JP4329644B2 - Market analyzer - Google Patents

Description

The present invention relates to the market analysis apparatus.

  In exchange trading, the basic principle of the content of orders placed by each of a plurality of companies participating in the exchange is to keep anonymous. A completely ideal market where exchange trading is sufficiently active, that is, the number and liquidity of market participants is sufficient, and there is no market control of a particular participant If so, there is no way to guess the contents of orders other than their own.

  In reality, however, many exchanges are not fully active as described above, but are incomplete markets. Participants with market dominance are generally identified, and the contents of the order may be roughly assumed. For example, in a situation where the relationship between order quantity and cost for order resources is roughly understood, it is possible to infer the order price and order quantity by a specific market participant from actual market information such as execution price and quantity. There is also. In such an environment, the above estimation is an important factor in examining the company's bid strategy. Conventionally, however, the above-mentioned guess has been left to human judgment, and analytical procedures have not been performed.

  Another approach for analyzing the market structure has been to simulate a virtual trading market. For example, as disclosed in Japanese Patent Application Laid-Open No. 2002-73985, let us analyze a macro market trend by simulating a virtual trading market in a computer and also simulating each participant's action method in the computer. And so on. However, in the case of the analysis by such virtual simulation, it is difficult to identify each model used for the simulation, for example, a participant's action rule, using the market information information actual value as teacher data.

JP 2002-73985 A

  In the incomplete market as described above, there is a problem that it is difficult to arithmetically estimate the order contents of a specific other participant.

  Means are provided for receiving public market information including contract prices and contract quantities in a plurality of time sections, and means for expressing estimated order contents as variables for order contents by other companies participating in exchange transactions. Using these as input information, for the objective function with an index that the result of the exchange contract calculation simulation based on the estimated order is consistent with the public market information, the objective function is to be optimized through the plurality of time sections. Perform optimization calculations. As a result of the optimization calculation, an arithmetic operation is performed to identify the variable as a most probable value.

  By the means for solving the above, it is possible to roughly estimate the ordering method of other companies participating in the exchange transaction. By estimating the order status of other companies, it is possible to obtain information that allows a quantitative analysis of the company's bid strategy. By simulating the contract calculation performed by the exchange based on the estimated order contents of other companies, for example, you can analyze the effect on the contract price when changing your own ordering method, and correlate with the profits of your company and other companies In-house bid strategies such as sensitivity analysis can be considered. In addition, by comparing a plurality of analysis results with respect to the analysis result, it becomes possible to recognize a change in the market situation based on a quantitative analysis.

  According to the present invention, it is possible to mathematically estimate the order contents of a specific other participant using the market information information actual value as input information.

  The present invention relates to a trading support system for participants in an exchange transaction in which a plurality of trading participants intervene, and in particular, the identity of other participants is specified, and the relationship between the order quantity and cost for ordering resources In the situation where the market price is roughly grasped, there is a means for simulating the market using the estimated order contents processed from the cost characteristics, and the actual market information such as the contract price and quantification that the market simulation result is released from the exchange. On the other hand, the transaction support method includes means for estimating order contents of other participants by making the difference small.

As an embodiment, a description will be given of a case where the present invention is applied to a transaction support apparatus in a wholesale power transaction. The transaction support apparatus analyzes the transaction strategy of other participants participating in the wholesale power transaction and provides decision support information according to the program processing procedure shown in FIG. First, as a first procedure, exchange market information is received by procedure 0101 and interconnected line congestion information is received by procedure 0102. These are acquired from the databases 0103 and 0104 that store information collected from the wholesale power exchange and the interconnected line management organization through the communication function.

Here, in order to explain the contents of the support apparatus, the wholesale power transaction system targeted by the support apparatus will be outlined, and the contents of the exchange market information and the interconnected line congestion information will be described. FIG. 2 shows the geographical state of power trading. A plurality of trading participants 0203 exist in a wide-area power system 0201, and each trading participant belongs to a partial power system (hereinafter referred to as an area). The own company 0204 using this support apparatus is one of the trading participants. These trading participants have power generation equipment (power supply) and consumption equipment (electric power demand) connected to the area to which they belong, and in a market environment where financial transactions are conducted for the transmission and reception of power through these facilities. is there. However, regarding power transmission / reception between areas, there is a physical restriction 0205 on the capacity of interconnection lines (power transmission facilities) connecting the areas. Even if one of the trading participants presents a lower unit price than the other participants and sells a large amount of power, if the connection capacity is exhausted (hereinafter referred to as “transmission line congestion”), This means that you cannot sell electricity. In the power transaction targeted by this embodiment, spot power is bought and sold the next day via the Internet. For example, tomorrow 12: 00-12: 30, the power of the thirty minutes time period (spot) is traded between participants. As shown in FIG. 3, each of the trading participants has a terminal 0302 connected to the Internet 0301, and performs a trading order operation 0303 through the terminal 0302. Similarly, the business owner has the terminal 0304 and can perform the ordering operation. These orders are collected in the power exchange server 0305 via communication. The order form is shown in FIG. The electric power trading order is made by a so-called boarding method in which orders of a plurality of traders bid by a specified deadline are matched at a time. The contract method at this time is defined in advance, and a so-called single-price type contract rule is adopted in which the contract unit price is single. The order will be bid for the desired amount and unit price for spot power the next day. The selling order can be indicated by the relationship between the amount of power sold and the unit price, such as 0401. The general idea is that orders increase in the desired unit price as the amount of electricity sold increases. In the case of a power purchase order, an order in which the desired unit price decreases as the power purchase amount increases, as in 0402. From the general concept of demand / supply curves, the place where the selling order curve 0401 and the buying order curve 0402 intersect indicates the contracted unit price and its contracted amount (transaction volume, volume). However, in the power trading as described above, when the connection line is congested, the area connected to both ends of the connection line has restrictions on delivery of demand and supply. there is a possibility. In this case, for example, as shown in FIG. 5, when the interconnected line congestion 0501 occurs, the contracted unit price may differ between the left end area and other central areas. The leftmost area can have a contract unit price 0502 and a contract amount 0503, and a group (hereinafter referred to as area group) 0504 to which other areas belong can have a contract unit price 0505 and a contract amount 0506. The exchange market information described above in the description of the processing procedure in FIG. 1 includes the contract unit prices and total contract quantities (volumes) such as the contract unit prices 0502 and 0505 for each area or area group in FIG. 5 and the approximately constants 0503 and 0506. ) Result information (market conditions). Further, the interconnected line congestion information indicates information on the amount of interconnected line usage and the presence / absence of congestion as indicated by 0501 in FIG. These exchange market information and interconnected line congestion information are made public by the organization operating the wholesale power exchange and the interconnecting line schedule, and the databases 0103 and 0104 are created by collecting these information. is doing.

  In procedures 0101 and 0102, a desired portion can be read from the databases 0103 and 0104. Specifically, the screen transitions from the console operation screen 0601 displayed when this support program is started as shown in FIG. 6 to the operation environment of FIG. 7 by pressing a button 0602. On the screen 0701, a display 0702 is displayed for a period during which the present support apparatus has already taken in data, and a function for taking in the latest updated information from the database by pressing a button 0703 is arranged. Moreover, it has a function that can confirm the captured contents at the bottom of the screen. When a time section to be viewed is designated by an operation of specifying a date by Torg 0704 and specifying a time zone by Torg 0705, exchange market information and interconnected line congestion information are displayed on display unit 0706. Information about the amount of use and occurrence of congestion is displayed 0706 for each interconnection line, and a contract price and a fixed amount are displayed 0707 for each area. A screen display as shown in 0708 is performed for a portion where a plurality of areas are in an area group without congestion. In addition, it is assumed that the contract price and contract quantity will not be disclosed due to some regulations. In such a case, as shown in FIG. 0709, the area or area group indicates that the information is not disclosed. With the screen having such a configuration, the operator acquires and confirms exchange market information and interconnected line congestion information.

  As the next procedure in the program processing in the transaction support apparatus, an analysis target spot is selected as procedure 0105 in FIG. As a screen operation, when the button 0603 in FIG. 6 is pressed, the screen transitions to the operation environment in FIG. On the screen 0801, when analyzing the strategies of other participants from now on, a selection operation of a spot set to be analyzed (hereinafter, analysis target spot selection) is performed. The analysis target spot selection here is an operation for designating a plurality of opportunities for spot trading in the past performance that other participants are supposed to have the same strategy as analysis targets. Specifically, the selection of the period by specifying the start date and the end date by the input unit 0802, the AND / OR specification by either or both of the day type and the day of the week by the input unit 0803, one by the input unit 0804 Consists of designations related to time zones in units of 30 minutes in a day. With respect to designation by period and designation by day type and day of the week, the designated content can be confirmed by the calendar display unit 0805. Users of the device can analyze the strategies of other participants during that period by specifying the designations that other participants are likely to take a similar strategy. For example, as shown in FIG. 8, six time zones before and after noon (11: 00 to 14:00) on weekdays in winter from December to January, especially on non-holiday days in the middle of the week except Monday and Friday. Can be used to analyze the performance of spot power products. By changing the designation target, analysis focusing on other spots, such as evenings during peak power demand in summer, can be performed.

  As a next procedure in the program processing in the transaction support apparatus, a cost profile is set as a procedure 0106 in FIG. As a screen operation, when the button 0604 in FIG. 6 is pressed, the screen transitions to the operation environment in FIG. On the screen 0901, by means of Torg 0902 and 0903, designation of other participants to be set and whether the setting is related to the selling order or the setting of the buy order are specified. The meaning of the cost profile set here will be described with reference to FIG. The cost profile indicates the amount of electric power sold (or the amount of electric power purchased) and its marginal cost as in the graph 1001. For example, in the power sale state 1002 (or the power purchase state), the cost increase amount 1003 generated in the next unit sale of power is grasped. The marginal cost characteristic is generally monotonically increasing. Since the true value is roughly determined by the cost characteristics of the operating generator group, if information such as the operating generator, the fuel consumption characteristics of the generator, and the procurement cost of the fuel is available, the marginal cost characteristics Can be calculated. The order price to the exchange is considered to have some margin for these marginal costs. Here, as one of the pricing method models, a method in which a relationship 1005 between the order unit price and the order quantity (hereinafter referred to as an estimated price curve) is expressed by adding a margin 1004 that does not depend on the amount of electricity purchased to the marginal cost. is doing. With regard to such marginal costs, accurate information for each other participant is not actually available, but a rough guess can be made from various publicly available information and analogy of in-house facilities and fuel costs. As described above, the cost characteristics estimated by means separate from the present apparatus can be set as a cost profile by the input to the numerical value input field 0904 on the screen 0901 and the drawing confirmation operation of the graph area 0905. This setting data is stored in the other company's cost profile database 0107 as the other company's cost characteristic at the analysis target spot designated in the procedure 0105.

  As the next procedure in the program processing in the transaction support apparatus, as shown in the procedure 0107 of FIG. The screen operation is acquired from the in-house order execution results database 0108 by pressing a button 0605 in FIG. Unlike the handling of other companies as described above, the order contents and contract contents for the past are managed in the database 0108 as confirmed information for the company. Therefore, it is only necessary to refer to the in-house order execution results at the analysis target spot designated in the procedure 0105 from the database.

  Based on the data prepared as described above, the transaction support apparatus performs market analysis calculation as procedure 0109 in FIG. 1 to calculate an estimated price curve of another company. Specifically, the optimization calculation is performed for a plurality of cross sections in the market, the most probable value is calculated (parameter identification) for the parameter of the margin 1004 of the pricing model, and the result and the cost profile are calculated. Create an estimated price curve. The contents of the market analysis calculation method will be described below.

  In the market analysis calculation, the problem is solved by quadratic programming by treating it as an optimization problem by the following formulation. In the following, as a formulation of the optimization problem of this embodiment, a description will be given assuming a situation in which one participant is participating in one area.

  First, constants and variables handled in this calculation will be described. Performance data obtained from exchange market information and interconnected line congestion information is expressed as the following constants.

  Information obtained as exchange market information is expressed by Equation 1. As described above, since the contract unit price and the contract amount (sell / buy) are presented for each area group based on the single price auction method, they are handled as a constant data group having a value for each analysis target spot. Also, depending on the area group, whether or not the information is disclosed is handled by the execution result disclosure flag in order to cope with the case where the contract unit price and the contract amount are not disclosed due to various circumstances.

  The content of the interconnected line congestion information is expressed by Equation 2. The presence / absence of congestion for each interconnection line connecting the areas is handled as a constant data group having a value for each spot to be analyzed. When the areas on both sides of the interconnection line belong to different area groups, that is, when the market is divided by the interconnection line, congestion occurs in either direction of the interconnection line Therefore, flag 1 is set as the constant. For convenience in formulating the problem, a non-existing interconnection line, that is, a connection between unconnected areas is always treated as a zero-capacity interconnection line in which congestion occurs.

  Equation 3 defines the areas that are grouped as a result of market segmentation. It can be uniquely determined from the exchange market information described above. Which area group each area belongs to is handled as flag information, and is handled as a constant data group having a value for each spot to be analyzed.

  The above formulas 1 to 3 are constants indicating actual data obtained from exchange market information and interconnected line congestion information. On the other hand, with regard to in-house orders, the unit price and quantity of sell / buy orders are obtained from the database 0108, which is confirmation information of past order contents, and are defined as constants as shown in Equation 4. Here, it is assumed that there are one way to order in-house through the analysis target spot. If the order unit price and the order quantity are not completely uniform within the range of the analysis target spot, the constant of the above equation 4 is determined after performing the process of averaging them.

  On the other hand, the order contents by other companies are not values that can be obtained from actual results, but are estimation targets that are expressed using variables that are optimized by this calculation. Here, it is expressed by a combination of a constant and a variable as shown in Equation 5. First, as in the above-described procedure 0106 and the setting in FIG. 9, cost profiles of respective competitors are assumed, and the amount of power that can be bought and sold and its cost are defined thereby. Then, it is assumed that the cost profile is corrected to become an estimated price curve by a margin value that is handled as a variable in this calculation. In other words, each of the purchase order and the sell order is handled as an order unit price represented by the sum of the order quantity, the cost and the margin.

  Of the order contents described in Equations 4 and 5 above, a flag variable like Equation 6 is introduced in order to define an actually executed order. When each order is executed, the variable becomes state 1.

  Finally, the interconnected power flow is expressed by the formulation of Equation 7. It is treated as a variable indicating the tidal flow of each interconnection line and a constant indicating the capacity.

  The state of maximum likelihood can be treated as an optimization problem aiming at minimizing the objective function of Expression 8 by expressing the variables and constants of Expression 1 to Expression 7 above. Hereinafter, each of the constituent terms of Expression 8 will be described.

  First, Expression 9 represents that the actual value of the actual value and the total amount of the estimated contracted order should match in each area group. Formulated for each of selling and buying.

  Equation 10 indicates that the total amount of buying and selling and the sum of the inflow and outflow from the interconnection line entering and exiting the area should be balanced in each area.

  Further, Expression 11 indicates that the tidal flow rate should be equal to the upper limit value of the interconnecting line capacity at a place where the interconnecting line is congested.

  Expression 12 indicates that the exchange is executing a contract process so that the sales order quantity is as large as possible in the entire market.

  On the other hand, Expression 13 represents a constraint condition, and restricts that the maximum estimated contract price does not exceed the contract price of the exchange market information in each area.

  By implementing quadratic programming for the optimization problem based on the above formulation, the size of each margin value, whether each of the assumed orders is filled or unfilled, and each The most probable estimate can be calculated for the interconnected tidal flow. That is, the most probable estimated price curve can be obtained at the analysis target spot designated in the procedure 0105.

  FIG. 11 shows an operation environment used in the procedure 0110 of FIG. 1 for drawing the analysis result after the market analysis calculation by the procedure 0109 is performed as described above. A screen 1101 is a screen for browsing analysis results. The display frame 1102 in the upper part of the screen describes the outline of the analysis target spot handled by the market analysis result, and the outline of the analysis result is shown in the lower part of the screen. The display at the bottom of the screen conforms to the operation environment of FIG. 7 described above, but the displayed value is not an actual value but an estimated value by market analysis calculation. In FIG. 1103, the average unit price and average estimated value in each area are displayed, and in FIG. 1104, the margin estimated value in each area is displayed. Further, from the result of the market analysis calculation, focusing on the convergence error of the partial function expression 10 configured for each area of the objective function of Expression 8, this is displayed together with 1104 in the figure as the estimated accuracy of each area. is doing. In addition, in FIG. 1105, information on the average tide flow rate for each interconnection line, the usage rate (congestion occurrence rate) of the interconnection line capacity due to the tidal flow, and the presence or absence of congestion is shown. With the above illustration, an outline of the analysis result can be grasped.

  Further, in order to obtain detailed information on the market analysis result, the user clicks on the screen portion indicating each area in FIG. Then, a transition is made to an operating environment in which detailed information such as an estimated price curve can be viewed for each area, as shown in FIG. On the screen 1201, when an estimated price curve to be displayed is specified by specifying a participant by a pull-down 1202 and specifying a trading direction by a pull-down 1203, a cost profile curve 1205 and an estimated margin 1205 are estimated on the graph. A price curve 1206 is drawn, and numerical information is displayed in a table 1207 and a frame 1208. In addition, information such as the amount of electric power purchased and sold, cost, transaction balance, etc. is presented for the estimated execution range by means of a graph drawing 1209 and a numerical display 1210. In addition to these, the above information can be compared and browsed by designating a comparison target on the right side of the screen. By designating a comparison target participant using the pull-down 1211, a cost profile curve 1212, an estimated margin 1213, an estimated price curve 1214 are drawn on the graph, and numerical information is displayed in a table 1215 and a frame 1216. The graph drawing 1217 is also performed on the range, so that the operation environment can be easily executed by comparing and viewing the two participants.

  As described above, the outline and detailed information of one market analysis result can be browsed by the operation environment of FIGS. 11 and 12.

  In addition to this, the present embodiment also provides a GUI capable of comparing and examining a plurality of market analysis results. The latest analysis result can be stored in the database 0111 by pressing a button 1106 in FIG. The stored analysis result is called a status. By calling the status saved in the past, it is possible to compare with the latest analysis result. When the button 1107 in FIG. 11 is pressed, the operation environment for comparing and viewing the analysis result in FIG. 13 is transitioned to.

  On the screen 1301, the latest analysis result is compared with one status. In the upper part 1302 of the screen, the outline of the analysis target spot of the latest analysis result is displayed as in the case of the screen 9. At the same time, the data name of the status to be compared is displayed on the display 1303. As data to be compared, data used by the previous operation of FIG. 13 is automatically selected unless otherwise operated. When the user wants to change, the user designates one of the statuses stored in the database 0111 after pressing the button 1304. The contents comparing the latest analysis result and the comparison target status are displayed at the bottom of the screen. As shown in FIG. 1305, the estimated content and the accuracy of each area can be browsed in a form divided into upper and lower stages. When an area where a deviation of a predetermined range or more occurs in the comparison contents, that is, when an area with a large difference in estimation margin or estimation accuracy occurs, a screen display that can be visually recognized by the operator as shown in 1305 is performed. . This makes it possible to recognize that a change has occurred in the newly executed analysis based on the comparison status.

  With the support device of this embodiment, the operator can roughly estimate the ordering method of other companies participating in the exchange transaction. For example, as in the above-described embodiment, the most probable numerical values regarding the order quantity / price of each participant can be obtained from the exchange market information regarding the wholesale power and the trade information of the interconnected line congestion information.

  Thus, by being able to estimate the order status of other companies, it is possible to obtain information that allows a quantitative analysis of the company's bid strategy. By simulating the contract calculation performed by the exchange based on the estimated order contents of other companies, for example, you can analyze the effect on the contract price when changing your own ordering method, and correlate with the profits of your company and other companies In-house bid strategies such as sensitivity analysis can be considered.

  Moreover, since it is possible to compare a plurality of analysis results (status) with respect to the analysis result, it is possible to recognize a change in the situation in the market based on a quantitative analysis. For example, in a wholesale power transaction, assuming a continuous work environment in which analysis is performed at the end of every month on weekday daytime spots for a month, a large change in state occurs in the estimation results or the estimation system. This indicates that any participant's ordering method may have changed from the previous month's thinking. It is possible to monitor the environment of the entire market by periodically performing the task of analyzing market conditions with this support device.

Processing procedure of this embodiment. Participation system for wholesale power transactions targeted by this example. The system configuration of wholesale power transactions targeted by this embodiment. The order contents in the wholesale power transaction targeted by this embodiment. The contents of an agreement in a wholesale power transaction targeted by this embodiment (example). Console screen of support device. Operation screen for importing exchange market information and interconnected line congestion information of support equipment. Operation screen for selecting a spot to be analyzed by the support device. Operation screen for setting the cost profile of the support device. Relationship between cost profile and estimated price curve. Operation screen for viewing analysis result of support device (part 1). Operation screen for viewing analysis results of support device (part 2). Operation screen for browsing the analysis result of the support device.

Explanation of symbols

0101 ... Reception of transaction information, 0102 ... Reception of interconnected line congestion information, 0103,0104 ... Database, 0201 ... Power system, 0203 ... Transaction participant, 0204 ... Own company, 0205 ... Physical constraints, 0301 ... Internet, 0302, 0304 ... terminal, 0303 ... trading order operation, 0305 ... power exchange server, 0401 ... sell order, 0402 ... buy order,
0501 ... Interconnection line congestion, 0502, 0505 ... Contract price, 0503, 0506 ... About fixed amount, 0504 ... Group, 0601 ... Console operation screen, 0602, 0703 ... button,
0701, 0801, 0901 ... screen, 0702, 0707 ... display, 0704, 0705, 0902, 0903 ... Torug, 0706 ... display section, 0708 ... screen display, 0802
0803, 0804: Input unit, 0805: Calendar display unit, 1001 ... Graph, 1002 ... Power sale state, 1003 ... Cost increase, 1004 ... Margin, 1005 ... Relationship between order unit price and order quantity.

Claims (1)

In a market analysis device that provides support information to businesses participating in exchange transactions,
Receive public market information including contract prices and contract quantities in multiple time sections from the power exchange server via the Internet,
Order amount for buy and sell orders by other companies participating in exchange trading, cost for order price for buy and sell orders, and margin for order price for buy and sell orders We handle estimated order details as the order price expressed as the sum of
Considering the contract status flag for buy orders and sell orders, and the tether flow rate and tether capacity for a given direction between areas,
Formula that indicates that the total amount of the contracted quantity and the estimated contracted order should match for selling and buying. When there is congestion in the interconnection line, the tidal flow should match the upper limit value of the interconnection line capacity. And an exchange based on the contents of an estimated execution order consisting of the sum of functions based on an expression that indicates that the exchange will execute as much processing as possible for the entire market as much as possible. The objective function with the index that the result of the contract processing calculation is consistent with the public market information is an index that satisfies the constraint that the maximum estimated contract price does not exceed the contract price of the exchange market information. Perform next programming, minimize by optimization calculations, perform market analysis calculations,
A market analysis apparatus comprising a terminal for displaying a result of market analysis calculation.

Images