Information processing program, information processing method, and information processing device

The information processing device addresses the increased workload in portfolio optimization by calculating weight parameters using a second-order polynomial model and Ising solver, enhancing efficiency and Sharpe ratio in investment pattern generation.

JP7583280B2Active Publication Date: 2024-11-14FUJITSU LTD
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
JP2021094800
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2021-06-04
Publication Date
2024-11-14
Estimated Expiration
2041-06-04

AI Technical Summary

Technical Problem

Conventional portfolio optimization techniques face increased workload when dealing with a large number of stocks, leading to complex optimization problems and difficulty in setting appropriate weight parameters.

Method used

An information processing device that calculates weight parameters using a second-order polynomial model and an Ising solver to optimize investment patterns, reducing the workload by identifying pairs of risk and return, and generating a QUBO formula to solve the optimization problem.

Benefits of technology

The device efficiently reduces the workload required for portfolio optimization by automatically setting appropriate weight parameters, enabling the generation of investment patterns with reduced calculation time and improved Sharpe ratio.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007583280000006
    Figure 0007583280000006
  • Figure 0007583280000007
    Figure 0007583280000007
  • Figure 0007583280000008
    Figure 0007583280000008
Patent Text Reader

Abstract

To reduce workload that is imposed when a portfolio is optimized.SOLUTION: An information processing apparatus 100 specifies a pair of a first risk and the maximum return to the first risk based on a secondary polynomial model 101. The information processing apparatus 100 calculates, based on the specified pair, a weight parameter λ that is included in a cost function 102 included in a mathematical model 103 and is applied to at least one item of a first item related to the risk and a second item related to the return. The information processing apparatus 100 generates the mathematical model 103 based on the calculated weight parameter λ. Based on the generated mathematical model 103, the information processing apparatus 100 solves an optimization problem represented by the mathematical model 103.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to an information processing program, an information processing method, and an information processing device. [Background technology]

[0002] Conventionally, there is a technology for optimizing a portfolio in the financial field. A portfolio is, for example, an asset allocation that indicates which stocks out of a plurality of stocks (financial trading targets) to invest in and which stocks out of a plurality of stocks to hold. Specifically, optimization is minimizing or maximizing an evaluation value calculated by weighting risk and return.

[0003] Prior art includes, for example, a technology that uses a problem model to design a conversion program that converts a first problem into a data format that can be input to a quantum-related computer. In addition, for example, there is a technology that formulates the processing content and processing time applied to each object as a QUBO model or an Ising model via a space-time network model. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2020-166802 A [Patent Document 2] JP 2020-140615 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, the conventional technology has a problem in that the workload required for optimizing a portfolio increases. For example, if the number of stocks is huge, the workload required for determining appropriate weights to be assigned to risk and return tends to increase, leading to an increase in the workload required for optimizing a portfolio.

[0006] In one aspect, the present invention aims to reduce the workload involved in optimizing a portfolio. [Means for solving the problem]

[0007] According to one embodiment, an information processing program, an information processing method, and an information processing device are proposed that identify a pair of a first risk and a first maximum return for the first risk based on a second-order polynomial model that indicates the relationship between a risk related to an investment in multiple trading objects and a maximum return for the risk, calculate a weight parameter for at least one of a first term related to risk and a second term related to return in a cost function including a first term related to risk and a second term related to return included in a mathematical model that represents an optimization problem related to an investment pattern for the multiple trading objects based on the identified pair, and generate the mathematical model based on the calculated weight parameter. Effect of the Invention

[0008] According to one embodiment, it is possible to reduce the workload involved in optimizing a portfolio. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating an example of an information processing method according to an embodiment. [Diagram 2] FIG. 2 is an explanatory diagram showing an example of an optimization processing system 200. As shown in FIG. [Diagram 3] FIG. 3 is a block diagram showing an example of a hardware configuration of the information processing device 100. As shown in FIG. [Figure 4] FIG. 4 is an explanatory diagram showing an example of a data structure 400 of an investment pattern. [Diagram 5] FIG. 5 is a diagram illustrating an example of the contents of the EF information management table 500. As shown in FIG. [Figure 6] FIG. 6 is a block diagram showing an example of a functional configuration of the information processing device 100. As shown in FIG. [Figure 7] FIG. 7 is an explanatory diagram (part 1) showing an operation example of the information processing device 100. [Figure 8] FIG. 8 is an explanatory diagram (part 2) showing an operation example of the information processing device 100. [Figure 9] FIG. 9 is an explanatory diagram (part 3) showing an operation example of the information processing device 100. [Figure 10] FIG. 10 is a flowchart illustrating an example of an overall processing procedure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of an information processing program, an information processing method, and an information processing device according to the present invention will be described in detail with reference to the drawings.

[0011] (An example of an information processing method according to an embodiment) 1 is an explanatory diagram showing an example of an information processing method according to an embodiment. The information processing device 100 is a computer for facilitating optimization of a portfolio. The information processing device 100 is, for example, a server or a PC (Personal Computer).

[0012] A portfolio is an asset allocation. A portfolio is, for example, an asset allocation indicating which stocks of a plurality of stocks to invest in and which stocks of a plurality of stocks to hold. A portfolio specifically indicates whether or not to invest in each of a plurality of stocks. A portfolio more specifically indicates whether or not to purchase each stock. A portfolio more specifically indicates whether or not to short sell each stock. A portfolio is specifically an investment pattern.

[0013] The optimization is performed by solving an optimization problem. The optimization problem includes a cost function for calculating an evaluation value of the portfolio. The evaluation value indicates, for example, the goodness of balance between risk and return of the portfolio. The cost function includes a first term related to risk and a second term related to return, both of which are multiplied by a weighting parameter. The cost function is defined, for example, by the following formula (1).

[0014] min(λΣ i Σ j w i w j Σ ij -(1-λ)(Σ i w i μ i )) ···(1)

[0015] λ is a weight parameter. λ∈[0,1]⊂R. w i is the investment ratio for the i-th stock. w j is the investment ratio for the jth stock. μ i is the forecast value of the return of the i-th stock in the T period. μ i is defined, for example, by the following formula (2): Σ ij is the variance-covariance matrix. Σ ij is defined, for example, by the following formula (3): Σ i Σ j w i w j Σ ij is the first term related to risk. Σ i w i μ i is the second term in terms of returns.

[0016] μ i =1 / T(Σ t=1 T R i,t ) ···(2)

[0017] Σ ij =1 / T(Σ t=1 T (R i,t -μ i )(Rj,t -μ j ))=1 / T(Σ t=1 T (R i,t R j,t -μ i μ j )) ···(3)

[0018] R i,t is the forecast value of the return for the i-th stock at time t. R i,t is defined by the following equation (4).

[0019] R i,t =(P i,t -P i,t-1 ) / P i,t-1 (4)

[0020] Optimization, for example, is to maximize a return from a portfolio within an acceptable risk. Optimization, for example, is to minimize a risk from a portfolio while obtaining a specified return. Specifically, optimization is to minimize or maximize an evaluation value calculated by a cost function. Specifically, optimization is to generate an appropriate portfolio in which an evaluation value calculated by a cost function is minimized or maximized.

[0021] In order to solve the optimization problem and optimize the portfolio, it is desirable to appropriately set weight parameters included in the cost function. Conventionally, the weight parameters have been set manually, for example.

[0022] Conventionally, there is a problem in that the workload required to solve the optimization problem and optimize the portfolio increases. For example, the more complex the optimization problem becomes, the more likely it is that the workload required to solve the optimization problem and optimize the portfolio increases.

[0023] For example, if the number of stocks is huge, the optimization problem tends to become complicated. For example, if the portfolio indicates both whether or not to buy a stock and whether or not to short sell a stock, the optimization problem tends to become complicated. For example, if the number of local solutions becomes huge, the optimization problem tends to become complicated.

[0024] Specifically, the more complex the optimization problem, the more likely it is that the optimization problem will fall into a local solution when solved, and therefore, unless a huge number of weight parameters are tried, the weight parameters cannot be set appropriately and the optimal solution cannot be obtained. In addition, when the optimization problem includes discrete constraints or constraint conditions related to short selling, etc., λ tends to become scattered.

[0025] In this embodiment, an information processing method that can reduce the workload involved in solving an optimization problem and optimizing a portfolio will be described.

[0026] In FIG. 1, the information processing device 100 has a template of a mathematical model 103 expressing an optimization problem regarding an investment pattern for a first group of stocks. The template is the mathematical model 103 for which various parameters are undetermined. The mathematical model 103 includes a cost function 102 including a first term related to risk and a second term related to return. In the cost function 102, a weight parameter is applied to at least one of the first term and the second term. The mathematical model 103 is, for example, a QUBO (Quadratic Unconstrained Binary Optimization) formula. The first group of stocks is, for example, a group of stocks corresponding to the present. Specifically, the first group of stocks is a group of stocks that currently exist.

[0027] The information processing device 100 acquires a quadratic polynomial model 101. The quadratic polynomial model 101 indicates the relationship between the risk associated with an investment in a second group of stocks and the maximum return for that risk. The second group of stocks is of the same type as the first group of stocks. The second group of stocks is, for example, the same group of stocks as the first group of stocks. The second group of stocks may be partially different from the first group of stocks. The second group of stocks corresponds to, for example, a period in the past that is relatively close to the present. Specifically, the second group of stocks is a group of stocks that existed in a period in the past that is relatively close to the present.

[0028] Specifically, the quadratic polynomial model 101 indicates an Efficient Frontier. The Efficient Frontier corresponds to a characteristic curve 111 in a graph 110 that indicates the relationship between the risk related to an investment and the maximum return for the risk. In the graph 110, the horizontal axis indicates the risk, and the vertical axis indicates the return. In the following description, the Efficient Frontier may be written as "EF". The information processing device 100 obtains the Efficient Frontier by, for example, generating the quadratic polynomial model 101.

[0029] (1-1) The information processing device 100 identifies a pair of a first risk and a maximum return for the first risk based on the quadratic polynomial model 101. The information processing device 100 identifies a pair of a first risk and a maximum return for the first risk, which maximizes the Sharpe ratio SR, based on, for example, the quadratic polynomial model 101. The Sharpe ratio SR is defined, for example, by the following formula (5).

[0030] SR=RR f / r ···(5)

[0031] R f is the risk-free return. R is the return due to the investment pattern. R is defined, for example, by the following formula (6). r is the risk due to the investment pattern. r is defined, for example, by the following formula (7) or formula (8).

[0032] return:R=Σi=1 N w i μ i (6)

[0033] risk: r = std(R) (7)

[0034] risk:r = var(R) = Σ i=1 N Σ j=1 N w i Σ ij w j (8)

[0035] (1-2) The information processing device 100 calculates a weighting parameter λ based on the identified pair. The information processing device 100 calculates a weighting parameter λ based on the identified pair by using, for example, the Lagrange equation.

[0036] (1-3) The information processing device 100 generates the mathematical expression model 103 based on the calculated weighting parameter λ. The information processing device 100 generates the mathematical expression model 103, for example, by setting the calculated weighting parameter λ in a template of the mathematical expression model 103.

[0037] (1-4) The information processing device 100 solves an optimization problem represented by the mathematical model 103 based on the generated mathematical model 103. The information processing device 100 uses, for example, an Ising solver to solve the optimization problem based on the generated mathematical model 103, and generates an appropriate investment pattern.

[0038] This allows the information processing device 100 to set appropriate weight parameters, thereby reducing the workload involved in solving the optimization problem and generating an appropriate investment pattern. The information processing device 100 can easily calculate useful weight parameters based on a pair of risk and maximum return on the EF, for example, without manually trying out various weight parameters.

[0039] A useful weight parameter is a weight parameter that makes it easier to obtain a useful investment pattern when solving an optimization problem using the weight parameter. A useful investment pattern is, for example, an investment pattern with a relatively small risk and a relatively large maximum return. As a result, the information processing device 100 can reduce the workload involved in solving an optimization problem to generate an appropriate investment pattern by, for example, reducing the workload involved in setting appropriate weight parameters.

[0040] Here, the case where the information processing device 100 operates independently has been described, but the present invention is not limited thereto. For example, the case where the information processing device 100 generates the quadratic polynomial model 101 and the case where the information processing device 100 solves an optimization problem related to an investment pattern have been described, but the present invention is not limited thereto.

[0041] For example, the information processing device 100 may cooperate with another computer. Specifically, the information processing device 100 may acquire the quadratic polynomial model 101 from another computer that generates the quadratic polynomial model 101. Specifically, the information processing device 100 may transmit the generated mathematical expression model 103 to another computer that solves an optimization problem related to an investment pattern. More specifically, an optimization processing system 200 described later in FIG. 2 in which the information processing device 100 cooperates with another computer is conceivable.

[0042] (An example of the optimization processing system 200) Next, an example of an optimization processing system 200 to which the information processing device 100 shown in FIG. 1 is applied will be described with reference to FIG.

[0043] 2 is an explanatory diagram showing an example of an optimization processing system 200. In FIG. 2, the optimization processing system 200 includes an information processing device 100, a generating device 201, a solution-finding device 202, and a client device 203.

[0044] In the optimization processing system 200, the information processing device 100 and the generating device 201 are connected via a wired or wireless network 210. The network 210 is, for example, a LAN (Local Area Network), a WAN (Wide Area Network), the Internet, etc. Furthermore, the information processing device 100 and the solution finding device 202 are connected via the wired or wireless network 210. Furthermore, the information processing device 100 and the client device 203 are connected via the wired or wireless network 210.

[0045] The information processing device 100 receives a request to generate an optimal investment pattern for a first group of stocks from the client device 203. The information processing device 100 receives a quadratic polynomial model indicating EF from the generation device 201. Upon receiving the request, the information processing device 100 calculates a weight parameter λ of the QUBO formula based on the quadratic polynomial model. The information processing device 100 generates a QUBO formula based on the calculated weight parameter λ.

[0046] The information processing device 100 transmits the generated QUBO formula to the solution-finding device 202. The information processing device 100 receives the investment pattern obtained by solving the QUBO formula from the solution-finding device 202. The data structure 400 of the investment pattern will be described later with reference to FIG. 4, for example. The information processing device 100 transmits the received investment pattern to the client device 203.

[0047] The generating device 201 is a computer that generates a quadratic polynomial model indicating EF. The generating device 201 stores an EF information management table 500, which will be described later in FIG. 5, in which a plurality of pieces of information indicating pairs of risk and maximum return for the risk in the second group of issues in the past are recorded. The generating device 201 generates a quadratic polynomial model based on the EF information management table 500, which will be described later in FIG. 5. The generating device 201 transmits the generated quadratic polynomial model to the information processing device 100. The generating device 201 is, for example, a server or a PC.

[0048] The solution-finding device 202 is a computer that solves the QUBO equation and generates an appropriate investment pattern. The solution-finding device 202 has an Ising solver. The solution-finding device 202 receives the QUBO equation from the information processing device 100. The solution-finding device 202 uses the Ising solver to solve the received QUBO equation and generate an appropriate investment pattern. The solution-finding device 202 transmits the generated investment pattern to the information processing device 100. The solution-finding device 202 is, for example, a server or a PC.

[0049] The client device 203 transmits a request to generate an optimal investment pattern to the information processing device 100 based on an operation input from the client. The client device 203 receives the investment pattern from the information processing device 100. The client device 203 outputs the received investment pattern so that the client can refer to it. The client device 203 is, for example, a PC, a tablet terminal, or a smartphone.

[0050] (Example of hardware configuration of information processing device 100) Next, an example of the hardware configuration of the information processing device 100 will be described with reference to FIG.

[0051] Fig. 3 is a block diagram showing an example of a hardware configuration of the information processing device 100. In Fig. 3, the information processing device 100 has a CPU (Central Processing Unit) 301, a memory 302, a network I / F (Interface) 303, a recording medium I / F 304, and a recording medium 305. In addition, each component is connected to each other via a bus 300.

[0052] Here, the CPU 301 is responsible for the overall control of the information processing device 100. The memory 302 has, for example, a read only memory (ROM), a random access memory (RAM), and a flash ROM. Specifically, for example, the flash ROM and the ROM store various programs, and the RAM is used as a work area for the CPU 301. The programs stored in the memory 302 are loaded into the CPU 301, causing the CPU 301 to execute the coded processes.

[0053] The network I / F 303 is connected to the network 210 through a communication line, and is connected to other computers via the network 210. The network I / F 303 manages an internal interface with the network 210, and controls input and output of data from other computers. The network I / F 303 is, for example, a modem or a LAN adapter.

[0054] The recording medium I / F 304 controls reading / writing data from / to the recording medium 305 under the control of the CPU 301. The recording medium I / F 304 is, for example, a disk drive, a solid state drive (SSD), a universal serial bus (USB) port, etc. The recording medium 305 is a non-volatile memory that stores data written under the control of the recording medium I / F 304. The recording medium 305 is, for example, a disk, a semiconductor memory, a USB memory, etc. The recording medium 305 may be detachable from the information processing device 100.

[0055] In addition to the above-mentioned components, the information processing device 100 may have, for example, a keyboard, a mouse, a display, a printer, a scanner, a microphone, a speaker, etc. Furthermore, the information processing device 100 may have a plurality of recording medium I / Fs 304 and recording media 305. Furthermore, the information processing device 100 may not have the recording medium I / Fs 304 and the recording media 305.

[0056] (Example of Hardware Configuration of Generation Device 201) Specifically, an example of the hardware configuration of the generating device 201 is similar to the example of the hardware configuration of the information processing device 100 shown in FIG. 3, and therefore a description thereof will be omitted.

[0057] (Example of Hardware Configuration of Solution Finding Device 202) Specifically, an example of the hardware configuration of the solution-finding device 202 is similar to the example of the hardware configuration of the information processing device 100 shown in FIG. 3, and therefore a description thereof will be omitted.

[0058] (Example of Hardware Configuration of Client Device 203) Specifically, an example of the hardware configuration of the client device 203 is similar to the example of the hardware configuration of the information processing device 100 shown in FIG. 3, and therefore a description thereof will be omitted.

[0059] In the following description, the case where the information processing device 100 operates independently without cooperating with other computers will be mainly described. For this reason, the information processing device 100 has an EF information management table 500 described later in FIG. 5, and generates a quadratic polynomial model. In addition, the information processing device 100 solves the generated QUBO equation and generates an appropriate investment pattern.

[0060] (Investment pattern data structure 400) Next, an example of the investment pattern data structure 400 will be described with reference to FIG.

[0061] Fig. 4 is an explanatory diagram showing an example of a data structure 400 of an investment pattern. As shown in Fig. 4, an investment pattern has a field for stock i, where i is an integer from 1 to N. An investment pattern is formed by setting information in each field for each stock.

[0062] In the field of stock i, flag information is set that indicates whether or not to purchase the i-th stock and whether or not to short sell it. For example, if the flag information is 1, it indicates that purchase will be made. For example, if the flag information is -1, it indicates that short selling will be made. For example, if the flag information is 0, it indicates that no investment will be made, and specifically, it indicates that neither purchase nor short selling will be made.

[0063] Here, buying is assumed to be equal investment. Short selling is assumed to be equal investment. Buying and short selling are assumed to be not carried out simultaneously for the same stock i. The investment pattern corresponds to the points shown on graph 410, for example. In graph 410, the horizontal axis represents risk and the vertical axis represents return. Risk and return arise depending on the investment pattern.

[0064] Here, the flag information is either +1, 0, or -1, and indicates whether or not to buy the i-th stock and whether or not to short sell the i-th stock, but this is not limiting. For example, the flag information may be either 1 or 0, and may indicate whether or not to invest in the i-th stock.

[0065] For example, an investment pattern may include flag information corresponding to stocks for which buying is available and flag information corresponding to stocks for which short selling is available, separately. Specifically, an investment pattern has a field for stock i, where i=1 to 2N. In the field for stock i, where i=1 to N, flag information is set indicating whether or not to invest in the i-th stock by buying. In the field for stock i, where i=N+1 to 2N, flag information is set indicating whether or not to invest in the i-th stock by short selling. The x-th stock and the N+x-th stock are, for example, the same stock.

[0066] (Storage contents of EF information management table 500) Next, an example of the contents stored in the EF information management table 500 will be described with reference to Fig. 5. The EF information management table 500 is realized by, for example, a storage area such as the memory 302 or the recording medium 305 of the information processing device 100 shown in Fig. 3.

[0067] Fig. 5 is an explanatory diagram showing an example of the contents stored in the EF information management table 500. As shown in Fig. 5, the EF information management table 500 has fields of return R_x and risk r_x, where x is 1 to Nd. The EF information management table 500 stores EF information by setting information in each field for each pair of risk and return. In the field of return R_x and risk r_x, a pair of a cost related to past investment in the second group of stocks and a maximum return for that cost is set.

[0068] (Example of functional configuration of information processing device 100) Next, an example of a functional configuration of the information processing device 100 will be described with reference to FIG.

[0069] 6 is a block diagram showing an example of a functional configuration of the information processing device 100. The information processing device 100 includes a storage unit 600, an acquisition unit 601, an identification unit 602, a generation unit 603, a calculation unit 604, and an output unit 605.

[0070] The storage unit 600 is realized by, for example, a storage area such as the memory 302 or the recording medium 305 shown in Fig. 3. In the following, a case where the storage unit 600 is included in the information processing device 100 will be described, but this is not limiting. For example, the storage unit 600 may be included in a device different from the information processing device 100, and the stored contents of the storage unit 600 may be accessible from the information processing device 100.

[0071] The acquiring unit 601 to the output unit 605 function as an example of a control unit. Specifically, the acquiring unit 601 to the output unit 605 realize their functions by, for example, causing the CPU 301 to execute a program stored in a storage area such as the memory 302 or the recording medium 305 shown in Fig. 3, or by the network I / F 303. The processing results of each functional unit are stored in, for example, a storage area such as the memory 302 or the recording medium 305 shown in Fig. 3.

[0072] The memory unit 600 stores various information that is referenced or updated in the processing of each functional unit. The memory unit 600 stores financial trading objects. Financial trading objects are, for example, stocks. Stocks are investment items. Stocks are shares. The memory unit 600 stores a first group of stocks corresponding to the present. The memory unit 600 stores a second group of stocks corresponding to the past. The second group of stocks is of the same type as the first group of stocks. The stocks are, for example, stocks listed on the First Section of the Tokyo Stock Exchange. It is preferable that the second group of stocks corresponds to a period in the past that is relatively close to the present.

[0073] The storage unit 600 stores information indicating pairs of risks related to investments in a plurality of stocks in the past and maximum returns for the risks. Investments include buying and short selling. The maximum return for the risk is a return from any investment pattern that can be realized with the risk. The maximum return for the risk is, for example, a return from an investment pattern that maximizes the return among the investment patterns that can be realized with the risk. Specifically, the storage unit 600 stores the EF information management table 500 in which a plurality of pieces of information indicating pairs of risks related to investments in a second group of stocks and maximum returns for the risks are recorded. The information is, for example, acquired by the acquisition unit 601 and stored in the storage unit 600.

[0074] The storage unit 600 stores a quadratic polynomial model. The quadratic polynomial model indicates a relationship between a risk associated with an investment in a plurality of stocks and a maximum return for the risk. The quadratic polynomial model indicates, for example, a relationship between a risk associated with an investment in a second group of stocks and a maximum return for the risk. The quadratic polynomial model specifically indicates EF. The storage unit 600 stores, for example, parameters of the quadratic polynomial model. The quadratic polynomial model is acquired by the acquisition unit 601, for example.

[0075] The memory unit 600 stores an investment pattern for a plurality of stocks. The investment pattern indicates buying or short selling for a stock. The memory unit 600 stores, for example, an investment pattern for a first group of stocks. The investment pattern indicates whether or not to invest in a stock included in the first group of stocks. If an investment is made in a stock included in the first group of stocks, the investment pattern indicates whether to buy or short sell the stock. Specifically, the memory unit 600 stores the investment pattern of the data structure 400 shown in FIG. 4.

[0076] The storage unit 600 stores a mathematical model expressing an optimization problem regarding an investment pattern. The mathematical model includes, for example, a cost function including a first term related to risk and a second term related to return. The mathematical model may also include, for example, a constraint condition indicating that overlapping buying and short selling of the same stock included in the first stock group is prohibited. The mathematical model may also include, for example, a constraint condition regarding the number of stocks to be bought in the first stock group. The mathematical model may also include, for example, a constraint condition regarding the number of stocks to be short sold in the first stock group. Specifically, the storage unit 600 stores a weight parameter related to at least one of the first and second terms in the cost function. The weight parameter is calculated, for example, by the generation unit 603 and stored in the storage unit 600.

[0077] The acquisition unit 601 acquires various information used for processing by each functional unit. The acquisition unit 601 stores the acquired various information in the storage unit 600 or outputs it to each functional unit. The acquisition unit 601 may also output the various information stored in the storage unit 600 to each functional unit. The acquisition unit 601 acquires various information based on, for example, an operation input by a user. The acquisition unit 601 may receive various information from, for example, a device different from the information processing device 100.

[0078] The acquiring unit 601 acquires a quadratic polynomial model. For example, the acquiring unit 601 acquires the quadratic polynomial model by receiving the quadratic polynomial model from another computer that generates the quadratic polynomial model. The other computer is, for example, the generating device 201. The acquiring unit 601 acquires the quadratic polynomial model by accepting an input of the quadratic polynomial model based on an operational input by a user.

[0079] The acquisition unit 601 acquires the parameters of the quadratic polynomial model by generating the quadratic polynomial model based on information indicating a pair of risk related to investments in a plurality of stocks in the past and a maximum return for the risk. Specifically, the acquisition unit 601 calculates parameters of the quadratic polynomial model based on a plurality of pieces of information indicating pairs of risk related to investments in a plurality of stocks in the past and a maximum return for the risk.

[0080] The acquiring unit 601 may receive a request to solve an optimization problem related to an investment pattern. The acquiring unit 601 receives the request to solve an optimization problem related to an investment pattern based on, for example, an operation input by a user. The acquiring unit 601 receives the request to solve an optimization problem related to an investment pattern from, for example, another computer. The other computer is, for example, the client device 203.

[0081] The optimization problem is to optimize an investment pattern. For example, optimization is to maximize a return within an acceptable risk. Specifically, optimization is to minimize or maximize an evaluation value calculated by weighting risk and return. More specifically, optimization is to minimize or maximize a value of a cost function.

[0082] The acquiring unit 601 may accept a designation of a return to be used by the identifying unit 602. The acquiring unit 601 accepts the designation of a return based on, for example, an operational input by a user. The acquiring unit 601 accepts the designation of a return from, for example, another computer. The other computer is, for example, the client device 203. The designated return may be included, for example, in a request to solve an optimization problem related to an investment pattern.

[0083] The acquisition unit 601 may receive a start trigger for starting processing of any of the functional units. The start trigger may be, for example, a predetermined operation input by a user. The start trigger may be, for example, reception of predetermined information from another computer. The start trigger may be, for example, output of predetermined information by any of the functional units.

[0084] Specifically, the acquiring unit 601 may accept the acquisition of a quadratic polynomial model as a start trigger for starting the processing of the identifying unit 602, the generating unit 603, and the calculating unit 604. Specifically, the acquiring unit 601 may accept the acquisition of a request to solve an optimization problem related to an investment pattern as a start trigger for starting the processing of the identifying unit 602, the generating unit 603, and the calculating unit 604. Specifically, the acquiring unit 601 may accept the acceptance of a designation of a return as a start trigger for starting the processing of the identifying unit 602, the generating unit 603, and the calculating unit 604.

[0085] The identification unit 602 identifies one or more pairs of a first risk and a maximum return for the first risk based on a quadratic polynomial model. For example, the identification unit 602 identifies a pair of a first risk and a maximum return for the first risk that maximizes the Sharpe ratio based on a quadratic polynomial model. This allows the identification unit 602 to obtain information that can serve as a guideline for calculating the weight parameter.

[0086] The identification unit 602, for example, based on a quadratic polynomial model and a specified return, identifies a first risk corresponding to the specified return, and identifies a pair of the identified first risk and the specified return. Specifically, the identification unit 602 inputs the specified return into the quadratic polynomial model, thereby identifying the first risk corresponding to the specified return, and identifying a pair of the identified first risk and the specified return. This allows the identification unit 602 to obtain information that can be a guide for calculating the weight parameter. The identification unit 602 may randomly identify one or more pairs of the first risk and the maximum return for the first risk, for example, based on the quadratic polynomial model.

[0087] The generating unit 603 calculates a weighting parameter for at least one of the first term and the second term in the cost function based on the identified pair. The generating unit 603 calculates the weighting parameter by using, for example, a Lagrangian function. This enables the generating unit 603 to generate a mathematical expression model.

[0088] The generating unit 603 generates a mathematical model based on the calculated weighting parameters. For example, the generating unit 603 generates a cost function including a first term and a second term multiplied by the calculated weighting parameters, and generates a mathematical model including the generated cost function. This enables the generating unit 603 to solve an optimization problem related to an investment pattern.

[0089] The generating unit 603 may, for example, assign a random number to the calculated weight parameter. Then, the generating unit 603 generates a cost function including a first term and a second term multiplied by the weight parameter to which the random number is assigned, and generates a mathematical model including the generated cost function. This allows the generating unit 603 to make it possible to solve an optimization problem related to an investment pattern.

[0090] The calculation unit 604 solves an optimization problem regarding an investment pattern based on the generated mathematical model. The calculation unit 604 solves the optimization problem regarding the investment pattern using, for example, DA / FDAS, and obtains an investment pattern. In this way, the calculation unit 604 can obtain an investment pattern that is determined to be appropriate.

[0091] The output unit 605 outputs the processing result of at least one of the functional units. The output format is, for example, display on a display, printout on a printer, transmission to an external device via the network I / F 303, or storage in a storage area such as the memory 302 or the recording medium 305. In this way, the output unit 605 can notify the user of the processing result of at least one of the functional units, thereby improving the convenience of the information processing device 100.

[0092] (Operation example of information processing device 100) Next, an example of the operation of the information processing device 100 will be described with reference to FIGS.

[0093] 7 to 9 are explanatory diagrams showing an example of the operation of the information processing device 100. In Fig. 7, the information processing device 100 has an EF information management table 500. The information processing device 100 has a next data input unit 700, an EF prediction unit 701, a QUBO generation unit 702, and an optimization unit 703.

[0094] The next data input unit 700 accepts input such as current stock prices and outputs them to the QUBO generation unit 702. Here, it is assumed that the risk associated with the investment and the maximum return for that risk have a relationship represented by a quadratic function. The EF prediction unit 701 generates a quadratic polynomial model that indicates the relationship between the risk associated with the investment and the maximum return for that risk. The EF prediction unit 701 outputs the generated quadratic polynomial model to the QUBO generation unit 702.

[0095] The QUBO generation unit 702 obtains pairs of investment costs and maximum returns for the costs based on a quadratic polynomial model. The QUBO generation unit 702 obtains one or more pairs that maximize the Sharpe ratio based on a quadratic polynomial model. The QUBO generation unit 702 calculates various parameters of a0, a1, and a2 that appear in the following formula (9) by the least squares method based on the one or more obtained pairs.

[0096] r=a0+a1R+a2R 2 (9)

[0097] r is the risk due to the investment pattern. R is the return due to the investment pattern. a0 is a parameter. a1 is a parameter. a2 is a parameter.

[0098] The QUBO generation unit 702 has a template of the QUBO formula. The template of the QUBO formula includes, for example, a template of a cost function and a constraint condition. The template of the cost function is a cost function with undetermined parameters. The template of the cost function is, for example, a cost function with undetermined weight parameters for adjusting a term related to a cost and a term related to a return. The template of the cost function is, for example, defined by the following formula (10).

[0099] 1 / 2Σ i Σ j w i w j Σ ij -θλ'(Σ i w i μ j -R)^2 (10)

[0100] w i w indicates whether to invest in the i-th stock. i If w is 1, it indicates that you will invest in the i-th stock. i If it is 0, it indicates that there is no investment in the i-th stock. i is between 1 and 2N. N is the number of stocks available for investment. Stocks with i=1 to N are stocks that are available for buying. Stocks with i=N+1 to 2N are stocks that are available for short selling. Stock with i=x and stock with i=N+x are the same stock.

[0101] w j w indicates whether to invest in the jth stock. j If w is 1, it indicates that the investment is made in the jth stock. j If μ is 0, it means that you do not invest in the jth stock. j is 1 to 2N. i is the forecast of the return of the i-th stock in the T-th period. Σ ij is the variance-covariance matrix for returns. Σ i Σ j w i w j Σ ij is the first term related to risk. (Σ i w i μ i -R)^2 is the second term related to the return. R is defined by the following formula (11). λ' is defined by the following formula (12).

[0102] R = 2a0 / a1 (11)

[0103] λ'=1 / 2(a1-4a0a2), θ~U(0.5,1.5) ···(12)

[0104] θ is a random number. The constraint condition is defined as a penalty term, for example, by the following formula (13), formula (14), and formula (15). For example, the penalty term is added to the cost function so as to be subtracted from the cost function. Formula (13) below defines the number of stocks to be short sold as N s The following equation (14) shows the constraints for approximating the number of stocks to be purchased, N b The following equation (15) shows the constraint condition for preventing simultaneous purchase and short selling of the same stock.

[0105] (Σ i=1 2N s i w i -N s )^2, s i =1 if N <i≦2N else 0 ···(13)

[0106] (Σ i=1 2N b i w i -N b )^2, b i =1 if i≦N else 0 (14)

[0107] Σ i=1 2N Σ j=1 2N c ij w i w j , c ij =1 if i=N+j or j=N+i else 0 ···(15)

[0108] The QUBO generation unit 702 determines the cost function by setting the calculated parameters a0, a1, and a2 in the template of the cost function defined by the above formula (10), and generates a QUBO formula including the determined cost function. The QUBO generation unit 702 outputs the current stock price and the generated QUBO formula to the optimization unit 703.

[0109] The optimization unit 703 is realized by an Ising solver. The optimization unit 703 receives the current stock price and the QUBO formula. The optimization unit 703 generates an appropriate investment pattern by solving the QUBO formula based on the current stock price. The optimization unit 703 outputs the generated investment pattern. Next, moving to the explanation of FIG. 8, an example of the effect of the information processing device 100 will be explained.

[0110] 8, graph 800 shows pairs of risk and return according to an investment pattern. In graph 800, the horizontal axis represents risk and the vertical axis represents return. Black circles in graph 800 correspond to investment patterns generated by the conventional method. On the other hand, diagonal lines in graph 800 correspond to investment patterns generated by the information processing device 100. Specifically, the diagonal lines in graph 800 correspond to a case where an investment pattern of 100 points is generated by the information processing device 100.

[0111] Similarly, the white circles in the graph 800 correspond to the investment patterns generated by the information processing device 100. Specifically, the white circles in the graph 800 correspond to the case where a 50-point investment pattern is generated by the information processing device 100. In other words, the white circles in the graph 800 correspond to the case where the calculation time is shortened compared to the case where a 100-point investment pattern is generated by the information processing device 100.

[0112] As shown in graph 800, for the same cost, the investment pattern indicated by the diagonal lines tends to produce a higher return than the investment pattern indicated by the black circles. Similarly, for the same cost, the investment pattern indicated by the white circles tends to produce a higher return than the investment pattern indicated by the black circles. Next, we move on to the explanation of FIG. 9.

[0113] 9, table 900 shows the calculation time and MaxSR of the traditional method and the calculation time and MaxSR of the information processing device 100. The column "traditional 100pt" shows the calculation time and MaxSR of the traditional method. In the traditional method, it takes more than one day to determine the weight parameter λ and 4.2 hours to solve the optimization problem, so the calculation time is 4.2 hours + 1 day. In addition, in the traditional method, MaxSR is 140.

[0114] The column "current 100pt" indicates the calculation time and MaxSR when the information processing device 100 generates an investment pattern of 100 points. According to the information processing device 100, it takes 4.2 hours to set the weight parameter λ and solve the optimization problem, so the calculation time is 4.2 hours. Also, according to the information processing device 100, MaxSR is 169. A larger MaxSR indicates a more appropriate investment pattern. The MaxSR by the information processing device 100 is, for example, 1.21 times larger than the MaxSR by the conventional method.

[0115] The column "current 50pt" indicates the calculation time and MaxSR when the information processing device 100 generates an investment pattern of 50 points. In other words, the column "current 50pt" indicates the calculation time and MaxSR when the information processing device 100 reduces the calculation time compared to when the information processing device 100 generates an investment pattern of 100 points. According to the information processing device 100, it takes 2.1 hours to set the weight parameter λ and solve the optimization problem, so the calculation time is 2.1 hours. The calculation time when generating an investment pattern of 50 points is, for example, 0.5 times the calculation time when generating an investment pattern of 100 points. Also, according to the information processing device 100, MaxSR is 161. The MaxSR by the information processing device 100 with the reduced calculation time is, for example, 1.15 times the MaxSR by the conventional method.

[0116] 8 and 9, the information processing device 100 can generate an appropriate investment pattern more easily than the conventional method. Even if the calculation time is shortened, the information processing device 100 can generate an appropriate investment pattern more easily than the conventional method.

[0117] (Properties that appear in the relationship between risk and return) Here, we will explain the properties that appear in the relationship between risk and return. For example, the cost function for fixing the return and minimizing the risk is defined by the following formula (16).

[0118]

number

[0119] W∈R N × 1 are the weights for N stocks. Σ∈R N × N is the variance-covariance matrix of the fluctuations of N stocks over the period. μ∈R N × 1 is the average fluctuation of N stocks over the period. R is the expected return. l∈R M × N are the other M score linear constraint coefficients. c∈R M × 1 are the other M constraint scores. Here, by applying the Lagrangian function L(·), the following equation (17) holds:

[0120]

number

[0121] λ is a weighting parameter. The Lagrangian function is expressed as W T and λ T By partially differentiating each of them and summarizing them, the following equation (18) is established. -1 A T ) -1 Let T be the equation (19) below, which is derived from the equation (18).

[0122]

number

[0123]

number

[0124] Here, according to the above formula (19), if the only constraint is the expected return, it is believed that risk can be expressed by a quadratic function of the return. Even if there are multiple constraints, it is believed that risk can be expressed by a quadratic function of the return, since the expected return is included in matrix b. Also, according to the above formula (18), λ is defined by the following formula (20), so it is believed that λ can be calculated based on the pair of risk and return.

[0125]

number

[0126] (Overall processing procedure) Next, an example of an overall processing procedure executed by the information processing device 100 will be described with reference to Fig. 10. The overall processing is realized by, for example, the CPU 301, storage areas such as the memory 302 and the recording medium 305, and the network I / F 303 shown in Fig. 3.

[0127] Fig. 10 is a flowchart showing an example of an overall processing procedure. In Fig. 10, the information processing device 100 generates a second-order polynomial model indicating an EF based on the EF information management table 500 (step S1001).

[0128] Next, the information processing device 100 identifies a combination of risk and return on the EF based on the quadratic polynomial model (step S1002). Then, the information processing device 100 calculates a weight parameter λ based on the identified combination using a Lagrangian function (step S1003).

[0129] Next, the information processing device 100 adds a random number θ within a certain range to the calculated weight parameter λ, and then generates a QUBO equation including the weight parameter λ to which the random number θ has been added (step S1004). Then, the information processing device 100 uses an Ising solver to solve the QUBO equation and generate an investment pattern (step S1005). After that, the information processing device 100 ends the entire process.

[0130] This allows the information processing device 100 to generate an appropriate investment pattern relatively easily. For example, the information processing device 100 uses a combination of risk and return on the EF when calculating weight parameters, and therefore can easily calculate useful weight parameters. For example, the information processing device 100 can eliminate the need to manually try out weight parameters without clues, and can reduce the workload involved in calculating weight parameters. For example, the information processing device 100 can generate an appropriate investment pattern based on useful weight parameters.

[0131] As described above, the information processing device 100 can identify a pair of a first risk and a maximum return for a first risk based on a quadratic polynomial model showing the relationship between a risk related to an investment and a maximum return for the risk. The information processing device 100 can calculate a weight parameter for at least one of the first and second terms in a cost function based on the identified pair. The information processing device 100 can generate a mathematical model including a cost function and expressing an optimization problem related to an investment pattern based on the calculated weight parameter. This allows the information processing device 100 to reduce the workload involved in solving an optimization problem.

[0132] According to the information processing device 100, it is possible to solve an optimization problem related to an investment pattern based on the generated mathematical model, and generate an investment pattern. In this way, the information processing device 100 can solve the optimization problem and generate an appropriate investment pattern. The information processing device 100 can enable a user to use an appropriate investment pattern. As a result, the information processing device 100 can make it easier for a user to obtain profits from investment.

[0133] According to the information processing device 100, it is possible to solve an optimization problem regarding an investment pattern based on a mathematical model including a constraint condition indicating that overlapping buying and short selling of the same stock included in the first stock group is prohibited, and generate an investment pattern. In this way, the information processing device 100 can generate an investment pattern taking into consideration the constraint condition. The information processing device 100 can improve the convenience of the user. For example, in a situation where the user does not want to overlap buying and short selling of the same stock, the information processing device 100 can generate an investment pattern adapted to the situation.

[0134] According to the information processing device 100, it is possible to solve an optimization problem regarding an investment pattern based on a mathematical model including a constraint condition regarding the number of stocks to be purchased from the first group of stocks, and generate an investment pattern. This allows the information processing device 100 to generate an investment pattern that takes into consideration the constraint condition. The information processing device 100 can improve the convenience of users. For example, in a situation where a user tries to bring the number of stocks to be purchased closer to a certain number, the information processing device 100 can generate an investment pattern adapted to the situation.

[0135] According to the information processing device 100, it is possible to solve an optimization problem regarding an investment pattern based on a mathematical model including a constraint condition regarding the number of stocks to be short-sold among the first group of stocks, and generate an investment pattern. This allows the information processing device 100 to generate an investment pattern that takes the constraint condition into consideration. The information processing device 100 can improve the convenience of users. For example, in a situation where a user tries to bring the number of stocks to be short-sold closer to a certain number, the information processing device 100 can generate an investment pattern adapted to the situation.

[0136] According to the information processing device 100, the weighting parameters can be calculated using the Lagrangian function. This allows the information processing device 100 to calculate weighting parameters whose calculation basis is indicated by the Lagrangian function and which have a relatively high reliability.

[0137] According to the information processing device 100, it is possible to specify a pair of a first risk and a maximum return for the first risk that maximizes the Sharpe ratio based on a quadratic polynomial model. This allows the information processing device 100 to accurately calculate weight parameters so as to facilitate the generation of an appropriate investment pattern. For example, the information processing device 100 uses a pair of a first risk and a maximum return for the first risk that maximizes the Sharpe ratio, and therefore makes it possible to easily calculate useful weight parameters.

[0138] According to the information processing device 100, it is possible to identify a first risk corresponding to a specified return based on a quadratic polynomial model and the specified return, and to identify a pair of the identified first risk and the specified return. This allows the information processing device 100 to accurately calculate weight parameters so as to facilitate the generation of an appropriate investment pattern. The information processing device 100 can facilitate the calculation of useful weight parameters, for example, based on the specified return.

[0139] According to the information processing device 100, a quadratic polynomial model can be generated based on information indicating a pair of risk related to investments in multiple stocks in the past and the maximum return for the risk. This allows the information processing device 100 to solve an optimization problem related to an investment pattern even if there is no other computer that generates a quadratic polynomial model. The information processing device 100 can easily generate a quadratic polynomial model with high accuracy based on information indicating a pair of risk related to investments in multiple stocks in the past relatively close to the present and the maximum return for the risk.

[0140] The information processing method described in this embodiment can be realized by executing a prepared program on a computer such as a PC or a workstation. The information processing program described in this embodiment is recorded on a computer-readable recording medium, and is executed by being read from the recording medium by the computer. The recording medium may be a hard disk, a flexible disk, a CD (Compact Disc)-ROM, an MO (Magneto Optical disc), a DVD (Digital Versatile Disc), or the like. The information processing program described in this embodiment may also be distributed via a network such as the Internet.

[0141] The following supplementary notes are further disclosed regarding the above-described embodiment.

[0142] (Appendix 1) Identifying a pair of a first risk and a first maximum return for the first risk based on a second-order polynomial model showing the relationship between a risk related to an investment in a plurality of trading objects and a maximum return for the risk; calculating a weighting parameter for at least one of a first term relating to risk and a second term relating to return in a cost function including a first term relating to risk and a second term relating to return, the first term being included in a mathematical model expressing an optimization problem relating to an investment pattern for the plurality of trading objects, based on the identified pair; generating the mathematical model based on the calculated weight parameters; An information processing program that causes a computer to execute a process.

[0143] (Appendix 2) The information processing program according to appendix 1, characterized in that an optimization problem regarding the investment pattern is solved based on the generated mathematical model.

[0144] (Appendix 3) The investment pattern indicates purchase or short selling of trading objects included in a first group of trading objects, The information processing program described in Appendix 1 or 2, characterized in that the mathematical model includes a constraint condition indicating that overlapping purchases and short sales of the same trading object included in the first group of trading objects is prohibited.

[0145] (Appendix 4) The investment pattern indicates purchase or short selling of trading objects included in a first group of trading objects, The information processing program according to any one of appendices 1 to 3, wherein the mathematical model includes a constraint on the number of trading objects to be purchased from the first group of trading objects.

[0146] (Appendix 5) The investment pattern indicates purchase or short selling of trading objects included in a first group of trading objects, The information processing program according to any one of appendices 1 to 4, wherein the mathematical model includes a constraint on the number of trading objects that are subject to short selling among the first group of trading objects.

[0147] (Appendix 6) The calculation process is 6. The information processing program according to any one of claims 1 to 5, further comprising calculating the weighting parameters by using a Lagrangian function.

[0148] (Appendix 7) The process to be specified is: The information processing program according to any one of appendices 1 to 6, further comprising identifying a pair of a first risk and a first maximum return for the first risk, which maximizes a Sharpe ratio based on the second-order polynomial model.

[0149] (Appendix 8) The process to be specified is: The information processing program according to any one of Appendices 1 to 7, characterized in that a first risk corresponding to a specified return is identified based on the second-order polynomial model and the specified return, and a pair of the identified first risk and the specified return is identified.

[0150] (Appendix 9) The second-order polynomial model is generated based on information indicating pairs of risks related to investments in multiple trading targets in the past and maximum returns for the risks. 9. The information processing program according to any one of claims 1 to 8, which causes the computer to execute a process.

[0151] (Appendix 10) Identifying a pair of a first risk and a first maximum return for the first risk based on a second-order polynomial model showing the relationship between a risk related to an investment in a plurality of trading objects and a maximum return for the risk; calculating a weighting parameter for at least one of a first term relating to risk and a second term relating to return in a cost function including a first term relating to risk and a second term relating to return, the first term being included in a mathematical model expressing an optimization problem relating to an investment pattern for the plurality of trading objects, based on the identified pair; generating the mathematical model based on the calculated weight parameters; An information processing method characterized in that the processing is executed by a computer.

[0152] (Appendix 11) Identifying a pair of a first risk and a first maximum return for the first risk based on a second-order polynomial model showing the relationship between a risk related to an investment in a plurality of trading objects and a maximum return for the risk; calculating a weighting parameter for at least one of a first term relating to risk and a second term relating to return in a cost function including a first term relating to risk and a second term relating to return, the first term being included in a mathematical model expressing an optimization problem relating to an investment pattern for the plurality of trading objects, based on the identified pair; generating the mathematical model based on the calculated weight parameters; An information processing device comprising a control unit. [Explanation of symbols]

[0153] 100 Information processing device 101 Second-degree polynomial model 102 Cost Function 103 Formula Model 110,410,800 graph 111 Characteristic curve 200 Optimization Processing System 201 Generation device 202 Solving device 203 Client device 210 Network 300 Bus 301 CPU 302 Memory 303 Network Interface 304 Recording media I / F 305 Recording media 400 Data Structures 500 EF information management table 600 Storage section 601 Acquisition Department 602 Specific part 603 Generation part 604 Calculation section 605 Output section 700 Next Data Entry Section 701 EF Prediction Department 702 QUBO generation part 703 Optimization Department 900 tables

Claims

1. Identifying a pair of a first risk and a first maximum return for the first risk based on a second-order polynomial model that indicates a relationship between a risk associated with an investment in a plurality of trading objects and a maximum return for the risk; calculating a weighting parameter for at least one of a first term relating to risk and a second term relating to return in a cost function including a first term relating to risk and a second term relating to return, the first term being included in a mathematical model expressing an optimization problem relating to an investment pattern for the plurality of trading objects, based on the identified pair; generating the mathematical model based on the calculated weight parameters; An information processing program that causes a computer to execute a process.

2. 2. The information processing program according to claim 1, further comprising: solving an optimization problem regarding said investment pattern based on said generated mathematical model.

3. the investment pattern indicates a purchase or a short sale of a trading object included in a first group of trading objects; 3. The information processing program according to claim 1, wherein the mathematical model includes a constraint condition indicating that overlapping purchases and short sales of the same trading object included in the first group of trading objects is prohibited.

4. the investment pattern indicates a purchase or a short sale of a trading object included in a first group of trading objects; 4. The information processing program according to claim 1, wherein the mathematical model includes a constraint on the number of trading objects to be purchased from the first group of trading objects.

5. the investment pattern indicates a purchase or a short sale of a trading object included in a first group of trading objects; 5. The information processing program according to claim 1, wherein the mathematical model includes a constraint on the number of trading objects that are subject to short selling among the first group of trading objects.

6. The process of specifying The information processing program according to any one of claims 1 to 5, further comprising identifying a pair of a first risk and a first maximum return for the first risk, which maximizes a Sharpe ratio based on the second-order polynomial model.

7. The process of specifying The information processing program according to any one of claims 1 to 6, characterized in that a first risk corresponding to a specified return is identified based on the second-order polynomial model and the specified return, and a pair of the identified first risk and the specified return is identified.

8. generating the second-order polynomial model based on information indicating pairs of risks related to investments in a plurality of trading objects in the past and maximum returns for the risks; 8. The information processing program according to claim 1, which causes the computer to execute a process.

9. Identifying a pair of a first risk and a first maximum return for the first risk based on a second-order polynomial model that indicates a relationship between a risk associated with an investment in a plurality of trading objects and a maximum return for the risk; calculating a weighting parameter for at least one of a first term relating to risk and a second term relating to return in a cost function including a first term relating to risk and a second term relating to return, the first term being included in a mathematical model expressing an optimization problem relating to an investment pattern for the plurality of trading objects, based on the identified pair; generating the mathematical model based on the calculated weight parameters; An information processing method characterized in that the processing is executed by a computer.

10. Identifying a pair of a first risk and a first maximum return for the first risk based on a second-order polynomial model that indicates a relationship between a risk associated with an investment in a plurality of trading objects and a maximum return for the risk; calculating a weighting parameter for at least one of a first term relating to risk and a second term relating to return in a cost function including a first term relating to risk and a second term relating to return, the first term being included in a mathematical model expressing an optimization problem relating to an investment pattern for the plurality of trading objects, based on the identified pair; generating the mathematical model based on the calculated weight parameters; An information processing device comprising a control unit.

Citation Information

Patent Citations

  • Optimum portfolio determination method and device

    JP2003132218A

  • Device, method, and program for calculating portfolio

    JP2004348530A

  • Server system and its control method, client terminals, recording media, and portfolio showing device

    JP2004348757A

  • INVESTMENT PORTFOLIO OPTIMIZATION SYSTEM AND METHOD AND COMPUTER PROGRAM PRODUCT THEREOF

    JP2006518522A

  • Production planning method and production planning system

    JP2020140615A