Index relationship analysis support device, and index relationship analysis support method

Abstract

To support estimating the causal relationship between indices including non-financial indices.SOLUTION: An index relationship analysis support device 101 comprises: a storage that stores causal relationship information including one or more pieces of information indicating the causal relationship between indices that are financial indices or non-financial indices, and the intensity of the causal relationship; and a control unit that acquires organization information that is information on values of a plurality of indices including the non-financial indices, and on the basis of the acquired organization information and the causal relationship information, executes relationship analysis processing of specifying the causal relationship between the indices set in the causal relationship information and the intensity of the causal relationship, and output processing of outputting information indicating the specified causal relationship and its intensity to an output device.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an index relationship analysis support device and an index relationship analysis support method. [Background technology]

[0002] Institutional investors, who manage funds, consider companies' disclosure of information from an environmental, social, and governance (ESG) perspective to be an important factor in their investment decisions. This is because they make investment decisions that take into account the sustainability of the company, not just the company's financial situation estimated from financial indicators. From this perspective, in recent years, investments in companies with good ESG status, known as "ESG investment," have been made.

[0003] In response to this, decision makers such as company executives need to determine the ESG materiality that is important to their business. However, since the impact of a company's ESG situation on financial indicators is often uncertain, data analysis is being carried out to determine important ESG indicators (hereinafter referred to as ESG indicators) from past data. For example, Patent Document 1 discloses a method for quantitatively analyzing ESG data to support ESG management and visualizing the results. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-088205 Summary of the Invention [Problem to be solved by the invention]

[0005] The technology disclosed in Patent Document 1 performs correlation analysis between ESG data and financial data and visualizes the results, making it possible to investigate management indicators (non-financial indicators) based on an ESG perspective from a quantitative perspective.

[0006] However, this technology only considers the correlation between ESG indicators and financial indicators, so ESG indicators that happen to be correlated with financial indicators may be extracted as important indicators even though they do not actually affect the financial indicators being analyzed.

[0007] Furthermore, the technology of Patent Document 1 does not allow decision makers such as managers to interpret the specific factors on which ESG indicators are correlated with financial indicators, making it difficult for them to make convincing decisions.

[0008] The present invention has been made in consideration of the above circumstances, and has an object to provide an index relationship analysis support device and an index analysis support method that are capable of supporting the estimation of causal relationships between indexes including non-financial indexes. [Means for solving the problem]

[0009] One of the present inventions for solving the above problems is an index relationship analysis support device that includes a storage device that stores causal relationship information including one or more pieces of information indicating a causal relationship between indexes that are financial indicators or non-financial indicators, and a control device that acquires organizational information that is information on the values ​​of a plurality of indicators including non-financial indicators, and executes a relationship analysis process that identifies, based on the acquired organizational information and the causal relationship information, the causal relationships that exist between each indicator set in the causal relationship information and the strength of the causal relationship, and an output process that outputs the identified causal relationships and information indicating the strength of the causal relationships to an output device. Effect of the Invention

[0010] According to the present invention, it is possible to assist in estimating causal relationships between indicators including non-financial indicators. Configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief description of the drawings]

[0011] [Figure 1]FIG. 1 is a diagram illustrating an example of a configuration of an index relationship analysis support system according to a first embodiment. [Diagram 2] FIG. 13 is a diagram illustrating an example of organizational non-financial information. [Diagram 3] FIG. 2 is a diagram showing an example of organizational financial information. [Figure 4] FIG. 11 is a diagram illustrating an example of causal relationship information. [Diagram 5] FIG. 11 is a diagram illustrating an example of index calculation formula information. [Figure 6] FIG. 11 is a flow diagram illustrating an example of a causal relationship analysis process. [Figure 7] FIG. 13 is a diagram showing an example of a data reading screen. [Figure 8] FIG. 13 is a diagram illustrating an example of a causal relationship setting screen. [Figure 9] FIG. 13 is a diagram showing an example of a screen displaying the contents of external information. [Figure 10] FIG. 11 is a flow diagram illustrating details of an evaluation index causal search process. [Figure 11] FIG. 11 is a diagram illustrating an example of analysis result information. [Figure 12] FIG. 13 is a diagram showing an example of an analysis result display screen. [Figure 13] FIG. 11 is a flow diagram illustrating details of index relationship preprocessing. [Figure 14] FIG. 11 is a flow diagram illustrating details of a causal relationship update process. [Figure 15] FIG. 13 is a diagram illustrating an example of causal relationship information updated by a causal relationship update process. [Figure 16] FIG. 13 is a diagram illustrating an example of causal relationship information updated by index relationship preprocessing. [Figure 17] FIG. 11 is a diagram showing an example of organizational non-financial information according to the second embodiment. [Figure 18] FIG. 11 is a diagram illustrating an example of configuration indicator information. [Figure 19] FIG. 13 is a diagram showing an example of an inter-company model comparison display screen according to the third embodiment. [Figure 20] FIG. 13 is a diagram showing an example of an inter-company analysis result comparison screen according to the fourth embodiment. [Figure 21]FIG. 13 is a diagram illustrating an example of the configuration of an index relationship analysis support system according to a fifth embodiment. [Figure 22] FIG. 11 is a diagram illustrating an example of analysis result information. [Figure 23] FIG. 11 is a diagram illustrating an example of an index value prediction process. [Figure 24] FIG. 11 is a diagram showing an example of policy information. [Diagram 25] FIG. 13 is a diagram showing an example of a policy information editing screen. [Figure 26] FIG. 11 is a flow diagram illustrating details of a future index value prediction process. [Figure 27] FIG. 13 is a diagram showing an example of a policy editing screen. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. <Example 1> 1 is a diagram showing an example of the configuration of an index relationship analysis support system 1 according to a first embodiment. The index relationship analysis support system 1 includes an index relationship analysis support device 101, an input / output device 103, and an external information DB 104. The index relationship analysis support device 101, the input / output device 103, and the external information DB 104 are communicatively connected to each other via wired or wireless communication networks 10, 102 such as the Internet, a LAN (Local Area Network), a WAN (Wide Area Network), a VPN (Virtual Private Network), or a dedicated line.

[0013] The index relationship analysis support device 101 analyzes the causal relationships between various indexes related to a business. The analysis result of the causal relationships includes identification of the index that is the cause of the causal relationship and the index that is the result, and the strength of the causal relationship between the indexes (hereinafter, referred to as the influence degree).

[0014] Indicators that are the subject of causal analysis include financial indicators and non-financial indicators. Financial indicators are, for example, stock prices and sales. Non-financial indicators are indicators of an organization from the perspective of ESG (Environmental-Social-Governance), such as the amount of waste generated, the ratio of female directors, and the level of active discussion within the organization.

[0015] The input / output device 103 is an information processing device such as a PC (Personal Computer) that is used by a user of a company or the like that uses the index relationship analysis support device 101. The input / output device 103 accepts input of information from the user and outputs information on the processing results by the index relationship analysis support device 101.

[0016] The external information DB 104 stores information useful for identifying indicators that are the subject of causal relationships, such as economic information, social conditions, changes in business forms of companies, and global information. The external information DB 104 is communicably connected to the index relationship analysis support device 101 via the communication network 10. The input / output device 103 can also modify the information in the external information DB 104 via the communication networks 10 and 102.

[0017] Next, the index relationship analysis support device 101 includes an arithmetic device 120 (control device) such as a CPU (Central Processing Unit), a DSP (Digital Signal Processor), a GPU (Graphics Processing Unit), an FPGA (Field-Programmable Gate Array), or an ASIC (Application Specific Integrated Circuit), a storage device 110 such as a RAM (Random Access Memory), a ROM (Read Only Memory), an HDD (Hard Disk Drive), or an SSD (Solid State Drive), and a communication device 130 configured with a NIC (Network Interface Card), a wireless communication module, a USB (Universal Serial Interface) module, or a serial communication module, etc.

[0018] Moreover, the index relationship analysis support device 101 stores each of the functional units (programs) of an organization information storage unit 111, a relationship information storage unit 112, a relationship analysis unit 121, and an index relationship preprocessing unit 122.

[0019] The organizational information storage section 111 stores organizational non-financial information 300, which is information including non-financial information of each organization, and organizational financial information 310, which is information including financial indicators of each organization.

[0020] The relationship information storage unit 112 stores causal relationship information 400, which is information on the causal relationships between indexes, and index calculation formula information 410, which is information on a calculation method for an index value. A calculation formula for calculating the value of a certain index (second index) from the value of another index (first index) is set in the index calculation formula information 410.

[0021] (Organizational non-financial information) 2 is a diagram showing an example of organizational non-financial information 300. Organizational non-financial information 300 includes data on names 302 of each non-financial indicator for each fiscal year 301 and values ​​303 of each non-financial indicator.

[0022] (Organizational non-financial information) 3 is a diagram showing an example of organizational financial information 310. The organizational financial information 310 includes data on the name 312 of each financial indicator for each fiscal year 311 and the value 313 of each financial indicator.

[0023] (Causal information) 4 is a diagram showing an example of causal relationship information 400. The causal relationship information 400 has data on an ID 401 of each causal relationship, an index of a cause in each causal relationship (influencing index 402), an index of a result in each causal relationship (influenced index 403), and a hypothesis 404 regarding the relationship between the influencing index and the influenced index in each causal relationship.

[0024] Hypothesis 404 is set to "positive" or "negative." "Positive" indicates that there is a positive correlation between the value of the affected index and the value of the influencing index. "Negative" indicates that there is a negative correlation between the value of the affected index and the value of the influencing index. For example, in a hypothesis that an increase in the ratio of female directors will improve corporate evaluation, "positive" is set. In a hypothesis that an increase in the ratio of female directors will decrease corporate evaluation, "negative" is set. Note that no value need be set in hypothesis 404.

[0025] The causal relationship information 400 may be set in advance by an administrator or the like using the input / output device 103, or may be set automatically or by the administrator or the like based on information acquired from the external information DB 40.

[0026] The indicators set in the influencing indicator 402 and the affected indicator 403 do not have to be composed only of indicators set in the organizational non-financial information 300 and the organizational financial information 310. In the following, the causal relationship identified by the influencing indicator 402 and the affected indicator 403 of the causal relationship information 400 may be referred to as a "direct causal relationship."

[0027] (Indicator calculation formula information) 5 is a diagram showing an example of index calculation formula information 410. The index calculation formula information 410 has data on the name 411 of each index, a list of other indexes (used indexes) used to calculate each index (list of used indexes 412), a list 413 of operators in the calculation formula of each index, and a list 414 of coefficients in the calculation formula of each index.

[0028] In this embodiment, the calculation formula for each index is expressed in a format such as, for example, "coefficient 1 × parameter 1 - coefficient 2 × parameter 2 +...", and these elements in the calculation formula are specified by a list of used indexes 412, a list of operators 413, and a list of coefficients 414. Index calculation formula information 410 is information on the calculation formula for indexes whose calculation formulas are known (particularly, financial indexes such as operating profit).

[0029] The index relation preprocessing unit 122 identifies an index to be corrected among the indexes in the causal relation information 400, and corrects the causal relation information 400 to new causal relation information that does not include the identified index by using a predetermined algorithm.

[0030] In this embodiment, the index relationship preprocessing unit 122 determines whether or not there is an index to be corrected, which is an index that is not set in the organizational information, among the indexes in the causal relationship information 400, to identify the index to be corrected. If there is an index to be corrected, the index relationship preprocessing unit 122 corrects the causal relationship information 400 to new causal relationship information in which the identified index is replaced with any of the indexes in the causal relationship information 400.

[0031] The relationship analysis unit 121 identifies the causal relationships between the indices set in the causal relationship information and the strength of the causal relationships based on the organizational information and the (corrected) causal relationship information. In this embodiment, the strength of the causal relationship is set as an influence level that indicates the magnitude of the influence of the causal index on the resulting index, but the expression of the strength of the causal relationship is not limited to this.

[0032] Each functional unit of the index relational analysis support device 101 described above is realized by the calculation device 120 reading and executing each program stored in the storage device 110. In addition, each program can be recorded on a recording medium and distributed, for example. Note that the index relational analysis support device 101 may be realized, in whole or in part, using virtual information processing resources provided using virtualization technology, process space separation technology, or the like, such as a virtual server provided by a cloud system. In addition, all or in part of the functions provided by the index relational analysis support device 101 may be realized, for example, by a service provided by a cloud system via an API (Application Programming Interface), or the like. Next, the processing performed in the index relationship analysis support system 100 will be described.

[0033] <Causal relationship analysis processing> FIG. 6 is a flow diagram illustrating an example of a causal relationship analysis process for analyzing a causal relationship between indexes.

[0034] First, the organizational information storage unit 111 reads the organizational information (the organizational non-financial information 300 and the organizational financial information 310) (step S202).

[0035] Moreover, the relationship information storage unit 112 reads the causal relationship information 400, which is information on a direct causal relationship, and the index calculation formula information 410 (step S203).

[0036] When the processes of steps S202 and S203 are performed, the input / output device 103 may display a data read screen for reading a file in which the organizational information and causal relationship information 400 of each company are stored.

[0037] (Data loading screen) 7 is a diagram showing an example of a data read screen 1300. The data read screen 1300 has a company selection field 1301 for selecting a company from which organizational information and causal relationship information 400 are read, an organization information path input field 1302 for inputting a path to a file of organizational information of the company selected in the company selection field 1301, a causal relationship information path input field 1303 for inputting a path to a file of causal relationship information 400 of the company selected in the company selection field 1301, and a revision execution field 1304 for executing the processes in and after step S204 described below.

[0038] Furthermore, the input / output device 103 may receive settings for the causal relationship information 400 from an administrator or the like in step S202 or step S203.

[0039] (Causal relationship setting screen) FIG. 8 is a diagram showing an example of a causal relationship setting screen displayed on the input / output device 103 when an administrator or the like sets the causal relationship information 400. As shown in FIG.

[0040] The causal relationship setting screen 500 has a data load field 501 for selecting a data file that will be the basis for the causal relationship information 400, a table editing field 511 that displays the contents of the data selected in the data load field 501 and accepts editing of the data from the user, a graph display field 513 that displays a graph representing the contents of the causal relationship edited in the table editing field 511, a legend field 520 that displays a legend for each figure displayed in the graph display field 513, and a data save field 531 for selecting a data file in which to store the created causal relationship information 400.

[0041] The data read field 501 has a file selection field 502 for selecting a data file, and a file read field 503 for reading the data file selected in the file selection field 502.

[0042] The table editing field 511 includes an editing table 512 which is a table for editing the data selected in the data reading field 501, an add relationship field 516 for adding a causal relationship, and a delete relationship field 517 for deleting a causal relationship.

[0043] The edit table 512 has a causal relationship selection field 514 for selecting a causal relationship, and a setting field 515 for setting an influencing index, an affected index, and a hypothesis in the causal relationship selected in the causal relationship selection field 514.

[0044] When the add relation field 516 is selected while the causal relation selection field 514 is selected, information on the causal relation related to the selected causal relation selection field 514 is added to the causal relation information 400. When the delete relation field 517 is selected while the causal relation selection field 514 is selected, information on the causal relation related to the selected causal relation selection field 514 is deleted from the causal relation information 400. An ID for the causal relation is automatically assigned.

[0045] Graph display field 513 displays node graphics 518 representing each index, and arrow graphics 519 representing the causal relationships between the indexes. Arrow graphics 519 may be displayed in graphics (arrows) of different formats depending on whether the hypothesis is "positive", "negative", or "unknown". Note that such a method of expressing the causal relationships between the indexes is merely an example, and may be changed to another appropriate method of expression.

[0046] Furthermore, the causal relationship setting screen 500 has an indicator addition column 521, which is a setting column for setting the causal relationship information 400 based on the graph display column 513, an impact change column 522, a shape deletion column 523, and an update column 524 for reflecting the setting contents in the graph display column 513 and the editing table 512.

[0047] The indicator addition field 521 is selected when adding a node graphic 518 representing an indicator to the graph display field 513. The influence change field 522 is selected when changing an arrow graphic 519 representing a hypothesis. The graphic deletion field 523 is selected when deleting a graphic selected by the user from among the graphics displayed in the graph display field 513.

[0048] The data save field 531 has a path input field 532 that accepts input of a path and file name of a data file, and a file save field 533 that is selected when recording the causal relationship information 400 in the data file input in the path input field 532.

[0049] By using the causal relationship setting screen 500, an administrator or the like can set and update the causal relationship information 400.

[0050] In step S203, the relationship information storage unit 112 may display the contents of the external information DB 104 on the input / output device 103 as the causal relationship information 400 or another screen.

[0051] 9 is a diagram showing an example of a screen displaying the contents of the external information DB 104 displayed on the input / output device 103. This screen 601 displays information indicating that the automobile exhaust gas problem is attracting attention in country A. The external information DB 104 includes information on factors that affect causal relationships and that differ from country or region to country, such as the business form of a company, the social situation, and the policies of the country where the head office is located.

[0052] The manager or the like can set an appropriate causal relationship by referring to the displayed contents of the external information DB 104. In the example shown in the figure, the manager or the like can make a hypothesis that greenhouse gas emissions, which are a non-financial indicator, may affect corporate evaluation, and set a causal relationship based on that hypothesis.

[0053] 6, the relationship analysis unit 121 identifies all indicators (indicators to be corrected) that are included in the causal relationship information 400 but not included in the organizational information (organizational non-financial information 300 and organizational financial information 310) (step S204). In this embodiment, it is assumed that set A: {increased discussion activity, corporate valuation} is identified as the indicators to be corrected.

[0054] The relationship analysis unit 121 determines whether or not the index to be corrected has been identified in step S204 (step S205). If the index to be corrected has been identified in step S204 (step S205: YES), the relationship analysis unit 121 executes the process of step S206, and if the index to be corrected has not been identified in step S204 (step S205: NO), the relationship analysis unit 121 executes the process of step S207.

[0055] In step S206, the relationship analysis unit 121 executes index relationship preprocessing S206 for creating new causal relationship information by correcting data related to the index to be corrected from the causal relationship information 400. Details of the index relationship preprocessing S206 will be described later.

[0056] Then, based on the new causal relationship information and organizational information created in the index relationship preprocessing S206, the relationship analysis unit 121 identifies the causal relationship between indexes that are directly causally related and the strength of the causal relationship (the degree of influence that the influencing index has on the affected index) (step S207).

[0057] For example, first, the relationship analysis unit 121 calculates the degree of influence by performing a regression analysis with each affected index in the new causal relationship information as a response variable and one or more influencing indexes corresponding to the affected index as explanatory variables. In the example of Fig. 4, when the affected index is a stock price, the influencing indexes used as explanatory variables are the amount of waste generated and the ratio of female directors. The relational equation (regression equation) in this regression analysis is expressed, for example, by equation (1).

[0058]

number

[0059] In formula (1), the objective variable yn (n=1 to N) is the predicted value of each affected indicator (N is the number of affected indicators, which in this embodiment is assumed to be N=4 (stock price, number of new businesses, cost of sales, sales)). The explanatory variable xm (m=1 to M) is the value of the influencing indicator (M is the number of affected indicators, which in this embodiment is assumed to be M=3 (proportion of female directors, amount of waste generated, number of new businesses)). Wnm is the coefficient of each explanatory variable, and this parameter corresponds to the degree of influence that the influencing indicator has on the affected indicator. Note that bn is the intercept.

[0060] In this embodiment, in formula (1), the coefficient of a term relating to an explanatory variable that is not a causal variable for a certain objective variable (an explanatory variable that does not affect the objective variable) is set to 0. For example, in a regression equation for stock prices, the coefficient of the number of new businesses is 0.

[0061] The relationship analysis unit 121 calculates the degree of influence by solving an optimization problem that minimizes the sum of the squares of the differences between the actual performance values ​​of the objective variables and the predicted values ​​calculated from formula (1) (the least squares method). At this time, the relationship analysis unit 121 sets the content (positive or negative) of the hypothesis in the new causal relationship information as a constraint on the coefficients in formula (1).

[0062] The relationship analysis unit 121 may add a predetermined coefficient to the formula (1) in order to make the scale of each index value (term related to each explanatory variable) uniform.

[0063] In this manner, in this embodiment, the causal relationship and the degree of influence between indicators are identified by regression analysis, but other algorithms such as multiple regression analysis, factor analysis, structural equation modeling, or covariance structure analysis may also be used.

[0064] Next, the relationship analysis unit 121 identifies the strength of the causal relationship (the degree of influence of the influencing index on the affected index) between indexes that are not directly causally related (hereinafter referred to as indirect causal relationship) based on the new causal relationship information and organization information created in the index relationship preprocessing S206, and stores the identified result together with the new causal relationship information in the analysis result information 1000 (step S208). As an example of an indirect causal relationship, when there is a direct causal relationship in which the number of new businesses directly affects sales, and a direct causal relationship in which the ratio of female directors directly affects the number of new businesses, the indirect causal relationship is that the ratio of female directors indirectly affects sales.

[0065] (Analysis result information) 10 is a diagram showing an example of analysis result information 1000 created by the processing of step S208. The analysis result information 1000 is data including an influence source index 1001 which is an index of the start point (ultimate cause) in the path of each indirect causal relationship, an influenced index 1002 which is an index of the end point (ultimate result) in the path of each indirect causal relationship, a path 1003 of each index in each indirect causal relationship, and a degree of influence 1004 of the start point index on the end index in each indirect causal relationship.

[0066] In the example shown in the figure, there are two paths of indirect causal relationship with the female director ratio as the starting point and sales as the end point: a path via corporate evaluation and a path via the number of new businesses. Therefore, the relationship analysis unit 121 calculates the influence of the female director ratio on sales by calculating the total value of the influence of each path.

[0067] Next, the relationship analysis unit 121 calculates, for each index, the degree of influence of that index on other non-financial indexes or financial indexes based on the index calculation formula information 410 (step S209).

[0068] Specifically, the relationship analysis unit 121 confirms that at least one of the used indicators set in the list of used indicators 412 of the index calculation formula information 410 is set in the affected indicator 1002 of any record in the analysis result information 1000 (causal relationship information), and calculates the index value indicated by the indicator name 411 of the index calculation formula information 410.

[0069] For example, if the index value is operating profit, the relationship analysis unit 121 creates the following influence degree calculation formula based on the list of used indexes 412, the list of operators 413, and the coefficients 414 of the index calculation formula information 410.

[0070] Impact on operating profit = (+1) x impact on sales + (-1) x impact on cost of sales + (-1) x impact on selling and administrative expenses

[0071] The relationship analysis unit 121 searches the analysis result information 1000 for influencing indicators 1001 (for example, the number of new businesses, the ratio of female directors, the amount of waste generated) that have the indices used (sales, cost of sales, selling and administrative expenses) as influenced indicators 1002 .

[0072] The relationship analysis unit 121 calculates the influence of each influencing indicator 1001 on each used indicator and substitutes it into the influence calculation formula. For example, the number of new businesses and female directors have influences on sales (+0.4, +0.1), so their influences on operating profit are (+1) x 0.4 = +0.4 and (+1) x 0.1 = +0.1, respectively. The amount of waste generated has influences on sales and influences on cost of sales (-0.02, +0.1), so its influence on operating profit is (-1) x (+0.1) + (+1) x (-0.02) = -0.12.

[0073] 11 is a diagram showing an example of analysis result information 1100 created by the processing of step S209. Compared to the analysis result information 1000 created by the processing of step S208, this analysis result information 1100 has an additional impact record 1105 in which the index calculated by the index calculation formula information 410 is used as the affected index. The input / output device 103 may display this analysis result information 1100 on an analysis result display screen.

[0074] (Analysis results display screen) 12 is a diagram showing an example of an analysis result display screen 1200. The analysis result display screen 1200 has a causal relationship display field 1201 that displays the causal relationships between the indexes shown in the analysis result information 1100 in a graph format, a correction target index display field 1211 that displays a list of names 1212 of indexes related to the causal relationships deleted in the index relationship preprocessing S206 (indexes for which the influence degree was not calculated (correction target index or set A)), a detailed information display selection field 1221 that has an index selection field 1222 for selecting an index 1225 that displays detailed information (the influence source index at the start, the influence destination index at the end, the path from the influence source index at the start to the influence destination index at the end, and the influence degree) from the analysis result information 1100, a detailed information display field 1223 that displays detailed information of the index selected in the index selection field 1222, and a reference status display field 1224 that displays the history of the values ​​of each index displayed in the detailed information display field 1223 in a graph or the like.

[0075] In the causal relationship display field 1201, figures 1202 and 1203 representing each index, arrow figures 1204 representing the causal relationship between the indexes, and the degree of influence of each index on other indexes 1205 are displayed. Note that the figures 1202 and 1203 representing the indexes may be displayed in different manners (e.g., solid lines or dashed lines) depending on whether the degree of influence on the index has been calculated or not.

[0076] In the causal relationship display field 1201, when there are two or more paths passing through the same index to be corrected, the arrow graphics 1204 relating to the index to be corrected may be displayed as a common path. In the example of Figure 12, the ratio of female directors affects corporate evaluation, which in turn affects sales and stock price, so two arrow shapes 1231 and 1232 are displayed from the ratio of female directors to corporate evaluation. These arrow shapes 1231 and 1232 indicate relationships with different impact indicators, and are displayed so that users can understand the differences. In the figure, the differences are expressed by the type of arrow (solid line or dotted line).

[0077] If the causal relationship display column 1201 contains information that each company wishes to keep secret (such as company know-how), information related to that information may be hidden.

[0078] By referring to the correction target index display field 1211, the user can determine data that should be acquired in the future in an analysis based on the current causal relationship.

[0079] In this way, the user can determine the non-financial indicators that affect each financial indicator by referring to the analysis result display screen 1200. Since the content displayed on the analysis result display screen 1200 is based on the causal relationship information 400 defined in advance based on a hypothesis, spurious correlations that are coincidental are not displayed, which provides a sense of satisfaction to the user in making decisions.

[0080] <Indicator related pre-processing> Next, FIG. 13 is a flow diagram illustrating details of the index relationship preprocessing S206.

[0081] The index relationship preprocessing unit 122 creates information (hereinafter referred to as graph information) indicating each causal relationship from each influencing index to each affected index based on each record of the causal relationship information 400 (step S702). For example, when the influencing index is the amount of waste generated and the affected index is the cost of sales, the index relationship preprocessing unit 122 creates data of "amount of waste generated -> cost of sales."

[0082] The index relation preprocessing unit 122 executes a causal relation update process S703 for deleting the index to be corrected in each causal relation and converting (replacing) it with an index other than the index to be corrected to create a new causal relation, and updates the causal relation information 900, and acquires the created and processed causal relation information. The causal relation update process S703 will be described later in detail.

[0083] The index relation preprocessing unit 122 deletes causal relations that have an index included in set A as an affected index or an influencing index from among the causal relations in the causal relation information updated in the causal relation update process S703 (step S704). This ends the index relation preprocessing S206.

[0084] <Causal relationship update process> 14 is a flow diagram for explaining details of the causal relationship update process S703. Here, the causal relationship (ID=3, influencing index is company evaluation, affected index is stock price) indicated by the record 405 of the causal relationship information 400 in FIG. 4 will be explained as an example.

[0085] The index relation preprocessing unit 122 confirms that the influencing index in a selected causal relation (hereinafter referred to as the target causal relation) is an index to be corrected (step S802). If the influencing index is not an index to be corrected, the index relation preprocessing unit 122 selects a different causal relation and confirms that the influencing index in the causal relation is an index to be corrected. Note that such a causal relation may be referred to as the first causal relation hereinafter.

[0086] For example, the causal relationship with ID=3 is an indicator to be corrected (set A) since the influencing indicator is the corporate evaluation.

[0087] Next, the index relation preprocessing unit 122 searches the causal relation information 400 for another causal relation (second causal relation) having the influencing index in the target causal relation as an influenced index (step S803).

[0088] For example, the source indicator of the causal relationship with ID=3 is corporate evaluation, and the causal relationship with corporate evaluation as the destination indicator is record 406 in the causal relationship information 400, where the destination indicator is the ratio of female directors, and record 407, where the source indicator is the amount of waste generated.

[0089] The index relation preprocessing unit 122 corrects the target causal relation by replacing the influencing indicator in the target causal relation with the influencing indicator in the causal relation identified in step S803 (step S804). That is, the index relation preprocessing unit 122 creates a new causal relation consisting of the result indicator in the first causal relation and the cause indicator in the second causal relation.

[0090] For example, the causal relationship for ID=3 is modified to two causal relationships: one in which the source indicator is the ratio of female directors and the destination indicator is stock price, and the other in which the source indicator is the amount of waste generated and the destination indicator is stock price.

[0091] The index relationship preprocessing unit 122 sets a hypothesis for the causal relationship corrected in step S804 (step S805).

[0092] For example, if the hypotheses for the target causal relationship and other causal relationships searched for in step S804 are identical (e.g., both are "positive"), the index relationship preprocessing unit 122 sets the hypothesis for the modified causal relationship to "positive" (for example, if it was "negative," it reverses it to "positive").

[0093] In addition, if the hypotheses for the target causal relationship and other causal relationships searched for in step S804 are different (for example, one is "positive" and the other is "negative"), the index relationship preprocessing unit 122 sets the hypothesis for the corrected causal relationship to "negative" (for example, if it was "positive", it is reversed to "negative").

[0094] For example, the hypothesis regarding the causal relationship in which the influencing indicator is the ratio of female directors is set to be positive, and the hypothesis regarding the causal relationship in which the influencing indicator is the amount of waste generated is set to be negative.

[0095] In addition, if at least one of the hypotheses for the target causal relationship and the other causal relationships searched for in step S804 has not been set, the index relationship preprocessing unit 122 sets the hypothesis set for the target causal relationship to “not set.”

[0096] The index relationship preprocessing unit 122 corrects the graph information created in step S702 based on each corrected causal relationship created in step S805 (step S806).

[0097] For example, since the names of the influencing indicators in the corrected causal relationship are "proportion of female directors" and "amount of waste generated," "corporate valuation → stock price" in the graph information is finally corrected to "proportion of female directors → corporate valuation → stock price" and "amount of waste generated → corporate valuation → stock price."

[0098] The index relationship preprocessing unit 122 recursively executes the above-mentioned processing up to step S806. That is, the index relationship preprocessing unit 122 recursively executes the processing up to step S806 until there is no causal relationship that has data of an influencing index or has an influencing index as an affected index in each causal relationship of the causal relationship information (step S807).

[0099] 15 is a diagram showing an example of causal relationship information 900 updated by the causal relationship update process S703. This causal relationship information 900 is data including an influence source index 901 for each causal relationship, an influence destination index 902 for each causal relationship, a hypothesis 903 of the direction of influence for each causal relationship, and a relationship path 904 for each causal relationship.

[0100] 16 is a diagram showing an example of causal relationship information 910 updated by index relationship preprocessing S206. This causal relationship information 910 has the same data configuration as the previous causal relationship information 900, but a record 905 in the previous causal relationship information 900 in which an index included in set A is set as an influencing index 912 or an influenced index 913 has been deleted.

[0101] By the above index relation preprocessing S206, causal relations including the index to be corrected are deleted from the causal relation information 400. This eliminates the need for the user to manually remove indices whose causal relations cannot be evaluated, and enables the processing of step S207 described above to be executed.

[0102] As described above, the index relationship analysis support device 101 of this embodiment identifies the causal relationships between each index and the degree of influence in the causal relationships set in the causal relationship information based on organizational information (organizational non-financial information 300 and organizational financial information 310), which is information on the values ​​of multiple indexes including non-financial indicators, and causal relationship information 400, which is information on the causal relationships between indicators that are financial indicators or non-financial indicators, and outputs information indicating the identified causal relationships and the degree of influence to the input / output device 103.

[0103] In this way, the index relationship analysis support device 101 of this embodiment can estimate the causal relationships between indexes including non-financial indexes based on organizational information of a company or the like.

[0104] In addition, the index relationship analysis support device 101 of this embodiment identifies indexes to be corrected among the indexes in the causal relationship information, corrects the causal relationship information 400 to new causal relationship information that does not include the identified indexes, and identifies the causal relationship between each index and the degree of influence of the causal relationship based on the organizational information and the new causal relationship information.

[0105] This allows the causal relationship information 400 to be corrected to include appropriate causal relationships, and the causal relationships can be analyzed.

[0106] For example, the index relationship analysis support device 101 of this embodiment identifies indexes to be corrected by determining whether or not there are any indexes to be corrected, which are indexes that are not set in the organizational information, among the indexes in the causal relationship information 400. If there are any indexes to be corrected, the causal relationship information 400 is corrected to new causal relationship information in which the above-mentioned identified index is replaced with any of the indexes in the causal relationship information 400.

[0107] As a result, even if there is an index that does not have the data (an index to be corrected), the causal relationship of each index can be specified by correcting the causal relationship information 400. It is expected that such a case will frequently occur in the analysis of non-financial indexes. This is because non-financial indexes may include indexes that are difficult to quantify (for example, corporate evaluations, etc.). In addition, such a situation occurs because the data held and the information disclosed differ for each company. In such a case, it is usually impossible to analyze the causal relationship regarding the index to be corrected, so it is necessary to manually check whether or not there is data regarding the index to be corrected and to reset the causal relationship information 400. Moreover, because the content of the causal relationship regarding other indexes may change due to business changes such as acquisitions and the acquisition of new ESG data, the data acquisition work of the causal relationship information 400 is also required in such a case. In addition, if it is desired to analyze the causal relationship for multiple companies, not limited to a specific company, this work is required for each target company, and it takes a lot of time and effort to perform it manually. The index relationship analysis support device 101 of this embodiment can analyze the causal relationship between each index even in such a situation.

[0108] Specifically, the index relationship analysis support device 101 of this embodiment identifies a first causal relationship in which the index to be corrected is the cause and a second causal relationship in which the index to be corrected is the result, and creates a new causal relationship consisting of an indicator of the result in the first causal relationship and an indicator of the cause in the second causal relationship, thereby correcting the causal relationship information 400 to new causal relationship information.

[0109] In this way, by using other causal relations set in the causal relation information 400, the causal relation information 400 can be appropriately corrected for the causal relation related to the index to be corrected.

[0110] In addition, the index relationship analysis support device 101 of this embodiment identifies a first causal relationship in which the index to be corrected is the cause and a second causal relationship in which the index to be corrected is the result, and creates a new causal relationship consisting of an indicator of the result in the first causal relationship and an indicator of the cause in the second causal relationship, thereby correcting the causal relationship information to new causal relationship information.

[0111] This allows the indices in the causal relationship information 400 to be composed of indices other than the correction target indices.

[0112] In addition, when the hypotheses on the positive or negative correlation between index values ​​in two causal relationships in the causal relationship information are different, the index relationship analysis support device 101 of this embodiment identifies the causal relationship between each index and the degree of influence in that causal relationship by reversing the cause and effect in one of the causal relationships.

[0113] This allows the degree of influence in the causal relationship information to be appropriately corrected.

[0114] Furthermore, the index relationship analysis support device 101 of this embodiment identifies the causal relationship between the indexes and the degree of influence in the causal relationship based on the index calculation formula information 410.

[0115] This makes it possible to identify the causal relationships and their strengths for the indices set in the causal relationship information 400.

[0116] In addition, the index relationship analysis support device 101 of this embodiment creates a predetermined relational equation (regression analysis equation) based on the causal relationship information, in which the value of the cause index and the value of the result index in each causal relationship are used as variables, and identifies the causal relationship between each index and the degree of influence in that causal relationship by substituting the value of each index in the organizational information into the created regression analysis equation.

[0117] This makes it possible to accurately identify the causal relationships between the indicators and their strength based on statistical processing.

[0118] <Example 2> In the first embodiment, among the indices registered in the causal relationship information 400, indices to be corrected that are not registered in the organization information (for example, indices that are difficult to quantify, such as active discussions) are excluded from targets for identifying causal relationships. Therefore, in the second embodiment, such indices to be corrected can also be used to identify causal relationships.

[0119] First, the configuration of the index relationship analysis support system 1 in the second embodiment is the same as that in the first embodiment.

[0120] However, the organizational information storage unit 111 in the second embodiment stores organizational non-financial information having a data structure different from that in the first embodiment. In addition, the relationship information storage unit 112 stores configuration indicator information 1500, which is information on a calculation method for calculating the value of a predetermined indicator (for example, an indicator to be corrected) from predetermined information.

[0121] Then, the relationship analysis unit 121 of the index relationship analysis support device 101 in the second embodiment accepts input of specified information of the configuration index information 1500, and based on the input information and the calculation method of the configuration index information 1500, identifies the causal relationship between each index including the specified index (index to be corrected) and the degree of influence of the causal relationship.

[0122] (Organizational non-financial information) 17 is a diagram showing an example of organizational non-financial information 1400 according to Example 2. This organizational non-financial information 1400 has various data values ​​1404 that can be used to quantify each correction target indicator, in addition to each data of each fiscal year 1401, each non-financial indicator name 1402, and each non-financial indicator value 1403 similar to Example 1.

[0123] In the example shown in the figure, the results (percentage of votes, etc.) of a survey conducted at a company asking questions such as "Is there an atmosphere in meetings where it is easy to speak up?" and "Does your boss listen to a variety of opinions?" are set as data value 1404 as data that can be used to quantify the indicator to be corrected, "active discussion."

[0124] (Component index information) 18 is a diagram showing an example of configuration indicator information 1500. It has data on the name 1501 of each indicator to be corrected and the calculation method 1502 of each indicator to be corrected. In the calculation method 1502, a calculation method (for example, data items or parameters used to calculate the indicator to be corrected, questionnaire items, etc.) is set for calculating the value of the indicator to be corrected using the data value 1404 of the organizational non-financial information 1400. Note that a predetermined calculation formula may also be set.

[0125] Next, the causal relationship analysis process according to the second embodiment will be described based on the causal relationship analysis process according to the first embodiment.

[0126] The process of step S202 is the same as that of embodiment 1. In step S203, the relationship information storage unit 112 reads the causal relationship information 400, the index calculation formula information 410, and the configured index information 1500.

[0127] In step S204, the relationship analysis unit 121 identifies all indices that are included in the causal relationship information 400 but are not included in the organization information or the configuration index information 1500 as indices to be corrected. In the above example, since “increased discussion” is an index included in the configuration index information 1500, it is not determined to be an index to be corrected.

[0128] The processes in steps S205 and S206 are the same as those in the first embodiment.

[0129] In step S207, the relationship analysis unit 121 identifies the causal relationship and the degree of influence between the indexes in the direct causal relationship, as in the first embodiment.

[0130] At this time, the relationship analysis unit 121 calculates the value of the indicator to be corrected among the indicators used in the formula (1) based on the data value 1404 of the organizational non-financial information 1400 .

[0131] For example, the relationship analysis unit 121 may set the average value of each item of the data value 1404 as the value of the index to be corrected, or may calculate the index value by other algorithms such as principal component analysis or factor analysis.

[0132] The processes in steps S208 and S209 are the same as those in the first embodiment.

[0133] Here, information on the configuration indicator information 1500 may be displayed on the analysis result display screen 1200. For example, when "Active discussion" is selected in the detailed information display selection field 1221, data on that indicator may be displayed.

[0134] As described above, the index relationship analysis support device 101 of the second embodiment accepts input of parameters in the configuration index information 1500, and based on the input parameters and the configuration index information 1500, identifies the causal relationship between each index and the influence degree of the causal relationship.

[0135] This makes it possible to identify causal relationships related to indicators that are not registered in organizational information, i.e., indicators that are usually difficult to quantify. For example, it is possible to quantify social-perspective indicators in non-financial indicators using other data such as the results of in-house employee surveys.

[0136] <Example 3> The index relationship analysis support system of the third embodiment is an information processing system for each company to analyze causal relationships based on organizational information of other companies including itself. The configuration and functions of the index relationship analysis support system 1 in the third embodiment are the same as those in the first embodiment.

[0137] However, the relationship analysis unit 121 according to the third embodiment creates a predetermined relational equation (regression analysis equation) based on the causal relationship information of a plurality of organizations, and identifies the causal relationship between each index and the degree of influence in the causal relationship by substituting the value of each index in the organizational information of the plurality of organizations into the created regression analysis equation.

[0138] Then, the relationship analysis unit 121 creates a regression analysis formula based on the causal relationship information of the specified organization, and identifies the causal relationship between each indicator and the degree of influence of the causal relationship by substituting the value of each indicator in the organizational information of the specified organization into the created regression analysis formula.

[0139] The process of the index relationship analysis support system 1 in the third embodiment will be described below.

[0140] First, the index relationship analysis support device 101 specifies a target company for which organizational information is to be used. The method of specifying the target company is not particularly limited, but for example, the index relationship analysis support device 101 may accept the selection of the target company from the input / output device 103, or may obtain a list of companies in a similar industry from a predetermined database.

[0141] Furthermore, the index relationship analysis support device 101 may specify companies with similar relationships between non-financial indicators. For example, the index relationship analysis support device 101 identifies companies with similar non-financial indicators by clustering the data of the organizational non-financial information 300 of each company. Also, for example, the index relationship analysis support device 101 specifies companies with similar trends in non-financial indicators by performing a predetermined statistical analysis.

[0142] In this embodiment, it is assumed that organizational information of five companies, companies A to E, is used, and company A is the company performing the causal relationship.

[0143] Next, the index relationship analysis support device 101 executes the same causal relationship analysis process as in the first embodiment. At this time, the index relationship analysis support device 101 creates a regression analysis formula (formula (1)) based on the organizational information (organizational non-financial information 300 and organizational financial information 310) of the above selected company, and identifies the causal relationship and the degree of influence. Hereinafter, the regression analysis formula used in this process is referred to as the company average model.

[0144] The company average model represents the average causal relationship among a group of companies. This company average model includes data from multiple companies, and is more reliable than the analysis results of a single company in terms of the number of data points. However, since the trends of multiple companies are mixed, it may not be suitable for use in determining important indicators for a specific company.

[0145] Next, the index relationship analysis support device 101 creates a regression analysis formula (hereinafter referred to as the target company model) based on the data of only company A in the organizational information (organizational non-financial information 300 and organizational financial information 310), and identifies the causal relationship and the degree of impact.

[0146] When solving this regression analysis formula, the indicator relationship analysis support device 101 sets the influence of the company average model, that is, the coefficient of the regression analysis formula (1), as the initial value of the regression analysis coefficient in the analysis of company A. In addition, when solving the optimization problem of the regression analysis formula (1), the indicator relationship analysis support device 101 limits the magnitude of change in the coefficient. This prevents the value from being significantly changed from the coefficient of the company average model.

[0147] The limit value for the change in the coefficients is not particularly limited, but for example, the square root of the sum of the squares of the differences between the coefficients can be set to 0.1 or less. In this case, the constraint condition is expressed by the following formula (2). The index relationship analysis support device 101 performs a regression analysis of formula (1) so as to satisfy formula (2), thereby realizing an analysis of the relationship between the indexes of company A based on the company average model.

[0148]

number

[0149] Here, wnm is the coefficient of the firm-average model, and w'nm is the coefficient of the regression analysis for firm A.

[0150] When data on an indicator held only by the company being analyzed (company A) is set in the organizational information and the indicator is included in the causal relationship information, values ​​regarding the initial value and the limit on the change in the coefficient may not be set, or any value may be set. On the other hand, when data on an indicator not held only by the company being analyzed (company A) is set in the organizational information and the indicator is included in the causal relationship information, regression analysis may be performed only on the indicators held by the company being analyzed.

[0151] Here, the input / output device 103 may display the results of the above processing on an inter-company model comparison display screen 1600.

[0152] <Company model comparison display screen> 19 is a diagram showing an example of an inter-company model comparison display screen 1600 according to Example 3. The inter-company model comparison display screen 1600 has a company average model summary display field 1601 that displays information about companies whose data was used in the calculation of the company average model, and a comparison display field 1611 that displays comparative information between the analysis results of the analysis target company and the analysis results of an average company.

[0153] A company average model summary display field 1601 displays a list of companies 1602 and a company selection method display field 1603. For example, when companies in a similar industry are selected, the company selection method display field 1603 outputs "Use companies in a similar industry (industry: manufacturing)," and when companies with similar relationships between non-financial indicators are selected, "Use companies with similar non-financial indicators." These outputs are changed appropriately depending on the selection method.

[0154] The comparison display field 1611 has an analysis result display field 1614 that displays the comparison results of the causal relationship between the company average model and the target company model, and a graph display field 1615 that displays the contents displayed in the analysis result display field 1614 in a graph.

[0155] The influencing index selection field 1612 is selected when limiting (filtering) the causal relationships displayed in the analysis result display field 1614 with respect to the influencing index. The influencing index selection field 1613 is selected when limiting (filtering) the causal relationships displayed in the analysis result display field 1614 with respect to the affected index.

[0156] By referring to the inter-company model comparison display screen 1600, the user can check the analysis results of the causal relationships of the target company as well as the difference in the degree of influence from the company average model, and can grasp the characteristics of the causal relationships between each indicator in the target company.

[0157] As described above, the index relationship analysis support device 101 of this embodiment creates a regression analysis based on causal relationship information of multiple organizations (e.g., companies A to E) including a specified organization (e.g., company A), and identifies and outputs the causal relationship between each index and the influence in the causal relationship by substituting the value of each index in the organizational information of the multiple organizations into the created relational expression. Also, the index relationship analysis support device 101 creates a regression analysis based on causal relationship information of a specified organization (e.g., company A), and identifies and outputs the causal relationship between each index and the influence in the causal relationship by substituting the value of each index in the organizational information related to the specified organization into the created relational expression.

[0158] This allows the user to obtain highly reliable results by comparing the analysis results from multiple organizations (e.g., companies A to E) even if there is little information about the organization (company A) for which causal relationships are to be analyzed.

[0159] That is, the standardization of items to be disclosed for non-financial information, including ESG information, has just begun, and disclosure obligations under laws and regulations are just beginning in Europe and other countries. Therefore, the disclosure of non-financial information of a company is limited to a period of 10 or 20 years at most, and the number of data is limited. Under such circumstances, if only the causal relationship analysis process of Example 1 is performed, in the case of a company that does not have much data on indicators, the reliability of the results of the causal relationship analysis may decrease. For example, when the number of data on indicators is small, even variables unrelated to the objective variable in the regression analysis are more likely to have a strong correlation with the objective variable by chance. As a result, the coefficient of the regression analysis, that is, the degree of influence, may be overestimated, making it difficult to extract important indicators.

[0160] Therefore, the index relationship analysis support device 101 of this embodiment uses organizational information of other companies to obtain the analysis result of, for example, average causal relationship of the companies, and then performs the analysis of the causal relationship of the company to be analyzed based on the result. In this way, by incorporating data of other companies into the analysis, the amount of data increases and the reliability of the analysis result improves, and at the same time, it is possible to confirm the difference in the degree of influence compared to other companies, which also contributes to confirming the position of the own company.

[0161] <Example 4> In the fourth embodiment, the index relation analysis support device 101 executes the causal relation analysis process according to the first embodiment etc. for each of a plurality of companies based on their own organization information.

[0162] That is, the relationship analysis unit 121 of the index relationship analysis support device 101 of this embodiment identifies, for each organization, the causal relationship between each index set in the causal relationship information and the degree of influence in the causal relationship, based on the organizational information of each of a plurality of organizations and the causal relationship information of each organization.

[0163] Then, the index relationship analysis support device 101 displays the results of the causal relationship analysis process for each of the multiple companies on an inter-company analysis result comparison screen of the input / output device 103 for each company.

[0164] (Inter-company analysis results comparison screen) 20 is a diagram showing an example of an inter-company analysis result comparison screen 1700 according to Example 4. The inter-company analysis result comparison screen 1700 has a company selection field 1701 for selecting a company for displaying the analysis result of the causal relationship, an industry selection field 1702 for selecting the industry of the company for displaying the analysis result, an analysis result display field 1705 for displaying the contents of the analysis result of the causal relationship of the companies specified in the company selection field 1701 and the industry selection field 1702 (company name, influence source index, influence destination index, relationship path, influence degree), a disclosure information display field 1706 for displaying organizational information of the companies specified in the company selection field 1701 and the industry selection field 1702, and a graph display field 1707 for displaying a graph showing the contents of the organizational information.

[0165] The influencing index selection field 1703 is selected when limiting (filtering) the causal relationships displayed in the analysis result display field 1705 with respect to the influencing index. The influencing index selection field 1704 is selected when limiting (filtering) the causal relationships displayed in the analysis result display field 1705 with respect to the affected index.

[0166] As described above, the index relationship analysis support device 101 of this embodiment identifies the causal relationships between indexes and the degree of influence of those causal relationships for each organization in a plurality of organizations, and outputs the identified causal relationships and the degree of influence for each organization to the inter-company analysis result comparison screen 1700.

[0167] This allows users to compare the degree of influence between indicators in multiple organizations, such as various companies or companies in similar industries. For example, by checking companies with similar degrees of influence, users can discover companies that are similar in terms of non-financial indicators such as ESG. In addition, for example, investors who make ESG investments can compare various companies or companies in similar industries when selecting companies to invest in.

[0168] <Example 5> In this embodiment, the index relationship analysis support device 101 executes a simulation to predict the future value of each index based on the causal relationship between the indexes analyzed in the first embodiment and the like. In addition, in this embodiment, the index relationship analysis support device 101 predicts the future index value when a specific measure is implemented. Note that the measure is a corporate activity that affects the index value, such as the introduction of waste recycling equipment to improve the amount of waste generated, which is a non-financial index, or the holding of a new business hackathon to increase the number of new businesses.

[0169] FIG. 21 is a diagram illustrating an example of the configuration of an index relationship analysis support system 1800 according to the fifth embodiment.

[0170] The index relationship analysis support system 1800 of this embodiment, like the first embodiment, is configured to include an index relationship analysis support device 1801, an input / output device 1803, and an external information DB 1804, and is communicatively connected via communication networks 180 and 1802.

[0171] Also, the index relationship analysis support device 1801 includes a calculation device 1820 (control device), a storage device 1810, and a communication device 1830, similar to the first embodiment. Also, the index relationship analysis support device 1801 includes functional units of an organization information storage unit 1811 and a relationship information storage unit 1812, similar to the first embodiment. Also, the index relationship analysis support device 1801 includes functional units of a relationship analysis unit 1821 and an index relationship preprocessing unit 1822, similar to the first embodiment.

[0172] Here, the index relationship analysis support device 1801 of the fifth embodiment further includes the functional units of an analysis result storage unit 1813, a policy information storage unit 1814, and a simulation execution unit 1823.

[0173] The analysis result storage unit 1813 stores analysis result information 2000, which is information on the results of the impact analysis process.

[0174] The policy information storage unit 1814 stores policy information 2200, which is information on the impact on a given index when a company implements a given policy.

[0175] The simulation execution unit 1823 predicts the index value of a specified index at a predetermined time point in the future based on information (analysis result information 2000) indicating the causal relationship identified by the relationship analysis unit 1821 and the degree of influence of that causal relationship.

[0176] (Analysis result information) 22 is a diagram showing an example of the analysis result information 2000. The analysis result information 2000 has data of an influencing index 2021 (including the case of an index in the calculation formula of the index calculation formula information 410) in each causal relationship, an affected index 2002 (including the case of an index in the calculation formula of the index calculation formula information 410. For example, sales, cost of sales) in each causal relationship, an influence degree 2003 from an influencing index to an affected index in each causal relationship (including a coefficient in the calculation formula of the index calculation formula information 410), an intercept 2004 of the regression equation in each causal relationship (in the case of a causal relationship calculated from the calculation formula of the index calculation formula information 410, 0 is set), and an operator 2005 in each causal relationship (in the case of a causal relationship calculated from a regression formula, "+" is set, and in the case of a causal relationship calculated from the calculation formula of the index calculation formula information 410, an operator in the calculation formula is set).

[0177] FIG. 23 is a diagram for explaining an example of a process related to index analysis and simulation of index values ​​(index value prediction process) performed in the fifth embodiment.

[0178] First, the same impact analysis process as in the first embodiment is executed (step S1901).

[0179] After that, the organization information storage unit 1811 reads the organization information (step S1902). Also, the analysis result storage unit 1813 reads the analysis result information 2000 (step S1903).

[0180] Then, the simulation execution unit 1823 executes a future index value prediction process S1904 for predicting the index value of each index at each future time point in the analysis result information 2000 up to a preset future time point (hereinafter referred to as a prediction target time point). The analysis result storage unit 1813 stores the prediction results of each index value in a predetermined database as an analysis result when no measure is applied (when the measure information 2200 is not taken into consideration). The future index value prediction process S1904 will be described in detail later.

[0181] In this embodiment, the simulation execution unit 1823 predicts each index for each fiscal year from the current fiscal year (2022) to the future fiscal year 2030.

[0182] Next, the policy information storage unit 1814 reads the policy information 2200 (step S1905).

[0183] (Policy information) 24 is a diagram showing an example of policy information 2200. The policy information 2200 has data on ID 2201 of each policy, name 2202 of each policy, price per unit 2203 of equipment used in each policy, useful life 2204 of the equipment, effect on index when the equipment is introduced in units 2205, number of installations of the equipment 2206, policy application year 2207 which is the timing of introducing the equipment (implementing the policy), and adjustable parameters 2209 for each policy. The adjustable parameters 2209 are a list of parameters whose values ​​are adjustable by the user among the data of the policy information 2200.

[0184] Note that, although the figure shows an example of the policy information 2200 for a policy to introduce waste recycling equipment as a policy for reducing the amount of waste generated, the policy information 2200 for other policies or other equipment may be set. Also, the data format of the policy information 2200 shown in the figure is one example, and other data formats may be adopted.

[0185] The policy information 2200 may be set in advance by the user through a predetermined screen (policy information editing screen) displayed by the input / output device 1803 .

[0186] (Policy information editing screen) FIG. 25 is a diagram showing an example of a policy information editing screen 2300 displayed on the input / output device 1803. As shown in FIG.

[0187] The policy information editing screen 2300 has a selection area 2301 for selecting the policy to be set, an overview display field 2302 in which an overview of the policy selected in the selection area 2301 is displayed, a policy information editing field 2303 in which current information of the policy selected in the selection area 2301 and which accepts editing input of the information, a parameter editing area 2304 in which adjustable parameters of the policy selected in the selection area 2301 are set, and a save field 2308 in which the contents set in the policy information editing field 2303 and the parameter editing field 2304 are saved in the policy information 2200.

[0188] The parameter editing area 2304 accepts input of parameter values ​​for each adjustable parameter set in the policy information editing column 2303. Specifically, the parameter editing area 2304 has an area 2305 for inputting the minimum value of the adjustable parameter, an area 2306 for inputting the maximum value, and an area 2307 for inputting the variation range. The variation range here is the variation range of the adjustable parameter in the simulation. For example, if the minimum value of the number of installed facilities, which is an adjustable parameter, is 3, the maximum value is 9, and the variation range is 2, a simulation is executed for the numbers of installed facilities of 3, 5, 7, and 9.

[0189] 23, the simulation execution unit 1823 executes a process similar to the future index value prediction process S1904 for each future time point from the present time point to the prediction target time point based on each index value obtained by adding the impact on the index value of each measure indicated by each measure information 2200 read in step S1905 to each index value calculated by the impact degree calculation process (step S1906). In this case, the simulation execution unit 1823 creates a pattern of each adjustment parameter for each measure indicated by the measure information 2200, and calculates the impact on the index value based on the created pattern. The analysis result storage unit 1813 stores the prediction result of each index value in a predetermined database as the analysis result when the measure is applied.

[0190] (Future index value prediction process) FIG. 26 is a flow diagram illustrating the details of the future index value prediction process S1904.

[0191] The simulation execution unit 1823 refers to the analysis result information 2000, and extracts all indices that are not included in the affected indices 2002 and are included in the influencing indices 2001 (step S2102). For example, the simulation execution unit 1823 extracts the female director ratio and the amount of waste generated.

[0192] The simulation execution unit 1823 sets an initial value for each index value extracted in step S2102, and records each index for which an initial value has been set in a predetermined database (hereinafter, referred to as a predicted list) (step S2103).

[0193] The simulation execution unit 1823 may randomly set the initial value of each index value. For example, the simulation execution unit 1823 acquires random numbers extracted from a normal distribution in which the actual measured value of each index at the time closest to the time to be predicted or the predicted value by the future index value prediction process S1904 is set as the average value and the standard deviation of the actual measured value of each index is set as the standard deviation. The acquired initial value becomes the predicted value of the corresponding index in the year in which the process is executed. In addition, the simulation execution unit 1823 may set the initial value based on the index that influences the index (for example, in the case of a causal relationship in which sales one year ago affect the number of new businesses in this year. In this case, for example, the actual measured value or predicted value of sales one year ago is set as the initial value). Note that the method of determining the initial value described here is one example, and prediction may be performed using a time series model such as an autoregressive model, for example.

[0194] The simulation execution unit 1823 identifies an index to be predicted (step S2104). For example, for each affected index 2002 in the analysis result information 2000, an index is extracted in which all of the influencing indexes 2001 corresponding to the affected index 2002 are included in the predicted list.

[0195] In the above example based on FIG. 22, the female director ratio and the amount of waste generated are set in the predicted list, so the stock price and the number of new businesses are extracted.

[0196] The simulation execution unit 1823 selects an index that is not included in the predicted list from among the indexes identified in step S2104, predicts a future value of the selected index, and stores the index name in the predicted list (step S2105).

[0197] In the example shown in FIG. 22 and described above, the stock price and the number of new businesses are not included in the predicted list, so predicted values ​​for the stock price and the number of new businesses are calculated.

[0198] The simulation execution unit 1823 predicts the future value of each index (for example, the value for the next year) by, for example, a regression analysis method.

[0199] In the case of prediction by regression analysis, the predicted value of each influencing indicator at each future point in time (forecast year), the degree of influence of each influencing indicator on the target influenced indicator, an intercept, and an operator are required.

[0200] Therefore, the simulation execution unit 1823 predicts the future value of the index by calculating a value obtained by multiplying the influence degree of each influencing index by the predicted value according to the target operator and adding an intercept to the result. For example, the prediction for the stock price is as follows:

[0201] (Predicted stock price) = (+0.4) x (Predicted ratio of female directors) + (+0.0) x (Predicted amount of waste generated) + (0.1)

[0202] It should be noted that the prediction algorithm described here is merely an example, and any other prediction algorithm may be adopted.

[0203] Thereafter, the simulation execution unit 1823 determines whether all index names in the analysis result information 2000 match the index names in the predicted list (step S2106).

[0204] If all index names in the analysis result information 2000 match the index names in the predicted list (step S2106: Yes), the future index value prediction process S1904 ends. If all index names in the analysis result information 2000 do not match the index names in the predicted list (step S2106: No), the simulation execution unit 1823 repeats the process of step S2104 to select another index for predicting a future value.

[0205] The above process is repeated for each future time point (each year) to be predicted up to the time point to be predicted, thereby making it possible to predict the index value at the time point to be predicted. The analysis result storage unit 1813 stores the prediction results of each index value in a predetermined database.

[0206] The simulation execution unit 1823 may modify the predicted index value.

[0207] When correcting the cost of sales and the amount of waste generated, the simulation execution unit 1823 executes the following processing, for example. That is, in the case of the cost of sales, since the equipment price is 100 million yen and the useful life is 5 years, the depreciation cost is 20 million yen for the number of installed equipment units. Therefore, the simulation execution unit 1823 sets the predicted value of the cost of sales to the value of the depreciation cost for five years from the policy application year 2207. Also, in the case of the amount of waste generated, the effect of the policy is 3% for the number of equipment units. Therefore, the simulation execution unit 1823 sets the predicted value of the amount of waste generated in the policy application year 2207 to a value obtained by reducing the predicted value of the amount of waste generated by the effect of the policy.

[0208] The measures to be used for predicting the index value may be set in advance by the user on the measure display screen 2440 displayed by the input / output device 1803. For example, the measure display screen 2440 displays a list of measures that can be used for predicting the index value, and provides a selection field for the user to select the measure to be used from the list.

[0209] (Policy display screen) 27 is a diagram showing an example of a measure display screen 2440. The measure display screen 2440 has a measure display field 2401 in which a list of measures set in the measure information 2200 is displayed, an index selection field 2402 in which one or more measures to be used for predicting an index value are selected, an analysis result display field 2403 in which the contents of the measure selected in the index selection field 2402 and the predicted values ​​of each index value calculated based on the measure are displayed and in which the selection of the measure is received, and a graph display field 2404 in which a graph of the index value displayed in the analysis result display field 2403 is displayed.

[0210] The analysis result display field 2403 displays the analysis result when each measure is not applied and the analysis result for each pattern of each adjustment parameter for each measure. Specifically, the analysis result display field 2403 displays the rank of each calculated index value, the pattern of adjustable parameters for each measure, and the predicted value of each index at the prediction target time.

[0211] When multiple measures are selected in the indicator selection field 2402, the analysis result display field 2403 may display information about the indicator values ​​in the order of the measures with the highest indicator values ​​(for example, in the order of the highest predicted value of the indicator at the time of prediction). This allows the user to preferentially check measures that are likely to improve the indicator (KPI). In the example shown in the figure, the amounts of waste generated are displayed in the analysis result display field 2403 in ascending order of amount, and the operating profits are displayed in descending order of value.

[0212] By referring to the above-mentioned measure display screen 2440, the user can predict each index value from the causal relationships between the indexes analyzed while taking into account the causal relationships between the indexes, and compare each index value for each measure. Furthermore, based on this result, the user can consider measures necessary to address ESG materiality. Furthermore, since the predicted value of the index value takes into account the causal relationships between the indexes, the user can interpret the prediction process and make a convincing decision.

[0213] As described above, the index relationship analysis support device 1801 of this embodiment predicts the index value of a specified index at a specified future point in time based on the analysis result information 2000, and outputs the predicted index value to the policy display screen 2440.

[0214] This allows the user to create future business plans from the perspective of various indicators.

[0215] In addition, the index relationship analysis support device 1801 of this embodiment identifies a causal indicator in a causal relationship in which the specified indicator is a result, and predicts an index value of the specified indicator based on the value of the identified indicator and the influence of the causal relationship related to that indicator.

[0216] In this way, by predicting the value of the index to be predicted based on the index that is the cause of the index to be predicted and the influence of the causal relationship based on this, it is possible to predict an appropriate index value that takes into account the causal relationship between the indexes.

[0217] Furthermore, the index relationship analysis support device 1801 of this embodiment predicts an index value based on the policy information 2200 of the influence of a policy on an index and the timing at which the policy affects the index.

[0218] This makes it possible to accurately predict the index values ​​that will be achieved through measures taken by companies and other organizations.

[0219] Although each embodiment has been described above, the present invention is not limited to the above-described embodiments, and can be implemented using any components within the scope of the gist of the present invention. The above-described embodiments and modifications are merely examples, and the present invention is not limited to these contents as long as the features of the invention are not impaired. In addition, although various embodiments and modifications have been described above, the present invention is not limited to these contents. Other aspects conceivable within the scope of the technical idea of ​​the present invention are also included in the scope of the present invention.

[0220] For example, part of the hardware included in each device of this embodiment may be provided in another device.

[0221] Furthermore, each program of each device may be provided in another device, a program may consist of multiple programs, or multiple programs may be integrated into one program.

[0222] Furthermore, the information storage format explained in each embodiment is merely an example and may be changed to other formats as appropriate.

[0223] In addition, in this embodiment, the indices to be corrected are indices that are included in the causal relationship information 400 but not included in the organizational information, but other indices may be used. The indices to be corrected may be, for example, indices designated by the user via the input device 103, or indices with a small amount of data in the organizational information. [Explanation of symbols]

[0224] 101 index relation analysis support device, 111 organization information storage unit, 112 relation information storage unit, 121 relation analysis unit, 122 index relation preprocessing unit

Claims

1. a storage device that stores causal relationship information including one or more pieces of information indicating causal relationships between indicators that are financial indicators or non-financial indicators; and a relationship analysis process for acquiring organizational information, which is information on the values ​​of a plurality of indicators including non-financial indicators, and identifying the causal relationships existing between the indicators set in the causal relationship information and the strength of the causal relationships, based on the acquired organizational information and the causal relationship information; a control device that executes an output process of outputting information indicating the specified causal relationship and the strength of the causal relationship to an output device. An index relationship analysis support device comprising:

2. The control device identifying an index to be corrected from among the indexes in the causal relationship information, and performing index relationship preprocessing to correct the causal relationship information into new causal relationship information that does not include the identified index using a predetermined algorithm; In the relationship analysis process, a causal relationship existing between the indicators and a strength of the causal relationship are identified based on the acquired organizational information and the new causal relationship information. The index relation analysis support device according to claim 1 .

3. The control device In the index relation preprocessing, the index to be corrected is identified by determining whether or not the indexes in the causal relationship information include an index to be corrected, which is an index not set in the organizational information, and if the index to be corrected is included, the causal relationship information is corrected to new causal relationship information in which the identified index is replaced with any of the indexes in the causal relationship information using a predetermined algorithm. The index relation analysis support device according to claim 2.

4. The control device In the index relationship preprocessing, a first causal relationship in which the index to be corrected is a cause and a second causal relationship in which the index to be corrected is a result are identified, and a new causal relationship consisting of an index of the result in the first causal relationship and an index of the cause in the second causal relationship is created, thereby correcting the causal relationship information to the new causal relationship information. The index relation analysis support device according to claim 3 .

5. the storage device stores causal relationship information further including information indicating whether a correlation between a value of an index of a cause and a value of an index of a result in the causal relationship is positive or negative; The control device In the relationship analysis process, when the correlation between the index values ​​of two causal relationships in the causal relationship information is different, the cause and effect in one of the causal relationships are reversed to identify the causal relationship between the indexes and the strength of the causal relationship. The index relation analysis support device according to claim 1 .

6. the storage device stores a calculation formula for calculating a value of a second index from a value of a first index registered in the causal relationship information; The control device In the relationship analysis process, a causal relationship existing between each index including the second index and a strength of the causal relationship are identified based on the calculation formula and the acquired organizational information. The index relation analysis support device according to claim 1 .

7. The control device In the relationship analysis process, a predetermined relational expression is created based on the causal relationship information, with the value of the cause index and the value of the result index in each causal relationship as variables, and the causal relationship between each index and the strength of the causal relationship are identified by substituting the value of each index in the organizational information into the created relational expression. The index relation analysis support device according to claim 1 .

8. the storage device stores a calculation method for calculating a value of a predetermined index from predetermined information; The control device In the relationship analysis process, an input of the predetermined information is accepted, and a causal relationship existing between each index including the predetermined index and a strength of the causal relationship are identified based on the input information and the calculation method. The index relation analysis support device according to claim 1 .

9. the storage device stores causal relationship information of a plurality of organizations including a designated organization; The control device In the relationship analysis process, creating the predetermined relational expression based on the causal relationship information of the plurality of organizations, and substituting the values ​​of each index in the organizational information of the plurality of organizations into the created relational expression to identify the causal relationships between each index and the strength of the causal relationships; creating the predetermined relational equation based on the causal relationship information of the specified organization, and substituting the value of each indicator in the organizational information of the specified organization into the created relational equation to identify the causal relationships between each indicator and the strength of the causal relationships; In the output process, each of the identified causal relationships and the strength of the causal relationships are output to an output device. The index relation analysis support device according to claim 7.

10. the storage device stores causal relationship information for a plurality of tissues; The control device In the relationship analysis process, Identifying, for each organization, a causal relationship between each indicator set in the causal relationship information and the strength of the causal relationship based on organizational information for each of the plurality of organizations and causal relationship information for each of the organizations; In the output process, each of the identified causal relationships and the strength of the causal relationships are output to an output device. The index relation analysis support device according to claim 1 .

11. The control device execute a simulation execution process to predict an index value of a designated index at a predetermined time point in the future based on the identified causal relationship and information indicating the strength of the causal relationship; In the output process, the predicted index value is output to an output device. The index relation analysis support device according to claim 1 .

12. The control device In the simulation execution process, a cause index in a causal relationship in which the specified index is a result is identified, and an index value of the specified index is predicted based on the value of the identified index and the strength of the causal relationship related to the identified index. The index relation analysis support device according to claim 11.

13. the storage device stores policy information including a predetermined policy, a value indicating the magnitude of the impact on the index, and information on the timing of the impact of the policy on the index; The control device In the simulation execution process, the predicted index value is corrected based on the policy information. The index relation analysis support device according to claim 11.

14. An index relationship analysis support method including a storage device that stores causal relationship information including one or more pieces of information indicating a causal relationship between indexes that are financial indexes or non-financial indexes, and a control device, The control device a relationship analysis process for acquiring organizational information, which is information on the values ​​of a plurality of indicators including non-financial indicators, and identifying the causal relationships existing between the indicators set in the causal relationship information and the strength of the causal relationships, based on the acquired organizational information and the causal relationship information; and executing an output process of outputting information indicating the identified causal relationship and the strength of the causal relationship to an output device. A method for supporting index relationship analysis.

Images