corresponding to the target asset pool of the asset securitization product initiated by the target initiator can be obtained from the designated database, including but not limited to: asset pool data of the asset securitization product, cash flow payment mechanism data, credit enhancement data, historical principal and interest payment data, historical data of similar assets of the initiator, etc.

[0059] Optionally, after extracting historical data of similar assets corresponding to the target asset pool of the object to be predicted, data caching can be performed, so that the cached data can be directly called in the subsequent data processing process to reduce calculation time.

[0060] Optionally, the data obtained from the specified database may be preprocessed, including but not limited to: data cleaning, conversion and formatting.

[0061] In this embodiment, after extracting historical data of similar assets corresponding to the target asset pool of the object to be predicted, the historical data of similar assets can be analyzed according to preset pressure parameters to obtain a pressure parameter benchmark sequence corresponding to the target asset, wherein the preset pressure parameters are parameters that affect the cash flow of the target asset.

[0062] Optionally, by parsing historical data of similar assets according to preset stress parameters, cash flow forecasts of the object to be forecasted under different stress scenarios can be achieved to assess its risk tolerance.

[0063] Optionally, preset stress parameters can be selected based on market analysis or risk management strategies. For example, for scenarios such as an increase in prepayment rate, an increase in default rate, and a decrease in recovery rate, the prepayment rate, default rate, and recovery rate can be selected as preset stress parameters. For another example, corresponding preset stress parameters can be selected for stress events that may occur within the asset pool, such as large-scale early repayments.

[0064] Optionally, by parsing the extracted historical data of similar assets, the expected cash flow of similar asset pools under different stress scenarios can be calculated.

[0065] Optionally, based on the parsed data above, a stress parameter benchmark sequence under each preset stress parameter can be generated. Here, since the cash flow performance of the underlying assets in the asset securitization product includes a dynamic process that changes over time, a sequence (i.e., a stress parameter benchmark sequence) can be used to represent the above parameters to show the changes in the above parameters at different time points during the life cycle of the asset securitization product.

[0066] Affected by factors such as the economic environment, the actual performance of the asset pool may differ from history to a certain extent. Therefore, the stress parameters need to be calibrated according to the actual performance. In this embodiment, the stress parameter benchmark sequence corresponding to the target asset can be calibrated based on the actual value of the parameter corresponding to the target asset and the preset stress parameter to obtain the stress parameter calibration sequence corresponding to the target asset, that is, the above-mentioned stress parameter benchmark sequence is aligned with the actual data, and the impact of the actual market situation is considered to generate a stress parameter calibration sequence for the target asset pool that is closer to the actual situation. For example, if the prepayment rate of the target asset pool in actual performance is significantly higher than the historical average level, the future prepayment rate forecast value of the stress parameter calibration sequence performance can also be adjusted to a higher level.

[0067] Optionally, after obtaining the stress parameter calibration sequence corresponding to the target asset, the stress parameter calibration sequence may be applied to the target cash flow aggregation table of the target asset pool to obtain a forecast cash flow aggregation table for redemption.

[0068] Here, the target cash flow collection table refers to a cash flow forecast table calculated based on the current status of the target asset pool and basic assumptions, which may include but is not limited to data such as expected principal recovery, interest payments, and possible credit losses, each of which can correspond to a specific collection period.

[0069] The asset pool of asset securitization products consists of a large number of assets of the same asset type that can generate cash income. The term, recovery amount and time point of each asset are different. Therefore, it is necessary to calculate the expected future cash flow recovery of each underlying asset, and then aggregate the cash flow of the asset pool assets according to the collection date of the asset securitization product to obtain the cash flow recovery of each collection period, which can then be used for the cash flow payment of the asset securitization product. The data after the cash flow of the asset pool assets is aggregated can be recorded in the form of a cash flow aggregation table, as shown in Table 1.

[0070] Table 1 Cash flow summary table

[0071]

[0072] Asset securitization products have great uncertainty in generating the continuous and stable cash flow expected by investors due to the possibility of early repayment and default recovery. Currently, only a small number of asset securitization products will publish asset pool cash flow aggregation tables under certain stress scenarios in their offering instructions, and some revolving structure asset securitization products do not even disclose the asset pool cash flow aggregation forecast after considering the continuous purchase of new assets.

[0073] In order to obtain a forecasted cash flow collection that is closer to the actual performance of the asset pool and ultimately used for redemption, in this embodiment, the cash flow of each collection period can be recalculated based on the calibrated pressure parameters, that is, the pressure parameter calibration sequence, to obtain a forecasted cash flow collection table for redemption.

[0074] Furthermore, asset securitization products have a structured feature, and each product shares the collection funds of the same asset pool. Therefore, when determining the cash flow of each product, it is necessary to allocate the collection cash flow of the entire asset pool. In this embodiment, a flow model can be adopted. Based on the cash flow aggregation table (i.e., the forecasted cash flow aggregation table for redemption) in which the cash flow of the underlying asset pool assets is aggregated, and under the premise of considering credit enhancement measures, the cash flow of the asset pool is split into each product of the asset securitization product according to its cash flow payment mechanism to obtain the expected cash flow of the asset securitization product.

[0075] Optionally, based on the forecast cash flow aggregation table, cash flows can be reallocated in combination with the credit enhancement measures adopted by the target sponsor. For example, for asset securitization products with structural designs, the principal and interest of priority securities can be paid first, followed by subordinated securities, thereby ensuring that priority securities are protected from credit losses in the asset pool.

[0076] Through the embodiments provided by the present application, historical data of similar assets corresponding to the target asset pool of the object to be predicted is extracted, wherein the data of similar assets is stored in a designated database, the object to be predicted is an asset securitization product initiated by a target initiator, the historical data of similar assets is historical data of the similar asset pool, the similar asset pool is a collection of similar assets of the target asset in the target asset pool, the asset type of the similar assets is the same as the asset type of the corresponding target asset and does not belong to the target asset pool; the historical data of similar assets is parsed according to preset pressure parameters to obtain a pressure parameter benchmark sequence corresponding to the target asset, wherein the preset pressure parameters are parameters that affect the cash flow of the target asset; based on the target asset and the preset pressure parameters, a pressure parameter benchmark sequence is obtained. The pressure parameter benchmark sequence corresponding to the target asset is calibrated with the true value of the parameter corresponding to the pressure parameter to obtain the pressure parameter calibration sequence corresponding to the target asset, and the forecast cash flow collection table for redemption is obtained by applying the pressure parameter calibration sequence to the target cash flow collection table of the target asset pool; based on the forecast cash flow collection table and the credit enhancement measures adopted by the target initiator, the cash flow of the target asset pool is allocated according to the cash flow payment mechanism of the object to be predicted to obtain the expected cash flow of the object to be predicted, which can solve the problem of low accuracy of data processing results due to the difficulty of data acquisition and complex data processing in the data processing method in the related technology, thereby improving the accuracy and reliability of data processing.

[0077] In an exemplary embodiment, the method further includes: calling a cash flow distribution sequence calculation method corresponding to the object to be predicted in a preset method module library under the account type corresponding to the target initiator to obtain a cash flow payment mechanism of the object to be predicted.

[0078] Since trust accounts include single trust accounts that do not distinguish between interest and principal (hereinafter referred to as single account) and sub-trust accounts that distinguish between interest and principal (hereinafter referred to as sub-accounts), and the repayment order of the cash flow distribution mechanism of different asset securitization products is relatively diverse and may change according to the status of the account, the key to handling cash flow distribution is to flexibly adapt to various repayment orders. In order to enhance adaptability and scalability, in this embodiment, a trust account framework and a module library for calculating the cash flow distribution order can be built, the calculation method of each unit can be agreed upon, and then the formulas of each link can be called in sequence for calculation according to the cash flow distribution order of the asset securitization product itself.

[0079] Correspondingly, for the object to be predicted, the cash flow distribution order calculation method corresponding to the object to be predicted in the preset method module library under the account type corresponding to the target initiator can be called, and the cash flow distribution order corresponding to the object to be predicted can be determined according to the called cash flow distribution order calculation method, so as to obtain the cash flow payment mechanism of the object to be predicted, that is, the cash flow payment mechanism of the object to be predicted indicates the cash flow distribution order corresponding to the object to be predicted. Among them, the account type corresponding to the target initiator may be a single account or a sub-account, and the cash flow distribution order calculation method module library can be configured under the trust account framework, so that the cash flow distribution order of different asset securitization products can be calculated by calling the cash flow distribution order calculation method in the cash flow distribution order calculation method module library.

[0080] For example, Figure 5 As shown, for an asset securitization product, based on its cash flow aggregation table, the corresponding fund account framework can be called according to its trust account type (for example, a single trust account that does not distinguish between interest and principal and a separate trust account that distinguishes between interest and principal, etc.), and the corresponding allocation order calculation method can be called according to its allocation order (for example, it can be allocated first to taxes / fees, or it can be allocated first to income within the limit, etc.) to complete the cash flow allocation calculation.

[0081] Optionally, the above trust account types and allocation order can be obtained when connecting to a designated database and acquiring relevant data of the object to be predicted.

[0082] Through this embodiment, by calling the configured method module library to perform cash flow distribution sequence calculation, the flexibility and reliability of the cash flow distribution sequence calculation can be improved, thereby improving the efficiency of data processing.

[0083] In an exemplary embodiment, the designated database further stores at least one of the following information: asset pool data of the target asset pool; cash flow payment mechanism data for indicating a cash flow payment mechanism; credit enhancement data for indicating credit enhancement measures; historical principal and interest payment data for indicating historical principal and interest payment conditions of the object to be predicted; the data used in the designated database for cash flow prediction of the object to be predicted is the underlying data, and the underlying data includes historical data of similar assets;

[0084] Correspondingly, the above method also includes: using a multi-threaded concurrent streaming query to specify whether the underlying data stored in the database has been updated; if the underlying data has been updated, synchronizing the updated underlying data to the cache of each node among the multiple nodes, wherein the underlying data used for cash flow prediction of the object to be predicted is the underlying data cached on one of the multiple nodes.

[0085] Similar to the aforementioned embodiment, a designated database can be connected and data in the designated database can be obtained, including but not limited to: asset pool data of the target asset pool; cash flow payment mechanism data for indicating the cash flow payment mechanism; and credit enhancement data for indicating credit enhancement measures.

[0086] Optionally, historical principal and interest repayment data indicating the historical principal and interest repayment status of the object to be predicted can be obtained from a designated database, that is, the principal recovery and interest payment status of the object to be predicted in each past period, including the deviation between the actual payment and the prediction. The above historical principal and interest repayment data can be used to provide a reference for the historical performance of the asset securitization product.

[0087] Optionally, the data in the designated database used for cash flow forecasting of the forecast object is underlying data, that is, data collected directly from the original source without any processing, which may include historical data of similar assets, and illustratively, may include the historical performance of similar assets in the asset pool.

[0088] Similar to the above-mentioned embodiment, after the required data is acquired from the specified database, data caching may be performed.

[0089] Optionally, in this embodiment, it is possible to check periodically or on demand whether the underlying data stored in the specified database has been updated, and a multi-threaded concurrent streaming query can be used to determine whether the underlying data stored in the specified database has been updated. That is, multi-threading technology can be used to make each thread responsible for querying a portion of the data, thereby processing multiple queries concurrently to improve the speed and efficiency of the query. At the same time, the data stream can be processed in a continuous manner instead of loading all data into the memory at one time, thereby avoiding memory overflow problems, and data updates in the database can be received and processed in real time.

[0090] Optionally, once it is detected that the underlying data is updated, data synchronization may be performed, that is, the updated underlying data is synchronized to the cache to ensure consistency between the database and the cache data during calculation.

[0091] Optionally, distributed caching technology can be used to achieve multi-node full data copy, that is, all data in the database is saved in a complete copy on each of the multiple nodes, so that data reading can be localized during data processing, without the need to read data across nodes, making it easier to process large amounts of data and cash flow integration logic.

[0092] Optionally, when synchronizing the underlying data, version control can be performed through serial numbers, that is, a serial number can be set for a specified database. When the underlying data in the specified database is updated, its serial number changes. Correspondingly, the serial number of the data cache will change as the serial number of the underlying database changes. By ensuring that the serial numbers of the specified database and the data cache are consistent, and updating and synchronizing the data when an inconsistency in the serial numbers is detected, it can be ensured that the cached data on each node is up to date and consistent, thereby preventing data conflicts.

[0093] Optionally, data synchronization can be triggered through a scheduling system, that is, when the underlying data in a specified database is updated, a data synchronization request can be initiated through an external scheduling system to synchronize the updated underlying data to the cache of each node, thereby reducing the access pressure on the database and improving the efficiency of data processing.

[0094] Through this embodiment, by acquiring the underlying data in the specified database and updating it in real time, the accuracy of the extracted data can be improved, thereby improving the accuracy of the cash flow forecast.

[0095] In an exemplary embodiment, the data used by the designated database for cash flow forecasting of the forecast object is the underlying data, and the underlying data is synchronized to the cache of each node among the multiple nodes; the underlying data includes historical data of the same type of assets; the amount of memory data on each node is less than or equal to a threshold of the amount of memory data corresponding to each node;

[0096] Correspondingly, the above method also includes: in response to the received target prediction request, according to the specified load balancing strategy, the target computing task corresponding to the object to be predicted is allocated to the target node among the multiple nodes, so as to execute the target computing task on the target node based on the underlying data cached on the target node, wherein the target prediction request is used to request cash flow prediction for the object to be predicted, the specified load balancing strategy is a strategy for balancing the computing tasks allocated to multiple nodes, the target computing task is a computing task for cash flow prediction of the object to be predicted, and on the target node, the target computing task is executed in a multi-threaded computing manner.

[0097] Similar to the above-mentioned embodiment, the underlying data of the object to be predicted in the specified database can be extracted, and the distributed cache technology can be used to completely synchronize the above underlying data to the cache of each node in multiple nodes, and the underlying data includes historical data of similar assets.

[0098] Here, since the memory resources of each node are limited, in order to prevent memory overflow problems, when synchronizing the underlying data cache, memory data volume threshold control can be performed. That is, the memory data volume threshold corresponding to each node can be obtained to ensure that the amount of data loaded into each node cache does not exceed this threshold, which can effectively solve the problem of business data jitter.

[0099] Optionally, during the data reading process, if a single node is found to have read too much data, the data reading of the node can be suspended. After other nodes have finished reading and processing the data, the data of another node can be cleared and the unread data of the node can be continued to be read through another node whose data has been cleared.

[0100] Optionally, the underlying data may be partitioned or sharded or compressed to ensure that the amount of data memory is within a safe range, thereby ensuring the stability and efficiency of data processing.

[0101] Correspondingly, in this embodiment, in response to a received target prediction request, the target computing task corresponding to the object to be predicted can be allocated to a target node among multiple nodes according to a specified load balancing strategy, so that the target computing task can be executed on the target node based on the underlying data cached on the target node, wherein the target prediction request is used to request a cash flow prediction for the object to be predicted.

[0102] Here, the specified load balancing strategy is a strategy for balancing the computing tasks assigned to multiple nodes. It can evenly distribute computing tasks to multiple computing nodes to prevent specific nodes from being overloaded, thereby improving the overall data processing capability and response speed.

[0103] Optionally, computing tasks can be allocated based on the computing power of each node among multiple nodes (including the configuration of the node, current load conditions, etc.), or based on the complexity of the computing tasks to be executed, they can be allocated to the nodes that are most suitable for processing this type of tasks. Tasks can also be preferentially allocated to nodes that have cached relevant underlying data, which is not limited in this embodiment.

[0104] Optionally, a specified load balancing strategy can be pre-set and stored, so that data can be quickly distributed when needed, improving the speed and accuracy of data processing.

[0105] Optionally, in order to accelerate the calculation, the cache performance of the distributed cache can be further explored, and the balance between the total amount of table data and the cache load can be obtained to obtain the data reading mode with the shortest time consumption, that is, the specified data reading mode. The specified data reading mode can be used to balance the total amount of table data and the cache load. It can be selected from a group of data reading modes through experiments, and the selection basis can be the total amount of table data and the cache load. Correspondingly, the forecast cash flow aggregation table can be cached according to the specified data reading mode.

[0106] Optionally, on the target node, a multi-threaded computing method can be used to execute the target computing task, that is, the computing task can be decomposed into multiple sub-tasks that can be processed in parallel, and the multiple sub-tasks can be assigned to different threads for simultaneous processing, thereby reducing the computing time and improving the efficiency of data processing. For example, when calculating the expected cash flow of assets in the asset pool in various future time periods, the computing task can be decomposed into calculations for each asset, and the calculations for different assets can be processed in parallel.

[0107] Optionally, in multi-threaded computing, computing tasks can be dynamically assigned to different threads according to the state of each thread to ensure load balancing of each thread. This can be achieved through a pre-set scheduling algorithm, and the scheduling algorithm is not specifically limited in this embodiment.

[0108] Through this embodiment, by allocating and managing computing tasks on multiple nodes according to a specified load balancing strategy, the computing efficiency and resource utilization of data processing can be improved.

[0109] In an exemplary embodiment, the method further includes: if there is an asset set of similar assets initiated by the target initiator, determining the asset set of similar assets initiated by the target initiator as a similar asset pool; if there is no similar asset initiated by the target initiator, determining the asset set of similar assets initiated by other initiators except the target initiator as a similar asset pool;

[0110] Correspondingly, after extracting historical data of similar assets corresponding to the target asset pool of the object to be predicted, the above method also includes: based on the cross-reference relationship between fields, standardizing the historical data of similar assets according to a set of specified fields to obtain standardized historical data of similar assets.

[0111] In this embodiment, the similar asset pool of the asset securitization product can be a collection of other assets initiated by the sponsor of the asset securitization product and of the same type as the underlying assets of the asset pool of the asset securitization product, but not the asset pool. That is, the similar asset pool of the asset securitization product can be a collection of other assets that have the same sponsor as the asset securitization product and of the same type as the underlying assets of the asset pool of the asset securitization product, but not the asset pool.

[0112] Optionally, if the target initiator does not initiate similar assets, similar assets of other initiators can be determined as a similar asset pool, that is, in the absence of similar assets initiated by the target initiator, the asset collection of similar assets initiated by other initiators except the target initiator can be determined as a similar asset pool.

[0113] Here, due to the irregularities in the data disclosure of historical data of similar assets from different initiators, including different disclosure calibers (i.e., different statistical methods or scopes of data, different initiators may use different standards or methods to report the same type of data. For example, for the default rate, the initiator may count the proportion of loans overdue for 90 days, or it may be the proportion overdue for 30 days. This leads to different actual meanings and application scopes of the same indicator in different data sets), different meanings of disclosed fields (i.e., the specific content of each column of data in the data set may be different, and fields with the same name represent different meanings in different data sets), etc., calculation and use are very difficult. Therefore, based on the cross-reference relationship between fields, the historical data of similar assets can be standardized according to a set of specified fields to obtain standardized historical data of similar assets.

[0114] Here, the cross-reference relationship between fields refers to the logical relationship that should be satisfied between multiple fields in the data. By checking this logical relationship, the integrity and accuracy of the data can be verified. If the cross-reference relationship between the fields does not hold, it can be determined that there is a problem in the data.

[0115] Optionally, the field names or meanings used by different initiators can be mapped to a unified standard field name according to a set of specified fields to ensure data consistency. Here, it can be pre-set that a set of specified fields can be represented in the form of a table, that is, the standard meanings and mapping relationships of the fields can be specified in the form of a table. Then, the above-mentioned acquired data can be identified, and if it meets the provisions of the standard meaning of the field, it can be stored in the corresponding position of the specified table.

[0116] Optionally, other standardization processing may be performed on the data. For example, for the same field, the acquired data may include its total value and its value for the same period, which can be uniformly converted into the value for the same period.

[0117] In addition, when the historical data of similar assets corresponding to the target asset pool to be predicted is incomplete, the statistical results of relevant fields of similar assets can be combined to give default values ​​or receive corrected values ​​input manually.

[0118] Optionally, after extracting historical data of similar assets corresponding to the target asset pool of the object to be predicted, the data can also be cleaned, for example, data that does not meet the calculation conditions can be eliminated, that is, due to special circumstances such as the small sample size of some data being unable to perform statistical calculations or the presence of outliers, these small sample combinations and outlier data can be eliminated.

[0119] Optionally, after extracting historical data of similar assets corresponding to the target asset pool to be predicted, the completeness of the data can also be verified. For example, a list of necessary data fields that affect the allocation and calculation of cash flow can be pre-defined, and the data extracted from the specified database can be tested. In the event that it is detected that a field in the above data field list is missing, the missing field information can be identified and the missing data field can be fed back through a feedback mechanism. Thus, the data can be updated and supplemented, and the supplemented data fields can be added to the existing asset pool data to ensure the integrity of all data fields necessary for data processing.

[0120] Through this embodiment, by performing standardization processing on the extracted data to obtain standardized historical data of assets of the same type, data standardization can be achieved and the accuracy and reliability of data processing can be improved.

[0121] In an exemplary embodiment, the same type of asset pool for asset securitization products may include a static structure asset pool or a revolving structure asset pool, wherein the static structure asset pool (hereinafter referred to as the static pool) may mean that once the asset pool is established, the assets in the asset pool will not be supplemented, and the revolving structure asset pool (hereinafter referred to as the dynamic pool) may mean that after the assets in the asset pool expire, assets of the same asset type will continue to be purchased to supplement the asset pool. Correspondingly, the historical data of similar assets include the historical data of the static structure asset pool and the historical data of the revolving structure asset pool (hereinafter referred to as the dynamic and static pool).

[0122] In this embodiment, the actual performance of similar assets (including early repayments - early repayment rate, default recovery - default rate and recovery rate, etc.) can be counted as the benchmark value of the stress parameter, and then calibrated according to the current actual performance of the asset securitization product. The purpose of introducing similar assets is to use their data to calculate the benchmark sequence of asset pool stress parameters of asset securitization products. There are three main reasons for doing this:

[0123] First, the stress parameters such as early repayment and default of the same originator and the same type of assets are homogeneous, that is, the stress parameters of the dynamic and static pools and the asset pools of asset securitization products are homogeneous;

[0124] Second, the dynamic and static pools contain more samples and richer historical information;

[0125] Third, the life cycle of static pool samples is generally longer and can be used to predict the future trend of asset pool stress parameters.

[0126] Similar to the aforementioned embodiment, in the case where there is an asset collection of similar assets initiated by a target initiator, the asset collection of similar assets initiated by the target initiator can be determined as a similar asset pool, and the actual historical performance of the similar asset pool can be used as its stress parameter; in the case where there is no similar asset initiated by a target initiator, the actual historical performance of the asset pool of asset securitization products of the same asset type can be statistically analyzed as its stress parameter.

[0127] As an optional implementation, in the case where the same type of asset pool includes a static structure asset pool, and the historical data of the same type of assets includes the historical data of the static structure asset pool, the historical data of the static structure asset pool can be parsed according to the first pressure parameter to obtain a first parameter benchmark sequence corresponding to the target asset. Parsing the historical data of the same type of assets according to the preset pressure parameter to obtain a pressure parameter benchmark sequence corresponding to the target asset includes: parsing the historical data of the static structure asset pool according to the first pressure parameter to obtain a first parameter benchmark sequence corresponding to the target asset, and the pressure parameter benchmark sequence corresponding to the target asset includes the first parameter benchmark sequence. Among them, the first pressure parameter includes the early repayment rate, the default rate and the recovery rate, and the first pressure parameter can be expressed as a time series.

[0128] As another optional implementation, in the case where the same type of asset pool includes a cyclic structure asset pool, and the historical data of the same type of assets includes the historical data of the cyclic structure asset pool, the historical data of the cyclic structure asset pool can be parsed according to the second pressure parameter to obtain a second parameter benchmark sequence corresponding to the target asset. Parsing the historical data of the same type of assets according to the preset pressure parameter to obtain a pressure parameter benchmark sequence corresponding to the target asset includes: parsing the historical data of the cyclic structure asset pool according to the second pressure parameter to obtain a second parameter benchmark sequence corresponding to the target asset, and the pressure parameter benchmark sequence corresponding to the target asset includes the second parameter benchmark sequence. Among them, the second pressure parameters include repayment rate, default rate, recovery rate, yield rate and purchase rate, and the second pressure parameters can all be expressed as a single value.

[0129] Optionally, the pressure parameters of the static structure asset pool and the pressure parameters of the cyclic structure asset pool may be calculated separately, and the algorithms thereof may be pre-set and directly called when calculating the corresponding pressure parameters.

[0130] For example, for the above pressure parameters, the following calculation algorithm may be used:

[0131] (1) Prepayment rate. Prepayment rate can usually be expressed by using the monthly prepayment ratio (SMM) and the conditional prepayment ratio (CPR), as shown in formulas (1) and (2), where CPR can be used to represent the prepayment rate:

[0132] (1)

[0133] (2)

[0134] (2) Default rate: Loans that are overdue for more than 90 days can be considered as default loans, as shown in formula (3):

[0135] (3)

[0136] (3) Recovery rate: If the underlying loans in the asset pool can be calculated as collateral value (such as residential mortgage loans, i.e., RMBS), the recovery rate can be calculated using the collateral value, as shown in formula (4):

[0137] (4)

[0138] If the value of the collateral cannot be calculated (for example, consumer loans, auto loans, and small and micro enterprise loans), the recovery rate is calculated based on the historical performance of the static pool, as shown in formulas (5) and (6):

[0139] (5)

[0140] (6)

[0141] The recovery rate discount coefficient of the collateral can be determined by the following methods: (1) calculation, which is to calculate the recovery rate discount coefficient in groups according to the characteristics of the collateral; (2) verification and adjustment, which is to verify and adjust the recovery rate discount coefficient with reference to actual recovery.

[0142] (4) Repayment rate. The repayment rate is calculated as shown in formula (7):

[0143] (7)

[0144] (5) Rate of return. The rate of return is calculated as shown in formula (8):

[0145] (8)

[0146] (6) Purchase rate. The purchase rate is calculated as shown in formula (9):

[0147] (9)

[0148] Through this embodiment, the accuracy and effectiveness of data processing can be improved by parsing historical data of similar assets to obtain a corresponding pressure parameter benchmark sequence.

[0149] In an exemplary embodiment, a pressure parameter baseline sequence corresponding to the target asset is calibrated based on the true values ​​of the parameters corresponding to the target asset and the preset pressure parameters to obtain a pressure parameter calibration sequence corresponding to the target asset, including: based on the difference between the true values ​​of the parameters corresponding to the target asset and the preset pressure parameters and the parameter values ​​of the same pressure parameters in the pressure parameter baseline sequence corresponding to the target asset, a translation adjustment is performed on the pressure parameter baseline sequence corresponding to the target asset to obtain a pressure parameter calibration sequence corresponding to the target asset.

[0150] Similar to the above-mentioned embodiment, due to the influence of various factors, the actual performance of the asset pool may be different from the history, so it is necessary to calibrate the pressure parameter according to the actual performance to obtain the pressure parameter calibration sequence. The basic logic of the calibration method is: using the deviation between the actual value of the parameter and the benchmark sequence, a translation adjustment is made to the benchmark sequence. For example, the benchmark sequence curve can be compared with the actual value curve of the parameter. For the non-overlapping parts, the benchmark sequence is translated to make it infinitely close to the actual value of the parameter.

[0151] Here, the stress parameter benchmark sequence may have 100 values, while the stress parameter actual value may only have 20 values. Therefore, the actual parameter value cannot be used directly. Instead, the benchmark sequence should be made close to the actual parameter value to obtain predictions for the remaining values. Thus, the calibrated stress parameter calibration sequence can reflect the impact of the actual value of the asset pool on the one hand, and retain the trend of the benchmark sequence on the other hand.

[0152] For example, the main ideas for calculating and calibrating stress parameters of asset securitization products are as follows: Figure 6 As shown, the static structure asset pool and the cyclic structure asset pool follow the same idea, but need to be processed separately, which is divided into three steps as follows:

[0153] Step 1: Calculation of stress parameter benchmarks. First, perform data processing on the dynamic and static pools (including but not limited to: determining whether there are matching dynamic and static pools, data standardization, etc.). Then, calculate the stress parameters of the dynamic and static pools and the average stress parameters of the dynamic and static pools of the same underlying asset type. If the asset securitization product has a matching dynamic and static pool, then the stress parameter benchmark of its ABS (Asset-Backed Securities) is the dynamic and static pool parameters. If there is no matching dynamic and static pool, then the stress parameter benchmark of its ABS is the average parameters of the dynamic and static pools of the same underlying asset type.

[0154] Step 2: Calculation of actual values ​​of ABS parameters can generate actual values ​​based on the actual performance of the asset pool of the asset securitization product and the detailed data of the underlying asset pool (trustee data parameters). For example, it can be determined whether there are new defaults, early repayments, etc. in the asset securitization product based on the actual performance of the asset pool disclosed in the "Trust Trustee Institution Report".

[0155] Step three, match the benchmark sequence of stress parameters of asset securitization products with the actual values ​​according to the asset securitization product asset pool, calibrate the stress parameters, and finally output them. Here, if there are periods, corresponding matching can be performed according to the periods.

[0156] Through this embodiment, by calibrating the acquired pressure parameter reference sequence according to actual values, the accuracy and reliability of data processing can be improved.

[0157] In an exemplary embodiment, before obtaining a forecasted cash flow aggregation table for redemption by applying the pressure parameter calibration sequence to the cash flow aggregation table of the target asset pool, the method further includes: calculating the expected cash flow recovery information of each target asset based on the asset pool data of the target asset pool, wherein the expected cash flow recovery information of each target asset is used to indicate the expected future cash flow recovery of each target asset; aggregating each target asset in the target asset pool according to each aggregation date of the object to be predicted, and generating a first candidate cash flow aggregation table based on the cash flow recovery information obtained on each aggregation date, wherein the cash flow recovery information on each aggregation date is used to indicate the cash flow recovery on each aggregation date; determining the cash flow aggregation table that best matches the latest balance of the target asset pool between the first candidate cash flow aggregation table and the second candidate cash flow aggregation table as the target cash flow aggregation table, wherein the second candidate cash flow aggregation table is a cash flow aggregation table integrated from the cash flow aggregation table of the target asset pool configured for the object to be predicted and the actual performance of the target asset pool that is statistically analyzed.

[0158] In this embodiment, the expected cash flow recovery information of each target asset can be calculated based on the asset pool data of the target asset pool, and each target asset in the target asset pool can be aggregated according to each aggregation date of the object to be predicted, and the first candidate cash flow aggregation table can be generated based on the cash flow recovery information obtained on each aggregation date.

[0159] For example, Figure 7 As shown, it can be divided into two steps:

[0160] The first step is to calculate the future repayment plan for each asset in the asset pool:

[0161] Based on the basic loan data such as the underlying asset type, loan repayment method, loan repayment frequency, loan issuance date, loan maturity date, and the current loan status data such as the remaining principal, current execution rate, current overdue days, and overdue principal of each underlying asset in the asset pool, the future loan repayment plan (which may include the repayment of principal and interest in each period) is calculated.

[0162] Different asset securitization products may have different loan repayment methods. Similar to the above-mentioned embodiment, the loan repayment method and the corresponding algorithm corresponding to the asset securitization product may be obtained from a designated database.

[0163] Exemplarily, the loan repayment method algorithm may include:

[0164] (1) Equal principal method: the principal repayment amount for each period is as shown in formula (10):

[0165] (10)

[0166] (2) Lump sum repayment method: the principal repayment amount for each period is as shown in formula (11):

[0167] (11)

[0168] (3) Periodic interest payment method: the amount of interest paid each period is as shown in formula (12):

[0169] (12)

[0170] (4) Lump sum payment method: the amount of interest paid each period is as shown in formula (13):

[0171] (13)

[0172] (5) Equal principal and interest method: the principal and interest payment amount for each period is as shown in formula (14):

[0173] (14)

[0174] The second step is to aggregate cash flows according to the aggregation date:

[0175] According to the collection date of the asset securitization product, the future repayment cash flow of each underlying asset obtained in step 1 is divided into each cash flow collection period, thereby obtaining the basic scenario asset pool cash flow collection table of the asset pool (i.e., the first candidate cash flow collection table).

[0176] Optionally, after obtaining the first candidate cash flow aggregation table, in order to improve the accuracy of the prediction and be closer to the actual cash flow performance of the asset pool, information from other sources can be comprehensively considered. For example, the asset pool cash flow aggregation table disclosed in the official documents of the object to be predicted (i.e., the asset securitization product initiated by the target initiator) at the time of issuance and the latest asset pool cash flow aggregation table disclosed in the above-mentioned "Trust Trustee Report" can be integrated to obtain the second candidate cash flow aggregation table, and compared with the above-mentioned first candidate cash flow aggregation table, the cash flow aggregation table that best matches the latest balance of the target asset pool is used as the target cash flow aggregation table, and the cash flow aggregation table that is closer to the actual situation of the asset pool is obtained as the target cash flow aggregation table.

[0177] For example, Figure 8 As shown, it can be divided into three steps:

[0178] Step 1: Update the disclosed cash flow aggregation table: Based on the asset pool cash flow aggregation table disclosed in the "Issuance Prospectus", update it according to the actual asset pool recovery disclosed in the "Trust Trustee Institution Report" and the updated asset pool cash flow recovery forecast table to obtain the latest disclosed cash flow aggregation table after integration (i.e., the second candidate cash flow aggregation table);

[0179] Step 2: Determine the latest cash flow aggregation table: Compare the latest disclosed cash flow aggregation table after integration calculated in step 1 (i.e., the second candidate cash flow aggregation table) with the above basic scenario cash flow aggregation table (i.e., the first candidate cash flow aggregation table), and determine the cash flow aggregation table that is closer to the actual situation of the asset pool according to the latest balance of the asset pool disclosed in the Trust Trustee Institution Report, i.e., the target cash flow aggregation table;

[0180] Step 3: Apply pressure to the cash flow aggregation table obtained in step 2, which is closer to the actual situation of the asset pool. That is, apply different pressure parameters to the asset pool to simulate the pressure scenarios that the asset securitization product may encounter, and obtain a cash flow aggregation table that takes pressure parameters into account, that is, a predicted cash flow aggregation table for redemption.

[0181] Through this embodiment, the expected cash flow recovery information is calculated based on the asset pool data and aggregated to generate a target cash flow aggregation table, which can improve the stability and reliability of data processing.

[0182] In an exemplary embodiment, the above method also includes: determining the information to be displayed of the object to be predicted based on the historical principal and interest payment data of the object to be predicted and the expected cash flow of the object to be predicted, and distributing the information to be displayed to the display interface for display, wherein the information to be displayed is used to describe the principal and interest payment cash flow of the object to be predicted during the entire life cycle.

[0183] In this embodiment, the information to be displayed of the object to be predicted can be determined based on the historical principal and interest payment data of the object to be predicted and the expected cash flow of the object to be predicted, wherein the information to be displayed is used to describe the principal and interest payment cash flow during the entire life cycle of the object to be predicted, that is, the predicted future redemption cash flow can be combined with the historical principal repayment cash flow to obtain a complete redemption cash flow table for each level of asset securitization products.

[0184] Here, the expected cash flow of the object to be predicted can be expressed in the form of a cash flow aggregation table, that is, it can be the above-mentioned predicted cash flow aggregation table for redemption.

[0185] Here, based on the historical principal and interest payment data of the object to be predicted, the historical principal repayment cash flow can be obtained, and can also be expressed in the form of a historical principal repayment cash flow table.

[0186] Optionally, the historical principal repayment data and the expected cash flow data can be aligned in time to form a complete cash flow time series. Thus, by integrating the historical principal repayment data of the object to be predicted with the calculated expected cash flow data, the principal repayment cash flow over the entire life cycle of the object to be predicted can be obtained, including but not limited to: principal recovery in each period, interest income and possible default losses, etc.

[0187] In order to facilitate understanding and viewing and provide a more comprehensive and transparent cash flow forecast information, the principal and interest payment cash flow during the complete life cycle of the above-mentioned object to be forecasted can be converted into information to be displayed, that is, visual data, such as charts, tables, graphs, etc.

[0188] Optionally, the information to be displayed can be distributed to a display interface for display, so that users can clearly view the cash flow performance of the asset securitization product throughout its life cycle through the display interface, which helps users make decisions.

[0189] Optionally, the display interface may be a display interface set up for the object to be predicted, which may be an interface supporting user interaction or may be used only for display, which is not limited in this embodiment.

[0190] Optionally, corresponding to the dynamic adjustment of the object to be predicted over time, the information to be displayed may also be dynamically updated.

[0191] Through this embodiment, historical principal and interest payment data is combined with predicted cash flow to generate information to be displayed and distributed to the display interface for display, which can improve the transparency of data processing.

[0192] The data processing method in the embodiment of the present application is explained below in conjunction with an optional example. The data processing method in this optional example can be applied to Fig. 9 The technical framework shown, such as Fig. 9 As shown, the technical framework includes:

[0193] Data service, used to implement data scheduling and data synchronization;

[0194] Interface service, used to provide interface call service to external scheduling system or other systems;

[0195] Computing services, used for internal system computing;

[0196] Cache service, used for data caching;

[0197] Distributed caching technology can be used to load the underlying database data into memory for calculation;

[0198] Basic framework: Provides distributed computing, can realize the encapsulation of the overall business logic, shield the internal logic according to the job, and realize the automation of computing. Among them, the framework support module can use Springboot for the overall framework technology support, the data synchronization module can use Mybatis-plus to connect to the database during data synchronization, and the external interface call module can use feign for external interface calls;

[0199] Database: used to provide detailed data on asset securitization products;

[0200] Persistence layer data: used for underlying data storage;

[0201] Data adaptation: Synchronize the underlying data (i.e., data in the persistence layer) to the distributed cache through API (Application Programming Interface) calls;

[0202] Distributed cache: used to load the underlying database data into memory for calculation;

[0203] Distributed computing engine: corresponds to the distributed cache and is used to perform corresponding data calculations;

[0204] Atomic formula: all data calculation formulas in this application (for example, data calculation formulas for recovery rate, early repayment rate, etc.), which are used to implement business logic based on the distributed computing engine;

[0205] Indicator calculation: refers to key indicators that reflect the performance of asset securitization products and asset pools, including but not limited to early repayment rate, recovery rate, etc.;

[0206] Business ETL (Extract, Transform, Load): data processing process used to extract data and perform necessary data preprocessing;

[0207] Business encapsulation: used to encapsulate data, including pressure parameters, aggregated cash flow, split cash flow, and cash flow by transaction;

[0208] Interface layer: used to provide interface call services, including RESTful (representational state transfer), Java SDK (Java software development kit), and Python SDK (Python software development kit).

[0209] Correspondingly, if Fig.10 As shown, the data processing system provided in the embodiment of the present application may include the following three modules: a data access and processing module, a model development module, and a data distribution module, wherein the processing flow of the data access and processing module can be divided into four stages: connecting to the database to obtain data in the database, data aggregation, data processing and data cleaning, and data storage; the processing flow of the model development module can be divided into four stages: statistics of asset pool stress scenario parameters, basic scenario asset pool cash flow forecast, asset pool cash flow calculation considering stress scenarios, and allocation of asset pool cash flow to asset securitization products; the processing flow of the data distribution module can be divided into three stages: generating future cash flow forecast values, converting them into complete redemption cash flows, and data distribution.

[0210] Through this optional example, after obtaining the required data, the data is parsed according to the preset pressure parameters to obtain the pressure parameter reference sequence and calibrate it, and finally the calibrated pressure parameter calibration sequence is applied to the corresponding data to obtain the final predicted data, which can improve the accuracy and reliability of data processing.

[0211] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0212] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM (Read-Only Memory, read-only memory) / RAM (Random Access Memory, random access memory), a disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0213] According to another aspect of the embodiment of the present application, a data processing device is also provided, which can be used to implement the data processing method provided in the above-mentioned embodiment, which has been described and will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware is also possible and conceived.

[0214] Fig.11 is a structural block diagram of an optional data processing device according to an embodiment of the present application, such as Fig.11 As shown in , the data processing device includes:

[0215] The extraction unit 1102 is used to extract the historical data of the same type of assets corresponding to the target asset pool of the object to be predicted, wherein the data of the same type of assets is stored in a designated database, the object to be predicted is an asset securitization product initiated by the target initiator, the historical data of the same type of assets is the historical data of the same type of asset pool, the same type of asset pool is a collection of the same type of assets of the target asset in the target asset pool, and the asset type of the same type of assets is the same as the asset type of the corresponding target asset and does not belong to the target asset pool;

[0216] The parsing unit 1104 is used to parse the historical data of the same type of assets according to the preset pressure parameters to obtain a pressure parameter benchmark sequence corresponding to the target asset, wherein the preset pressure parameters are parameters that affect the cash flow of the target asset.

[0217] The first execution unit 1106 is used to calibrate the pressure parameter reference sequence corresponding to the target asset based on the actual value of the parameter corresponding to the target asset and the preset pressure parameter, to obtain the pressure parameter calibration sequence corresponding to the target asset, and to obtain the predicted cash flow collection table for redemption by applying the pressure parameter calibration sequence to the target cash flow collection table of the target asset pool.

[0218] The first allocation unit 1108 is used to allocate the cash flow of the target asset pool according to the cash flow payment mechanism of the object to be predicted based on the predicted cash flow aggregation table and the credit enhancement measures adopted by the target initiator to obtain the expected cash flow of the object to be predicted.

[0219] It should be noted that the extraction unit 1102 in this embodiment can be used to execute the above step S202, the parsing unit 1104 in this embodiment can be used to execute the above step S204, the first execution unit 1106 in this embodiment can be used to execute the above step S206, and the first allocation unit 1108 in this embodiment can be used to execute the above step S208.

[0220] Through the embodiments provided by the present application, historical data of similar assets corresponding to the target asset pool of the object to be predicted are extracted, wherein the data of similar assets are stored in a designated database, the object to be predicted is an asset securitization product initiated by a target initiator, the historical data of similar assets are historical data of the similar asset pool, the similar asset pool is a collection of similar assets of the target assets in the target asset pool, the asset type of the similar assets is the same as the asset type of the corresponding target asset and does not belong to the target asset pool; the historical data of similar assets are parsed according to preset pressure parameters to obtain a pressure parameter benchmark sequence corresponding to the target asset, wherein the preset pressure parameters are parameters that affect the cash flow of the target asset; based on the target asset and the preset pressure parameters, a pressure parameter benchmark sequence is obtained. The true values ​​of the parameters corresponding to the parameters are used to calibrate the stress parameter benchmark sequence corresponding to the target assets to obtain the stress parameter calibration sequence corresponding to the target assets, and the stress parameter calibration sequence is applied to the target cash flow aggregation table of the target asset pool to obtain the forecast cash flow aggregation table for redemption; based on the forecast cash flow aggregation table and the credit enhancement measures adopted by the target initiator, the cash flow of the target asset pool is allocated according to the cash flow payment mechanism of the object to be predicted to obtain the expected cash flow of the object to be predicted, which solves the technical problem of low accuracy of data processing results due to the difficulty of data acquisition and complex data processing in the data processing method in the related technology, and improves the accuracy and reliability of data processing.

[0221] In an exemplary embodiment, the designated database also stores at least one of the following information: asset pool data of the target asset pool; cash flow payment mechanism data for indicating the cash flow payment mechanism; credit enhancement data for indicating credit enhancement measures; historical principal and interest payment data for indicating the historical principal and interest payment of the object to be predicted; the data in the designated database used for cash flow prediction of the object to be predicted is the underlying data, and the underlying data includes historical data of similar assets.

[0222] In this embodiment, the above-mentioned device also includes: a query unit, which is used to use a multi-threaded concurrent streaming query to specify whether the underlying data stored in the database has been updated; a synchronization unit, which is used to synchronize the updated underlying data to the cache of each node among the multiple nodes when the underlying data is updated, wherein the underlying data used for cash flow prediction of the object to be predicted is the underlying data cached on one of the multiple nodes.

[0223] In an exemplary embodiment, the data used by the designated database for cash flow forecasting of the forecast object is the underlying data, and the underlying data is synchronized to the cache of each node among multiple nodes; the underlying data includes historical data of similar assets; the amount of memory data on each node is less than or equal to the memory data amount threshold corresponding to each node.

[0224] In this embodiment, the above-mentioned device also includes: a second allocation unit, which is used to respond to the received target prediction request and allocate the target computing task corresponding to the object to be predicted to a target node among multiple nodes according to the specified load balancing strategy, so as to execute the target computing task on the target node based on the underlying data cached on the target node, wherein the target prediction request is used to request cash flow prediction for the object to be predicted, the specified load balancing strategy is a strategy for balancing the computing tasks allocated to multiple nodes, the target computing task is a computing task for cash flow prediction of the object to be predicted, and the target computing task is executed on the target node using multi-threaded computing.

[0225] In an exemplary embodiment, the above-mentioned device also includes: a determination unit, which is used to determine the asset set of similar assets initiated by the target initiator as a similar asset pool when there is an asset set of similar assets initiated by the target initiator; when there is no similar asset initiated by the target initiator, determine the asset set of similar assets initiated by other initiators except the target initiator as the similar asset pool; a second execution unit, which is used to, after extracting the historical data of similar assets corresponding to the target asset pool of the object to be predicted, standardize the historical data of similar assets according to a set of specified fields based on the cross-reference relationship between the fields to obtain standardized historical data of similar assets.

[0226] In an exemplary embodiment, the parsing unit includes one of the following: a first parsing module, which is used to parse the historical data of the static structure asset pool according to a first stress parameter to obtain a first parameter benchmark sequence corresponding to the target asset when the same type of asset pool includes a static structure asset pool and the historical data of the same type of assets includes the historical data of the static structure asset pool, wherein the first stress parameter includes the early repayment rate, the default rate and the recovery rate, and the stress parameter benchmark sequence corresponding to the target asset includes the first parameter benchmark sequence; a second parsing module, which is used to parse the historical data of the cyclical structure asset pool according to a second stress parameter to obtain a second parameter benchmark sequence corresponding to the target asset when the same type of asset pool includes a cyclical structure asset pool and the historical data of the same type of assets includes the historical data of the cyclical structure asset pool, wherein the second stress parameter includes the repayment rate, the default rate, the recovery rate, the yield and the purchase rate, and the stress parameter benchmark sequence corresponding to the target asset includes the second parameter benchmark sequence.

[0227] In an exemplary embodiment, the first execution unit includes: a translation module, which is used to translate and adjust the pressure parameter baseline sequence corresponding to the target asset based on the difference between the actual value of the parameter corresponding to the target asset and the preset pressure parameter and the parameter value of the same pressure parameter in the pressure parameter baseline sequence corresponding to the target asset, so as to obtain a pressure parameter calibration sequence corresponding to the target asset.

[0228] In an exemplary embodiment, the above-mentioned device also includes: a calculation unit, which is used to calculate the expected cash flow recovery information of each target asset according to the asset pool data of the target asset pool before obtaining the predicted cash flow collection table for redemption by applying the pressure parameter calibration sequence to the cash flow collection table of the target asset pool, wherein the expected cash flow recovery information of each target asset is used to indicate the expected future cash flow recovery of each target asset; a collection unit, which is used to collect each target asset in the target asset pool according to each collection date of the object to be predicted, and generate a first candidate cash flow collection table based on the cash flow recovery information obtained on each collection date, wherein the cash flow recovery information on each collection date is used to indicate the cash flow recovery on each collection date; a determination unit, which is used to determine the cash flow collection table that best matches the latest balance of the target asset pool from the first candidate cash flow collection table and the second candidate cash flow collection table as the target cash flow collection table, wherein the second candidate cash flow collection table is a cash flow collection table integrated by the cash flow collection table of the target asset pool configured for the object to be predicted and the actual performance of the target asset pool statistically.

[0229] In an exemplary embodiment, the forecast cash flow aggregation table is cached according to a specified data reading mode, and the specified data reading mode is used to balance the total amount of table data and the cache load. The above device also includes: a calling unit, which is used to call the cash flow distribution sequence calculation method corresponding to the object to be predicted in the preset method module library under the account type corresponding to the target initiator, and obtain the cash flow payment mechanism of the object to be predicted.

[0230] In an exemplary embodiment, the above-mentioned device also includes: a third execution unit, which is used to determine the information to be displayed of the object to be predicted based on the historical principal and interest payment data of the object to be predicted and the expected cash flow of the object to be predicted, and distribute the information to be displayed to the display interface for display, wherein the information to be displayed is used to describe the principal and interest payment cash flow of the object to be predicted during the entire life cycle.

[0231] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0232] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein the steps of any of the above method embodiments are executed when the program is run.

[0233] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, a ROM, a RAM, a mobile hard disk, a magnetic disk, or an optical disk.

[0234] According to another aspect of the embodiments of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to execute the steps in any of the above method embodiments through the computer program. In an exemplary embodiment, the electronic device may further comprise a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0235] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0236] According to another aspect of the embodiment of the present application, a computer program product is also provided, which includes a computer program / instruction, and the computer program / instruction contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1209, and / or installed from the removable medium 1211. When the computer program is executed by the central processing unit 1201, various functions provided by the embodiment of the present application are executed. The above-mentioned serial numbers of the embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments.

[0237] Fig.12 The computer system structure block diagram of the electronic device used to implement the embodiment of the present application is schematically shown. Fig.12 As shown, the computer system 1200 includes a CPU (Central Processing Unit) 1201, which can perform various appropriate actions and processes according to the program stored in the ROM 1202 or the program loaded from the storage part 1208 to the RAM 1203. Various programs and data required for system operation are also stored in the random access memory 1203. The central processor 1201, the read-only memory 1202 and the random access memory 1203 are connected to each other through a bus 1204. An I / O (Input / Output) interface 1205 is also connected to the bus 1204.

[0238] The following components are connected to the I / O interface 1205: an input section 1206 including a keyboard, a mouse, etc.; an output section 1207 including a CRT (Cathode Ray Tube), an LCD (Liquid Crystal Display), etc., and a speaker, etc.; a storage section 1208 including a hard disk, etc.; and a communication section 1209 including a network interface card such as a LAN card, a modem, etc. The communication section 1209 performs communication processing via a network such as the Internet. A drive 1210 is also connected to the input / output interface 1205 as needed. A removable medium 1211, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1210 as needed so that a computer program read therefrom is installed into the storage section 1208 as needed.

[0239] In particular, according to an embodiment of the present application, the process described in each method flow chart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer readable medium, and the computer program contains a program code for executing the method shown in the flow chart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1209, and / or installed from the removable medium 1211. When the computer program is executed by the central processor 1201, various functions defined in the system of the present application are executed.

[0240] It should be noted that Fig.12 The computer system 1200 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0241] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0242] The above are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A data processing method, characterized in that: include: Extracting historical data of similar assets corresponding to the target asset pool of the object to be predicted, wherein the data of similar assets is stored in a designated database, the object to be predicted is an asset securitization product initiated by a target initiator, the historical data of similar assets is historical data of a similar asset pool, the similar asset pool is a collection of similar assets of the target asset in the target asset pool, the asset type of the similar asset is the same as the asset type of the corresponding target asset and does not belong to the target asset pool; Parsing the historical data of the same type of assets according to preset pressure parameters to obtain a pressure parameter benchmark sequence corresponding to the target asset, wherein the preset pressure parameter is a parameter that affects the cash flow of the target asset; Calibrate the pressure parameter reference sequence corresponding to the target asset based on the actual value of the parameter corresponding to the preset pressure parameter of the target asset to obtain the pressure parameter calibration sequence corresponding to the target asset, and obtain the forecast cash flow collection table for redemption by applying the pressure parameter calibration sequence to the target cash flow collection table of the target asset pool; Based on the forecast cash flow aggregation table and the credit enhancement measures adopted by the target initiator, the cash flow of the target asset pool is allocated according to the cash flow payment mechanism of the object to be forecasted to obtain the expected cash flow of the object to be forecasted.

2. The method according to claim 1, characterized in that The designated database also stores at least one of the following information: asset pool data of the target asset pool; cash flow payment mechanism data for indicating the cash flow payment mechanism; credit enhancement data for indicating the credit enhancement measures; historical principal and interest payment data for indicating the historical principal and interest payment of the object to be predicted; the data used in the designated database for cash flow prediction of the object to be predicted is the underlying data, and the underlying data includes the historical data of the same type of assets; The method further comprises: Using a multi-threaded concurrent streaming method to query whether the underlying data stored in the specified database is updated; When the underlying data is updated, the updated underlying data is synchronized to the cache of each node among the multiple nodes, wherein the underlying data used for cash flow prediction of the object to be predicted is the underlying data cached on one of the multiple nodes.

3. The method according to claim 1, characterized in that The data used by the designated database for cash flow forecasting of the object to be forecasted is the underlying data, and the underlying data is synchronized to the cache of each node in the multiple nodes; the underlying data includes the historical data of the same type of assets; The amount of memory data on each node is less than or equal to a threshold value of the amount of memory data corresponding to each node; The method further comprises: In response to the received target prediction request, the target computing task corresponding to the object to be predicted is allocated to a target node among the multiple nodes in accordance with a specified load balancing strategy, so that the target computing task is executed on the target node based on the underlying data cached on the target node, wherein the target prediction request is used to request a cash flow prediction for the object to be predicted, the specified load balancing strategy is a strategy for balancing the computing tasks allocated to the multiple nodes, the target computing task is a computing task for cash flow prediction for the object to be predicted, and the target computing task is executed on the target node using a multi-threaded computing method.

4. The method according to claim 1, characterized in that: The method further includes: if there is an asset set of the same type of assets initiated by the target initiator, determining the asset set of the same type of assets initiated by the target initiator as the same type of asset pool; if there is no asset of the same type initiated by the target initiator, determining the asset set of the same type of assets initiated by other initiators except the target initiator as the same type of asset pool; After extracting the historical data of similar assets corresponding to the target asset pool of the object to be predicted, the method further includes: based on the cross-reference relationship between fields, standardizing the historical data of similar assets according to a set of specified fields to obtain the standardized historical data of similar assets.

5. The method according to claim 1, characterized in that: The step of parsing the historical data of the same type of assets according to the preset pressure parameters to obtain a pressure parameter benchmark sequence corresponding to the target asset includes one of the following: In a case where the same asset pool includes a static structure asset pool, and the same asset historical data includes historical data of the static structure asset pool, the historical data of the static structure asset pool is parsed according to a first stress parameter to obtain a first parameter benchmark sequence corresponding to the target asset, wherein the first stress parameter includes a prepayment rate, a default rate, and a recovery rate, and the stress parameter benchmark sequence corresponding to the target asset includes the first parameter benchmark sequence; In the case that the similar asset pool includes a cyclic structure asset pool, and the similar asset historical data includes the historical data of the cyclic structure asset pool, the historical data of the cyclic structure asset pool is parsed according to the second stress parameter to obtain a second parameter benchmark sequence corresponding to the target asset, wherein the second stress parameter includes a repayment rate, a default rate, a recovery rate, a yield rate and a purchase rate, and the stress parameter benchmark sequence corresponding to the target asset includes the second parameter benchmark sequence.

6. The method according to claim 1, characterized in that The step of calibrating the pressure parameter reference sequence corresponding to the target asset based on the parameter true value corresponding to the preset pressure parameter of the target asset to obtain the pressure parameter calibration sequence corresponding to the target asset includes: Based on the difference between the true value of the parameter corresponding to the target asset and the preset pressure parameter and the parameter value of the same pressure parameter in the pressure parameter benchmark sequence corresponding to the target asset, the pressure parameter benchmark sequence corresponding to the target asset is shifted and adjusted to obtain the pressure parameter calibration sequence corresponding to the target asset.

7. The method according to claim 1, characterized in that Before obtaining a forecast cash flow aggregation table for redemption by applying the stress parameter calibration sequence to the cash flow aggregation table of the target asset pool, the method further includes: Calculating the expected cash flow recovery information of each target asset according to the asset pool data of the target asset pool, wherein the expected cash flow recovery information of each target asset is used to indicate the expected future cash flow recovery of each target asset; According to each aggregation date of the object to be predicted, each target asset in the target asset pool is aggregated, and a first candidate cash flow aggregation table is generated based on the cash flow recovery information obtained on each aggregation date, wherein the cash flow recovery information on each aggregation date is used to indicate the cash flow recovery status on each aggregation date; The cash flow aggregation table that best matches the latest balance of the target asset pool among the first candidate cash flow aggregation table and the second candidate cash flow aggregation table is determined as the target cash flow aggregation table, wherein the second candidate cash flow aggregation table is a cash flow aggregation table that is integrated from the cash flow aggregation table of the target asset pool configured for the object to be predicted and the actual performance of the target asset pool calculated by statistics.

8. The method according to any one of claims 1 to 7, characterized in that The predicted cash flow aggregation table is cached according to a specified data reading mode, and the specified data reading mode is used to balance the total amount of table data and the cache load; The method further comprises: The cash flow distribution sequence calculation method corresponding to the object to be predicted in the preset method module library under the account type corresponding to the target initiator is called to obtain the cash flow payment mechanism of the object to be predicted.

9. The method according to any one of claims 1 to 7, characterized in that The method further comprises: Based on the historical principal and interest payment data of the object to be predicted and the expected cash flow of the object to be predicted, the information to be displayed of the object to be predicted is determined, and the information to be displayed is distributed to a display interface for display, wherein the information to be displayed is used to describe the principal and interest payment cash flow during the complete life cycle of the object to be predicted.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program implements the steps of any one of the methods of claims 1 to 8 when executed by a processor.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images