TWM613536U - Investment risk scoring system for fund commodities - Google Patents

Abstract

In the investment risk scoring system of a fund product, the data server collects historical financial behavior data associated with at least one of customer and fund transaction behavior, fund profit and loss records, other asset status, digital browsing behavior, and credit card consumption behavior, as well as basic personal information. Reference risk attribute data related to data, financial situation and investment experience; the scoring server uses machine learning algorithms to analyze and calculate multiple sample data sets collected by the data server for model training to establish corresponding target parameters The historical financial behavior data is fed into the scoring model to obtain a model score, and based on the model score and the reference score corresponding to the reference risk attribute data, the risk score of the customer corresponding to the target parameter is generated.

Description

Investment risk scoring system for fund commodities

This new model relates to investment risk assessment, especially refers to an investment risk scoring system for fund commodities.

With the explosion of investment information and the introduction of new investment and wealth management products, investors’ investment behaviors have become more diversified and more difficult to grasp. Therefore, if investors are only evaluated based on their existing basic information such as gender, age, annual income, etc., Especially for the investment risk grade or preference of fund products, the evaluation results obtained in this way often cannot fully reflect the investment risk grade or preference of investors for fund products.

Therefore, in order to effectively promote financial services such as fund products, how to develop a scoring method that can effectively reflect the investment risk of customers on fund products has become one of the issues that financial institutions urgently need to address.

Therefore, the purpose of the present invention is to provide an investment risk scoring system for fund commodities, which can overcome at least one shortcoming of the prior art.

Therefore, an investment risk scoring system for fund products provided by the present invention is used to evaluate a customer's risk when investing in fund products, and includes a data server and a scoring server.

The data server operates to collect historical financial behavior data and reference risk attribute data of the customer, as well as multiple sample data sets for model training. The historical financial behavior data is associated with at least one of fund transaction behavior, fund profit and loss records, other asset status, digital browsing behavior, and credit card consumption behavior. The reference risk attribute information is related to personal basic information, financial situation and investment experience. Each sample data set contains multiple parameter values corresponding to M characteristic parameters related to a target parameter. These characteristic parameters are related to fund transaction behavior, fund profit and loss records, other asset status, digital browsing behavior, and credit card consumption behavior United.

The scoring server is connected to the data server to receive the historical financial behavior data, the reference risk attribute data and the sample data sets, and includes a modeling module and a scoring module. The modeling module operates to analyze and calculate the sample data sets using machine learning algorithms to establish a scoring model corresponding to the target parameter. The scoring module is connected to the modeling module and operates to feed the historical financial behavior data into the scoring model to estimate a model score, and at least according to the model score and the reference score corresponding to the reference risk attribute data , Generate the customer's risk score corresponding to the target parameter.

In the new investment risk scoring system for fund commodities, the scoring server also includes an output module. The input module is connected to the scoring module, and operates to output a risk assessment result corresponding to the customer and containing the risk score.

In the new investment risk scoring system for fund commodities, the machine learning algorithm includes a random forest algorithm, and the scoring model includes a plurality of decision trees, and each decision tree is based on the corresponding m selected from the M feature parameters The characteristic parameters are used for decision-making calculation and m<M.

In the new investment risk scoring system for fund commodities, the target parameter is the proportion of assets allocated to fund commodities.

In the new investment risk scoring system for fund commodities, the scoring module also generates the risk score according to a predetermined first weight and a predetermined second weight respectively assigned to the reference score and the model score, so that the risk Score=(the reference score×the predetermined first weight)+(the model score×the predetermined second weight).

The function of this new model is to use historical financial behavior data associated with fund transaction behavior, fund profit and loss records, other asset status, digital browsing behavior, and/or credit card consumption behavior, and use machine learning algorithms to establish the corresponding target The model score estimated by the parameter scoring model; then the risk score generated based on the model score and the reference score corresponding to the customer's reference risk attribute data, compared with the existing risk scoring method, can fully and truly reflect the customer For the risk appetite that the fund product can bear, this risk score can provide clients with follow-up fund management and/or investment planning services on the premise of effectively avoiding investment risks.

Before the present invention is described in detail, it should be noted that in the following description, similar elements are represented by the same numbers.

Referring to FIG. 1, an investment risk scoring system 100 for fund products according to a new embodiment of the present invention is used to evaluate the risk of a client when investing in fund products, and includes a data server 1 and a scoring server 2.

In this embodiment, the data server 1 can be implemented by, for example, a computer device, and can be connected to an external database (not shown), such as a database of a financial service institution, and used for data collection, especially for collection such as The relevant data of the (to be scored) customer of the financial service institution of the banking institution, as well as all the data required for modeling.

In this embodiment, the scoring server 2 can also be implemented by a computer device, which is connected to the data server 1, and may include, for example, a modeling module 21 and a scoring module connected to the modeling module 21 22. And an output module 23 connected to the scoring module 22. Each of the modeling module 21 and the scoring module 22 can be implemented by hardware, software, firmware or a combination thereof, and the output module 23 can be implemented as but not limited to a display, for example.

Hereinafter, referring to FIG. 1 and FIG. 2, how the investment risk scoring system 100 executes an investment risk scoring procedure for, for example, a customer who belongs to a banking institution and wants to invest in fund products will be described in detail. It is worth noting that the external database connected to the data server 1 here is, for example, a database provided by the banking institution. Preferably, this database can provide richer and more diversified financial information about all customers of the banking institution. The client personal information management program may include the following steps S21 to S25, for example.

First, in step S21, the data server 1 searches, for example, the database of the banking institution to collect historical financial behavior data and reference risk attribute data of the customer, as well as multiple sample data sets for model training, and The collected historical financial behavior data, the reference risk attribute data and the sample data sets are sent to the scoring server 2. It is worth noting that the historical financial behavior data is related to at least one of fund transaction behavior, fund profit and loss records, other asset status, digital browsing behavior, and credit card consumption behavior. The reference risk attribute data is related to personal basic information, financial status, and Related to investment experience, each sample data set contains multiple parameter values corresponding to M characteristic parameters related to a target parameter. These characteristic parameters are related to fund transaction behavior, fund profit and loss records, other asset status, and digital browsing behavior Related to credit card consumption behavior.

More specifically, the historical financial behavior data may include relevant information about the fund products purchased by the client and their profits and losses (such as the average number of days of holding fund products, the type of fund products preferred, the maximum amount of loss, and the degree of volatility of the return ), the customer holds other financial products issued by the banking institution and relevant information about the account balance of the financial account, and the customer’s browsing footprint on the banking institution’s service websites such as fund management (for example, long-term and short-term browsing The number of times and the time of each visit), and the relevant information of the customer's use of the credit card issued by the banking institution (for example, the credit card status and whether to use the revolving interest rate). The reference risk attribute data is, for example, a risk attribute questionnaire answered by the customer through manual filling or input before the first purchase of fund products by the customer. The content of this risk attribute questionnaire may include, for example, basic information (such as age, occupation), Financial background data (such as income, source of funds) and investment experience data (such as financial products purchased), and based on the content of the risk attribute questionnaire through a predetermined scoring method, a reference score for the customer can be obtained. In this embodiment, the target parameter is, for example, "proportion of assets allocated to fund commodities", and for example, M=35, but it is not limited to this example. Therefore, each sample data set can contain, for example, 35 parameter values, including: 16 related to fund trading behaviors: "the number of bond funds held in the past two years", "the number of stock funds held in the past two years" "Number", "Number of multiple asset funds held in the past two years", "Number of currency funds held in the past two years", "Number of balanced funds held in the past two years", ``Hold portfolios in the past two years "Number of funds", "Number of ETF funds held in the past two years", "Average purchase amount", "Average purchase amount to total assets ratio", "Average holding days", "Number of purchases as a percentage of all transactions ", "Subscription times as a percentage of all transaction activities", "Subscription times/redemption times", "Maximum profit amount", "Maximum profit amount-maximum loss amount", "Asset management scale (AUM) ranking classification" Parameter values of other characteristic parameters; 7 related to the fund’s profit and loss records: "Average rate of return of historical trading funds", "Volatility of average rate of return of historical trading funds", "Number of fund products held in history", " Parameter values of characteristic parameters such as the volatility of the historical maximum loss return, the volatility of the historical maximum loss, the volatility of the historical maximum loss, and the profit factor; 5 related to other asset conditions "Taiwan Dollar Parameter values of characteristic parameters such as “Live Deposit Balance”, “Taiwan Dollar Fixed Deposit Balance”, “Check Deposit Balance”, “Foreign Currency Live Deposit Balance”, and “Foreign Currency Fixed Deposit Balance”; 4 of them are related to the digital browsing behavior. The number of visits to the fund management network in a month", the "average time spent on the fund management network in the last month", "the number of visits to the fund management network in the last five months", and the ``average stay time on the fund management network in the last five months" and other characteristic parameters Parameter value; and 3 parameter values of characteristic parameters such as "Credit Limit/Annual Income", "Whether to Use Revolving Rate", and "Average Credit Limit Utilization Rate" associated with credit card consumption behavior.

When the scoring server 2 receives the historical financial behavior data, the reference risk attribute data, and the sample data sets from the data server 2, in step S22, the modeling module 21 operates according to the For the sample data set, a scoring model corresponding to the target parameter is established by using a machine learning algorithm. More specifically, in this embodiment, the machine learning algorithm includes a random forest algorithm, and the scoring model includes a plurality of decision trees, and each decision tree is based on m corresponding feature parameters selected from the M feature parameters Perform decision calculus and m<M. It is worth noting that the random forest algorithm is a kind of "supervised learning" algorithm. Compared with the neural network citation algorithm of deep learning, it usually has a higher model interpretation ability, and the target parameter has been defined before modeling. (That is, the "proportion of assets allocated to fund commodities"), so that these sample data sets can be used to continuously modify in the learning process to develop a relatively correct scoring model.

More specifically, referring to FIG. 3, the modeling module 21 uses a bagging algorithm to randomly select a predetermined number of sample data sets from the sample data sets each time as the training data set for this time, and a total of Perform n times, so that n training data sets can be obtained. For each training data set, in order to generate different random vectors θi, i=1,...,n, perform node segmentation on each of the m feature parameters randomly selected from the M feature parameters to select the smallest The division method of the Gini coefficient is split and finally a decision tree is generated. In this case, each decision tree can be used as a classifier, and can generate a classification result in response to the input data; then, the modeling module 21 also uses classification or regression algorithms to aggregate and calculate The n classification results generated by a decision tree are used to determine a target classification level from a plurality of predetermined classification levels, and output a score corresponding to the target classification level. It is worth noting that because the scoring model combines bagging algorithm and classification or regression algorithm, it can improve the accuracy and stability of the model, while avoiding the occurrence of overfitting and maintaining model integration (Ensemble) characteristic.

Then, in step S23, the scoring module 22 operates to feed the historical financial behavior data as input data into the scoring model to estimate a model score corresponding to the customer. Since the target parameter of the scoring model is "the proportion of assets allocated to fund products", if the higher the ratio, the higher the customer’s preference for fund products and the higher the risk of holding fund products, the model scores The higher it will be, the lower it will be otherwise.

Then, in step S24, the scoring module 22 generates the customer's corresponding target parameter based on the model score, the reference score corresponding to the reference risk attribute data, a predetermined first weight, and a predetermined second weight. Risk score. More specifically, the predetermined first weight and the predetermined second weight are respectively assigned to the reference score and the model score, so that the risk score = (the reference score × the predetermined first weight) + (the model score ×The predetermined second weight).

Finally, in step S25, the output module 23 outputs the risk score result containing the risk score. For example, the output module 23 may display the risk score result, or in other embodiments, the output module 23 may be a communication module to transmit the risk score result to a specific user terminal (such as a bank A computer terminal of the organization or a mobile phone held by the customer).

Therefore, the banking institution can provide the customer with suggestions on fund product allocation (such as the ratio of stocks and debts) during the subsequent fund management or investment planning based on the customer's risk score.

In summary, since the scoring model is trained using sample data sets related to fund transaction behavior, fund profit and loss records, other asset status, digital browsing behavior, and credit card consumption behavior, the model score obtained can be realistic Reflect the risk appetite that clients can bear for fund commodities. Therefore, the risk score obtained by combining the (objective) model score with the (subjective) reference score, compared to the score obtained by the existing risk scoring method, can provide customers with follow-up fund management and financial management under the premise of effectively avoiding investment risks. /Or investment planning services. Therefore, the new investment risk scoring system 100 can indeed achieve the purpose of a new model.

However, the above are only examples of the present model. When the scope of implementation of the present model cannot be limited by this, all simple equivalent changes and modifications made in accordance with the patent scope of the present model application and the contents of the patent specification still belong to This new patent covers the scope.

100: Investment risk scoring system 1: Data server 2: Scoring server 21: Modeling module 22: Scoring module 23: output module S21-S25: steps

The other features and functions of the present invention will be clearly presented in the embodiments with reference to the drawings, in which: Fig. 1 is a block diagram illustrating the structure of the investment risk scoring system for fund commodities according to the embodiment of the present invention; 2 is a flowchart illustrating how the embodiment performs an investment risk scoring procedure for a customer who wants to invest in fund products; and Figure 3 is a schematic diagram illustrating the scoring model established by the embodiment of the present invention. Architecture.

100: Investment risk scoring system

1: Data server

2: Scoring server

21: Modeling module

22: Scoring module

23: output module

Claims (5)

An investment risk scoring system for fund products, used to assess the risk of a customer when investing in fund products, and includes: a data server that operates to collect the customer’s historical financial behavior data and reference risk attribute data, and use In the model training multiple sample data sets, the historical financial behavior data is associated with at least one of fund transaction behavior, fund profit and loss records, other asset status, digital browsing behavior, and credit card consumption behavior. The reference risk attribute data is related to personal basic Data, financial situation and investment experience are related. Each sample data set contains multiple parameter values corresponding to M characteristic parameters related to a target parameter. The M characteristic parameters are related to fund transaction behavior, fund profit and loss records, and others. Asset status, digital browsing behavior and credit card consumption behavior; and a scoring server connected to the data server to receive the historical financial behavior data, the reference risk attribute data and the sample data sets, and include a modeling model Group, operate to use machine learning algorithms to analyze and calculate the sample data sets to establish a scoring model corresponding to the target parameter, and a scoring module, connect the modeling module and operate to the historical financial behavior The data is fed into the scoring model to estimate a model score, and at least the model score and the reference score corresponding to the reference risk attribute data are used to generate the risk score of the customer corresponding to the target parameter. For the investment risk scoring system for fund commodities as described in claim 1, wherein the scoring server further includes: an output module, connected to the scoring module, and operated to output a risk assessment corresponding to the client and containing the risk score result. The investment risk scoring system for fund commodities as described in claim 1, wherein the machine learning algorithm includes a random forest algorithm, and the scoring model includes a plurality of decision trees, and each decision tree is based on m selected from the M The corresponding characteristic parameters of the two characteristic parameters are subjected to decision calculus and m<M. For the investment risk scoring system for fund commodities as described in claim 3, the target parameter is the proportion of assets allocated to fund commodities. For the investment risk scoring system for fund products as described in claim 1, wherein the scoring module further generates the risk score according to a predetermined first weight and a predetermined second weight respectively assigned to the reference score and the model score , So that the risk score=(the reference score×the predetermined first weight)+(the model score×the predetermined second weight).

Images