Real estate score calculation system, real estate score calculation method, and program

Abstract

To provide a system for proposing a measurement model that measures stability of rent of investment real estate.SOLUTION: A system is configured to include a user terminal 20 of a user, and a real estate score calculation device 10 that is connected with the user terminal 20 in a data transmittable / receivable manner and calculates a P score as a measurement model that measures stability of rent of investment real estate. The real estate score calculation device 10 is configured to: input real estate information collected from the outside to a machine learning model; extract a plurality of feature quantities greatly affecting a rent index by analyzing the machine learning model; define the feature quantities as a plurality of parameters; set the maximum point and a point calculation reference for each parameter based on the rent index and a correlation level; calculate points for each parameter based on the point calculation reference, when parameter information is input as the real estate information from the user terminal 20; and calculate the P score from the calculated points for respective parameters.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a system, method, and program for proposing an econometric model for measuring rental stability of investment real estate. [Background technology]

[0002] Generally known methods for valuing investment real estate include the income capitalization method and comparative transaction method. However, in Japan, actual transaction data for real estate transactions is not publicly disclosed, and there are no generally available statistical models for determining prices. This has made it difficult for individual investors to judge the appropriateness of the price of investment real estate.

[0003] For example, Patent Document 1 discloses a real estate investment analysis support device that supports the analysis of the performance of investment real estate from various perspectives. In addition, MFS Corporation offers a real estate investment service (INVASE) called a "voucher service" that determines the amount of real estate investment loan that can be borrowed, and a "refinancing service" that introduces financial institutions to which real estate investment loans can be refinanced.

[0004] However, there is no service available that quantitatively evaluates the appropriateness of the price of investment real estate. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2020-126522 A Summary of the Invention [Problem to be solved by the invention]

[0006] The present invention aims to provide a real estate score calculation system that calculates a P score (a numerical value that quantitatively indicates the degree of possibility that the expected future rent will not be obtained) as a quantitative model to measure the stability of rental income for investment real estate. [Means for solving the problem]

[0007] The real estate score calculation system according to the first aspect of the present invention is A system comprising a user terminal of a user, and a real estate score calculation device connected to the user terminal so as to be capable of transmitting and receiving data, and which calculates a P-score as a quantitative model for measuring the stability of rental income of an investment real estate, The real estate score calculation device is A machine learning unit processes information collected from real estate information sources (external or internal real estate information sites, real estate companies, and other information sources) into real estate information targeted at rent indicators that are indicators of rent stability (e.g., price per square meter, vacancy rate, average period from when a vacancy occurs until a tenant is decided, etc.), and inputs the processed real estate information into a machine learning model; a P-score model creation unit that analyzes the machine learning model to extract multiple feature amounts that have a high influence on the rent index, defines these feature amounts as multiple parameters, sets a weight for each of the parameters based on the degree of correlation with the rent index, and sets a maximum point and a point calculation standard for each of the parameters; A P score calculation unit that calculates a point for each parameter based on a point calculation standard for each parameter set in the P score model creation unit when the parameter information is input as real estate information from the user terminal, and calculates the sum of the calculated points for each parameter as the P score; The present invention is characterized by comprising:

[0008] The real estate score calculation system according to the second aspect of the present invention is the real estate score calculation system according to the first aspect of the present invention, In the real estate score calculation system according to claim 1, The present invention is characterized by comprising a base cap rate calculation unit that calculates a base cap rate as an appropriate yield using an approximation formula obtained from the P score calculated by the P score calculation unit and an actual yield distribution chart.

[0009] The real estate score calculation system according to the third aspect of the present invention is the first or second aspect of the present invention, The P score model creation unit is characterized in that it sets a maximum point for each parameter based on the rent index and the degree of correlation so that the total value of the maximum points for each parameter is a constant value, and sets a point calculation standard to calculate points according to the parameter information input from the user terminal.

[0010] The real estate score calculation system according to the fourth aspect of the present invention is the first or second aspect of the present invention, The P score model creation unit is characterized in that it analyzes the relative degree of influence between the parameters and sets the proportion of each parameter to the whole as a weight so that the total value of the weights for each parameter becomes 100%.

[0011] The real estate score calculation system according to the fifth aspect of the present invention is the first or second aspect of the present invention, The parameters set by the P score model creation unit are characterized by including all or part of the following: "location," "nearest station," "distance from station," "year built," "exclusive floor area," "rental status," "floor," and "brand."

[0012] The real estate score calculation method according to the sixth aspect of the present invention is A method for calculating a P-score as a quantitative model for measuring the stability of rental income of an investment property based on real estate information input from a user's user terminal, comprising: A machine learning process in which information collected from real estate information sources (external or internal real estate information sites, real estate companies, and other sources) is processed into real estate information targeting rent indices, which are indicators of rent stability, and the processed real estate information is input into a machine learning model; A P-score model creation process for analyzing the machine learning model to extract multiple feature quantities that have a high influence on the rent index, defining these feature quantities as multiple parameters, setting the ratio of each of the parameters to the total parameters based on the degree of correlation with the rent index as a weight, and setting the maximum points and point calculation criteria for each of the parameters; A P-score calculation step of calculating points for each parameter based on the point calculation criteria for each parameter set in the P-score model creation step when the parameter information is input as real estate information from the user terminal, and calculating the sum of the calculated points for each parameter as the P-score; The present invention is characterized by comprising:

[0013] The real estate score calculation program according to the seventh aspect of the present invention is A program that calculates a P-score as a quantitative model for measuring the stability of rental income for investment real estate based on real estate information input from a user's user terminal, On the computer, A machine learning means for processing information collected from real estate information sources (external or internal real estate information sites, real estate companies, and other information sources) into real estate information targeted at rent indexes, which are indicators of rent stability, and inputting the processed real estate information into a machine learning model; a P-score model creation means for analyzing the machine learning model to extract multiple feature quantities that have a high influence on the rent index, defining these feature quantities as multiple parameters, setting a weight for each of the parameters based on the degree of correlation with the rent index as a proportion of the total parameters, and setting a maximum point and a point calculation standard for each of the parameters; A P score calculation means for calculating points for each of the parameters based on a point calculation standard for each of the parameters set by the P score model creation means when the parameter information is input as real estate information from the user terminal, and calculating the sum of the calculated points for each of the parameters as the P score; The present invention is characterized in that the above-mentioned is executed. Effect of the Invention

[0014] According to the present invention, it is possible to quantitatively evaluate the appropriateness of the price of an investment property by using the P score as a quantitative model for measuring the stability of the rental of the investment property. For example, a user can grasp the fair price of the property by selecting eight items of information of the property he / she wishes to purchase and inputting the rent, and can also understand the extent to which the changes in the location, age, rental situation, etc. affect the price by changing the conditions. The yield against the P score is updated monthly based on the market situation, so the price information is always up to date, ensuring the transparency of the price of the investment property, promoting buying and selling, and as a result, it is expected to lead to an improvement in the liquidity of the investment property. [Brief description of the drawings]

[0015] [Figure 1] 1 is a system configuration diagram showing an example of the configuration of a real estate score calculation system according to an embodiment of the present invention. [Diagram 2] 1 is a functional block diagram illustrating a real estate score calculation device according to an embodiment of the present invention. [Diagram 3] FIG. 2 is a flowchart illustrating a P-score calculation process procedure in the real estate score calculation device according to the embodiment of the present invention. [Figure 4] FIG. 2 is a diagram showing an example of each parameter definition required for P score calculation. [Figure 5-1] FIG. 13 is a diagram showing an example of point calculation criteria set for each parameter required for P score calculation. [Figure 5-2] FIG. 13 is a diagram showing an example of point calculation criteria set for each parameter required for P score calculation. [Figure 5-3] FIG. 13 is a diagram showing an example of point calculation criteria set for each parameter required for P score calculation. [Figure 5-4] FIG. 13 is a diagram showing an example of point calculation criteria set for each parameter required for P score calculation. [Figure 5-5]FIG. 13 is a diagram showing an example of point calculation criteria set for each parameter required for P score calculation. [Figure 5-6] FIG. 13 is a diagram showing an example of point calculation criteria set for each parameter required for P score calculation. [Figure 5-7] FIG. 13 is a diagram showing an example of point calculation criteria set for each parameter required for P score calculation. [Figure 5-8] FIG. 13 is a diagram showing an example of point calculation criteria set for each parameter required for P score calculation. [Figure 6] FIG. 13 is a diagram showing an example of calculating a base cap rate. [Figure 7] FIG. 1 is a diagram showing an example of calculation criteria for a base cap rate. [Figure 8] This is a diagram showing an example of calculating P score and base cap rate. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] Hereinafter, the embodiment of the present invention will be described in detail with reference to the drawings. Note that the embodiment described below is merely an example given to facilitate understanding of the present invention, and does not limit the present invention. In other words, the present invention may be modified or improved from the embodiment described below without departing from the spirit of the present invention.

[0017] The usage environment of the real estate score calculation system according to the embodiment of the present invention will be described with reference to FIG. 1. FIG. 1 is a system configuration diagram showing an example of the configuration of the real estate score calculation system. As shown in FIG. 1, the real estate score calculation device 10 is a device that calculates a P score as a quantitative model for measuring the stability of the rent of an investment property, and is connected to a user terminal 20 via a communication network. Here, the P score is a numerical value that quantitatively indicates the degree of possibility that the expected rent will not be obtained in the future, and the numerical range of the P score in this embodiment is set to a range of 0 to 5, and the lower the numerical value of the P score, the higher the risk of not obtaining the rent.

[0018] For example, a user who wants to obtain information on investment real estate accesses the real estate score calculation device 10 from the user terminal 20. When the user inputs real estate information of the investment real estate that is the target property from the user terminal 20, the real estate score calculation device 10 calculates the P score of the target property based on the real estate information and displays the result on the user terminal 20. Here, the above real estate information includes information such as "location", "nearest station", "distance from the station", "age of construction", "floor area", "rental situation", "floor number", "apartment brand", etc.

[0019] In addition, the real estate score calculation device 10 can calculate the P score of the target property based on the above real estate information, and can also calculate the appropriate return rate and appropriate price of the investment real estate that is the target property.

[0020] Next, the functional configuration of the real estate score calculation device according to the embodiment of the present invention will be described with reference to FIG. 2. FIG. 2 is a functional block diagram for explaining the real estate score calculation device. As shown in FIG. 2, the real estate score calculation device 10 includes a machine learning unit 11, a P score model creation unit 12, a P score calculation unit 13, and a base cap rate calculation unit 14. Each functional unit will be described in detail below.

[0021] <Machine learning unit> The machine learning unit 11 performs a process of processing the information collected from an external real estate company (real estate information source) into real estate information targeted at rental indicators, and inputs the processed real estate information into a machine learning model (hereinafter, it will be described using the unit price per square meter, which is one of the representative examples of rental indicators).

[0022] The P score model creation unit 12 analyzes the machine learning model, extracts a plurality of features having a high degree of influence on the unit price per square meter, defines these features as a plurality of parameters, sets the ratio of each parameter to the total parameter as the specific gravity based on the degree of correlation with the unit price per square meter, and sets the maximum point and point calculation criteria for each parameter.

[0023] In addition, the P score model creation unit 12 sets the maximum points for each parameter based on the price per square meter and the degree of correlation so that the total value of the maximum points for each parameter is a constant value, and sets a point calculation standard to calculate points according to the parameter information input from the user terminal.

[0024] In addition, the P score model creation unit 12 analyzes the relative influence between the parameters and sets the percentage of each parameter to the whole as a weight so that the total value of the weights for each parameter is 100%. The larger the weight value, the greater the influence that parameter has on the square meter price. The maximum points for each parameter are set in accordance with the weight value.

[0025] Next, the maximum points and point calculation criteria and weightings set by the P score model creation unit 12 will be described with reference to FIG. 4 and FIG. 5-1 to FIG. 5-8.

[0026] Fig. 4 is a diagram showing an example of the parameter definition. In the example of Fig. 4, eight types of items are defined as the parameters: "Location", "Nearest station", "Distance from station", "Age of building", "Exclusive floor area", "Rental status", "Floor", and "Condominium brand".

[0027] In the example of Figure 4, a "maximum point" and a "weight" are set for each parameter, the total of the "maximum points" is "5.0", and the total of the "weights" is "100%". For the "location" parameter, "maximum point = 0.94" is set, and furthermore, the percentage of this parameter in the total is set to "weight = 19%". The other parameters are as shown in Figure 4.

[0028] Figs. 5-1 to 5-8 are diagrams showing an example of point calculation criteria set for each parameter. Here, for each parameter of "location", "nearest station", "distance from the station", "age of construction", "floor area", "rental situation", "floor number", and "apartment brand", the point calculation criteria are set to calculate points according to the parameter information input from the user terminal 20.

[0029] In the example of Figs. 5-1 to 5-8, for the "location" parameter, according to the parameter information input from the user terminal 20, for example, when it is in Chiyoda Ward, "point = 0.87" is calculated, and when it is in Minato Ward, "point = 0.94" (maximum point) is calculated. Thus, the point calculation criteria are set.

[0030] Also, for the "nearest station" parameter, according to the parameter information input from the user terminal 20, for example, when it is Shinjuku, the point calculation criteria are set so that "point = 0.64" is calculated.

[0031] For other parameters, the point calculation criteria are set so that they are calculated as shown in Figs. 5-1 to 5-8.

[0032] When the parameter information is input as real estate information from the user terminal 20, the P score calculation unit 13 calculates the points for each parameter based on the point calculation criteria for each parameter set by the P score model creation unit 12, and calculates the sum of the calculated points for each parameter as the P score.

[0033] Next, a calculation example of the P score will be explained using a concrete example. For example, when "Location = Shinjuku-ku", "Nearest station = Shinjuku", "Distance from station = 10 minutes", "Age of building = 2 years", "Exclusive floor area = 20 m", "Rental status = Vacant", "Location floor = 3rd floor", and "Condominium brand = Gala" are input as the parameter information, the points for each parameter are calculated as follows based on the point calculation criteria for each parameter. Location points = 0.73 "Nearest station" points = 0.64 "Distance from station" points = 0.18 "Age of building" points = 1.41 "Exclusive area" points = 0.21 "Rental Status" points = 0.25 "Floor location" points = 0.06 "Condominium brand" points = 0.16

[0034] Next, the points calculated for each of the parameters are added together to calculate the P score. Therefore, the P score is calculated as follows: P-score = 0.73 + 0.64 + 0.18 + 1.41 + 0.21 + 0.25 + 0.06 + 0.16 = 3.64

[0035] In this embodiment, the numerical range of the P score is set to 0 to 5, and in this example, a numerical value of P score = 3.64 was obtained. The user can quantitatively evaluate the appropriateness of the price of the investment real estate from the numerical value of the P score. As mentioned above, the P score is a numerical value that quantitatively indicates the degree of possibility that the expected rent will not be obtained in the future, and the lower the numerical value of the P score, the higher the risk of a decrease in rent or not being able to obtain rent.

[0036] <Base Cap Rate Calculation Department> The base cap rate calculation unit 14 calculates a base cap rate as an appropriate yield using an approximation formula obtained from the P score calculated by the P score calculation unit and an actual yield distribution map.

[0037] Figure 6 shows an example of calculating the base cap rate. The base cap rate (BCR) is a value that is reviewed every month. For example, if the P score is 5.0, the BCR is calculated as 3.28% (October 2022).

[0038] Figure 7 shows an example of the calculation criteria for the base cap rate. As shown in Figure 7, the P score is calculated from the real estate information collected every month, and the P score and actual yield are plotted on a scatter plot, and the approximation formula is used as the calculation formula for the base cap rate.

[0039] Next, a data processing procedure when real estate information is input from a user terminal in the real estate score calculation system according to the embodiment of the present invention will be described with reference to the flowchart of FIG.

[0040] <Data processing procedure> In step S10, real estate information of the investment property that is the target property is input from the user terminal 20. Here, the parameter information of this real estate information includes information such as "location," "nearest station," "distance from station," "age of building," "exclusive floor area," "rental status," "floor," and "apartment brand."

[0041] Next, in step S20, points for each parameter are calculated from the parameter information input as real estate information. That is, based on the point calculation standard for each parameter set in the P score model creation unit 12, points for each parameter are calculated from the parameter information input as real estate information from the user terminal 20.

[0042] Next, in step S30, the P score is calculated using the points for each parameter calculated in step S20. In other words, the P score is calculated by adding up the points for each parameter. In addition, although not shown in this flowchart, it is also possible to calculate the appropriate yield (base cap rate) and the appropriate price of the investment property that is the target property.

[0043] Next, in step S40, the calculation result calculated in step S30 is transmitted to the user terminal 20. This calculation result may include the fair yield (base cap rate) and fair price of the target property in addition to the P score of the target property.

[0044] The data processing procedure described above may be realized as a method executed by a computer, or may be realized as a program to be executed by a computer.

[0045] Finally, we will explain the case where P score and base cap rate (BCR) were calculated as the verification result of this system. Figure 8 shows the example of MFS Otemachi Mansion as the target property, and the P score was calculated as 3.31 and BCR as 3.89%.

[0046] From the above, it can be seen that using the P-score and base cap rate (BCR) as a quantitative model to measure the stability of rental rates for investment real estate is effective, and it is expected that this will ensure the transparency of investment real estate prices, promote buying and selling, and ultimately lead to improved liquidity of investment real estate.

[0047] In this embodiment, the rent index, which is an index showing the stability of rent, is explained using the square meter price, but it is not limited to this. The above explanation is merely an example of the square meter price, because a certain correlation was observed between the P score generated by targeting the square meter price and the actual yield as a result of analysis based on the hypothesis that "a highly stable property that can continue to receive high rent = there should be a correlation with the square meter price, based on the formula "rent / yield = price". Therefore, in addition to or instead of the square meter price, a certain data (item, feature amount) that is considered to be (most) indicative of the stability of rent, such as the vacancy rate or the average period from the occurrence of vacancy to the decision of a tenant, may be targeted. [Explanation of symbols]

[0048] 10. Real estate score calculation device 11…Machine Learning Department 12…P score model creation section 13…P score calculation section 14…Base cap rate calculation section 20...User terminal

Claims

1. A system comprising a user terminal of a user, and a real estate score calculation device connected to the user terminal so as to be capable of transmitting and receiving data, and which calculates a P-score as a quantitative model for measuring the stability of rental income of an investment real estate, The real estate score calculation device is a machine learning unit that processes information collected from real estate information sources into real estate information targeting a rent index, which is an index showing the stability of rent, and inputs the processed real estate information into a machine learning model; a P-score model creation unit that analyzes the machine learning model to extract multiple feature amounts that have a high influence on the rent index, defines these feature amounts as multiple parameters, sets a weight for each of the parameters based on the degree of correlation with the rent index, and sets a maximum point and a point calculation standard for each of the parameters; a P-score calculation unit that calculates a point for each parameter based on a point calculation standard for each parameter set in the P-score model creation unit when the parameter information is input as real estate information from the user terminal, and calculates the sum of the calculated points for each parameter as the P-score; A real estate score calculation system comprising:

2. In the real estate score calculation system according to claim 1 , A real estate score calculation system characterized by having a base cap rate calculation unit that calculates a base cap rate as an appropriate yield using an approximation formula obtained from the P score calculated by the P score calculation unit and an actual yield distribution chart.

3. In the real estate score calculation system according to claim 1 or 2, A real estate score calculation system characterized in that the P score model creation unit sets a maximum point for each parameter based on the rent index and the degree of correlation so that the total value of the maximum points for each parameter is a constant value, and sets a point calculation standard to calculate points according to the parameter information input from the user terminal.

4. In the real estate score calculation system according to claim 1 or 2, The P score model creation unit analyzes the relative degree of influence between the parameters and sets the proportion of each parameter to the whole as a weight so that the total value of the weights for each parameter becomes 100%, thereby achieving a real estate score calculation system.

5. In the real estate score calculation system according to claim 1 or 2, A real estate score calculation system characterized in that the parameters set by the P score model creation unit include all or part of the following: "location," "nearest station," "distance from station," "year built," "exclusive floor area," "rental status," "floor," and "brand."

6. A method for calculating a P-score as a quantitative model for measuring the stability of rental income of an investment property based on real estate information input from a user's user terminal, comprising: A machine learning process in which the information collected from the real estate information source is processed into real estate information targeting a rent index, which is an index showing the stability of rent, and the processed real estate information is input into a machine learning model; A P-score model creation process for analyzing the machine learning model to extract multiple feature quantities that have a high influence on the rent index, defining these feature quantities as multiple parameters, setting the ratio of each of the parameters to the total parameters based on the degree of correlation with the rent index as a weight, and setting the maximum points and point calculation criteria for each of the parameters; a P-score calculation step of calculating points for each parameter based on the point calculation criteria for each parameter set in the P-score model creation step when the parameter information is input as real estate information from the user terminal, and calculating the sum of the calculated points for each parameter as the P-score; A real estate score calculation method comprising:

7. A program that calculates a P-score as a quantitative model for measuring the stability of rental income for investment real estate based on real estate information input from a user's user terminal, On the computer, A machine learning means for processing the information collected from the real estate information source into real estate information targeting a rent index, which is an index showing the stability of rent, and inputting the processed real estate information into a machine learning model; a P-score model creation means for analyzing the machine learning model to extract a plurality of feature quantities having a high influence on the rent index, defining these feature quantities as a plurality of parameters, setting a weight for each of the parameters based on the degree of correlation with the rent index as a proportion of the total parameters, and setting a maximum point and a point calculation standard for each of the parameters; a P-score calculation means for calculating points for each of the parameters based on a point calculation standard for each of the parameters set by the P-score model creation means when the parameter information is input as real estate information from the user terminal, and calculating the sum of the calculated points for each of the parameters as the P-score; A real estate score calculation program characterized by executing the above.

Images