JP6381844B1 - Computer system, method, and program for accumulating assets whose value varies over time - Google Patents

Abstract

A computer capable of predicting a specific value such as a maximum value or a minimum value (that is, a maximum value or a minimum value) in a predetermined period of the value of an asset whose value fluctuates irregularly over time. Providing systems, methods, and programs. A computer system for accumulating a foreign currency whose value fluctuates as time passes predicts a time when the value of the foreign currency becomes the lowest value within a predetermined period (step S501), and The foreign currency is purchased (step S502) at a time when the predicted value of the foreign currency is the lowest price. [Selection] Figure 5

Description

  The present invention relates to a computer system, a method, and a program for accumulating assets whose values change over time.

  Regularly funded foreign exchange deposit transaction system that regularly purchases foreign currency denominated in various currencies (hereinafter also referred to as “foreign currency”), accumulates the foreign currency purchased, and sends profit and loss evaluation information of the accumulated foreign currency to the terminal Is known (see, for example, Patent Document 1).

Japanese Patent Laid-Open No. 2004-192587

  However, the conventional system does not analyze whether the value of the foreign currency to be purchased is the optimal price at the time of purchase or whether the purchase time is optimal when purchasing foreign currency regularly. There has been a problem that purchase optimization has not been achieved for buyers.

  The present invention has been made in view of such a problem, and is the highest value or the lowest value (that is, the maximum value or the minimum value) of the asset value whose value fluctuates irregularly with time. ) Or a specific ranking value (eg, second lowest value, tenth highest value, etc.), and the like. Objective.

  The present invention provides a computer system for accumulating a foreign currency whose value fluctuates over time, and the computer system predicts when the value of the foreign currency becomes a low value of a specific rank within a predetermined period. And purchasing the foreign currency at the predicted time.

  In one embodiment, the predicting includes dividing the predetermined period into a plurality of periods, and, among the plurality of periods after the division, a time when the value of the foreign currency becomes a low price of a specific rank. Determining the period to which it belongs.

  In one embodiment, the predicting includes: (1) further dividing a period in which a time when the value of the foreign currency is a low price of a specific order belongs into a plurality of periods; Of the plurality of periods after the division, determining a period to which a period when the value of the foreign currency is a low price of a specific rank belongs, and when a period when the value of the foreign currency is a low price of a specific rank belongs It further includes repeating until the determined period becomes the same length as the time when the value of the foreign currency becomes the low price of a specific rank.

  In one embodiment, the determining includes determining a period in which an average value of the foreign currency value is low among the plurality of divided periods.

  In one embodiment, the determining uses each of a plurality of prediction models to determine a period to which a period when the value of the foreign currency becomes a low price of a specific rank among a plurality of periods after the division. And, among the plurality of periods after the division, a period that is determined to be a period to which a time when the value of the foreign currency is a low value of a specific rank belongs by a larger number of prediction models, And determining that it is the period to which the time when the value is at a low value of a specific rank belongs.

  In one embodiment, the determining uses each of a plurality of prediction models to determine a period to which a period when the value of the foreign currency becomes a low price of a specific rank among a plurality of periods after the division. And, for each of the first half of the plurality of periods after the division, the number of prediction models determined that the period belongs to a period in which the value of the foreign currency becomes a low value of a specific rank And determining that a period in which the number of the prediction models is equal to or greater than a predetermined number is a period to which a period when the value of the foreign currency is at a low value of a specific rank belongs.

  In one embodiment, the determination further includes changing the predetermined number according to a relative position of the divided period with respect to the entire predetermined period.

  In one embodiment, the plurality of periods are two periods.

  In one embodiment, the predicting includes suppressing the prediction of a time within a predetermined suppression period from a start point of the predetermined period as a time when the value of the foreign currency becomes a low price of a specific rank. .

  In one embodiment, the prediction and the purchasing are repeated.

  In one embodiment, the purchasing is performed automatically without a user input operation for purchasing the foreign currency at a time when the predicted value of the foreign currency is at a certain low price. .

  In one embodiment, the low price of the specific rank is the lowest price.

  In one embodiment, the low price of the specific rank is the n-th lowest value (n is an integer of 2 or more).

  The present invention provides a method for accumulating a foreign currency whose value fluctuates over time, said method being executed in a computer system, wherein said method is characterized in that the value of said foreign currency is specified within a predetermined period of time. Predicting a time when the ranking is low, and purchasing the foreign currency at the predicted time.

  The present invention provides a program for accumulating a foreign currency whose value fluctuates over time. When the program is executed in a computer system, the value of the foreign currency has a specific rank within a predetermined period. The computer system is caused to perform processing including predicting a time when the price is low and purchasing the foreign currency at the predicted time.

  The present invention is configured to predict the time when the value of the foreign currency becomes a value of a specific rank within a predetermined period based on the past value fluctuation of the foreign currency whose value varies with time. Provide a computer system.

  In one embodiment, the computer system of the present invention is further configured to issue the foreign currency transaction request at the predicted time.

  The present invention provides a program executed in a computer system, and the program executes the foreign currency within a predetermined period based on a past value fluctuation of a foreign currency whose value changes with time. The computer system is caused to perform processing including predicting a time when the value of the currency becomes a value of a specific rank.

  In one embodiment, the process further includes issuing a transaction request for the foreign currency at the predicted time.

  According to the present invention, the value of an asset whose value fluctuates irregularly over time, the highest value or the lowest value (that is, the maximum value or the minimum value) or the value of a specific rank (for example, 2) in a predetermined period. It is possible to provide a computer system, method, and program that make it possible to predict a specific value such as the second lowest value, the tenth highest value, etc.

  Furthermore, according to the present invention, a computer system that makes it possible to increase the evaluation value of the asset by periodically purchasing or accumulating the asset based on the prediction of the time when the specific value is reached, Methods and programs can be provided.

The figure which shows an example of the data flow between the user in the AI foreign currency reserve deposit, a bank, and a foreign exchange market. The figure which shows an example of a structure of the computer system 200 for implement | achieving a new financial product. The figure which shows an example of the data structure of the information regarding the accumulation | storage of the value fluctuation asset stored in the database part 220. FIG. The figure which shows an example of the multilayer perceptron 450 which specific value prediction AI utilizes. The flowchart which shows an example of the process 500 in the computer system 200 for implement | achieving a new financial product. The flowchart which shows an example of the process which estimates the time when the value of specific value prediction AI becomes the cheapest value within a predetermined period in step S501. The flowchart which shows an example of a process in the case of performing the process of step S602 using a some prediction model. A process of performing the process of step S602 using a plurality of prediction models and restricting when the first half of the predetermined period is predicted as the time when the value of the value-changing asset is the lowest price within the predetermined period. The flowchart which shows an example. The graph which shows an example of the time fluctuation of a US dollar price (yen). The graph which shows an example of the time fluctuation of a US dollar price (yen). The graph which shows an example of the time fluctuation of a US dollar price (yen). The graph which shows an example of the time fluctuation of a US dollar price (yen). The graph which shows an example of the time fluctuation of the Australian dollar price (yen). The graph which shows an example of the time fluctuation of the Australian dollar price (yen). The graph which shows an example of the time fluctuation of the Australian dollar price (yen). The graph which shows an example of the time fluctuation of the Australian dollar price (yen). The figure which shows an example of computer system 200 'which is an alternative structure of computer system 200 for implement | achieving a new financial product.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

1. AI foreign currency reserve deposit The inventor of the present invention has developed a new financial product called “AI foreign currency reserve deposit”. AI foreign currency reserve deposits use artificial intelligence (AI) to predict the day when the target foreign currency will be the lowest price within a given period, and the target foreign currency will be the day when the target foreign currency is predicted to be the lowest price. It is a reserve deposit that accumulates the target foreign currency for each predetermined period by purchasing. According to the AI foreign currency reserve deposit, it is possible to purchase foreign currency at an exchange rate more advantageous than the conventional foreign currency reserve deposit. As a result, the risk due to exchange rate fluctuations can be further reduced. Here, AI is an abbreviation of “Artificial Intelligence”, and is a technique for artificially implementing human intelligence (or intelligence that surpasses human intelligence) on a computer system, or such a technique. Refers to a computer system in which is implemented.

  FIG. 1 shows an example of a data flow between a user, a bank, and a foreign exchange market in an AI foreign currency reserve deposit. In the example shown in FIG. 1, it is assumed that the target foreign currency is accumulated every month from April. Hereinafter, each step shown in FIG. 1 will be described.

  Step S101: The user 10 applies to the bank 20 for an AI foreign currency reserve deposit. Thereby, the data flow of the AI foreign currency reserve deposit is started. The application for the AI foreign currency reserve deposit is performed, for example, when the user 10 accesses the bank system of the bank 20 using a user device (for example, a smartphone), and the user operates the user device according to a predetermined application procedure. Alternatively, such application may be performed by the user 10 submitting an application for AI foreign currency reserve deposit to the bank 20 counter. In the application, the user 10 sets, for example, a reserve currency, a reserve amount, a target amount, and the like. For example, in the application, the user 10 may set to accumulate the Australian dollar until the amount of 10,000 yen is accumulated for 1 million yen. At this time, the user 10 cannot set the accumulation timing (for example, which day the Australian dollar is to be purchased every month).

  The bank 20 receives an application for AI foreign currency reserve deposit and performs a procedure for the application. When the procedure by the bank 20 is completed, the bank system of the bank 20 executes processing for the AI foreign currency reserve deposit.

  For example, the bank system of the bank 20 predicts the day when the target foreign currency is the lowest price in April on the first day of April. Such prediction is performed using AI (artificial intelligence). For example, it is assumed that April 14 is predicted to be the day when the target foreign currency is the lowest price.

  Step S102: On April 14, the bank 20 transmits a purchase request to the foreign exchange market 30 to purchase the target foreign currency using the Japanese yen amount set by the user 10.

  Step S103: In response to the purchase request from the bank 20, the foreign exchange market 30 sells the target foreign currency to the bank 20 at the exchange rate of April 14. The bank 20 accumulates the target foreign currency purchased from the foreign exchange market 30 for April in the account of the user 10.

  Step S104: The bank 20 reports to the user 10 that the target foreign currency purchased for April has been accumulated in the account of the user 10.

  For example, the bank system of the bank 20 predicts the day when the target foreign currency is the lowest price in May on the first day of May. Such prediction is performed using AI (artificial intelligence). For example, suppose that May 2 is predicted to be the day when the target foreign currency is the lowest price.

  Step S105: On May 2, the bank 20 transmits a purchase request to the foreign exchange market 30 to purchase the target foreign currency using the Japanese yen of the amount set by the user 10.

  Step S106: In response to the purchase request from the bank 20, the foreign exchange market 30 sells the target foreign currency to the bank 20 at the exchange rate of May 2. The bank 20 accumulates the target foreign currency purchased from the foreign exchange market 30 for May in the account of the user 10.

  Step S107: The bank 20 reports to the user 10 that the target foreign currency purchased for May has been accumulated in the account of the user 10.

  For example, the bank system of the bank 20 predicts the day when the target foreign currency is the lowest price in June on the first day of June. Such prediction is performed using AI (artificial intelligence). For example, suppose that June 28 is predicted to be the day when the target foreign currency is the lowest price.

  Step S108: On June 28, the bank 20 transmits a purchase request to the foreign exchange market 30 to purchase the target foreign currency using the Japanese yen of the amount set by the user 10.

  Step S109: In response to the purchase request from the bank 20, the foreign exchange market 30 sells the target foreign currency to the bank 20 at the exchange rate of June 28. The bank 20 accumulates the target foreign currency purchased from the foreign exchange market 30 for June in the account of the user 10.

  Step S110: The bank 20 reports to the user 10 that the target foreign currency purchased for June has been accumulated in the account of the user 10.

  In this way, in the AI foreign currency reserve deposit, the day when the target foreign currency becomes the lowest price within a predetermined period is predicted using AI (artificial intelligence), and the day when the target foreign currency is predicted to become the lowest price. The target foreign currency is purchased. In this way, foreign currency is accumulated every predetermined period. Thereby, according to the AI foreign currency reserve deposit, it is possible to purchase and accumulate foreign currency at an exchange rate more advantageous than the conventional foreign currency reserve deposit that purchases foreign currency on a certain date every month. As a result, the risk due to exchange rate fluctuations can be further reduced. This makes it possible to increase the value of foreign currency assets evaluated by mark-to-market at the end of any foreign currency purchase cycle.

  Note that the prediction of the day when the target foreign currency is at the lowest price by the AI is only a prediction, and does not guarantee that it will be the lowest price on that day. However, by improving the prediction accuracy by AI, it is possible to more reliably perform accumulation at a more advantageous exchange rate.

  In the embodiment described above with reference to FIG. 1, the example in which the prediction of the day when the target foreign currency is the lowest price by the AI is performed on the first day of the transaction month is described, but the timing at which the prediction is performed is not limited to this. . For example, the prediction may be performed before the transaction period is reached (for example, the day before the transaction month).

  In the embodiment described above with reference to FIG. 1, the AI is used to predict the day when the target foreign currency is the lowest price within a predetermined period, but the present invention is not limited to this. Using AI, a day other than the day when the target foreign currency is the lowest price within the predetermined period, for example, the day when the target foreign currency is the second lowest value within the predetermined period, and the fifth lowest price Day, 10th lowest value, 14th lowest value, etc. may be predicted. This is a more effective method for accumulating foreign currency that is highly random and difficult to predict. For foreign currencies that are highly variable in value and difficult to predict, by not aiming for the lowest price, it is possible to avoid the highest price without taking the lowest price, thus improving the final expected value.

  In the embodiment described above with reference to FIG. 1, the bank 20 reports to the user 10 every time it accumulates the target foreign currency, but the necessity and timing of reporting to the user 10 are arbitrary. For example, the report may not be actively reported from the bank 20 to the user 10, or may be reported to the user after a certain number of accumulations have been completed.

  In the embodiment described above with reference to FIG. 1, the AI foreign currency reserve deposit is described as an example of a new financial product, but the new financial product is not limited to this. For example, transactions are not limited to purchases or funding. The transaction may be a sale. At this time, using AI (artificial intelligence), the day when the target foreign currency becomes the highest value within a predetermined period is predicted, and the target foreign currency is sold on the day when the target foreign currency is predicted to become the highest value. For example, the object of transaction is not limited to foreign currency. The target of the transaction can be any asset whose value varies over time (hereinafter also referred to as “value-changing asset”). Examples of value-changing assets include, but are not limited to, foreign currencies, investment trusts, gold, commodities (crude oil, corn, etc.). In other words, the new financial product uses artificial intelligence (AI) and the target value fluctuation asset within a predetermined period is a specific value (the highest value, the lowest price, the 14th lowest value, the 5th highest value, etc.) (E.g. second, minute, hour, day, or week), and within the time when the target value-changing asset is predicted to be the lowest price or the lowest price A portfolio of value-changing assets by purchasing the value-changing assets or selling the value-changing assets within the time when the value-changing assets are predicted to be the highest value or high of a certain rank. It is a product that makes it possible to form. With this new financial product, it is possible to purchase or sell value-changing assets at a price that is more advantageous than traditional funded financial products. As a result, it is possible to further reduce the risk due to value fluctuations and increase the evaluation value.

2. Configuration of Computer System for Realizing New Financial Product FIG. 2 shows an example of the configuration of a computer system 200 for realizing the new financial product described above.

  The computer system 200 is a computer system used in the bank 20, for example.

  The computer system 200 is connected to the asset trading computer system 300 via the network 400. Here, the type of the network 400 is not limited. The network 400 may be, for example, the Internet or a LAN.

  The asset trading computer system 300 is a computer system that includes a server device or the like that can execute processing for trading value-changing assets. The asset trading computer system 300 is a computer system used in the foreign exchange market 30, for example. The asset trading computer system 300 may be, for example, a computer system under the jurisdiction of a trading brokerage company such as a money brokerage company participating in the foreign exchange market 30 or a trading party. The asset trading computer system 300 is configured to perform processing for selling value-changing assets in response to a request to purchase value-changing assets. In addition, the asset trading computer system 300 is configured to perform processing for purchasing value-changing assets in response to a request to sell value-changing assets. Since the computer system 200 is connected to the asset trading computer system 300, it is possible to trade value-changing assets.

  The computer system 200 includes a server device 210 and a database unit 220 connected to the server device 210. The server device 210 includes an interface unit 211, a processor unit 212, and a memory unit 213.

  The interface unit 211 is configured to control communication via the network 400. The interface unit 211 can transmit information to the outside of the server device 210, and can receive information from the outside of the user device 210. For example, the interface unit 211 transmits a request for purchase of a variable value asset to the asset trading computer system 300 via the network 400. For example, the interface unit 211 transmits a request for selling the value-changing asset to the asset trading computer system 300 via the network 400. For example, the interface unit 211 transmits information to the database unit 220 connected to the server device 210 and receives information from the database unit 220.

  The processor unit 212 controls the overall operation of the server device 210. The processor unit 212 reads a program stored in the memory unit 213 and executes the program. Thereby, it is possible to make the server apparatus 210 function as an apparatus that executes a desired step. For example, the processor unit 212 can perform part or all of the process of accumulating value-changing assets. The processor unit 212 may be implemented by a single processor or a plurality of processors.

  The memory unit 213 stores a program for executing processing in the server device 210, data necessary for executing the program, and the like. The memory unit 213 includes, for example, a program for trading value-changing assets or a program for accumulating value-changing assets (for example, a program that realizes the processes shown in FIGS. 5 and 6A to 6C described later) ) Is stored in part or in whole. Here, it does not matter how the program is stored in the memory unit 213. For example, the program may be preinstalled in the memory unit 213. Alternatively, the program may be installed in the memory unit 213 by being downloaded via the network 400. The memory unit 213 can be implemented by any storage unit.

  The database unit 220 stores information related to accumulation of value fluctuation assets for each user. The information related to the accumulation of value-changing assets includes information on the accumulation destination of value-changing assets. For example, information on the value-changing asset accumulated in the value-change asset (for example, the purchase time of the value-changing asset, the purchase price of the value-changing asset, the purchase amount of the value-changing asset) , A reserve amount of value fluctuation assets, etc.). The information on the value variable asset reserve is information related to the value variable asset sold from the value variable asset reserve (for example, the sale time of the value change asset, the sale price of the value change asset, the amount sold of the value change asset, etc.) May also be included. When the value-changing asset is in a foreign currency, the information on the storage destination of the value-changing asset is information on the foreign currency funding account, and information on the foreign currency accumulated in the foreign currency funding account (for example, purchase date and time of foreign currency, purchase price of foreign currency) , Foreign currency purchase amount, total foreign currency reserve, etc.) and information on foreign currency sold from the foreign currency reserve account (eg, foreign currency sale date, foreign currency sale price, foreign currency sale amount, etc.).

  The database unit 220 can also store past value fluctuation data, for example. The past value fluctuation data can be used for learning a specific value prediction AI, which will be described later, in order to predict future value fluctuation. The historical value fluctuation data used for learning is the value fluctuation data of the target asset against the currency for purchase (for example, Japanese yen) (for example, if the US dollar is purchased and accumulated in the Japanese yen, the value fluctuation of the US dollar against the Japanese yen) Data). The historical value fluctuation data used for learning is the target asset value fluctuation data for other currencies (eg US dollar, Australian dollar, Euro, NZ dollar, etc.) in addition to the target asset value fluctuation data for the purchase currency. Data, or data such as S & P 500 index, Nikkei Stock Average, German Stock Index, or WTI crude oil price may be included. The database unit 220 can also store learned value fluctuation trends, for example. When a specific value prediction AI described later uses a multilayer perceptron, the database unit 220 may store the weight vector of each node established by learning.

  FIG. 3 shows an example of a data configuration of information related to accumulation of value-changing assets stored in the database unit 220.

  In the example illustrated in FIG. 3, an example of a data configuration of information stored in the database unit 220 when the value fluctuation asset is a foreign currency is illustrated.

  The database unit 220 stores foreign currency accumulation account information for each user. The database unit 220 includes a database 221 for storing information on the first user's foreign currency reserve account and a database 222 for storing information on the second user's foreign currency reserve account. The database 221 for storing the information of the first user's foreign currency reserve account is, for example, the first user's foreign currency reserve account number, the foreign currency purchase date and time, which started the dollar-denominated deposit from April 2017, Includes purchase price of foreign currency and total accumulated foreign currency. For example, the database 221 stores information that the first user purchased $ 113 on April 14, 2017, and the accumulated amount became $ 113, and the first user entered May 2, 2017. Stores information that $ 119 was purchased and the total funded amount was $ 232, and that the first user purchased $ 125 on June 28, 2017 and stored the total funded value was $ 357 Then, information that the first user purchased 110 dollars on July 10, 2017, and the accumulated amount becomes 467 dollars may be stored. The database 222 for storing the information of the second user's foreign currency reserve account includes, for example, the second user's foreign currency reserve account number, the foreign currency purchase date and time, which started the euro-denominated reserve deposit from July 2016, Includes purchase price of foreign currency and total accumulated foreign currency. For example, the database 221 may store information that the second user purchased 520 euros on July 12, 2016, and the total accumulated amount is 520 euros.

  Time (for example, seconds) when the value of the value-changing asset becomes a specific value (for example, a value of a specific rank such as the lowest price, the highest price, the 14th lowest value, or the 5th highest value) within a predetermined period. , Minutes, hours, days, weeks, etc.) (hereinafter also referred to as “specific value prediction AI”) can be realized by the computer system 200 having the above-described configuration. The specific value prediction AI uses a multilayer perceptron 450 as a prediction model.

  FIG. 4 shows an example of the multilayer perceptron 450 used by the specific value prediction AI.

  Multilayer perceptron 450 has an input layer, a hidden layer, and an output layer. In the example shown in FIG. 4, the multilayer perceptron 450 is shown as having two hidden layers, but the number of hidden layers is not limited to this. The multi-layer perceptron 450 can comprise one or more hidden layers. In addition, each of the input layer, hidden layer, and output layer of the multilayer perceptron 450 can include any number of nodes. For example, each layer may include nodes so that nodes decrease from the input layer toward the output layer, or each hidden layer may include the same number of nodes.

  The specific value prediction AI can predict the future change in the value of the value change asset by learning the past value change data using the multilayer perceptron 450. The past value fluctuation data to be learned includes, for example, value fluctuation data of the asset to be predicted with respect to the currency for purchase (for example, Japanese yen). In addition, historical value change data learned may include value change data for the asset being predicted for currencies other than the currency for purchase (eg, US dollar, Australian dollar, euro, NZ dollar, etc.), or , S & P 500 index, Nikkei average stock price, German stock index, or asset price data other than forecasted assets such as WTI crude oil price. The specific value prediction AI can establish a weight vector of each node of the multilayer perceptron 450 by learning past value fluctuation data as teacher data. For example, the specific value prediction AI performs learning by using, as teacher input data, data for a certain period of at least one type of value fluctuation data, and using the value fluctuation data of a prediction target asset after the certain period as teacher output data. be able to.

  The teacher output data may be, for example, an average value of a predetermined period of assets to be predicted after a certain period. The multi-layer perceptron 450 having the weight vector of each node established by learning using the teacher output data outputs a prediction of an average value of a predetermined period in the future when the value fluctuation data of the immediately preceding period is input. Will be able to.

  The output data for teachers is, for example, that the asset to be predicted becomes the lowest price in the first half of a certain period (for example, January, 10 business days, 6 business days, 2 business days, etc.) It may be a binary value (0 or 1) indicating whether it is the lowest price. When the multi-layer perceptron 450 having the weight vector of each node established by learning using the teacher output data is input with the value fluctuation data of the immediately preceding period, a certain period in the future (for example, January, 10 Business day, 6 business days, 2 business days, etc.), it is possible to output a prediction of whether it will be the lowest price in the first half or the lowest price in the second half.

  The teacher output data includes, for example, a period in which the asset to be predicted is divided into a plurality of periods after a certain period (for example, January, 10 business days, 6 business days, 2 business days). It may be a value indicating the period during which the price is lowest. When the multi-layer perceptron 450 having the weight vector of each node established by learning using the teacher output data is input with the value fluctuation data of the immediately preceding period, a certain period in the future (for example, January, 10 Business days, 8 business days, 4 business days, etc.) can be output in which of the periods divided into a plurality of periods is predicted to be the lowest price.

  The specific value prediction AI uses the multi-layer perceptron 450 and based on past value fluctuation data, for example, “a month with a large uptrend tends to be the lowest in the first half of the month and the highest in the second half of the month. "In a month with a big downtrend, the second half of the month is likely to be the lowest, and the first half of the month is likely to be the highest." It is possible to learn a value fluctuation tendency based on a trend such as “It is easier for the weekend to be at the lowest price than at the beginning of the week when it is a rising trend”. The specific value prediction AI is based on past value fluctuation data, for example, based on a fluctuation amount such as “definitely descends after a day when it rises above a certain threshold”, “always rises after a day when it falls above a certain threshold” Can learn the trend of value fluctuation. The learning result is stored in the database unit 220 together with past fluctuation data.

  The specific value prediction AI uses the multi-layer perceptron 450, and based on the learning result, the value of the value change asset within a predetermined period (for example, the lowest value, the highest value, the 14th lowest value, or A time (for example, second, minute, hour, day, week, or the like) that is the fifth highest value is predicted. The input to the multi-layer perceptron 450 for the prediction process may be value fluctuation data for the period immediately preceding the predetermined period to be predicted. The input to the multi-layer perceptron 450 for processing the forecast may be the value fluctuation data of the asset to be forecasted against the purchase currency (eg, Japanese yen), but the currency other than the purchase currency (eg, US dollar, Australian dollar, (E.g., Euro, NZ dollars, etc.) or asset price data other than forecasted assets such as S & P 500 index, Nikkei Stock Average, German Stock Price Index, or WTI crude oil price. It is good also as an input in the process of prediction. The input to the multilayer perceptron 450 may be, for example, value fluctuation data for the last 16 days, value fluctuation data for the last 32 days, or value fluctuation data for the last 64 days of a predetermined period to be predicted. It should be noted that the more data input to the multi-layer perceptron 450, the better the prediction accuracy, but the amount of calculation increases as the data increases. In addition, it should be noted that the more past value fluctuation data learned, the better the prediction accuracy, but this also increases the amount of computation for learning as the data increases. In addition, the more types of past value fluctuation data learned, the better the prediction accuracy, but if irrelevant data is learned for prediction, the specific value prediction AI will be confused and the prediction accuracy will decrease. It should be noted that there is a risk of losing.

  The multi-layer perceptron 450 is an example of a prediction model, and the specific value prediction AI can predict the time when the value of the value fluctuation asset becomes the specific value using any other prediction model. For example, the specific value prediction AI can predict the time when the value of the value fluctuation asset becomes the specific value using CNN (convolutional neural network), recurrent neural network (RNN), or the like as a prediction model. In general, CNN has better prediction accuracy than multilayer perceptron 450.

  The specific value prediction AI can use a plurality of prediction models to predict the time when the value of the value fluctuation asset becomes a specific rank value. The plurality of prediction models may be, for example, a plurality of prediction models having different types or amounts of past value variation data used for learning. The plurality of prediction models may be, for example, a plurality of prediction models having different types or amounts of value fluctuation data in the immediately preceding period input for prediction. The plurality of prediction models may be a plurality of prediction models having different types of neural networks to be used, such as a prediction model using the multilayer perceptron 450, a prediction model using CNN, a prediction model using RNN, and the like. The specific value prediction AI uses each of a plurality of prediction models, takes the majority of the respective prediction results, and predicts the time when the value of the value fluctuation asset becomes a value of a specific rank, thereby improving the prediction accuracy. It is possible.

  In the example shown in FIG. 2, each component of the server apparatus 210 is provided in the server apparatus 210, but the present invention is not limited to this. Any one of the components of the server device 210 may be provided outside the server device 210. For example, when each of the processor unit 212 and the memory unit 213 is configured by separate hardware components, the hardware components may be connected via an arbitrary network. At this time, the type of network is not limited. Each hardware component may be connected via a LAN, wirelessly connected, or wiredly connected, for example.

  In the example shown in FIG. 2, the database unit 220 is provided outside the server device 210, but the present invention is not limited to this. It is also possible to provide the database unit 220 inside the server device 210. At this time, the database unit 220 may be mounted by the same storage unit as the storage unit that mounts the memory unit 213, or may be mounted by a storage unit that is different from the storage unit that mounts the memory unit 213. In any case, the database unit 220 is configured as a storage unit for the server device 210. The configuration of the database unit 220 is not limited to a specific hardware configuration. For example, the database unit 220 may be configured with a single hardware component, or may be configured with a plurality of hardware components. For example, the database unit 220 may be configured as an external hard disk device of the server device 210 or may be configured as storage on a cloud connected via the network 400.

3. Process diagram in a computer system for implementing the new financial instruments 5 shows an example of a process 500 in a computer system 200 for implementing the new financial instruments. Here, a process 500 for accumulating assets whose values vary with time will be described as an example.

  In step S501, the specific value prediction AI realized by the computer system 200 predicts the time when the value of the value fluctuation asset becomes the lowest value within a predetermined period. The specific value prediction AI predicts, for example, a time when the value of the value-changing asset becomes the lowest value within a predetermined period by a process described later with reference to FIG. Here, the predetermined period may be an interval for accumulating value fluctuation assets. For example, when accumulating monthly variable value assets, the predetermined period may be January (or 23 business days). For example, when accumulating weekly value-changing assets, the predetermined period may be one week (or 5 business days). For example, in the case of accumulating daily variable asset, the predetermined period may be 1 day (or 9 hours). The time when the value of the value-changing asset becomes the lowest value within the predetermined period can be any time unit such as seconds, minutes, hours, days, or weeks.

  When the time when the value of the value fluctuation asset predicted in step S501 reaches the lowest value, in step S502, the processor unit 212 of the server apparatus 210 executes a process of purchasing the value fluctuation asset. Specifically, the processor unit 212 transmits a purchase request for a variable value asset to the asset trading computer system 300 via the interface unit 211, and executes a process of purchasing a desired variable value asset from the asset trading computer system. . The processor unit 212 may store information on the purchased value-changing assets in the database unit 220.

  By repeating step S501 and step S502, value fluctuation assets are accumulated every predetermined period. Information on the accumulated value variation assets is stored in the database unit 220.

  In addition, after step S501 and before step S502, input is made to the user to confirm with the user that the value-changing asset predicted at step S501 will be purchased at the time when the value of the value-changing asset is the lowest price. An operation may be performed. However, step S502 is preferably performed automatically without any input operation by the user. This is because it is bothersome for the user and impairs convenience that the user needs to confirm each time the value-changing asset is purchased. Regarding the automatic execution of step S502, for example, the user's consent may be obtained at the time of application for value change asset accumulation by the user.

  In the example described above, the time when the value of the value-changing asset is the lowest value within the predetermined period is predicted in Step S501. However, the value of the value-changing asset is determined within the predetermined period in Step S501 ′ instead of Step S501. Is the lowest price (for example, the second cheapest value, the fifth cheapest value, the tenth cheapest value, the 14th cheapest value, that is, the nth cheapest value (n is an integer of 2 or more)) You may make it predict the time which becomes. This is a more effective method when the value fluctuation of the value fluctuation asset is highly random and difficult to predict. By not aiming for the lowest price, it is possible to avoid the highest price without taking the lowest price, thus improving the final expected value. Thereby, stable asset purchase can be achieved.

  Note that the processing in the computer system 200 may be processing for selling an asset whose value varies with time. In this case, in step S501, when the specific value prediction AI predicts the time when the value of the value-changing asset becomes the highest value within a predetermined period, and reaches the time when the value of the predicted value-changing asset becomes the highest value. In step S502, the processor unit 212 of the server apparatus 210 may execute a process of selling the value-changing asset. Even in this case, as in the case of step S501 ′ described above, the value of the value-changing asset within a predetermined period has a high value (for example, the second highest value, the fifth highest value, and the tenth highest value). Value, the 14th highest value, that is, the time when it becomes the nth highest value (n is an integer of 2 or more), which is highly random in the value fluctuation of the value fluctuation asset, This method is more effective when it is difficult to predict.By not aiming for the highest value, the lowest expected value can be avoided without taking the highest value, so the final expected value is improved. Can achieve more stable asset sales.

  That is, in the processing in the computer system 200, in step S501, the specific value prediction AI indicates that the value of the value fluctuation asset is a specific value (highest value, lowest price, nth lowest value, or nth value within a predetermined period). When a time when a high value (a value of a specific rank such as n is an integer of 2 or more) is predicted and the time when the value of the predicted value-changing asset becomes a specific value is reached, in step S502, the server device 210 The processor unit 212 may execute a process of trading (purchasing or selling) a value-change asset.

  Instead of predicting the time when the lowest price / highest value is predicted, the time when the value of a specific rank is predicted may be switched every predetermined period, for example. For example, when it is predicted that a trend will be strong in one month (value fluctuation is linear), it may be possible to predict when the price will be the lowest / highest price, or in another month , When it is predicted that no trend will occur (value fluctuations will go up and down in a curve), instead of predicting when it will be the lowest / highest price, it will be possible to predict when it will be the value of a specific rank. May be. This prediction may be performed using, for example, any trend prediction method and / or volatility prediction method.

  Instead of predicting the time when the lowest price / highest value is predicted, the timing when the value of a specific rank is predicted may be switched for each target value fluctuation asset. For example, when analyzing past value fluctuation data for a certain value-changing asset and determining that there is no trend in more than 9/12 months in a year, Instead of predicting the time when the lowest price / highest price is obtained, the time when the value of a specific rank is obtained may be predicted.

  When predicting a time when a value of a specific rank is to be predicted, the specific rank may be set by the specific value prediction AI in consideration of price fluctuation of the value fluctuation asset, or may be manually set. Further, the specific order may be set individually for each target value fluctuation asset. For example, the optimal ranking may be verified by using a back test.

  FIG. 6A shows an example of a process for predicting the time when the value of the value-changing asset becomes the lowest value within a predetermined period in step S501. As described below, the specific value prediction AI predicts a time when the value of the value-changing asset becomes the lowest value within a predetermined period by a binary search. In the binary search, since an appropriate prediction model can be used according to the search width, prediction can be performed with high accuracy. For example, when the search range is long-term, a prediction model for predicting long-term fluctuation can be used, and as the search width becomes short-term, a prediction model for predicting short-term fluctuation can be used. In general, prediction accuracy is better for models that predict short-term fluctuations than for models that predict long-term fluctuations.

  In step S601, the specific value prediction AI divides a predetermined period into two periods. For example, when the predetermined period is February (46 business days), the specific value prediction AI includes 46 business days, the first period including the first business day to the 23rd business day, and the 24th business day to It can be divided into a second period including the 46th business day. For example, when the predetermined period is January (23 business days), the specific value prediction AI includes 23 business days, the first period including the first business day to the 12th business day, and the 13th business day to It can be divided into a second period including the 23rd business day. For example, when the predetermined period is one week (5 business days), a first period including the first business day to the third business day, and a second period including the fourth business day to the fifth business day, Can be divided into Thus, the two periods after the division may be the same length or different lengths. Each of the two periods after the division may include a period of any length.

  In step S602, the specific value prediction AI determines the period to which the time when the value of the value-changing asset is the lowest value belongs among the two periods after the division in step S601. The period to which the time when the value of the value fluctuation asset becomes the lowest price can be determined using, for example, a prediction model that predicts the fluctuation of the prediction target period. In order to determine the period to which the time when the value of the value-changing asset is the lowest value belongs, the specific value prediction AI, for example, determines the period during which the average value of the value-changing asset is low among the two periods after the division To do. This is statistically based on the fact that the time when the value of the value-changing asset is at its lowest value is likely to belong to the period when the average value of the value-changing asset is lower. At this time, the specific value prediction AI learns past value fluctuation data as teacher data so that a prediction of an average value in a predetermined period in the future can be output.

  In step S602, the specific value prediction AI uses a plurality of prediction models to determine a period to which the period when the value of the value-changing asset is the lowest value belongs among the two periods after the division in step S601. It may be. At this time, for example, the processing shown in FIG. 6B may be performed.

  FIG. 6B shows an example of processing when the processing in step S602 is performed using a plurality of prediction models.

  In step S602-1, the specific value prediction AI uses each of a plurality of prediction models, and among the two periods after the division in step S601, the period to which the time when the value of the value-changing asset is the lowest value belongs. judge. For example, when the specific value prediction AI uses five prediction models, the time when the value of the value-changing asset becomes the lowest value in the two periods after the division in step S601 using the first prediction model. The period to which the value change asset belongs is determined from the two periods after the division in step S601 using the second prediction model. Among the two periods after the division in step S601, the period to which the time when the value of the value fluctuation asset becomes the lowest price is determined using the prediction model 5. The determination of the period to which the time when the value of the value-changing asset is at the lowest price belongs, for example, by learning past value change data as teacher data so that a prediction of the average value of a predetermined period in the future can be output. The specific value prediction AI may be performed by determining a period in which the average value of the value-changing assets is low among the two periods after the division. The determination of the period to which the time when the value of the value-changing asset is at the lowest price can be made, for example, so that a prediction of whether the price becomes the lowest price in the first half of the certain period in the future or the lowest price in the second half can be output. You may perform by the specific value prediction AI which is learning the past value fluctuation data as teacher data.

  Next, in step S602-2, the specific value prediction AI is a period to which the period when the value of the value-changing asset becomes the lowest value by more prediction models among the two periods after the division in step S601 belongs. Is determined to be a period to which the time when the value of the value fluctuation asset becomes the lowest value belongs. By using a plurality of prediction models and determining the period to which the time when the value of the value-changing asset is the lowest value belongs, the prediction accuracy can be improved. For example, the first prediction model determines that the first half of the two periods after the division in step S601 is the period to which the time when the value of the value-changing asset is the lowest value belongs, It is determined by the prediction model that the first half of the two periods after the division in step S601 is a period to which the time when the value of the value-changing asset is the lowest value belongs. It is determined that the latter half of the two periods after the division in S601 is a period to which the time when the value of the value-changing asset is the lowest value belongs, and after the division in step S601 by the fourth prediction model The first half of the two periods is determined to be the period to which the time when the value of the value-changing asset is the lowest value belongs, and the fifth prediction model If it is determined that the latter half of the two periods after the division at 601 belongs to the period when the value of the value-changing asset is the lowest price, the two periods after the division at step S601 The number of prediction models determined that the first half of the period is the period to which the time when the value of the value-changing asset is the lowest value belongs is the latter half of the two periods after the division in step S601. Since there are more than the number of prediction models determined to be the period to which the time when the value of the value-changing asset is the lowest value belongs, the specific value prediction AI is a majority decision, and is one of the two periods after the division in step S601. The first half period is determined to be a period to which the time when the value of the value-changing asset is the lowest value belongs.

  For example, the number of forecast models determined that the period when the value of the value-changing asset is at the lowest price belongs to the first half of the period, and the period to which the period when the value of the value-changing asset is the lowest is the latter half If the number of prediction models determined to be the same, the predetermined period may be determined as the period to which the period when the value of the value-changing asset is the lowest price belongs. It is preferable to determine that this period is a period to which the period when the value of the value fluctuation asset is the lowest value belongs. This is because the potential risk is greater in the first half, as will be described later.

  Returning to FIG. 6A again, in step S603, the specific value prediction AI has the lowest value of the value-changing asset during the period in which the time when the value of the value-changing asset is the lowest value belongs in step S602. It is determined whether or not the length is the same as the time. When it is determined in step S602 that the period in which the time when the value of the value-changing asset is the lowest value belongs is not the same length as the time when the value of the value-changing asset is the lowest value (in the case of No), The process proceeds to step S604. When it is determined in step S602 that the period in which the day when the value of the value-changing asset is at the lowest price belongs is the same length as the period when the value of the value-changing asset is the lowest (if Yes) , The process proceeds to S606. This is because it can be determined in step S602 that the period in which the time when the value of the value-changing asset is the lowest belongs belongs to the time when the value of the value-changing asset is the lowest.

  In step S604, the specific value prediction AI further divides the period in which it is determined that the period when the value of the value-changing asset is the lowest price belongs into two periods. For example, when the period in which the time when the value of the value-changing asset is at the lowest price belongs is 11 business days, the specific value prediction AI includes 11 business days and includes the first business day to the sixth business day. It can be divided into a first period and a second period including the seventh business day to the eleventh business day. For example, when the period in which the time when the value of the value-changing asset is at the lowest price belongs is 6 business days, the specific value prediction AI includes 6 business days, the first business day to the third business day. It can be divided into a first period and a second period including the fourth business day to the sixth business day. Thus, the two periods after the division may be the same length or different lengths. Each of the two periods after the division may include a period of any length.

  In step S605, the specific value prediction AI determines a period to which the time when the value of the value-changing asset is the lowest value belongs among the two periods after the division in step S604. The period to which the time when the value of the value fluctuation asset becomes the lowest price can be determined using, for example, a prediction model that predicts the fluctuation of the prediction target period. In order to determine the period to which the time when the value of the value-changing asset is the lowest value belongs, the specific value prediction AI, for example, determines the period during which the average value of the value-changing asset is low among the two periods after the division To do. This is statistically based on the fact that the time when the value of the value-changing asset is at its lowest value is likely to belong to the period when the average value of the value-changing asset is lower. At this time, the specific value prediction AI learns past value fluctuation data as teacher data so that a prediction of an average value in a predetermined period in the future can be output.

  In step S605, the specific value prediction AI uses a plurality of prediction models to determine a period to which the period when the value of the value-changing asset is the lowest value belongs among the two periods after the division in step S604. It may be. At this time, for example, a process similar to the process shown in FIG. 6B may be performed.

  In step S605-1, the specific value prediction AI uses each of the plurality of prediction models, and among the two periods after the division in step S604, the period to which the time when the value of the value-changing asset is the lowest value belongs. judge. For example, when the specific value prediction AI uses five prediction models, the time when the value of the value-changing asset becomes the lowest value in the two periods after the division in step S604 using the first prediction model. The period to which the value change asset belongs is determined from among the two periods after the division in step S604, using the second prediction model, Using the prediction model 5, the period to which the period when the value of the value-changing asset becomes the lowest value among the two periods after the division in step S <b> 604 is determined. The determination of the period to which the time when the value of the value-changing asset is at the lowest price belongs, for example, by learning past value change data as teacher data so that a prediction of the average value of a predetermined period in the future can be output. The specific value prediction AI may be performed by determining a period in which the average value of the value-changing assets is low among the two periods after the division. The determination of the period to which the time when the value of the value-changing asset is at the lowest price can be made, for example, so that a prediction of whether the price becomes the lowest price in the first half of the certain period in the future or the lowest price in the second half can be output. You may perform by the specific value prediction AI which is learning the past value fluctuation data as teacher data.

  Next, in step S605-2, the specific value prediction AI is a period to which the period when the value of the value fluctuation asset becomes the lowest value by more prediction models among the two periods after the division in step S604 belongs. Is determined to be a period to which the time when the value of the value fluctuation asset becomes the lowest value belongs. For example, it is determined by the first prediction model that the latter half of the two periods after the division in step S604 is a period to which the time when the value of the value-changing asset is the lowest value belongs. It is determined by the prediction model that the first half of the two periods after the division in step S604 is a period to which the time when the value of the value-changing asset is the lowest value belongs. The latter half of the two periods after the division in S604 is determined to be a period to which the time when the value of the value-changing asset is the lowest value belongs, and after the division in step S604 by the fourth prediction model The first half of the two periods is determined to be the period to which the time when the value of the value-changing asset is the lowest value belongs, and the fifth prediction model If it is determined that the latter half of the two periods after the division at 604 belongs to the period when the value of the value-changing asset is the lowest value, the two periods after the division at step S604 The number of prediction models determined that the latter half of the period is the period to which the time when the value of the value-changing asset is the lowest value belongs is the first half of the two periods after the division in step S604. Since there are more than the number of prediction models determined to be the period to which the time when the value of the value fluctuation asset has the lowest price belongs, the specific value prediction AI is a majority decision, and is one of the two periods after the division in step S604. The latter half of the period is determined as the period to which the time when the value of the value-changing asset is the lowest value belongs.

  For example, the number of forecast models determined that the period when the value of the value-changing asset is at the lowest price belongs to the first half of the period, and the period to which the period when the value of the value-changing asset is the lowest is the latter half If the number of prediction models determined to be the same, the predetermined period may be determined as the period to which the period when the value of the value-changing asset is the lowest price belongs. It is preferable to determine that this period is a period to which the period when the value of the value fluctuation asset is the lowest value belongs. This is because the potential risk is greater in the first half, as will be described later.

  After step S605, the process proceeds again to step S603. In step S603, the specific value prediction AI has the same length as the period in which the time when the value of the value-changing asset is the lowest value belongs in step S605 to the time when the value of the value-changing asset is the lowest. It is determined whether or not. When it is determined in step S605 that the period when the time when the value of the value-changing asset is at the lowest price belongs is not the same length as the time when the value of the value-changing asset is at the lowest value (in the case of No), Proceed to step S604 again. Step S603 is performed until it is determined in step S605 that the period in which the time when the value of the value-changing asset is the lowest value belongs is the same length as the time when the value of the value-changing asset is the lowest. Repeat step S605. When it is determined in step S605 that the period during which the time when the value of the value-changing asset is at the lowest price belongs is the same length as the time when the value of the value-changing asset is at the lowest value (Yes) , The process proceeds to S606. This is because it can be determined that the period in which the time when the value of the value-changing asset is the lowest belongs belongs to the time when the value of the value-changing asset is the lowest.

  The timing predicted by the specific value prediction AI in step S501 (step S601 to step S605) is the first half of the predetermined period, in particular, the first period of the predetermined period (for example, the beginning of the month), and is predicted in step S502. When the process of purchasing value-changing assets is executed at a certain time, there is a risk that the value fluctuates in the remaining period of the predetermined period. In other words, if you purchase value-changing assets in the first half of a given period, especially at the beginning of a given period (for example, at the beginning of the month), the uncertainty of value fluctuations in the remaining period of the given period is large, The return may not be commensurate. This is particularly noticeable for assets with strong randomness in value fluctuations. For example, if it is predicted that the price will be the lowest price at the beginning of the month and the value-changing asset is purchased at the beginning of the month, if the trend of the value-changing asset changes in the middle of the month, However, it may result in the highest value. As a result, the prediction result by the specific value prediction AI becomes unstable, and as a result, the accumulation result of the value fluctuation asset also becomes unstable.

  In order to avoid such a case where the risk and the return are not matched, in step S501, the specific value prediction AI sets the first half of the predetermined period, in particular, the beginning of the predetermined period within the predetermined period. When predicting the time when the value of the value-changing asset is the lowest value, it is preferable to limit the value when the confidence level of the prediction model is high. This is because if the confidence level of the prediction model is high, it can be said that the risk that the trend of value-changing assets changes in the middle of the month, and as a result, the maximum value is grasped is extremely low. Thereby, even if the asset has a strong randomness of value fluctuation, it is possible to avoid a case where the risk and the return are not matched, and to obtain a stable prediction result.

  The case where the degree of confidence by the prediction model is high is, for example, a case where a predetermined number or more of the plurality of prediction models perform the same prediction. This is because the accuracy is high when more prediction models make the same prediction. For example, in step S602, when the first half of the predetermined period is predicted as the time when the value of the value-changing asset is the lowest price within the predetermined period, the case where the degree of confidence by the prediction model is high is limited. You may perform the process shown by these.

  FIG. 6C restricts the time when the process of step S602 is performed using a plurality of prediction models, and the first half of the predetermined period is predicted as the time when the value of the value-changing asset is the lowest price within the predetermined period. An example of processing in the case is shown.

  Step S602-1 is the same process as step S602-1 shown in FIG. 6B. The specific value prediction AI uses each of the plurality of prediction models to determine a period to which the period when the value of the value-changing asset is the lowest value belongs among the two periods after the division in step S601.

  In step S602-2 ′, the specific value prediction AI determines that the first half of the two periods after the division in step S601 is a period to which the time when the value of the value-changing asset is the lowest value belongs. Determine the number of models. For example, when each of the five prediction models is used, the first half of the two periods after the division in step S601 belongs to the time when the value of the value-changing asset is the lowest price, according to the first prediction model. It is determined that it is a period, and it is determined by the second prediction model that the first half of the two periods after the division in step S601 is a period to which the time when the value of the value-changing asset is the lowest value belongs. The third prediction model determines that the latter half of the two periods after the division in step S601 is a period to which the time when the value of the value-changing asset is the lowest value belongs. According to the prediction model, it is determined that the first half of the two periods after the division in step S601 is a period to which the time when the value of the value-changing asset is the lowest value belongs. If it is determined by the fifth prediction model that the latter half of the two periods after the division in step S601 is a period to which the time when the value of the value-changing asset becomes the lowest value belongs, The number of prediction models determined that the first half of the two periods after the division in S601 is a period to which the time when the value of the value-changing asset is the lowest value belongs is determined to be three.

  Next, in step S602-3, the specific value prediction AI determines whether or not the number of prediction models determined in step S602-2 'is greater than or equal to a predetermined number. If the number of prediction models is greater than or equal to the predetermined number, the process proceeds to step S602-4. When the number of prediction models is less than the predetermined number, the process proceeds to step S602-5. Here, the predetermined number may be any number larger than half of the total number of prediction models used. The predetermined number may be, for example, the total number of prediction models used or 80% of the total number.

  In step S602-4, the specific value prediction AI determines that the first half of the two periods after the division in step 601 is the period to which the time when the value of the value-changing asset is the lowest value belongs. This is because the degree of confidence by a plurality of prediction models is high and the risk of selecting the first half period is low.

  In step S602-5, the specific value prediction AI determines that the latter half of the two periods after the division in step 601 is the period to which the time when the value of the value-changing asset is the lowest value belongs. This is because the degree of confidence by a plurality of prediction models is low and the risk of selecting the first half period is high.

  FIG. 6C also shows a case where the time when the first half of the predetermined period is predicted as the time when the value of the value-changing asset is the lowest price within the predetermined period is limited to the case where the confidence level based on the prediction model is high. The same processing as that to be performed may be performed.

  Step S605-1 is the same processing as step S605-1 mentioned above. The specific value prediction AI uses each of the plurality of prediction models to determine a period to which the period when the value of the value-changing asset becomes the lowest value belongs among the two periods after the division in step S604.

  In step S605-2 ′, the specific value prediction AI determines that the first half of the two periods after the division in step S604 is a period to which the time when the value of the value-changing asset is the lowest value belongs. Determine the number of models. For example, when each of the five prediction models is used, the first period of the two periods after the division in step S604 belongs to the period when the value of the value-changing asset is the lowest value, according to the first prediction model. It is determined that it is a period, and it is determined by the second prediction model that the first half of the two periods after the division in step S604 is a period to which the time when the value of the value-changing asset is the lowest value belongs. The third prediction model determines that the latter half of the two periods after the division in step S604 is a period to which the time when the value of the value-changing asset is the lowest value belongs. According to the prediction model, it is determined that the first half of the two periods after the division in step S604 is a period to which the time when the value of the value-changing asset is the lowest value belongs. If it is determined by the fifth prediction model that the first half of the two periods after the division in step S604 is a period to which the time when the value of the value-changing asset is the lowest value belongs, The number of prediction models determined that the first half of the two periods after the division in S604 is a period to which the time when the value of the value-changing asset is the lowest value belongs is determined to be four.

  Next, in step S605-3, the specific value prediction AI determines whether or not the number of prediction models determined in step S605-2 'is greater than or equal to a predetermined number. If the number of prediction models is greater than or equal to the predetermined number, the process proceeds to step S605-4. When the number of prediction models is less than the predetermined number, the process proceeds to step S605-5. Here, the predetermined number may be any number larger than half of the total number of prediction models used. The predetermined number may be, for example, the total number of prediction models used or 80% of the total number.

  For example, the predetermined number may be changed according to the position of the divided period with respect to the entire predetermined period. For example, the two periods after the division in step S604 are one of the first half of the predetermined period (for example, the predetermined period is 23 business days, and the two periods after the division in step S604 are the first period) In the case of the first period including the first business day to the sixth business day and the second period including the seventh business day to the eleventh business day), the predetermined number may be high. In particular, the closer the divided period is to the start point of the predetermined period, the higher the predetermined number is preferable. This is because the closer the period after the division is to the start point of the predetermined period, the greater the risk to be borne. For example, when the predetermined period is 23 business days and the two periods after the division in step S604 are the first period including the first business day and the second period including the second business day, The predetermined number is preferably the maximum. For example, when the divided period is somewhat apart from the starting point of the predetermined period (for example, the predetermined period is 23 business days, and the two periods after the division in step S604 are the third business day to Even in the first period including the fourth business day and the second period including the fifth business day to the sixth business day), the predetermined number may be maximized in consideration of the risk. For example, the two periods after the division in step S604 are one of the latter half of the predetermined period (for example, the predetermined period is 23 business days, and the two periods after the division in step S604 are In the case of the first period including 13 business days to 15 business days and the second period including 16th business day to 18th business day), the predetermined number may be low. If the predetermined number is the minimum value, the process is equivalent to step S602-2.

  The predetermined number may be set by the specific value prediction AI in consideration of the price fluctuation of the value fluctuation asset, or may be set manually. The predetermined number may be individually set for each target value fluctuation asset.

  In step S605-4, the specific value prediction AI determines that the first half of the two periods after the division in step 604 is a period to which the time when the value of the value-changing asset is the lowest value belongs. This is because the degree of confidence by a plurality of prediction models is high and the risk of selecting the first half period is low.

  In step S605-5, the specific value prediction AI determines that the latter half of the two periods after the division in step 604 is a period to which the time when the value of the value-changing asset is the lowest value belongs. This is because the degree of confidence by a plurality of prediction models is low and the risk of selecting the first half period is high.

  In order to avoid the case where the risk and the return are not matched, in step S501, the specific value prediction AI additionally or alternatively indicates the first half of the predetermined period, in particular, the timing at the beginning of the predetermined period. The prediction of the time when the value of the value fluctuation asset becomes the lowest value within a predetermined period may be suppressed. This is achieved, for example, by providing a predetermined suppression period and suppressing the prediction of the time within the predetermined suppression period as the time when the value of the value fluctuation asset becomes the lowest value within the predetermined period.

  The predetermined suppression period is a period starting from the start point of the predetermined period, and the length of the period may be an arbitrary length. For example, the length of the predetermined suppression period may be changed according to the target asset. For example, in the case of an asset whose trend is easy to read (for example, when euro is purchased in Japanese yen), the length of the predetermined suppression period can be shortened. For example, in the case of an asset whose trend is difficult to read (for example, when purchasing NZ dollars in Japanese yen), the length of the predetermined suppression period can be increased.

The following table is an example of the length of the predetermined suppression period when the predetermined period is January.

  The length of the predetermined suppression period may be set by the specific value prediction AI in consideration of the price fluctuation of the value fluctuation asset, or may be set manually.

  Here, as an example, the process of step S501 in the computer system 200 in the case of accumulating US dollars every month will be described with reference to FIGS. 7A to 7D. The predetermined period is assumed to be January (23 business days). The period when the value of the value-changing asset is at its lowest value is in units of days. 7A to 7D are graphs showing an example of time fluctuation of the US dollar price (yen). A solid line shows an actual fluctuation value, and a broken line shows a fluctuation value predicted using a prediction model that predicts fluctuation in the prediction target period.

  In step 601, the specific value prediction AI divides a predetermined period (23 business days) into two periods on the first day of the month. As shown in FIG. 7A, the specific value prediction AI includes a predetermined period (23 business days), a first period 701 including the first business day to the 12th business day, and a 13th business day to the 23rd business day. Dividing into a second period 702 of days.

  In step S602, the specific value prediction AI determines a period to which the day when the US dollar price is the lowest price belongs among the first period 701 and the second period 702. The specific value prediction AI uses a prediction model that predicts fluctuations for 23 days, and determines a period in which the average value of the US dollar price is low in the first period 701 and the second period 702, and determines the average of the US dollar price. The period in which the value is low is determined to be the period to which the day when the US dollar price is the lowest price belongs. In this example, the specific value prediction AI uses an average value 710 of the first period 701 predicted using a prediction model that predicts fluctuations for 23 days and a prediction model that predicts fluctuations for 23 days. The average value 720 of the predicted second period 702 is compared, and it is determined that the second period 702 is a period to which the day when the US dollar price is the lowest price belongs.

  As an example, a multi-layer perceptron including an input layer, three hidden layers, and an output layer, each node being 31, 16, 8, 4, 1 and decreasing from the input layer toward the output layer. In the first period 701 and the second period 702, the period in which the average value of the US dollar price is low is determined by the specific value prediction AI using the multi-layer perceptron as a prediction model for predicting the fluctuation for 23 days. When the process of determining that the period when the average price is low is the period to which the day when the US dollar price is the lowest belongs, the following maximum prediction accuracy was obtained. Note that the value fluctuation data used for learning the specific value prediction AI is only the value fluctuation data of the US dollar against the Japanese yen.

Here, “random” is a model that predicts the day when the price is the lowest at random, and was used as a comparison target. Also from this result, it can be seen that the more data input to the multilayer perceptron, the better the prediction accuracy.

  Further, as an example, a specific value prediction AI using CNN as a prediction model determines a period in which the average value of the US dollar price is low in the first period 701 and the second period 702, and the average value of the US dollar price is low When the process for determining that the period is the period to which the day when the US dollar price is the lowest belongs, when the value fluctuation data used for learning is changed, the following maximum prediction accuracy is obtained.

Thus, the prediction model to be used and the value fluctuation data to be learned are closely related to the prediction accuracy. By selecting an appropriate prediction model, value variation data to be learned, a prediction accuracy exceeding 70% can be achieved.

  In step S603, the specific value prediction AI determines whether or not the second period 702 is one day. Since the second period 702 is a period of 11 days consisting of the 13th business day to the 23rd business day, the process proceeds to step S604.

  In step S604, the specific value prediction AI divides the second period 702 into two periods on the first day of the second period 702. As shown in FIG. 7B, the specific value prediction AI includes the second period 702, the third period 703 including the 13th business day to the 17th business day, and the 18th business day to the 23rd business day. The period is divided into a fourth period 704.

  Note that the example shown in FIG. 7B shows that step S604 is executed on the first day of the second period 702, but the timing at which each step of the process of step S501 is performed is arbitrary. For example, on the first day of the month, all of the processing in step S501 (processing for predicting the day when the value of the value-changing asset becomes the lowest price within a predetermined period) may be completed. A part of the process of step S501 (a process for predicting the day when the value of the value-changing asset becomes the lowest price within a predetermined period) is performed every time it is determined that the day with the lowest price belongs. Also good.

  In step S605, the specific value prediction AI determines a period to which the day when the US dollar price is the lowest price belongs among the third period 703 and the fourth period 704. The specific value prediction AI uses a prediction model that predicts fluctuations for 13 days, determines a period in which the average value of the US dollar price is low in the third period 703 and the fourth period 704, and determines the average of the US dollar price. The period in which the value is low is determined to be the period to which the day when the US dollar price is the lowest price belongs. In this example, the specific value prediction AI uses an average value 730 of the third period 703 predicted using a prediction model that predicts fluctuations for 13 days and a prediction model that predicts fluctuations for 13 days. The average value 740 of the predicted fourth period 704 is compared, and it is determined that the third period 703 is a period to which the day when the US dollar price is the lowest price belongs.

  Returning to step S603, the specific value prediction AI determines whether or not the third period 703 is one day. Since the third period 703 is a period of five days including the thirteenth business day to the seventeenth business day, the process proceeds to step S604 again.

  In step S604, the specific value prediction AI divides the third period 703 into two periods. As shown in FIG. 7C, the specific value prediction AI includes the third period 703, the fifth period 705 including the thirteenth business day to the fourteenth business day, and the fifteenth business day to the seventeenth business day. The period is divided into a sixth period 706.

  In step S605, the specific value prediction AI determines a period to which the day when the US dollar price is the lowest price belongs among the fifth period 705 and the sixth period 706. The specific value prediction AI uses a prediction model that predicts fluctuations for five days, and determines a period in which the average value of the US dollar price is low in the fifth period 705 and the sixth period 706, and the average of the US dollar price The period in which the value is low is determined to be the period to which the day when the US dollar price is the lowest price belongs. In this example, the specific value prediction AI uses an average value 750 of the fifth period 705 predicted using a prediction model that predicts fluctuations for 5 days and a prediction model that predicts fluctuations for 5 days. The predicted average value 760 of the sixth period 706 is compared, and it is determined that the fifth period 705 is a period to which the day when the US dollar price is the lowest price belongs.

  Returning to step S603, the specific value prediction AI determines whether or not the fifth period 705 is one day. Since the fifth period 705 is a period of two days including the thirteenth business day to the fourteenth business day, the process proceeds to step S604 again.

  In step S604, the specific value prediction AI divides the fifth period 705 into two periods. As shown in FIG. 7D, the specific value prediction AI divides the fifth period 705 into a seventh period 707 composed of the 13th business day and an eighth period 708 composed of the 14th business day.

  In step S605, the specific value prediction AI determines a period to which the day when the US dollar price is the lowest price belongs among the seventh period 707 and the eighth period 708. The specific value prediction AI uses a prediction model that predicts fluctuations for two days, and determines a period in which the average value of the US dollar price is low in the seventh period 707 and the eighth period 708, and the average of the US dollar price The period in which the value is low is determined to be the period to which the day when the US dollar price is the lowest price belongs. In this example, the specific value prediction AI uses an average value 770 of the seventh period 707 predicted using a prediction model that predicts fluctuations for two days and a prediction model that predicts fluctuations for two days. The predicted average value 780 of the eighth period 708 is compared, and it is determined that the eighth period 708 is a period to which the day when the US dollar price is the lowest price belongs.

  Returning to step S603, the specific value prediction AI determines whether or not the eighth period 708 is one day. Since the eighth period 708 is a period of one day consisting of the 14th business day, the process proceeds to step S606 and ends.

  With the above processing, it is predicted that the day when the US dollar is the lowest price in this month is the 14th business day. When the 14th business day is reached, in step S502, the processor unit 212 purchases US dollars and accumulates them.

  Here, as a further example, the processing in step S501 in the computer system 200 when accumulating the Australian dollars every month will be described with reference to FIGS. 8A to 8D.

8A to 8D are graphs showing an example of time fluctuation of the Australian dollar price (yen). A solid line shows an actual fluctuation value, and a broken line shows a fluctuation value predicted using a prediction model that predicts fluctuation in the prediction target period. The predetermined period is assumed to be January (23 business days). The period when the value of the value-changing asset is at its lowest value is the daily unit. The length of the predetermined suppression period is 4 days. The specific value prediction AI is predicted using five prediction models. The predetermined number shall be changed as follows.

  In step 601, the specific value prediction AI divides a predetermined period (23 business days) into two periods on the first day of the month. As shown in FIG. 8A, the specific value prediction AI includes a predetermined period (23 business days), a first period 801 composed of the first business day to the twelfth business day, and a thirteenth business day to the 23rd business day. Dividing into a second period 802 of days.

In step S602-1, the specific value prediction AI determines, using each of the five prediction models, a period to which the day when the Australian dollar price is the lowest price belongs among the first period 801 and the second period 802. To do. The specific value forecast AI is a period to which the day when the Australian dollar price is the lowest among the first period 801 and the second period 802 belongs to each of the five forecast models for forecasting the fluctuation for 23 days. Determine. The prediction model may be, for example, a plurality of prediction models with different types or amounts of past value change data used for learning. For example, the type or amount of past value change data input for prediction may be A plurality of different prediction models may be used. For example, a plurality of prediction models having different types of neural networks to be used, such as a prediction model using a multilayer perceptron 450, a prediction model using a CNN, a prediction model using an RNN, etc. It may be. In this example, it is assumed that each prediction model predicts as follows.

  In step S602-2 ′, the specific value prediction AI is a period to which a day in which the first period 801, which is the first half of the first period 801 and the second period 802, has the lowest Australian dollar price, belongs. The number of prediction models determined to be is determined. From the results in Table 5, the number of prediction models is determined to be 3.

  In step S602-3, the specific value prediction AI determines whether or not the number of prediction models determined in step S602-2 'is greater than or equal to a predetermined number. Since the predetermined number from the first business day to the twelfth business day is 3, and the number of prediction models determined in step S602-2 'is three or more, the process proceeds to step S602-4.

  In step S602-4, the specific value prediction AI determines that the first period 801 is a period to which the day when the Australian dollar price becomes the lowest price belongs.

  In step S603, the specific value prediction AI determines whether or not the first period 801 is one day. Since the first period 801 is a period of 12 days including the first business day to the twelfth business day, the process proceeds to step S604.

  In step S604, the specific value prediction AI divides the first period 801 into two periods. As shown in FIG. 8B, the specific value prediction AI includes a first period 801, a third period 803 composed of the first business day to the sixth business day, and a seventh business day to the twelfth business day. The period is divided into a fourth period 804.

  Note that the example shown in FIG. 8B shows that step S604 is executed on the first day of the third period 803, but the timing at which each step of the process of step S501 is performed is arbitrary. For example, on the first day of the month, all of the processing in step S501 (processing for predicting the day when the value of the value-changing asset becomes the lowest price within a predetermined period) may be completed. A part of the process of step S501 (a process for predicting the day when the value of the value-changing asset becomes the lowest price within a predetermined period) is performed every time it is determined that the day with the lowest price belongs. Also good.

In step S605, the specific value prediction AI determines a period to which the day when the Australian dollar price becomes the lowest price belongs in the third period 803 and the fourth period 804. The specific value prediction AI uses five prediction models for predicting fluctuations for 13 days, and determines the period to which the day when the Australian dollar price is the lowest price belongs among the third period 803 and the fourth period 804 To do. In this example, it is assumed that each prediction model predicts as follows.

  In step S605-2 ′, the specific value prediction AI is a period to which the day in which the third period 803, which is the first half of the third period 803 and the fourth period 804, has the lowest Australian dollar price belongs. The number of prediction models determined to be is determined. From the results in Table 6, it is determined that the number of prediction models is four.

  In step S605-3, the specific value prediction AI determines whether or not the number of prediction models determined in step S605-2 'is greater than or equal to a predetermined number. Since the predetermined number of the first business day to the sixth business day is 5, and the number of prediction models determined in step S605-2 'is not 5 or more, the process proceeds to step S605-5.

  In step S605-5, the specific value prediction AI determines that the fourth period 804, which is the latter half of the period, is a period to which the day when the Australian dollar price is the lowest price belongs.

  Returning to step S603, the specific value prediction AI determines whether or not the fourth period 804 is one day. Since the fourth period 804 is a period of six days consisting of the seventh business day to the twelfth business day, the process proceeds again to step S604.

  In step S604, the specific value prediction AI divides the fourth period 804 into two periods. As shown in FIG. 8C, the specific value prediction AI includes the fourth period 804, the fifth period 805 including the seventh business day to the ninth business day, and the tenth business day to the twelfth business day. This is divided into a sixth period 806.

In step S605, the specific value prediction AI determines a period to which the day when the Australian dollar price is the lowest price belongs in the fifth period 805 and the sixth period 806. The specific value prediction AI uses five prediction models for predicting fluctuations for five days, and determines the period to which the day when the Australian dollar price is the lowest among the fifth period 805 and the sixth period 806 belongs. To do. In this example, it is assumed that each prediction model predicts as follows.

  In step S605-2 ′, the specific value prediction AI is a period to which the day in which the fifth period 805, which is the first half of the fifth period 805 and the sixth period 806, has the lowest Australian dollar price belongs. The number of prediction models determined to be is determined. From the results in Table 7, the number of prediction models is determined to be 5.

  In step S605-3, the specific value prediction AI determines whether or not the number of prediction models determined in step S605-2 'is greater than or equal to a predetermined number. Since the predetermined number of the seventh business day to the ninth business day is 4, and the number of prediction models determined in step S605-2 'is four or more, the process proceeds to step S605-4.

  In step S605-4, the specific value prediction AI determines that the fifth period 805 is a period to which the day when the Australian dollar price becomes the lowest price belongs.

  Returning to step S603, the specific value prediction AI determines whether or not the fifth period 805 is one day. Since the fifth period 805 is a period of three days including the seventh business day to the ninth business day, the process proceeds to step S604 again.

  In step S604, the specific value prediction AI divides the fifth period 805 into two periods. As shown in FIG. 8D, the specific value prediction AI includes a fifth period 805, a seventh period 807 composed of the seventh business day, and an eighth period 808 composed of the eighth business day to the ninth business day. And split into

In step S605, the specific value prediction AI determines the period to which the day when the Australian dollar price becomes the lowest price belongs in the seventh period 807 and the eighth period 808. The specific value prediction AI uses five prediction models for predicting fluctuations for two days, and determines the period to which the day when the Australian dollar price is the lowest among the seventh period 807 and the eighth period 808 belongs. To do. In this example, it is assumed that each prediction model predicts as follows.

  In step S605-2 ′, the specific value prediction AI is a period to which a day in which the seventh period 807, which is the first half of the seventh period 807 and the eighth period 808, has the lowest Australian dollar price belongs. The number of prediction models determined to be is determined. From the results in Table 8, the number of prediction models is determined to be 4.

  In step S605-3, the specific value prediction AI determines whether or not the number of prediction models determined in step S605-2 'is greater than or equal to a predetermined number. Since the predetermined number of the seventh business day is 4, and the number of prediction models determined in step S605-2 'is 4 or more, the process proceeds to step S605-4.

  In step S605-4, the specific value prediction AI determines that the seventh period 807 is a period to which the day when the Australian dollar price becomes the lowest price belongs.

  Returning to step S603, the specific value prediction AI determines whether or not the seventh period 807 is one day. Since the seventh period 807 is a period of one day consisting of the seventh business day, the process proceeds to step S606 and ends.

  With the above processing, it is predicted that the day when the Australian dollar is the lowest price in this month is the seventh business day. When the seventh business day is reached, in step S502, the processor unit 212 purchases Australian dollars and accumulates them.

  When predicting the time when the value of the specific rank is predicted instead of predicting the time when the price is the lowest, in steps S602 and S605, the period to which the time when the value of the asset becomes the specific rank belongs is determined. Also good. For example, a case will be described in which a period to which the time when the asset value is the 14th lowest value belongs is determined in 23 business days.

  In step S601, after 23 business days are divided into two periods, in step S602, the period in which the lowest price does not exist is selected from the two periods after the division. This is because 14> 23/2.

  Next, after step S603, 11 or 12 business days are divided into two periods in step S604, and in step S605, the period in which the low price (minimum value) exists among the two periods after the division is determined. select. This is because (14-23 / 2) <23/4.

  Next, after step S603, 5 or 6 business days are divided into two periods in step S604, and then in step S605, the period in which the low price (minimum value) exists among the two periods after the division is determined. select. This is because (14-23 / 2) <23/8.

  Next, after step S603, after 2 or 3 business days are divided into two periods in step S604, in step S605, the period in which there is no low (minimum) value among the two periods after the division is determined. select. This is because (14-23 / 2)> 23/16.

  This is repeated until it is determined Yes in step S603.

  In this way, in the same manner as the binary search described above, it is possible to determine the period to which the period when the asset value is in a specific order belongs.

  In the example described above, it has been described that the time when the value of the value-changing asset becomes a value of a specific rank within a predetermined period is described by a binary search that divides the target period into two periods. It is not limited to binary search. It is also within the scope of the present invention to divide the target period into more than two periods and predict when the value of the value-changing asset will have a specific rank value within a predetermined period.

  For example, in step S601, the specific value prediction AI is divided into m periods (m is an integer greater than 2) instead of dividing a predetermined period into two periods, and in step S602, the specific value prediction AI Is the period to which the time when the value of the value-changing asset becomes a value of a specific rank (for example, the lowest price, the highest value, the nth lowest value, the nth highest value) among the m periods after the division May be determined. When determining the period to which the time when the value of the value-changing asset is at the lowest price belongs, this determination is performed by, for example, teaching the past value fluctuation data so that a prediction of an average value in a predetermined period in the future can be output. It may be performed by determining a period in which the average value of the value-changing assets is the lowest among the m periods after the division by the specific value prediction AI learned as data. The determination of the period to which the time when the value of the value-changing asset is at the lowest price belongs, for example, to output a prediction of which period among the periods obtained by dividing a certain period in the future into the lowest price It may be performed by specific value prediction AI that learns past value fluctuation data as teacher data so that it can be performed. When determining the period to which the time when the value of the value-changing asset has the highest value belongs, this determination is performed by, for example, teaching the past value fluctuation data so that a prediction of an average value in a predetermined period in the future can be output. It may be performed by determining a period in which the average value of the value-changing assets is the highest among the m periods after the division by the specific value prediction AI learned as data. The determination of the period to which the value of the value-changing asset has the highest value belongs, for example, outputting a prediction of which period among the periods obtained by dividing a certain period in the future into the highest value It may be performed by specific value prediction AI that learns past value fluctuation data as teacher data so that it can be performed. When determining the period to which the period when the value of the value-changing asset becomes a value of a specific rank belongs, this determination can be performed by, for example, past value fluctuations so that a forecast of the average value of a predetermined period in the future can be output. It may be performed by determining a period in which the average value of the value-changing assets has a specific rank among the m periods after the division by the specific value prediction AI learning the data as teacher data. . The determination of the period to which the time when the value of the value-changing asset has a value of a specific rank belongs, for example, in which period of a period obtained by dividing a certain period in the future into a value of a specific rank May be performed by a specific value prediction AI in which past value fluctuation data is learned as teacher data so that the prediction of the current value can be output.

  For example, in step S604, the specific value prediction AI further divides the period in which the time when the value of the value fluctuation asset has a specific rank value belongs into m periods, and in step S605, the specific value prediction The AI belongs to a period in which the value of the value-changing asset becomes a value of a specific rank (for example, the lowest price, the highest price, the nth lowest value, the nth highest value) among the m periods after the division. The period may be determined. When determining the period to which the time when the value of the value-changing asset is at the lowest price belongs, this determination is performed by, for example, teaching the past value fluctuation data so that a prediction of an average value in a predetermined period in the future can be output. It may be performed by determining a period in which the average value of the value-changing assets is the lowest among the m periods after the division by the specific value prediction AI learned as data. The determination of the period to which the time when the value of the value-changing asset is at the lowest price belongs, for example, to output a prediction of which period among the periods obtained by dividing a certain period in the future into the lowest price It may be performed by specific value prediction AI that learns past value fluctuation data as teacher data so that it can be performed. When determining the period to which the time when the value of the value-changing asset has the highest value belongs, this determination is performed by, for example, teaching the past value fluctuation data so that a prediction of an average value in a predetermined period in the future can be output. It may be performed by determining a period in which the average value of the value-changing assets is the highest among the m periods after the division by the specific value prediction AI learned as data. The determination of the period to which the value of the value-changing asset has the highest value belongs, for example, outputting a prediction of which period among the periods obtained by dividing a certain period in the future into the highest value It may be performed by specific value prediction AI that learns past value fluctuation data as teacher data so that it can be performed. When determining the period to which the period when the value of the value-changing asset becomes a value of a specific rank belongs, this determination can be performed by, for example, past value fluctuations so that a forecast of the average value of a predetermined period in the future can be output. It may be performed by determining a period in which the average value of the value-changing assets has a specific rank among the m periods after the division by the specific value prediction AI learning the data as teacher data. . The determination of the period to which the time when the value of the value-changing asset has a value of a specific rank belongs, for example, in which period of a period obtained by dividing a certain period in the future into a value of a specific rank May be performed by a specific value prediction AI in which past value fluctuation data is learned as teacher data so that the prediction of the current value can be output.

  At this time, for example, in step S602-2 or step S605-2, the specific value prediction AI is a value with a specific rank in which the value of the value fluctuation asset is determined by more prediction models in the m periods after the division. It is determined that a period determined to be a period to which a certain period belongs is a period to which a period in which the value of the asset has a specific rank value belongs.

  For example, in step S602-2 ′ or step S605-2 ′, the specific value prediction AI is the rank in which the value of the value fluctuation asset has a specific rank for each of the first half of the m periods after the division. The number of prediction models determined to be the period to which the time of the value belongs is determined. Next, in step S602-3 or step S605-3, the specific value prediction AI is determined for each of the first half periods. It is determined whether the number of predicted models is equal to or greater than a predetermined number. When the number is greater than or equal to the predetermined number, the period is determined to be a period to which a period in which the value of the value-changing asset becomes a specific rank value belongs.

  For example, when m = 6, in step S602-2 ′ or step S605-2 ′, the specific value prediction AI is the value of the value-changing asset for each of the first three periods of the six divided periods. The number of prediction models determined to be a period to which the period when the value becomes a value of a specific rank belongs is determined. Next, in step S602-3 or step S605-3, the specific value prediction AI determines whether or not the number of prediction models determined for each of the first three periods is a predetermined number, and is greater than or equal to the predetermined number. If it is, the period is determined to be a period to which the period when the value of the value-changing asset becomes a specific rank value belongs. For example, each of the number of prediction models determined for the first period of the first three periods and the number of prediction models determined for the third period of the first three periods are less than a predetermined number. When the number of prediction models determined for the second period of the first three periods is equal to or greater than the predetermined number, the value of the value-changing asset is specified in the second period of the first three periods. It is determined that it is a period to which the period of the rank value belongs.

  At this time, if there is no period in which the number of prediction models is equal to or greater than the predetermined number, for example, the entire period in the latter half of the plurality of periods is a period to which the period when the value of the value-changing asset is a specific rank value belongs. It may be determined that there is, and the process may proceed to step S603, or the entire period excluding the earliest period among the plurality of periods may be a period to which the period when the value of the value-changing asset has a specific ranking value belongs. It may be determined to proceed to step S603. As a result, it is possible to avoid selecting the first period of the predetermined period when the confidence level of the prediction model is low, and as a result, it is possible to avoid the case where the risk and the return are not matched. .

4). Alternate configuration diagram 9 of a computer system for implementing the new financial instruments shows an example of a computer system 200 'which is an alternative configuration of a computer system 200 for implementing the new financial instruments described above. In FIG. 9, the same elements as those shown in FIG. 2 are denoted by the same reference numerals, and description thereof is omitted here. In the computer system 200 ′, the user device 100 predicts the time when the value of the value-changing asset becomes a specific value within a predetermined period, and the value change occurs when the value of the predicted value-changing asset becomes the specific value. In this configuration, the server device 210 performs an asset transaction.

  The computer system 200 ′ includes a server device 210, a database unit 220 connected to the server device 210, and the user device 100. Server device 210, user device 100, and asset transaction computer system 300 are connected to each other via network 400. Here, the type of the network 400 is not limited. The network 400 may be, for example, the Internet or a LAN.

  The user device 100 is a terminal used by the user 10 who uses a service for accumulating value-changing assets, and may be any terminal device such as a smartphone, a tablet, or a personal computer. The user device 100 includes an interface unit 111, a processor unit 112, and a storage unit 113.

  The interface unit 111 is configured to control communication via the network 400. The interface unit 111 can transmit information to the outside of the user apparatus 100, and can receive information from the outside of the user apparatus 100. For example, the interface unit 111 transmits, via the network 400, the prediction of the date when the value of the value-changing asset becomes the lowest value within a predetermined period to the server apparatus 210.

  The processor unit 112 controls the overall operation of the user device 100. The processor unit 112 reads a program stored in the storage unit 113 and executes the program. Thereby, it is possible to make the user apparatus 100 function as an apparatus that executes a desired step. For example, the processor unit 112 can perform a part of the process of accumulating value fluctuation assets. The processor unit 112 may be implemented by a single processor or may be implemented by a plurality of processors.

  The storage unit 113 stores a program for executing processing in the user device 100, data necessary for executing the program, and the like. In the storage unit 113, for example, a program for trading value-changing assets or a program for accumulating value-changing assets (for example, a program that realizes the processes shown in FIGS. 5 and 6A to 6C described above) ) Is stored. Here, it does not matter how the program is stored in the storage unit 113. For example, the program may be preinstalled in the storage unit 113. Alternatively, the program may be installed in the storage unit 113 by being downloaded via the network 400. The storage unit 113 can be implemented by any storage unit.

  For example, past value fluctuation data can also be stored in the storage unit 113. Past value change data may be used to predict future value changes. Historical value fluctuation data used for learning is for other currencies (for example, US dollar, Australian dollar, Euro, NZ dollar, etc.) in addition to the target asset value fluctuation data for purchase currency (for example, Japanese yen). Value fluctuation data of the target asset, or data such as S & P 500 index, Nikkei average stock price, German stock price index, or WTI crude oil price may be included. The database unit 220 can also store learned value fluctuation trends, for example. The database unit 220 may store the weight vector of each node of the multilayer perceptron 450 established by learning.

  The specific value prediction AI can be realized by the user apparatus 100 having the above-described configuration. As described above, the specific value prediction AI can predict the time when the value of the value-changing asset is the lowest value using an arbitrary prediction model.

  In the example shown in FIG. 9, each component of the user apparatus 100 is provided in one user apparatus 100, but the present invention is not limited to this. Any one of the components of the user device 100 may be provided outside the user device 100. For example, when each of the processor unit 112 and the storage unit 113 is configured by separate hardware components, the hardware components may be connected via an arbitrary network. At this time, the type of network is not limited. Each hardware component may be connected via a LAN, wirelessly connected, or wiredly connected, for example.

  The processing in the computer system 200 ′ is the same as the processing 500 in the computer system 200. However, the process of step S501 is different from the process 500 in the computer system 200 in that the specific value prediction AI realized by the user device 100 performs the process and the process of step S502 is performed by the processor unit 212 of the server apparatus 210. .

  In the above-described example, it has been described that the day when the value of the value-changing asset is the lowest value is predicted. However, the present invention predicts a time unit shorter than “day” (for example, hour, minute, second, etc.). It is also within the scope of the present invention to predict or a unit of time longer than “day” (eg, week, month, etc.). For example, in the case of accumulating daily variable value assets, the computer system 200 or 200 ′ performs the same process as described above with reference to FIG. 5 and FIGS. Can be predicted when the price becomes the lowest price, and the value-change asset can be purchased when the value of the predicted value-change asset becomes the lowest price. Here, when the value of the value fluctuation asset becomes the lowest value, it can be an arbitrary unit of time. When the value of the value-changing asset is the lowest value, for example, it may be in seconds, minutes, or hours.

  The present invention is not limited to the embodiment described above. It is understood that the scope of the present invention should be construed only by the claims. It is understood that those skilled in the art can implement an equivalent range based on the description of the present invention and the common general technical knowledge from the description of specific preferred embodiments of the present invention.

  The present invention provides the highest or lowest value (ie, maximum or minimum value) or a value of a specific rank (for example, the second value) of an asset whose value fluctuates irregularly over time. The present invention is useful for providing a computer system, a method, and a program that make it possible to predict the time when a specific value such as a cheap value, a 10th highest value, or the like is reached. Furthermore, the present invention provides a computer system, a method, and a computer system that can increase the value of the asset by purchasing or accumulating the asset periodically or continuously based on the prediction of a specific value. It is also useful for providing a program.

10 user 20 bank 30 foreign exchange market 100 user device 200, 200 ′ computer system 210 server device 220 database unit 300 asset transaction computer system 400 network

Claims (17)

A computer system including a processor unit and a memory unit, wherein a program is stored in the memory unit, and when the program is executed,
Predicting when the value of a foreign currency falls to a certain low price within a predetermined period, said predicting,
Dividing the predetermined period into a plurality of periods;
Determining a period to which a period in which the value of the foreign currency becomes a low value of a specific rank among a plurality of periods after the division based on the past value fluctuation of the foreign currency ;
(1) further dividing the period in which it is determined that the time when the value of the foreign currency is a low price of a specific order belongs, into a plurality of periods;
(2) determining a period to which a period in which the value of the foreign currency is at a low value of a specific rank belongs among a plurality of periods after the division based on past value fluctuations of the foreign currency;
Repeating the period in which it is determined that the time when the value of the foreign currency is at a low price of a specific rank belongs to the same length as the time when the value of the foreign currency is at a low price of a specific rank , That,
Issuing a purchase request to purchase the foreign currency at the predicted time period, causing the processor unit to perform processing. It said determining, among the plurality of divided periods, including the average value of the foreign currency to determine low period, the computer system according to claim 1. The determination is as follows.
Using each of a plurality of prediction models, a period to which a period in which the value of the foreign currency becomes a low value of a specific rank among a plurality of periods after the division based on past value fluctuations of the foreign currency Judging,
Among the plurality of periods after the division, a period determined by a larger number of prediction models as a period to which the time when the value of the foreign currency becomes a low price of a specific rank belongs, and the value of the foreign currency is specified The computer system according to claim 1, further comprising: determining that the period to which the low rank of the order belongs belongs. The determination is as follows.
Using each of a plurality of prediction models, a period to which a period in which the value of the foreign currency becomes a low value of a specific rank among a plurality of periods after the division based on past value fluctuations of the foreign currency Judging,
For each of the first half of the plurality of periods after the division, determining the number of prediction models determined that the period belongs to a period in which the value of the foreign currency is at a low value of a specific rank. When,
3. The method according to claim 1, further comprising: determining a period in which the number of the prediction models is equal to or greater than a predetermined number as a period to which a period when the value of the foreign currency is a low price of a specific rank belongs. Computer system. The computer system according to claim 4 , wherein the determination further includes changing the predetermined number in accordance with a relative position of the divided period with respect to the entire predetermined period. Wherein the plurality of periods are two periods, according to any one of claims 1-5 computer system. The predicting includes suppressing the prediction of a time within a predetermined suppression period from a start point of the predetermined period as a time when the value of the foreign currency becomes a low price of a specific rank. 6 computer system according to any one of. Wherein the predicting, the thing is to be issued repeatedly carried out, the computer system according to any one of claims 1-7. The issuing is performed automatically without a user input operation for purchasing the foreign currency at a time when the predicted value of the foreign currency becomes a low price of a specific rank. 8 computer system according to any one of. The price of a particular rank is lowest price, computer system according to any one of claims 1-9. The price of a particular rank, the n-th a cheap value (n is an integer of 2 or more), the computer system according to any one of claims 1-9. A method executed in a computer system comprising a processor unit, the method comprising:
The processor unit predicts a time when the value of the foreign currency is a low price of a specific rank within a predetermined period, and the prediction includes:
Dividing the predetermined period into a plurality of periods;
Determining a period to which a period in which the value of the foreign currency becomes a low value of a specific rank among a plurality of periods after the division based on the past value fluctuation of the foreign currency ;
(1) further dividing the period in which it is determined that the time when the value of the foreign currency is a low price of a specific order belongs, into a plurality of periods;
(2) determining a period to which a period in which the value of the foreign currency is at a low value of a specific rank belongs among a plurality of periods after the division based on past value fluctuations of the foreign currency;
Repeating the period in which it is determined that the time when the value of the foreign currency is at a low price of a specific rank belongs to the same length as the time when the value of the foreign currency is at a low price of a specific rank , That,
Issuing said purchase request to purchase said foreign currency at said predicted time. A program executed in a computer system including a processor unit and a memory unit, wherein the program is stored in the memory unit, and when the program is executed in the computer system,
Predicting when the value of a foreign currency falls to a certain low price within a predetermined period, said predicting,
Dividing the predetermined period into a plurality of periods;
Determining a period to which a period in which the value of the foreign currency becomes a low value of a specific rank among a plurality of periods after the division based on the past value fluctuation of the foreign currency ;
(1) further dividing the period in which it is determined that the time when the value of the foreign currency is a low price of a specific order belongs, into a plurality of periods;
(2) determining a period to which a period in which the value of the foreign currency is at a low value of a specific rank belongs among a plurality of periods after the division based on past value fluctuations of the foreign currency;
Repeating the period in which it is determined that the time when the value of the foreign currency is at a low price of a specific rank belongs to the same length as the time when the value of the foreign currency is at a low price of a specific rank , That,
A program causing the processor unit to perform processing including issuing a purchase request to purchase the foreign currency at the predicted time. A computer system including a processor unit and a memory unit, wherein a program is stored in the memory unit, and when the program is executed,
Predicting the time when the value of the foreign currency becomes a value of a specific rank within a predetermined period based on the past value fluctuation of the foreign currency whose value fluctuates over time, the prediction Is
Dividing the predetermined period into a plurality of periods;
Determining a period to which a period when the value of the foreign currency becomes a value of a specific rank among a plurality of periods after the division based on the past value fluctuation of the foreign currency ;
(1) further dividing a period determined to belong to a period when the value of the foreign currency becomes a value of a specific rank;
(2) determining a period to which a period in which the value of the foreign currency becomes a value of a specific rank among a plurality of periods after the division based on a past value fluctuation of the foreign currency;
Repeating the period in which it is determined that the period when the value of the foreign currency has a value of a specific rank belongs to the same length as the period when the value of the foreign currency has a value of a specific rank , That
A computer system that causes the processor unit to perform processing including: The process is
The computer system according to claim 14 , further comprising issuing a transaction request to trade the foreign currency at the predicted time. A program executed in a computer system including a processor unit and a memory unit, wherein the program is stored in the memory unit, and when the program is executed,
Predicting the time when the value of the foreign currency becomes a value of a specific rank within a predetermined period based on the past value fluctuation of the foreign currency whose value fluctuates over time, the prediction Is
Dividing the predetermined period into a plurality of periods;
Determining a period to which a period when the value of the foreign currency becomes a value of a specific rank among a plurality of periods after the division based on the past value fluctuation of the foreign currency ;
(1) further dividing a period determined to belong to a period when the value of the foreign currency becomes a value of a specific rank;
(2) determining a period to which a period in which the value of the foreign currency becomes a value of a specific rank among a plurality of periods after the division based on a past value fluctuation of the foreign currency;
Repeating the period in which it is determined that the period when the value of the foreign currency has a value of a specific rank belongs to the same length as the period when the value of the foreign currency has a value of a specific rank A program that causes the processor unit to perform processing including the above. The process is
The program according to claim 16 , further comprising: issuing a transaction request for trading the foreign currency at the predicted time.