Server, method, and computer program for providing an investment guide

Abstract

The server that provides investment guides includes a collection unit that collects investment-related information related to investment experts, a learning data generation unit that generates learning data for multiple learning models based on the collected investment-related information, a learning unit that trains the multiple learning models based on the generated learning data, an ensemble model generation unit that generates investment replication ensemble models related to the investment experts based on the multiple learned learning models, and an investment guide information provision unit that provides investment guide information to investor terminals using the investment replication ensemble models, and the multiple learning models may include a stock price model, a financial model, and an economic model.

Description

Technical Field

[0001] The present invention relates to a server, a method, and a computer program for providing investment guidance.

Background Art

[0002] As the interest in fintech increases, the number of users who invest using investment products (e.g., stocks, funds, etc.) is gradually increasing.

[0003] In the case of stock investment, high returns may be generated depending on information collection and analysis capabilities, but losses may also occur due to judgment errors or market conditions.

[0004] Most individual investors who conduct stock trading are increasingly finding it difficult to achieve high returns due to the lack of professional knowledge and information, insufficient funds, etc., unlike investment institutions that specialize in investment.

[0005] On the other hand, most individual investors tend to follow the investment styles (e.g., trading techniques, etc.) of investment experts (e.g., Warren Buffett, etc.). However, the trading techniques of investment experts are difficult to define in terms of a few elements.

[0006] For example, although the trading techniques of investment experts can be quantified to some extent through a rule-based model or a quant model, information loss may occur.

[0007] For example, it is also possible to analyze the profit and loss factors for each trading element of investment experts through a portfolio analysis tool, but this is an indirect method and it is difficult to accurately derive the trading techniques of investment experts.

Prior Art Documents

Patent Documents

[0008]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0009] The present invention is for solving the problems of the above-described prior art, and aims to provide investment guidance information through an investment imitation ensemble model based on a plurality of learning models learned based on investment-related information about investment experts.

[0010] However, the technical problems to be solved by this embodiment are not limited to the technical problems as described above, and other technical problems may exist.

Means for Solving the Problems

[0011] As means for solving the above-described technical problems, a server that provides investment guidance according to a first aspect of the present invention includes a collection unit that collects investment-related information about investment experts, a learning data generation unit that generates learning data for a plurality of learning models based on the collected investment-related information, a learning unit that learns the plurality of learning models based on the generated learning data, an ensemble model generation unit that generates an investment imitation ensemble model about the investment experts based on the learned plurality of learning models, and an investment guidance information providing unit that provides investment guidance information to an investor terminal using the investment imitation ensemble model, and the plurality of learning models may include a stock price model, a financial model, and an economic model.

[0012] A method for providing an investment guide executed by an investment guide providing server according to a second aspect of the present invention includes the steps of collecting investment-related information about investment experts, generating learning data for a plurality of learning models based on the collected investment-related information, training the plurality of learning models based on the generated learning data, generating an investment imitation ensemble model for the investment experts based on the trained plurality of learning models, and providing investment guide information to an investor terminal using the investment imitation ensemble model, and the plurality of learning models may include a stock price model, a financial model, and an economic model.

[0013] A computer program stored in a computer-readable recording medium including a sequence of instruction codes for providing an investment guide according to a third aspect of the present invention, when executed by a computer device, collects investment-related information about investment experts, generates learning data for a plurality of learning models based on the collected investment-related information, trains the plurality of learning models based on the generated learning data, generates an investment imitation ensemble model for the investment experts based on the trained plurality of learning models, provides investment guide information to an investor terminal using the investment imitation ensemble model, and the plurality of learning models may include a sequence of instruction codes including a stock price model, a financial model, and an economic model.

[0014] The means for solving the above-described problems are merely illustrative and should not be construed as limiting the present invention. In addition to the above-described exemplary embodiments, there may be additional embodiments described in the drawings and the detailed description of the invention.

Effect of the Invention

[0015] According to any one of the means for solving the problems of the present invention described above, the present invention can train a plurality of learning models based on investment-related information about investment experts and provide investment guide information through an investment imitation ensemble model based on the trained plurality of learning models.

[0016] As a result, since the investment imitation ensemble model of the present invention provides a judgment similar to the judgment considered by investment experts at the time of buying and selling, even individual investors with little investment knowledge can, through the investment imitation ensemble model that has learned the investment style of investment experts, be helped to invest in a manner similar to the investment power of investment experts.

Brief Description of the Drawings

[0017]

Figure 1

Figure 2a

Figure 2b

Figure 3

Modes for Carrying Out the Invention

[0018] Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those having ordinary knowledge in the technical field to which the present invention pertains can easily implement it. However, the present invention can be embodied in various different forms and is not limited to the embodiments described herein. And in the drawings, in order to clearly explain the present invention, parts not related to the explanation are omitted, and similar reference numerals are given to similar parts throughout the specification.

[0019] Throughout the specification, when a part is described as "connected" to another part, this includes not only the case where it is "directly connected", but also the case where it is "electrically connected" with other elements interposed therebetween. Also, when a part is described as "including" a certain component, this means that, unless otherwise stated to the contrary, it may further include other components, rather than excluding other components.

[0020] As used herein, the term "unit" includes a unit implemented by hardware, a unit implemented by software, and a unit implemented using both. Also, one unit may be implemented using two or more pieces of hardware, and two or more units may be implemented by one piece of hardware.

[0021] In this specification, some of the operations or functions described as being performed by a terminal or device may instead be performed by a server connected to the terminal or device. Similarly, some of the operations or functions described as being performed by a server may also be performed by a terminal or device connected to the server.

[0022] Hereinafter, specific details for implementing the present invention will be described with reference to the attached configuration diagrams or process flowcharts.

[0023] FIG. 1 is a block diagram of an investment guide providing server 10 according to an embodiment of the present invention.

[0024] Referring to FIG. 1, the investment guide providing server 10 may include a collection unit 100, a learning data generation unit 110, a learning unit 120, an ensemble model generation unit 130, an investment guide information providing unit 140, and an ensemble model providing unit 150. However, the investment guide providing server 10 shown in FIG. 1 is merely an embodiment of the present invention, and various modifications are possible based on the components shown in FIG. 1.

[0025] The collection unit 100 may collect investment-related information about investment experts (such as fund managers, etc.) from the servers of securities companies.

[0026] Here, the investment-related information may include asset information regarding at least one investment brand held by the investment expert, and transaction details and transaction index information regarding the investment brands held. Here, the asset information regarding at least one investment brand may include the holding ratio information regarding the investment brand on a daily basis.

[0027] For example, the collection unit 100 may collect the first investment-related information about the first investment expert and the first investment-related information about the second investment expert.

[0028] The learning data generation unit 110 may generate learning data for a plurality of learning models based on the collected investment-related information about the investment experts. Here, the plurality of learning models may include a stock price model, a financial model, and an economic model.

[0029] The learning data generation unit 110 may respectively generate learning data for a plurality of learning models for each investment expert based on the investment-related information collected for each investment expert. For example, the learning data generation unit 110 may use the investment-related information of the first investment expert to generate learning data for a plurality of learning models for learning the trading style (such as trading positions and holding ratios of investment brands, etc.) of the first investment expert, and use the investment-related information of the second investment expert to generate learning data for a plurality of learning models for learning the trading style of the second investment expert.

[0030] The learning data generation unit 110 may derive stock price related index information from the collected investment related information about investment experts, and generate first learning data for learning the stock price model based on the derived stock price related index information. Here, the stock price related index information may include, for example, price earnings ratio, trading volume, standard deviation of stock price, and moving average line (MV, Moving Average) of stock price. Alternatively, the stock price related index information may further include, for example, 5-day earnings ratio (e.g., 10-day, 20-day, 60-day, 120-day earnings ratio), moving average line (MA5(20, 60, 120)) information, exponential moving average line (EMA5(20, 60, 120)) information, trading volume information, OBV (On Balance Volume) information, PVT (Price Volume Trend) information, AD (Accumulation Distribution) information, separation degree information, RSI (Relative Strength Index) information, etc.

[0031] The learning data generation unit 110 may derive financial related ratio information from the collected investment related information, and generate second learning data for learning the financial model based on the derived financial related ratio information. Here, the financial related ratio information may include, for example, information between price earnings ratio (PER, Price Earning Ratio), return on total assets (ROA, Return On Asset), enterprise value (EV, Enterprise Value), and earnings before interest, tax, depreciation and amortization (EBITDA, Earnings Before Interest&Tex, Depreciation and Amortization). Also, the financial related ratio information may further include, for example, return on equity (ROE, Return On Equity) information, price to book value ratio (PBR, Price Book-value Ratio) information, price to sales ratio (PSR, Pice Sales Ratio) information, operating profit margin information, net profit margin information, sales growth rate information, debt ratio information, dividend payout ratio information, etc.

[0032] The learning data generation unit 110 may derive macroeconomic indicator-related information from the collected investment-related information and generate third learning data for learning an economic model based on the derived macroeconomic indicator-related information. Here, the macroeconomic indicator-related information may include, for example, exchange rate information, oil price information, interest rate information, price information, raw material information, and GOD economic growth rate information.

[0033] The learning unit 120 may cause a plurality of learning models to learn the trading styles of investment experts based on the learning data about the investment experts.

[0034] For example, referring to FIG. 2a, the learning unit 120 may input first learning data (data generated based on stock price-related index information) about a first investment expert into a stock price model 201 and cause the stock price model 201 to learn so as to predict the trading position information and the holding ratio information of investment stocks of the first investment expert through the stock price model 201.

[0035] Also, the learning unit 120 may input second learning data (data generated based on financial ratio information) about the first investment expert into a financial model 203 and cause the financial model 203 to learn so as to predict the trading position information and the holding ratio information of investment stocks of the first investment expert through the financial model 203.

[0036] Also, the learning unit 120 may input third learning data (data generated based on macroeconomic indicator-related information) about the first investment expert into an economic model 205 and cause the economic model 205 to learn so as to predict the trading position information and the holding ratio information of investment stocks of the first investment expert through the economic model 205.

[0037] The ensemble model generation unit 130 may generate an investment imitation ensemble model related to an investment expert based on the plurality of learned learning models.

[0038] For example, referring to FIG. 2b, the ensemble model generation unit 130 may generate an investment imitation ensemble model 207 for the first investment expert by using the stock price model 201, the financial model 203, and the economic model 205 in which the investment style of the first investment expert is learned.

[0039] The ensemble model generation unit 130 may set different weighting values for each of the plurality of learning models based on the investment orientation of the investment expert. The investment characteristics of the investment expert can be grasped through the weighting values given to each learning model.

[0040] For example, when the weighting value of the stock price model is set higher than the weighting values of other learning models (financial model, economic model), it can be grasped that the investment expert mainly determines whether to invest while checking the movement of the stock price.

[0041] The ensemble model generation unit 130 may generate an investment imitation ensemble model that imitates the trading style of the investment expert based on the plurality of learning models and the weighting values set for each of the plurality of learning models.

[0042] The ensemble model generation unit 130 may generate an investment imitation ensemble model that imitates the trading style of each investment expert for each of the plurality of investment experts.

[0043] Here, the investment imitation ensemble model may be expressed as in [Equation 1].

[0044] [Equation 1] Investment imitation ensemble model = Stock price model * W1 + Financial model * W2 + Economic model * W3

[0045] When current stock price data, financial data, and economic data are input to the plurality of learning models, the output values (trading position information and holding ratio information of investment stocks) derived from each learning model may be used as input values of the investment imitation ensemble model.

[0046] For example, referring to FIG. 2b, the first result value of the multiplication operation between the current stock price data and the first weighting value set in the stock price model 201 is input to the investment simulation ensemble model 207, and the second result value of the multiplication operation between the current financial data and the second weighting value set in the financial model 203 is input to the investment simulation ensemble model 207. The third result value of the multiplication operation between the economic data and the third weighting value set in the economic model 205 may also be input to the investment simulation ensemble model 207.

[0047] Based on the first result value, the second result value, and the third result value, the investment simulation ensemble model 207 may derive the final trading position information reflecting the investment style of the first investment expert and the holding ratio information of the investment stocks.

[0048] The investment guidance information providing unit 140 may provide investment guidance information to the investor terminal by using the investment simulation ensemble model. Here, the investment guidance information is a guide that mimics the investment of the investment expert in the transaction of the investment stock, and may include the transaction position information regarding the investment stock and the holding ratio information of the investment stock.

[0049] For example, based on the final trading position information and the holding ratio information derived from the investment simulation ensemble model 207 that mimics the trading style of the first investment expert, the investment guidance information providing unit 140 may derive investment guidance information (guidance information that presents the investment stocks with high delivery / acquisition probability and proposes the holding ratio of the investment stocks).

[0050] The ensemble model providing unit 150 may provide a plurality of investment simulation ensemble models regarding a plurality of investment experts to the investor terminal.

[0051] The investment guidance information providing unit 140 may be selected from the investor terminal at least one of the plurality of investment simulation ensemble models, and provide investment guidance information to the investor terminal based on the selected investment simulation ensemble model.

[0052] The investment guidance information providing unit 140 may derive investment guidance through the selected investment simulation ensemble model by inputting current stock price data, financial data, and economic data into a plurality of learning models included in the investment simulation ensemble model selected by the investor terminal, and using the output values derived from each learning model as input values for the selected investment simulation ensemble model.

[0053] When at least two or more investment simulation ensemble models are selected from the investor terminal, the investment guidance information providing unit 140 may provide the investor terminal with investment guidance information for each of the selected two or more investment simulation ensemble models derived through each of the selected two or more investment simulation ensemble models.

[0054] When the transaction position information regarding the investment brands included in the investment guidance information for each of the selected two or more investment simulation ensemble models is mutually contradictory, the investment guidance information providing unit 140 may calculate the probabilities for each of the selected two or more investment simulations using the softmax function. Here, the softmax function is an activation function used in multi-class classification for classifying three or more classes. When the number of classes to be classified is N, an N-dimensional vector is input to estimate the probability of belonging to each class. The output value of the softmax function is a value between 0 and 1, all of which are normalized values, and the sum of the output values is always 1.

[0055] When the transaction position information regarding the investment brands included in the investment guidance information for each of the selected two or more investment simulation ensemble models is mutually contradictory, the investment guidance information providing unit 140 may provide the investor terminal with the investment guidance information derived through the investment simulation ensemble with a high probability among the selected two or more investment simulations.

[0056] For example, when the first investment simulation ensemble model predicts an acquisition with a probability of 0.5 and the second investment simulation ensemble model predicts a delivery with a probability of 0.7, the investment guide information providing unit 140 may select the second investment simulation ensemble model that predicted the highest probability among the first investment simulation ensemble model and the second investment simulation ensemble model, and provide the investment guide information derived through the second investment simulation ensemble model to the investor terminal.

[0057] When the transaction position information regarding the investment brands included in the investment guide information for two or more selected investment simulation ensemble models is mutually contradictory, the investment guide information providing unit 140 may select, from the investor terminal, the item preferred by the investor among the stock price item, financial item, and economic item, select an investment simulation ensemble based on the selected item, and provide the investment guide information derived through the selected investment simulation ensemble model to the investor terminal.

[0058] On the other hand, those skilled in the art should be able to fully understand that each of the collection unit 100, learning data generation unit 110, learning unit 120, ensemble model generation unit 130, investment guide information providing unit 140, and ensemble model providing unit 150 can be embodied separately, or one or more of them can be integrated and embodied.

[0059] FIG. 3 is a flowchart showing a method for providing an investment guide according to an embodiment of the present invention.

[0060] Referring to FIG. 3, in step S301, the investment guide providing server 10 may collect investment-related information regarding investment experts.

[0061] In step S303, the investment guide providing server 10 may generate learning data for a plurality of learning models based on the collected investment-related information. Here, the plurality of learning models may include a stock price model, a financial model, and an economic model.

[0062] In step S305, the investment guidance providing server 10 may train a plurality of learning models based on the generated learning data.

[0063] In step S307, the investment guidance providing server 10 may generate an investment imitation ensemble model regarding investment experts based on the trained plurality of learning models.

[0064] In step S309, the investment guidance providing server 10 may provide investment guidance information to the investor terminal by using the generated investment imitation ensemble model.

[0065] In the above description, steps S301 to S309 may be further divided into additional steps or combined into fewer steps according to embodiments of the present invention. Also, some steps may be omitted as necessary, and the order between steps may be changed.

[0066] An embodiment of the present invention may also be embodied in the form of a recording medium including computer-executable instructions such as program modules executed by a computer. The computer-readable medium may be any available medium accessible by a computer, including all volatile and non-volatile media, and all separable and non-separable media. Also, the computer-readable medium may include all computer storage media. The computer storage media includes all volatile and non-volatile, separable and non-separable media embodied in any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data.

[0067] The above description of the present invention is for illustrative purposes, and those with ordinary knowledge in the technical field to which the present invention pertains should be able to understand that it can be easily deformed into other specific forms without changing the technical idea and essential features of the present invention. Therefore, it should be understood that the above-described embodiments are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented dispersedly, and similarly, the components described as being dispersed may also be implemented in a combined form.

[0068] The scope of the present invention is indicated by the claims described below rather than the above detailed description, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be construed as being included in the scope of the present invention.

Claims

1. In a server that provides an investment guide, a collection unit that collects investment-related information about investment experts; a learning data generation unit that generates learning data for a plurality of learning models based on the collected investment-related information; a learning unit that learns the plurality of learning models based on the generated learning data; an ensemble model generation unit that generates an investment imitation ensemble model for the investment experts based on the learned plurality of learning models; an ensemble model providing unit that provides a plurality of the investment imitation ensemble models for the investment experts to an investor terminal; an investment guide information providing unit that provides investment guide information to the investor terminal using the investment imitation ensemble model; comprising the plurality of learning models include a stock price model, a financial model, and an economic model; the investment guide information providing unit when two or more of the plurality of investment imitation ensemble models are selected from the investor terminal and the trading position information regarding the investment stocks included in the investment guide information is contrary in each of the two or more investment imitation ensemble models, calculates the prediction probability for each of the selected two or more investment imitation ensemble models, and provides the investment guide information derived through one investment imitation ensemble model selected based on the calculated probability to the investor terminal. An investment guide providing server.

2. The investment-related information includes asset information regarding at least one investment stock held by the investment expert, the transaction breakdown regarding the held investment stock, and transaction index information. The investment guide providing server according to claim 1, wherein the asset information regarding the at least one investment stock includes the holding ratio information regarding the investment stock on a daily basis.

3. The learning data generation unit derives stock price related index information from the collected investment related information, and generates first learning data for learning the stock price model based on the derived stock price related index information. The investment guide providing server according to claim 1, wherein the stock price related index information includes a price earnings ratio, a trading volume, a standard deviation of the stock price, and a moving average line (MV, Moving Average) of the stock price.

4. The learning data generation unit derives financial ratio related information from the collected investment related information, and generates second learning data for learning the financial model based on the derived financial ratio related information. The investment guide providing server according to claim 1, wherein the financial ratio related information includes information on a price earnings ratio (PER, Price Earning Ratio), a return on total assets (ROA, Return On Asset), an enterprise value (EV, Enterprise Value), and earnings before interest, taxes, depreciation, and amortization (EBITDA, Earnings Before Interest & Tax, Depreciation and Amortization).

5. The learning data generation unit derives macroeconomic index related information from the collected investment related information, and generates third learning data for learning the economic model based on the derived macroeconomic index related information. The investment guide providing server according to claim 1, wherein the macroeconomic index related information includes exchange rate information, oil price information, interest rate information, price information, and GOD economic growth rate information.

6. The ensemble model generation unit generates the investment simulation ensemble model based on the plurality of learning models and the weighted values set for each of the plurality of learning models, When current stock price data, financial data, and economic data are input into the plurality of learning models, output values derived from each learning model are used as input values of the investment simulation ensemble model. The investment guide providing server according to claim 1.

7. The investment guide information providing unit inputs current stock price data, financial data, and economic data into a plurality of learning models included in the investment simulation ensemble model selected by the investor terminal, uses the output values derived from each learning model as input values of the selected investment simulation ensemble model to derive the investment guide information through the selected investment simulation ensemble model, The investment guide information includes transaction position information regarding the investment brand and holding ratio information of the investment brand. The investment guide providing server according to claim 1.

8. In a method of providing an investment guide executed by an investment guide providing server, collecting investment-related information regarding investment experts; generating learning data for a plurality of learning models based on the collected investment-related information; training the plurality of learning models based on the generated learning data; generating an investment simulation ensemble model regarding the investment experts based on the trained plurality of learning models; providing a plurality of the investment simulation ensemble models regarding the plurality of the investment experts to an investor terminal; and providing investment guide information to the investor terminal by using the investment simulation ensemble model, The plurality of learning models include a stock price model, a financial model, and an economic model, In the step of providing the investment guide information to the investor terminal, When two or more of the plurality of investment simulation ensemble models are selected from the investor terminal, and when the trading position information regarding the investment stocks included in the investment guidance information is contrary in each of the two or more investment simulation ensemble models, calculate the prediction probability for each of the selected two or more investment simulation ensemble models, and provide the investment guidance information derived through one investment simulation ensemble model selected based on the calculated probability to the investor terminal. An investment guidance providing method.

9. The step of generating learning data for the plurality of learning models includes: deriving macroeconomic indicator related information from the collected investment related information; generating third learning data for learning the economic model based on the derived macroeconomic indicator related information. The investment guidance providing method according to claim 8, wherein the macroeconomic indicator related information includes exchange rate information, oil price information, interest rate information, price information, and GOD economic growth rate information. The step of generating an investment simulation ensemble model for the investment expert includes:

10. generating the investment simulation ensemble model based on the plurality of learning models and the weighting values set for each of the plurality of learning models. When current stock price data, financial data, and economic data are input into the plurality of learning models, the output values derived from each learning model are used as input values for the investment simulation ensemble model. The investment guidance providing method according to claim 8. The step of providing investment guidance information to the investor terminal includes: inputting current stock price data, financial data, and economic data into a plurality of learning models included in the investment simulation ensemble model selected by the investor terminal;

11. The step of providing investment guidance information to the investor terminal includes: inputting current stock price data, financial data, and economic data into a plurality of learning models included in the investment simulation ensemble model selected by the investor terminal; Deriving the investment guidance information through the selected investment simulation ensemble model by using the output values derived from the respective learning models as input values of the selected investment simulation ensemble model, The investment guidance information includes trading position information regarding an investment brand and holding ratio information of the investment brand. The investment guidance providing method according to claim 8.

12. In a computer program stored in a computer-readable recording medium including a sequence of instruction words for providing investment guidance, When the computer program is executed by a computer device, Collecting investment-related information regarding investment experts, Generating learning data for a plurality of learning models based on the collected investment-related information, Training the plurality of learning models based on the generated learning data, Generating an investment simulation ensemble model regarding the investment experts based on the trained plurality of learning models, Providing a plurality of the investment simulation ensemble models regarding the plurality of the investment experts to an investor terminal, Providing investment guidance information to the investor terminal by using the investment simulation ensemble model, The plurality of learning models include a stock price model, a financial model, and an economic model, When providing the investment guidance information to the investor terminal, A computer program stored in a computer-readable recording medium, including a sequence of instruction words, wherein two or more of the plurality of investment simulation ensemble models are selected from the investor terminal, and when the trading position information regarding the investment targets included in the investment guidance information is contrary in each of the two or more investment simulation ensemble models, calculating the prediction probability for each of the two or more selected investment simulation ensemble models, and providing the investment guidance information derived through one investment simulation ensemble model selected based on the calculated probability to the investor terminal.

Images