JP4463027B2 - Payment / payment condition setting system and method in risk exchange related to weather - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To conduct a risk exchange between entrepreneurs by using weather allowed to observe a third point different from the locations of the entrepreneurs trying to conduct the risk exchange relative to the weather. <P>SOLUTION: This system comprises: a weather data acquisition section 22 acquiring weather data indicative of the past weather histories on two points A, B in which the entrepreneurs exist, and the third point C, respectively ; a regression analysis section 26 conducting a regression analysis for the weather data on the third point C based on the weather data on the points A, B, respectively; and a regression error distribution calculation section 30 calculating the distribution of a regression error based on a regression analysis result; and a dead zone decision section 34 deciding a dead zone relative to the weather in the case of deciding income and expenditure amounts in accordance with the weather based on the calculated regression error distribution; and also a third point selection section selecting the third point C based on the correlation between the weather data on the points A, B and the weather data on each of the points. <P>COPYRIGHT: (C)2006,JPO&amp;NCIPI

Description

  In the present invention, two operators existing at different points set the payment / payment amount conditions when exchanging risks related to the weather using measurable weather at a third point different from these points as an index. It relates to the system and method used.

  For electric power companies and gas companies, changes in temperature are a variable factor for earnings. For example, in an electric power company, the higher the summer temperature, the higher the demand for power due to cooling, and the higher the profit. Conversely, the lower the summer temperature, the worse the profit. On the other hand, in the gas company, the higher the summer temperature, the lower the demand for hot water supply gas and the higher the profit. On the other hand, the lower the summer temperature, the worse the profit. As described above, the electric power company and the gas company generate a risk of fluctuation in profitability in the opposite direction depending on the temperature.

  Therefore, in order to hedge the risk of fluctuations in earnings due to temperature as described above (referred to as “temperature risk” in this specification), there is a case where temperature risk is exchanged between the electric power company and the gas company. That is, when the temperature is higher than usual, the hedge amount is paid from the power company whose profit is improved to the gas company whose profit is worsened. On the contrary, when the temperature is lower than normal, the gas whose profit is improved is paid. The company pays the hedge amount to a power company whose profits deteriorate.

In relation to this exchange of temperature risk, for example, Patent Literature 1 discloses a system for enabling exchange of temperature risk between businesses existing in different regions. The temperature risk exchange method disclosed in this document is, for example, when a Tokyo electric power company and an Osaka gas company exchange temperature risks, and when both the Tokyo temperature and the Osaka temperature are higher than usual, Tokyo Pay the hedge amount from the power company in Osaka to the gas company in Osaka and pay the hedge amount from the gas company in Osaka to the power company in Tokyo when the temperature in Tokyo and Osaka are both lower than usual, otherwise The hedge amount is not paid or received.
JP 2004-78903 A

  As described above, in the temperature risk exchange method disclosed in Patent Document 1, since the temperatures of the two regions (Tokyo and Osaka in the above example) where each business operator is located are used, the payment amount is determined from the temperature of each point. The scheme for doing so becomes complicated. Therefore, a simple method capable of determining the payment amount using only the temperature at one point as an index is desired. In this connection, the applicant pays attention to the fact that there is a correlation in the weather between the locations of the two parties that are trying to exchange risks, and the payment conditions are based on the weather at one of these locations as an indicator. Is proposed (Japanese Patent Application No. 2004-175887). Furthermore, the present applicant has conceived that a third point different from each location of the parties performing the temperature risk exchange is used as an index, and thereby confirms that a more effective risk exchange can be performed.

  The present invention has been made under the circumstances as described above, and uses a measurable weather at a third point different from the location of the business operator who wants to exchange risks related to the weather. The purpose is to enable risk exchange.

In order to achieve the above-described object, the present invention is based on the assumption that two operators existing at different points A and B can use meteorological weather at a third point C different from those points A and B as an index. A system used to set the conditions for the amount of money to be paid when exchanging the risks involved,
The point A, B and for each point k that are candidates for the third point C, a weather acquisition unit for acquiring table to meteorological data history of past weather,
Correlation coefficients ρ _{A, k} , ρ _{B, k} between the respective weather data at the points A 1, B and the weather data at the respective points _k are calculated, and a composite correlation coefficient ρ _k, which is the product of them. = Ρ _{A, k} · ρ _{B, k} is the third point selecting unit for selecting the point _k where the maximum is k as the third point C;
The point A, the respective weather data B, a regression analysis section for performing a regression analysis of the meteorological data of the third point C,
Based on the regression analysis result by the regression analysis unit, the regression error for calculating the distribution of the respective regression errors ε _A and ε _B between the meteorological data at the points A and B and the meteorological data at the third point C 1. A distribution calculator;
The respective standard deviations σ _εA and σ _εB of the distributions of the calculated regression errors ε _A and ε _B are calculated, and the distribution function is assumed to be a predetermined probability distribution. ± x · σ _εA and ± x · , using a value x such that a value obtained by integrating ± with x is (1−predetermined allowable probability) and the calculated standard deviations σ _εA and σ _εB. a dead zone determining unit that determines a wider range of σ _εB as a dead zone when determining the amount of payment according to the weather;
It is characterized by providing.

  According to the present invention, the dead zone when determining the payment amount according to the weather based on the distribution of the regression error when the regression analysis between the weather at the points A and B and the weather at the third point C is performed. To decide. Here, the dead zone is a range in which the amount received and paid is zero regardless of the deviation from the annual value of the weather. Therefore, even if the weather fluctuates within the range corresponding to the regression error distribution, no money is paid or received. For this reason, payment is reversed between the case where the risk exchange related to the weather is performed using the weather at either of the two points A and B as an index, and the case where the risk exchange related to the weather is performed using the weather at the third point C as an index. Can be prevented from occurring. For example, temperature and precipitation data can be used as the weather data in the present invention.

Further, by selecting the third point C based on the correlation between the meteorological data for the points A and B and the meteorological data at each point, the third is highly correlated with the weather between the points A and B. A point can be selected. The higher the weather correlation, the smaller the regression error in the regression analysis described above (and the smaller the distribution of the regression error distribution), so narrow the dead zone by using the weather data at the third point selected as described above. This enables effective risk exchange.

  According to the present invention, it is possible to exchange risks between these operators using measurable weather at a third point different from the location of the operator who wants to exchange risks related to the weather.

  Hereinafter, a system according to an embodiment of the present invention will be described. The system of the present embodiment is for supporting the setting of the condition for receiving and paying the hedge amount when two companies existing in different regions exchange risks related to the weather (temperature risk in the present embodiment).

First, the principle of temperature risk exchange according to the present embodiment will be described.
As described in the above-mentioned section of the prior art, for example, by paying the hedge amount according to the level of temperature between companies that suffer from the risk of fluctuations in profitability due to high and low temperatures, such as power companies and gas companies , Temperature risks are being exchanged. Conventionally, when the temperature risk is exchanged only between businesses existing in the same region, or when businesses in different regions exchange temperature risk as in Patent Document 1, these two regions are exchanged. Both temperatures are being used. This is because the temperature fluctuation trend varies from region to region, so when the payment conditions for operators in different regions are determined using the temperature in one region, the following unreasonable results occur: Because it will do. Hereinafter, an example in which a business operator α (for example, a gas company) in Tokyo and a business company β (for example, an electric power company) in Fukuyama exchange temperature risks will be described in detail with reference to FIGS.

  FIG. 1 is a diagram showing annual changes in average temperatures in Tokyo and Fukuyama in the summer (July to September). As shown in the figure, although the temperature fluctuations at both locations show similar trends, it can be seen that the temperature levels are reversed in some years.

  FIG. 2 is a diagram illustrating an example of a payment / payment condition for setting a payment / payment amount based on a deviation of the temperature from the annual value. As can be seen from FIG. 1, since the temperature at each point has a gradually increasing trend, according to this trend, for example, an annual reference temperature μ is obtained by a least square line, and this reference temperature μ is used as an annual value. To do. In addition, since the standard deviation of the temperature probability distribution varies from region to region, as shown in Fig. 2, the horizontal axis represents the value obtained by standardizing the deviation from the annual value with the standard deviation of the temperature probability distribution of the deviation from the annual value. The amount is to be determined.

  In this example, the operator α in Tokyo is, for example, a gas company, and the profit deteriorates as the summer temperature increases. The operator β in Fukuyama, for example, is an electric power company, and the income increases as the summer temperature increases. It is supposed to be a company that improves. In the example of FIG. 2, the payment amount when the payment from the business operator α to the business operator β occurs is represented by a positive value, and the payment amount when the payment from the business operator β to the business operator α occurs is It is expressed as negative. Therefore, under the payment / payment conditions shown in the figure, when the temperature is higher than usual (the payment / payment amount is negative), payment from the operator β to the operator α is performed.

  FIG. 3 shows the amount of payment received when the payment amount is determined using the temperature in Tokyo according to the payment conditions shown in FIG. 2, and when the payment amount is determined using the temperature of Fukuyama in (b) Tokyo. Show. As mentioned above, there is a certain degree of correlation between the temperature in Tokyo and the temperature in Fukuyama, so almost the same payments occur regardless of the calculation, but the year surrounded by an ellipse in the figure In the case of (a) and the case of (b), the sign of the payment amount is reversed. For example, the payment amount determined using the temperature in Tokyo in 1983 is positive (payment from operator α to operator β), while the payment amount determined using the temperature in Fukuyama is negative (from operator β). Payment to operator α). In this case, if the payment amount is set using the temperature in Tokyo, the Fukuyama operator β should pay the hedge amount in order to compensate for the deterioration in profits due to the low temperature in Fukuyama, which is originally low. It will be necessary. In this way, if the amount of payment is reversed depending on the temperature at which point is used, the place where you should receive payment (or pay to the other party) using the temperature at your location is conversely paid to the other party ( In other words, an unreasonable situation (hereinafter referred to as “reverse payment”) in which the payment is received from the other party is invited.

  The reverse payment is caused by the fact that the temperatures in Tokyo and Fukuyama are not completely correlated, and a regression error occurs when the temperatures at both locations are regressed.

  On the other hand, in this embodiment, the third point C having a correlation higher than the temperature correlation between Tokyo and Fukuyama is selected between the temperatures of Tokyo and Fukuyama, and the temperature of the third point C is determined. By using it, operators in Tokyo and Fukuyama will be able to exchange risks that are unlikely to cause reverse payment. However, since it is impossible to select the third point C that has a complete correlation with the temperatures in Tokyo and Fukuyama, the reverse payment occurs even if the temperature at the third point C selected as described above is used. there is a possibility. Therefore, a dead zone related to the temperature is provided in the money payment conditions as illustrated in FIG. 2 (that is, a range in which the payment amount is zero even if the temperature changes), and the deviation from the annual value of the temperature is within the range of this dead zone. In such a case, the reverse payment is prevented by preventing the payment from being generated.

FIG. 4 is a graph showing an example of a payment / payment condition in which a dead zone is set. Similarly to FIG. 2, the horizontal axis represents the deviation from the annual value of the temperature at the third point C (that is, the reference temperature μ _C ). It represents a value standardized on the basis of the standard deviation σ _C of the deviation from the annual value of the temperature at three points C. By using the temperature normalized by the standard deviation σ _C in this way, the temperature is appropriate without being affected by the value of the standard deviation σ _C of the temperature at the third point C (that is, the degree of temperature fluctuation at the third point C). Payment amount can be determined. In the payment / payment condition of FIG. 4, the basic payment / payment condition is as shown in FIG. 2, and the payment / payment amount within the dead zone is zero in the payment / payment condition of FIG. However, as a payment / payment condition including the dead band, for example, as shown in FIG. 5, a payment / payment condition in which the basic payment / payment condition is shifted to both sides of the dead band can be used.

Hereinafter, the processing procedure of the air temperature risk exchange according to the present embodiment will be specifically described.
6 and 7 are flowcharts showing the processing procedure of the temperature risk exchange according to the present embodiment.

  First, past temperature data at each location including the locations A and B (in this example, A is Tokyo and B is Fukuyama) of the businesses α and β who are going to exchange the temperature risk and the candidate site of the third location C ( For example, the average value of average temperatures from July to September of the past year is acquired (S100 in FIG. 6).

Next, the change trend of the temperature data of each region is obtained by, for example, linear approximation by the least square method, and the reference temperature μ _k at each point _k is calculated using this linear approximation formula (for example, the following formula (1)) ( S102).
_{μ k = -P k + Q k} × ( year) ... (1)

Next, the annual actual temperature T _k at each point k is normalized using the reference temperature μ _k . That is, for each point k, for example, assuming that the probability distribution of (T _k −μ _k ) is a normal distribution, the normalized temperature tk is _obtained by the following equation (2) (S104).
t _k = (T _k −μ _k ) / σ _k (2)
However, σ _k is the standard deviation of the probability distribution of _{(T k -μ} k).

Then, the point A, the normalized temperature _t A of B, a _{t B,} the correlation coefficient [rho _{A, k,} [rho _{B, k} between normalized temperature _{t k} of each point k calculated (S106), Their combined correlation coefficient ρ _k is calculated by the following equation (3) (S108). In general, the correlation coefficient is calculated by ρ _{x, y} = Cov (x, y) / (σ _x · σ _y ) and is affected by the standard deviations σ _x and σ _y of the temperature distribution. So, by using the normalized temperature t _k normalized by the standard deviation sigma _k, the standard deviation of the normalized temperature t _k as 1, and remove the effect of the standard deviation.
ρ _k = (ρ _{A, k} × ρ _{B, k} ) ^1/2 (3)
Then, the point k at which the value of the combined correlation coefficient ρ _k calculated by Expression (3) is the largest is selected as the third point C (S110).

  FIG. 8 shows an example of the correlation coefficient of the temperatures between the point A (Tokyo) and the point B (Fukuyama) and each point, and the composite correlation coefficient calculated from these correlation coefficients. In the example of the figure, Toyama is selected as the third point C because the combined correlation coefficient of Toyama is 0.92, which is the largest value.

Next, regression coefficients r _A and r _{B are} performed by performing regression analysis of the standardized temperature t _C at the selected third point C, the standardized temperature t A at the point _A, and the standardized temperature t _B at the point _{B, respectively.} Is obtained (S112).
Next, regression errors ε _A and ε _B in the regression analysis are calculated by the following equations (4) and (5) (S114).
ε _A = t _C −r _A · t _A (4)
ε _B = t _C −r _B · t _B (5)
That is, the regression errors ε _A and ε _B are errors between the temperatures of the points A and B estimated by the regression analysis using the temperature of the third point C and the actual temperatures of the points A and B. This occurs because the temperature at point C and the temperatures at points A and B are not completely correlated.

Next, a dead zone in the payment / payment condition is set based on the probability distribution of the regression errors ε _A and ε _B (S116 in FIG. 7). This dead zone is set by assuming that the distribution of the regression errors ε _A and ε _B is, for example, a t distribution, and the reverse _payment described above is performed based on the standard deviations σ _εA and σ _εB of the regression errors ε _A and ε _B. The occurrence probability is set to be equal to or less than the specified allowable probability P. For example, when the allowable probability P is 5% (that is, a level once every 20 years), the occurrence probability 5% value in the t distribution (the value obtained by integrating the distribution function of the t distribution in the range of ± x becomes 0.95. since the value of becomes like x) is 1.7, the dead zone of _± 1.7σ εA, set the _± 1.7σ εB. Incidentally, as described with reference to FIG 4, receipts and payments amount, because it is determined on the basis of the temperature normalized by the standard deviation sigma _C of temperature of the third point C, as a unit standard deviation sigma _C In other _words , the dead zone is ± 1.7 · (σ _εA / σ _C ) · σ _C , ± 1.7 · (σ _εB / σ _C ) · σ _C.

  As described above, two dead zones are set corresponding to the points A and B. However, in order for the operators α and β at the points A and B to exchange risks, it is necessary to share the dead zone at both points. There is. Therefore, in the present embodiment, the wider one of the two dead zones is used as the dead zone for the payment / payment condition.

For example, when calculation was performed using past temperature data of Tokyo, Fukuyama, and Toyama, the standard deviation σ _εA of the regression error between Toyama and Tokyo was 0.31, and the regression error between Toyama and Fukuyama The standard deviation σ _εB was 0.44, and the standard deviation σ _C of Toyama's temperature was 0.953. In this case, the dead zone set based on the regression error between Tokyo and Toyama is ± 1.7 × (0.31 / 0.953) σ _C = ± 0.55σ _C , based on the regression error between Fukuyama and Toyama. The set dead zone is ± 1.7 × (0.44 / 0.953) σ _C = ± 0.78σ _{C. Since the} latter is wider than the former, ± 0.78σ _C is determined as the dead zone. If this is, the use of different dead zone at the β Tokyo operators α and Fukuyama of business, Toyama of the normalized temperature t _C is the inside of the outer and ± 0.78σ _C of ± 0.55σ _C, This is because it is necessary to prevent such a discrepancy from occurring because there is a discrepancy in payment / payment in which the payment / delivery occurs in the former dead zone but no payment / payment occurs in the latter dead zone. If the dead zone with a narrower range (in this example, the dead zone based on the regression error between Tokyo and Toyama ± 0.55σ _C ) is used, the Fukuyama operator who has a large regression error with Toyama Since there is a high possibility that the reverse payment with the case of using the air temperature will occur, it is preferable to use the wider dead zone as described above.

  Using the dead zone set as described above, for example, as shown in FIG. 4 or FIG. 5, a payment / payment condition including the dead zone is determined (S118). Then, the information indicating the determined payment / payment condition and the selected third point C is transmitted to the operators α and β (S120), and if the consent of both is obtained (S122), the payment / payment condition is fixed. On the other hand, when consent is not obtained, it returns to S100 and repeats a process on another condition.

  In the temperature risk exchange process described above, the system 10 according to the present embodiment is from the input of the temperature data at S100 to the setting of the dead zone at S116, the setting of the payment / payment conditions at S118, or the transmission / reception conditions at S118 and S120. The processing is performed.

  FIG. 9 is a hardware configuration diagram of the system 10 according to the present embodiment. As shown in the figure, a system 10 includes a CPU 100, a memory 102, a storage device 104 such as a hard disk device, a drive device 106 such as a CD-ROM or DVD-ROM, a display device 108, an input device 110 such as a keyboard or a mouse, The computer system includes a communication interface device 112 and the like.

  FIG. 10 is a functional block diagram of the system 10 of the present embodiment. As shown in the figure, the system 10 includes a point acquisition unit 20, a weather data acquisition unit 22, a data normalization unit 24, a third point selection unit 26, a regression analysis unit 28, a regression error distribution calculation unit 30, and an allowable probability input unit. 32, a dead zone determining unit 34, a dead zone output unit 36, and a third point output unit 38. These functional units 20 to 38 are realized by the CPU 100 reading a program installed in the storage device 104 into the memory 102 and executing it.

  Further, the system 10 is connected to a temperature database server 12 that stores past temperature data of each region in the country as, for example, a value of average temperature every month. The temperature database server 12 provides the system 10 with past temperature data of the area requested by the system 10. Further, as shown in the figure, the system 10 may be connected to user computers 14 and 16 of business operators α and β who are going to exchange temperature risks via a network. The user computers 14 and 16 may be computer systems such as personal computers and workstations, or may be mobile terminals such as mobile phones and PDAs. In short, any information processing apparatus that can communicate with the system 10 may be used. Communication between the system 10 and the temperature database server 12 and the user computers 14 and 16 is controlled by the communication interface device 112.

  The point acquisition unit 20 acquires point designation information indicating the locations A and B of the two companies that are to exchange the temperature risk. This information may be acquired by receiving an input from the input device 110 of the system 10, or may be performed by receiving information input by the user computer 14 or 16 via a network.

The weather data acquisition unit 22 is online by requesting, for example, the temperature database server 12 for temperature data representing the past actual temperature T _k of each point k including the acquired points A and B and the candidate point of the third point C. Get in. However, the present invention is not limited to this. For example, the data may be read from a storage medium such as a CD-ROM or DVD-ROM in which the temperature data of each place is stored.

The data normalization unit 24 calculates the reference temperature μ _k for each year by performing, for example, a linear approximation calculation by the least square method for the temperature T _k of each point k acquired by the weather data acquisition unit 22 and (T _k The standard deviation σ _k of the distribution of −μ _k ) is calculated. Then, using these calculated values, the (2) the temperature was normalized according to equation to calculate the normalized temperature t _k. The calculation of the reference temperature mu _k is not limited to the linear approximation, for example, it may be performed by a secondary or curve approximation of a cubic or the like. In short, the reference temperature μ _k may be calculated by performing an appropriate approximation according to the trend of the temperature.

The third point selecting section 26, point A, calculated and the normalized temperature _t A, _{t B} of B, and correlation coefficient [rho _{A, k,} [rho _{B, k} between normalized temperature _{t k} of each point k Then, the composite correlation coefficient ρ _k is calculated according to the above equation (3). Then, the point k at which the calculated value of the combined correlation coefficient ρ _k is the largest is selected as the third point C. The selected third point C is appropriately displayed and output by the third point output unit 38 or transmitted to the user computers 14 and 16 of the businesses α and β as indicated by broken lines in FIG.
The regression analysis unit 28 performs a regression analysis between the standardized temperature t _C at the selected third point C and the standardized temperatures t _A and t _B at the points A and B, and the regression coefficients r _A and r _B are obtained. calculate.
The regression error distribution calculation unit 30 calculates the regression errors ε _A and ε _B according to the above equations (4) and (5) using the calculated regression coefficients r _A and r _B to obtain respective probability distributions. The standard deviations σ _εA and σ _εB of the regression errors ε _A and ε _B thus obtained are calculated.

  The allowable probability input unit 32 receives an input of an allowable value (allowable probability) P of the probability of reverse payment depending on which temperature is used. For example, when the occurrence of reverse payment is set to a level once every 20 years, 5% is input as the allowable probability P.

The dead zone determination unit 34 determines the dead zone based on the input allowable probability P and the standard deviations σ _εA and σ _{εB of} the regression errors ε _A and ε _B. That is, as described above, for example, if the allowable probability P is 5%, the 1.7% which is the 95% value of the t distribution is used, and ± 1.7 · (σ _εA / σ _C ) · σ _C , Of ± 1.7 · (σ _εB / σ _C ) · σ _C , the wider one is _defined as the dead zone. In this embodiment, the probability distribution of the regression errors ε _A and ε _B is assumed to be a t distribution. However, the present invention is not limited to this, and the dead zone may be determined using a normal distribution or other appropriate distribution.

  The dead zone output unit 36 displays and outputs the determined dead zone on the display device 108 or transmits it to the user computers 14 and 16 of the business operators α and β as indicated by broken line arrows in the figure. Businesses α and β refer to this dead zone to determine payment / payment conditions as shown in FIG. 4 or 5, for example.

  FIG. 11 shows, for example, when a Fukuyama operator exchanges temperature risks with the same Fukuyama operator using the temperature of Fukuyama (Case (1)), When the temperature risk is exchanged using the temperature in the case (case (2)), and when the temperature risk exchange is conducted with the temperature in Toyama, which is the third point, with the business operator in Tokyo (that is, In the case of the present embodiment: the result of evaluating the annual average amount of payment for case (3)) is shown. In the case (2), based on the probability distribution of the regression error between the temperature in Tokyo and the temperature in Fukuyama, the same allowable probability P as in the case (3) is used in the same manner as S116 in the above embodiment. A dead zone was set.

  As shown in FIG. 11, according to the present embodiment (Case (3)), an annual average payment amount of 70 million yen is realized, and 0 in the case of using the temperature in Tokyo (Case (2)). More than 0.6 billion yen, when payment is exchanged between Fukuyama companies (case (1)), the payment amount close to 80 million yen has been realized, more effective than case (2) It can be seen that the temperature risk exchange can be performed. This is because the correlation between the temperature between Fukuyama and Toyama is higher than the correlation between the temperatures between Fukuyama and Tokyo, and the dead zone can be narrowed by using the temperature in Toyama.

  As described above, according to the present embodiment, even when an operator existing at different points exchanges temperature risk, while using a simple scheme based on the temperature of a third point different from those points, By providing a dead band in the payment condition, it is possible to prevent reverse payment.

  In addition, by selecting a point having a large temperature correlation with the location points A and B of the business operator performing the temperature risk exchange as the third point C, a minimum dead zone is set to prevent reverse payment. Therefore, an effective temperature risk exchange can be performed without setting an excessively large dead zone. That is, by selecting the third point C having a large correlation between the temperatures of the points A and B, as described with reference to FIG. 11 above, the temperature at one of the points A and B is used. Compared to, the dead zone can be narrowed to increase the amount of payment. Therefore, even if the temperature correlation between the points A and B is not so large, an effective temperature risk exchange can be performed between the operators at both points.

  In the above embodiment, the system 10 executes the process until the determined dead zone is output. However, the present invention is not limited to this. For example, the gradient of the payment amount with respect to the temperature in the payment condition is registered in the system 10 in advance. The payment conditions as shown in FIG. 4 and FIG. 5 are set based on the determined dead zone and this gradient, and this is output on the screen, or the user computers 14 and 16 of the operators α and β are used. You may make it transmit to online.

  In addition, if the acceptance / rejection of the set payment conditions is received online from the user computers 14 and 16 of the business operators α and β, and if the conditions are not accepted, various conditions such as the above-mentioned gradient of payment amount and allowable establishment P are changed online. By doing so, a contract for temperature risk exchange may be established on the system 10.

As described above, the dead zone in the _{payment / payment} condition is determined by the larger value of the standard deviations σ _εA and σ _εB of the regression errors ε _A and ε _B between the points A and B and the third point C. . Since the standard deviations σ _εA and σ _εB of the regression errors ε _A and ε _B become smaller as the temperature correlation with the third point C increases, the third point C out of the points A and B is eventually obtained. The dead zone is determined according to the temperature at the point where the correlation of the temperature between and is small. In the example of the above embodiment, the correlation coefficient between Toyama and Fukuyama, which is the third point C, is 0.90, and the correlation coefficient between Toyama and Tokyo is 0.94. The dead zone will be determined accordingly. Therefore, if the correlation coefficient with the temperature of Toyama is larger than 0.90, risk exchange can be performed in the same dead zone (that is, the same payment and payment conditions). In other words, the Fukuyama business operator can exchange the temperature risk with the business operator at the point D using the payment / payment conditions set for exchanging the temperature risk with the business operator in Tokyo as it is.

  As described above, when the operators at the points A and B perform the temperature risk exchange, if the third point C is selected and the dead zone is determined by the method of the present embodiment, the temperatures of the points A and B and the third point C are determined. As the point X is the smaller of the temperature correlations between the two points, the operator at this point X determines both the point D and the operator whose temperature correlation with the third point C is greater than the point X as described above. It is possible to exchange temperature risks using the dead zone as it is. As described above, the method of the present embodiment also has an effect that one business operator can easily find another partner who can exchange the temperature risk under the same payment and payment conditions.

  By the way, in the above embodiment, the case where the exchange of the temperature risk is hedged to hedge the profit risk due to the temperature fluctuation, but the profit of the operator may be influenced by other weather conditions such as precipitation in addition to the temperature. The present invention can also be applied to risk exchange based on precipitation. That is, for the two operators that are adversely affected by profits due to the magnitude of precipitation, the second calculation is based on the combined correlation coefficient of precipitation data between points A and B and each point k in the same manner as in the above embodiment. By selecting three points, performing regression analysis of precipitation data between points A and B and third point C, and setting a dead zone based on the regression error, precipitation data at third point C is obtained. It is possible to exchange risks appropriately while using them.

It is a figure which shows annual transition of the average temperature of Tokyo and Fukuyama in the summer (July to September). It is a figure showing an example of the payment conditions for setting a payment amount based on the deviation from the annual value of the temperature. It is a figure which shows each payment amount when the payment amount is determined using (a) the temperature of Tokyo according to the payment condition shown in FIG. 2, and (b) the payment amount is determined using the temperature of Fukuyama. It is a graph showing an example of the payment conditions which set the dead zone. It is a graph showing another example of the payment conditions which set the dead zone. It is a flowchart (the 1) which shows the process sequence of the air temperature risk exchange concerning this embodiment. It is a flowchart (the 2) which shows the process sequence of the air temperature risk exchange concerning this embodiment. It is a figure which shows the example of the synthetic | combination correlation coefficient calculated from the correlation coefficient of the temperature between point A (Tokyo) and the point B (Fukuyama), and each point, and those correlation coefficients. It is a hardware block diagram of the system of this embodiment. It is a functional block diagram of the system of this embodiment. When a Fukuyama operator exchanges temperature risks with the same Fukuyama operator using the temperature of Fukuyama (Case (1)), the temperature of the Tokyo operator is used for the temperature. When risk exchange is performed (case (2)), and when temperature risk exchange is performed with the temperature in Toyama, which is the third point, with a business operator in Tokyo (that is, in this embodiment: It is a figure which shows the result of having evaluated the annual average amount of payment amount about case (3)).

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 System 12 Temperature database server 14,16 User computer 20 Point acquisition part 22 Weather data acquisition part 24 Data normalization part 26 Third point selection part 28 Regression analysis part 30 Regression error distribution calculation part 32 Permissible probability input part 34 Dead zone determination Part 36 dead zone output part 38 third point output part 100 CPU
102 memory 104 storage device

Claims (10)

Amount of money received and paid when two companies located at different points A and B exchange risks related to weather using measurable weather at a third point C different from those points A and B A system used to set
The point A, B and for each point k that are candidates for the third point C, a weather acquisition unit for acquiring table to meteorological data history of past weather,
Correlation coefficients ρ _{A, k} , ρ _{B, k} between the meteorological data at each of the points A and B and the meteorological data at each point _k are calculated, and a composite correlation coefficient ρ _k that is a product of them is calculated. = Ρ _{A, k} · ρ _{B, k} is the third point selecting unit for selecting the point _k where the maximum is k as the third point C;
The point A, the respective weather data B, a regression analysis section for performing a regression analysis of the meteorological data of the third point C,
Based on the regression analysis result by the regression analysis unit, the regression error for calculating the distribution of the respective regression errors ε _A and ε _B between the meteorological data at the points A and B and the meteorological data at the third point C 1. A distribution calculator;
The respective standard deviations σ _εA and σ _εB of the distributions of the calculated regression errors ε _A and ε _B are calculated, and the distribution function is assumed to be a predetermined probability distribution. ± x · σ _εA and ± x · , using a value x such that a value obtained by integrating ± with x is (1−predetermined allowable probability) and the calculated standard deviations σ _εA and σ _εB. a dead zone determining unit that determines a wider range of σ _εB as a dead zone when determining the amount of payment according to the weather;
A payment and payment condition setting system for risk exchange related to the weather characterized by comprising: The data acquisition unit further includes a data normalization unit that normalizes the weather data acquired by the weather acquisition unit based on the probability distribution, and the regression analysis unit performs a regression analysis based on the normalized weather data. The system of claim 1 . The data acquisition unit further includes a data normalization unit that normalizes the weather data acquired by the weather acquisition unit based on the probability distribution, and the third point selection unit is configured to standardize the points A and B and the points k . 3. The system according to claim 1, wherein the third point C is selected based on the correlation of weather data. An allowable probability input unit that receives an input of an allowable value of the probability that the payment when the weather at one of the points A and B is used as an index and the payment when the weather at the third point C is used as an index; Prepared,
The dead zone determining section, a pre-Symbol input tolerance as the predetermined permission probability, the system of any one of claims 1 to 3, characterized in that to determine the dead band.   The system according to claim 1, further comprising a dead zone output unit that outputs information representing the determined dead zone. The system according to claim 5 , wherein the dead band output unit transmits information representing the dead band to each of the computers of the two operators. The system according to claim 1 , further comprising a third point output unit that outputs information representing the selected third point C. The system according to claim 7, wherein the third point output unit transmits information representing the third point C 1 to each of the computers of the two operators. Any one of claims 1-8, characterized in that it comprises the receipts and payments condition setting unit sets the receipts and payments conditions based on dead band with the determined transmit information representative of the receipts and payments conditions set to the two operators of the computer The system according to claim 1. Amount of money received and paid when two companies located at different points A and B exchange risks related to weather using measurable weather at a third point C different from those points A and B Is a method for setting
And step computer, the point A, B and for each point k that are candidates for the third point C, to obtain the table to meteorological data history of past weather,
A computer calculates a correlation coefficient ρ _{A, k} , ρ _{B, k} between the weather data at each of the points A 1, B and the weather data at each point k, and a composite phase relationship that is a product of them Selecting a point k at which the number ρ _k = ρ _{A, k} · ρ _{B, k} is the maximum as the third point C;
A computer performing a regression analysis between the weather data of each of the points A 1 and B and the weather data of the third point C;
Computer, on the basis of the regression by analysis regression analysis, the point A, the respective weather data B, the third point C respectively of the regression error epsilon _A and weather _data, the distribution of epsilon _B A calculating step;
The computer calculates the standard deviations σ _εA and σ _εB of the distributions of the calculated regression errors ε _A and ε _B , and assumes that the distribution is a predetermined probability distribution. Using a value x such that a value obtained by integrating the distribution function with ± x becomes (1−predetermined allowable probability) and the calculated standard deviations σ _εA and σ _εB , ± x · σ _εA and A step of determining a wider range of ± x · σ _εB as a dead zone when determining a payment amount according to the weather;
A payment / payment condition setting method in risk exchange related to weather, characterized by comprising: