JP5715989B2 - Method and computer system for predicting the value of structured financial products - Google Patents

Description

  The present invention relates to a method and computer system for predicting the value of a structured financial product. In particular, the present invention relates to a computer-implemented method and computer system for predicting the value of weather-based structured financial products.

  In the insurance industry, insurance policies are provided to insure against the occurrence of certain weather conditions. Similarly, insurance companies and financial service companies offer financial products based on weather conditions, in particular financial derivatives (commonly referred to as structured financial products). Weather-based structured financial products are investment products whose value is based on specific weather measures such as temperature, precipitation, sunshine hours, heating days, cooling days, and wind speed.

  US Pat. No. 6,418,417 discloses a method for evaluating weather-based financial instruments. For financial instruments that have an execution date and maturity date and are specified for a specific region and at least one weather condition, the value of the weather-based financial instrument is the historical weather data and the value between the execution date and the maturity date. Determined based on future weather data predicted for the period of time.

  US Patent Application Publication No. 2004/0230519 discloses a method for generating a pricing model for weather derivatives. The pricing model is based on past weather data and current weather data. Such models utilize deviations of current weather data from past weather data or from a predictive measure.

  It is an object of the present invention to provide a computer-implemented method and computer system for predicting the value of weather-based structured financial products based on predicted weather data. It is a further object of the present invention to provide a computer-implemented method and computer system for predicting the value of weather-based structured financial products based on predicted weather data. The quality of forecast weather data is taken into account.

  According to the invention, the above object is achieved in particular by the structure of the independent claims. Further advantageous embodiments arise from the dependent claims and the description.

  According to the present invention, when the aforementioned object is to predict the value of a weather-based structured financial product, the predicted value is calculated based on predicted weather data for a predetermined period and a predetermined region. Reference weather data is calculated from past weather data for a predetermined period and a predetermined region. Based on the predicted weather data and the reference weather data, a quality index indicating the predicted quality associated with the predicted weather data is calculated. By calculating the reference weather data from past weather data and calculating the predicted quality index from the reference weather data, a quality measure of the predicted value of the financial product is provided. Quality measures allow investors and providers of financial products to make better decisions about the value of financial products.

  In a preferred embodiment, the reference value is calculated based on the reference weather data, and the value of the financial product is calculated from the reference value and from the predicted value weighted by the quality index. By calculating the value of the financial product from the reference value and the predicted value and weighting the predicted value with the quality index, it becomes possible to adjust the influence of the predicted value on the calculated value of the financial product. The impact of forecast value depends on the quality of forecast weather data. Preferably, the weight of the predicted value increases as the accuracy of the predicted weather data with respect to the reference weather data increases. Therefore, the probability that the calculated value of the financial product is accurate is improved. In addition, the new approach presented in this application helps to create and direct an optimal weather derivative portfolio.

  Preferably, the predicted weather index is obtained from the predicted weather data, and the predicted value is calculated by applying the structure parameter of the financial product to the predicted weather index. Similarly, the reference weather index is obtained from the reference weather data, and the reference value is calculated by applying the structure parameter of the financial product to the reference weather index. For example, the predicted weather data, the reference weather data, and the past weather data include temperature data. Similarly, the predicted weather index and the reference weather index include at least one of an average temperature, a cumulative temperature, a heating degree day, and a cooling degree day in a predetermined period and a predetermined region.

  In the embodiment, the step of calculating the quality index includes the step of calculating the ranking probability score of the predicted weather data, the step of calculating the ranking probability score of the reference weather data, and the ranking probability score of the predicted weather data and the reference weather. Calculating a quality index as a ranking probability skill score from the ranking probability score of the data.

  In an embodiment, the forecast weather data is calculated from weather data over multiple years in the past and long-term weather forecast data that includes one or more months.

  In a further preferred embodiment, the step of calculating the forecast weather data comprises determining a first cumulative distribution function of weather data over a plurality of past years for a predetermined period of time, and a quartile of the first cumulative distribution function. A step of calculating an accumulated value included, and a step of obtaining a second cumulative distribution function of the predicted weather data for a predetermined period. The second cumulative distribution function scales the first cumulative distribution function downward using the quartile level obtained from the long-term weather forecast data of the cumulative value included in the quartile of the first cumulative distribution function. Can be obtained.

  In another embodiment, calculating the forecast weather data comprises determining reference climatology information comprising a deterministic component from past weather data for a predetermined period, reference meteorology information, and an ensemble of the predetermined period. Calculating predicted weather data from the prediction.

  In an embodiment, multiple sets of predicted weather data for subsequent periods are assigned to and stored in their respective periods. The predicted weather data is calculated from a plurality of sets of predicted weather data. Each set of predicted weather data is weighted by a weighting factor having a value that increases between one period and a subsequent period. By including a weighted set of predicted weather data from the preceding period, the quality (ie, accuracy) of the predicted weather data is improved.

  In the embodiment, the predicted weather data includes a temperature data over a plurality of past years and a first temperature data cumulative distribution function obtained from long-term temperature predicted data including one or more months. The step of calculating the reference weather data includes a step of obtaining a second cumulative distribution function of the temperature data by applying a probabilistic time series model to the past temperature data. The predicted value is calculated by applying the structure parameter of the financial product to the predicted weather index obtained from the first cumulative distribution function. The reference value is calculated by applying the structure parameter of the financial product to the reference weather index obtained from the second cumulative distribution function. The step of calculating the quality index includes a step of calculating a first ranking probability score based on the first cumulative distribution function, and a step of calculating a second ranking probability score based on the second cumulative distribution function. And calculating a quality index as a ranking probability skill score from the first ranking probability score and the second ranking probability score.

  In addition to computer-implemented methods and computer systems for predicting the value of weather-based structured financial products, the present invention also provides for one or more computers to perform a method for predicting the value of weather-based structured financial products. It also relates to a computer program product comprising computer program code means for controlling a plurality of processors. In particular, the computer program product includes a computer readable medium having computer program code means therein.

FIG. 6 is a block diagram schematically illustrating an exemplary configuration of a computer system implementing an embodiment of the present invention, comprising a computer with a display and data input means. FIG. 6 is a flow diagram illustrating an example of a series of steps performed by the present invention for predicting the value of a weather-based structured financial product. FIG. 6 is a block diagram schematically showing a probabilistic time series model for generating reference weather data based on past weather data. It is a block diagram which shows briefly the daily abnormal method which produces | generates forecast weather data based on the past weather data and long-term weather forecast data. It is a figure which shows the cumulative distribution function of the past weather data which shows a cumulative value about the quartile of a distribution function. FIG. 5b is a diagram illustrating a cumulative distribution function of predicted weather data scaled downward based on the cumulative distribution function shown in FIG. 5a. It is a figure which shows the cumulative distribution function which shows calculation of the ranking probability score of prediction weather data calculated | required by the daily abnormal method. It is a figure which shows the cumulative distribution function which shows calculation of the ranking probability score of prediction weather data calculated | required by the quartile method.

  The invention will now be described in more detail, by way of example, with reference to the accompanying drawings, in which:

  In FIG. 1, reference numeral 1 represents a computer system. The computer system 1 includes one or more computers 1 ′ (eg, personal computers) that include one or more processors. As shown schematically, the computer system 1 includes a display and data input means (eg a keyboard 17 and a pointing device 2 in the form of a computer mouse). The computer system 1 includes a memory, a database 16 and various functional modules (ie, a weather prediction module 11, a weather prediction module 12 with a weighting module 121, a reference module 13, a value prediction module 14 with a weighting module 141, and It further includes a quality indicator module 15). Preferably, the functional module and database 16 are implemented as programmed software modules. The computer program code of the software module is inserted (preferably on a computer-readable medium (in a memory integrated in the computer 1 'of the computer system 1 or in the computer 1' of the computer system 1). Realized as a computer program product (stored on a data carrier). In a variant, the computer system 1 is connected to the external data source 5, for example via a telecommunications network.

  The weather reference module 11 is configured to establish reference weather data based on past weather data. Past weather data is stored in the database 16 or retrieved from the external data source 5. As shown in FIG. 3, the past weather data shown in block 31 as a time series encompassing many years is divided into parts having deterministic data shown in blocks 32 and 33, and probabilistic data shown in block 34. Disassembled into prepared parts. The deterministic part includes past trend data shown in block 32 and seasonal pattern data shown in block 33. The reference weather data is obtained for a predetermined period and a predetermined area. The period and region are defined corresponding to the parameters of the target structured financial product to be predicted. The reference weather data shown in block 38 is simulated by applying a probabilistic time series model to past weather data. In particular, the reference weather data is established from deterministic data applicable for a given period and from probabilistic data determined for said period via autoregression. No forecast weather data is used to determine the reference weather data.

  The weather forecast module 12 is configured to establish forecast weather data based on past weather data over multiple years and long-term weather forecast data that includes one or more months. Past weather data and long-term weather forecast data are provided by a forecasting service provider such as the European Medium Term Weather Forecast Center (ECMWF) for a predetermined period and region, for example. Long-term weather forecast data is stored in the database 16 or retrieved from the external data source 5.

  In the first variation, the predicted weather data is obtained using the daily abnormal method shown in FIG. In the daily anomaly method, long-term weather forecast data is provided in the form of so-called ensemble prediction (abnormality) shown in block 41. The ensemble prediction is synthesized with the reference climatology data shown in blocks 42 and 43. The reference climatology data represents past data for a specific year, for example, past trend data shown in block 43 and seasonal pattern data shown in block 42. The predicted weather data shown in block 44 is thus generated by reconstruction of deterministic weather data and some possible predictions (ensemble predictions) for a given period of time.

  In a second preferred variant, the forecast weather data is determined using the tertile method shown in FIGS. 5a and 5b. As shown in FIG. 5a, for past weather data, a cumulative distribution function 50 is determined for a predetermined period (e.g. (daily) cumulative distribution of temperature during a predetermined period). The accumulated value (for example, 11850, 12100, 13600) included in the tertiles 51, 52, 53 is calculated for each tertile 51, 52, 53 of the cumulative distribution function 50 of past weather data. As shown in FIG. 5b, a cumulative distribution function 57 is determined for a predetermined period of the predicted weather data. The quantile level (for example, 5%) corresponding to the cumulative value (for example, 11850, 12100, and 13600) included in the third quantiles 51, 52, and 53 of the cumulative distribution function 50 of the past weather data from the long-term weather forecast data. , 12% and 83%). The cumulative distribution function 57 of the forecasted weather data is obtained by scaling down the cumulative distribution function 50 of the past weather data using the quantile level (for example, 5%, 12%, and 83%) obtained from the long-term weather forecast data. Established by.

  Multiple sets of forecast weather data for subsequent periods are stored in the database 16 assigned to their respective periods. The weighting module 121 is configured to calculate, from a plurality of sets of predicted weather data stored in the database 16, predicted weather data for a target used for further calculations. Each set of predicted weather data is weighted with a weighting factor having a value that increases from one period to the next subsequent period. For example, the length of the period is 1 month, and the set of forecast weather data includes forecast weather data for the current month and periods with delays of 1, 2, 3, and 4 months. For example, the forecast weather data for the object used for further calculations is 60%, 20%, 10%, 7.5 for the current month and periods with 1 month, 2 months, 3 months and 4 months respectively. %, 2.5% using the weighting coefficient.

  In the following paragraphs, an example of a series of steps performed by the present invention for predicting the value of a weather-based structured financial product will be described with reference to FIG.

In step S1, the value prediction module 14 calculates the predicted value of the structured financial product. First, the predicted weather index is calculated from the predicted weather data calculated by the daily abnormal method or the tertile method described above with reference to FIG. 4 or FIGS. 5a and 5b. For example, past weather data, and thus predicted weather data, includes temperature data, and the predicted weather index includes average temperature, cumulative temperature, heating degree days, or cooling degree days for a predetermined period and a predetermined area. The index type is defined by the individual parameters of the financial instrument to be predicted. The predicted weather index can be calculated from the weather data by the abnormal method shown in FIG. Alternatively, the tertiles shown in FIGS. 5a and 5b can be applied to calculate an average temperature index or a cumulative temperature index. It is possible to extend the quartile method to additional indexes such as cooling days or heating days. This can be achieved by calculating a conditional distribution of the desired temperature index, assuming a predicted tertile of average temperature. Those skilled in the art will appreciate that although the methods and systems proposed in this application are not limited to temperature data, the weather data can also be represented by precipitation or, for example, wind speed and direction. Therefore, in step S12, the predicted value of the structured financial product is calculated by applying the structured parameter of the financial product to the predicted weather index calculated in step S11. For example, in the case of a financial instrument with a put or call deal structure, in addition to region, weather-based index type, and duration, the structure parameters include strike (P _strike ), tick (P _tick ), and limit (P _tick ). _limit ) parameter value. In the case of a put or call deal structure, the predicted value of a structured financial instrument is calculated by formula (1) or (2), respectively.

V _{forwarded-put} = E [min (max ((P _strike −I _forwarded ) · P _tick , 0), P _limit )] (1)
V _{forwarded-call} = E [min (max ((I _forwarded− P _strike ) · P _tick , 0), P _limit )] (2)
I _forecasted is a random variable that follows the prediction index distribution.

  In step S13, the predicted value of the structured financial product is displayed on the display 3 as an output.

  In step S2, the reference module 13 calculates the standard value of the structured financial product. First, the reference weather index is calculated in S21 from the reference weather data calculated as described above with reference to FIG. Corresponding to past weather data, the reference weather data includes temperature data, and the reference weather index includes the average temperature, cumulative temperature, heating degree days, or cooling degree days of a predetermined period and a predetermined area. The reference weather index is calculated from the reference weather data as described above for the predicted weather index. Further, in step S22, the reference value of the structured financial product is calculated by applying the structured parameter of the financial product to the reference weather index calculated in step S21 as described above for the predicted value of the financial product.

In step S3, the quality index module 15 calculates a quality index indicating the predicted quality of the predicted weather data. The quality index is calculated based on the predicted weather data and the reference weather data. In step S31, a ranking probability score (RPS _forecasted ) is calculated according to formula (3) for the predicted weather data. Here, CDF _forecasted is the cumulative distribution function of the predicted weather data, and CD _Factual is the cumulative distribution function of the actual realization (ie, weather data representing the actual weather conditions that occurred).

_{RPS forecasted = ∫ (CDF forecasted (} x) -CDF actual (x)) 2 dx (3)
In FIGS. 6a and 6b, reference numeral 60 represents the actual implementation, reference numeral 61 represents the cumulative distribution function CDF _reference of the standard weather data, and reference numeral 62 represents the cumulative prediction data derived by the tertile method. This represents the distribution function CDF _forcasted . In FIG. 6a, reference numeral 63 represents an area representing the ranking probability score of the standard weather data. In FIG. 6b, reference numeral 64 represents a region representing the ranking probability score of the predicted weather data derived by the tertile. Obviously, the tertile method gives results closer to the actual realization.

In step S32, a ranking probability score (RPS _reference ) is calculated according to formula (4) for the reference weather data. Here, CDF _reference is a cumulative distribution function of reference weather data.

RPS _reference = ∫ (CDF _reference (x) −CDF _actual (x)) ² dx (4)
In step S33, the quality index is calculated by the formula (5) as a ranking probability skill score (RPSS) calculated from the average ranking probability score of the predicted weather data and the average ranking probability score of the reference weather data for several years. Calculated.

ランク付け確率スキル・スコアは、基準天候データに対する、天候データの予測の精度を示す。ランク付け確率スキル・スコアは、基準シミュレーションに対する予測の精度の向上のパーセンテージを示す。スキル・スコアは、単に統計シミュレーションによって導き出された基準の精度と等しい精度を有する予測の場合、０％の値を有する。正のスコアは、予測精度が基準の精度に対して向上していることを示す。負のスコアは、予測精度が基準の精度より悪いことを示す。品質指標の算出は、ランク付け確率スキル・スコアのみに限定されるものでない。別の品質尺度も本願で提示した方法論で使用できる。

The ranking probability skill score indicates the accuracy of weather data prediction relative to the reference weather data. The ranking probability skill score indicates the percentage improvement in prediction accuracy relative to the reference simulation. The skill score has a value of 0% for predictions that have an accuracy equal to the accuracy of the criteria simply derived by statistical simulation. A positive score indicates that the prediction accuracy is improved relative to the reference accuracy. A negative score indicates that the prediction accuracy is worse than the reference accuracy. The calculation of the quality index is not limited to the ranking probability skill score. Other quality measures can also be used with the methodology presented herein.

  In step S34, the quality index is displayed on the display 3 as an output.

In a preferred option, in step S4, the value prediction module 14 calculates the predicted value of the structured financial product from the reference value calculated in step S2 and the predicted value calculated in step S1. The predicted value of the structured financial product is calculated from the reference value and the predicted value using the quality index calculated in step S3 as a weighting coefficient. The predicted value V _forecasted of the structured financial product is calculated by, for example, Expression (6). Here, α (s) is a function of the quality index, V _forecast is the predicted value calculated in step S1, and V _reference is the reference value calculated in step S2.

_Vforecasted = α (0) · _Vforecast + (1−α (s)) · _Vreference (6)
In step S41, the (weighted) predicted value V _{forecasted of} the structured financial product is displayed on the display 3 as an output.

  Different process sequences are possible without departing from the scope of the invention. For example, step S32 can be performed prior to step S31.

Claims (34)

A computer-implemented method for predicting the value of weather-based structured financial products,
A value prediction module calculating a predicted value based on predicted weather data for a predetermined period and a predetermined region;
A reference module calculating reference weather data from past weather data for the predetermined period and the predetermined area;
A quality index module calculating a quality index indicating a predicted quality associated with the predicted weather data based on the predicted weather data and the reference weather data;
The value prediction module calculating by computer a prediction value of the weather-based structured financial product based on the prediction value weighted by the quality indicator and a reference value calculated from the reference weather data;
The predicted value is a put option value V _{forwarded-put} defined by V _{forwarded-put} = E [min (max ((P _strike −I _forwarded ) · P _tick , 0), P _limit )],
P _strike is the strike price value of the weather-based structured financial product,
P _tick is the tick price value of the weather-based structured financial product,
P _limit is the limit price value of the weather-based structured financial product,
I _forecasted is a random variable that follows the distribution of the predicted weather index.   2. The method of claim 1, wherein the method further comprises the reference module calculating the reference value based on the reference weather data.   2. The method of claim 1, wherein the value prediction module calculates the predicted value by applying a structure parameter of the weather-based structured financial product to a predicted weather index determined from the predicted weather data. The method wherein the reference module calculates the reference value by applying the mechanism parameter of the weather-based structured financial product to a reference weather index determined from the reference weather data.   4. The method according to claim 3, wherein the predicted weather data, the reference weather data, and the past weather data include air temperature data, and the predicted weather index and the reference weather index are the predetermined period and the predetermined weather data. A method comprising one of an average temperature, a cumulative temperature, a number of heating days, and a number of days of cooling.   5. The method according to claim 1, wherein the step of calculating the quality indicator by the quality indicator module is a step of calculating a ranking probability score of the predicted weather data by the quality indicator module. 6. The quality indicator module calculates a ranking probability score of the reference weather data; and the quality indicator module ranks from the ranking probability score of the predicted weather data and the ranking probability score of the reference weather data. Calculating the quality index as an attached probability skill score.   6. The method according to any one of claims 1 to 5, wherein the predicted weather data is calculated from past years of weather data and long-term weather forecast data over one or more months. A sixth aspect of the method, the step of weather prediction module calculates the forecasted weather data, the step of the weather prediction module asks for the first predetermined period of time the cumulative distribution function of the historical weather data The weather prediction module calculating a cumulative value included in the quartile of the first cumulative distribution function; and the weather prediction module included in the tertile of the first cumulative distribution function. Obtaining a second cumulative distribution function of the predicted weather data for the predetermined period of time by scaling down the first cumulative distribution function using a quantile level obtained from the long-term weather forecast data of the cumulative value. And a method comprising. A sixth aspect of the method, the step of weather prediction module calculates the forecasted weather data, the weather forecast module deterministic component comprises reference climatology information of said predetermined from the historical weather data A method comprising: determining a time period; and wherein the weather prediction module calculates the predicted weather data from the reference climatology information and an ensemble prediction for the predetermined period. A The method of claim 1, the database stores a plurality of sets of forecasted weather data, each of the plurality of predicted weather data set is assigned to the period subsequent to said predetermined period of time, weather forecast A method wherein the module calculates the predicted weather data from the plurality of sets of predicted weather data, each set of predicted weather data being weighted with a weighting factor having a value that increases between one period and a subsequent period. The method of claim 1, wherein the forecast weather data includes a first cumulative distribution function of temperature data determined from past multi-year temperature data and long-term temperature forecast data over one or more months, step of calculating the reference weather data, weather reference module, by applying a stochastic time series model the historical temperature data, comprising the step of determining a second cumulative distribution function of the temperature data, the value predicted The module calculates the predicted value by applying a structure parameter of the weather-based structured financial product to a predicted weather index obtained from the first cumulative distribution function, and the reference module calculates the reference value, By applying the structure parameter of the weather-based structured financial product to the reference weather index obtained from the second cumulative distribution function Calculating the quality index, wherein the quality index module calculates a first ranking probability score based on the first cumulative distribution function; and Calculating a second ranking probability score based on a second cumulative distribution function; and the quality indicator module calculates a ranking probability skill from the first ranking probability score and the second ranking probability score. Calculating the quality indicator as a score. A computer system that predicts the value of a weather-based structured financial product,
A computer-implemented value prediction module for calculating a predicted value based on predicted weather data for a predetermined period and a predetermined region;
A reference module for calculating reference weather data from past weather data of the predetermined period and the predetermined area;
A quality index module that calculates a quality index indicating the predicted quality associated with the predicted weather data based on the predicted weather data and the reference weather data;
The value prediction module implemented on the computer further calculates a predicted value of the weather-based structured financial product based on the predicted value weighted by the quality index and a reference value calculated from the reference weather data And
The computer system is
An output module implemented on the computer for performing at least one of outputting and displaying the predicted value of the weather-based structured financial product;
The predicted value is a put option value V _{forwarded-put} defined by V _{forwarded-put} = E [min (max ((P _strike −I _forwarded ) · P _tick , 0), P _limit )],
P _strike is the strike price value of the weather-based structured financial product,
P _tick is the tick price value of the weather-based structured financial product,
P _limit is the limit price value of the weather-based structured financial product,
I _forecasted is a computer system that is a random variable that follows the distribution of the forecast weather index.   12. The computer system of claim 11, wherein the reference module is configured to calculate the reference value based on the reference weather data.   12. The computer system according to claim 11, wherein the value prediction module calculates the predicted value by applying a structure parameter of the weather-based structured financial product to a predicted weather index obtained from the predicted weather data. The computer system is configured to calculate the reference value by applying the mechanism parameter of the weather-based structured financial product to a reference weather index determined from the reference weather data.   14. The computer system according to claim 13, wherein the predicted weather data, the reference weather data, and the past weather data include temperature data, and the predicted weather index and the reference weather index include the predetermined period and A computer system comprising one of an average temperature, a cumulative temperature, a heating degree days, and a cooling degree days in the predetermined area.   15. The computer system according to claim 11, wherein the quality indicator module calculates a ranking probability score of the predicted weather data and calculates a ranking probability score of the reference weather data. And a computer system configured to calculate the quality index as a ranking probability skill score from the ranking probability score of the predicted weather data and the ranking probability score of the reference weather data.   16. The computer system according to claim 11, further comprising a weather forecasting module that calculates the forecasted weather data from weather data for a plurality of years in the past and long-term weather forecast data for one or more months. Computer system.   17. The computer system according to claim 16, wherein the weather prediction module determines a first cumulative distribution function of the past weather data for the predetermined period, and a quartile of the first cumulative distribution function. The cumulative value included in the first cumulative distribution function is calculated using the quantile level obtained from the long-term weather forecast data of the cumulative value included in the third quantile of the first cumulative distribution function. A computer system configured to determine a second cumulative distribution function of the predicted weather data for the predetermined time period by scaling down.   17. The computer system according to claim 16, wherein the weather prediction module obtains reference climatology information having a deterministic component from the past weather data for the predetermined period, and the reference climatology information and the A computer system configured to calculate the forecast weather data from an ensemble forecast for a predetermined time.   12. The computer system of claim 11, wherein the system includes a database that stores a plurality of sets of predicted weather data, each of the plurality of sets being assigned to a period subsequent to the predetermined period. The system includes a weather prediction module that calculates the predicted weather data from the plurality of sets of predicted weather data, each set of predicted weather data having a value that increases between one period and a subsequent period. A computer system that is weighted by a weighting factor. 12. The computer system according to claim 11, wherein the forecasted weather data includes a first cumulative distribution function of temperature data obtained from temperature data of a plurality of past years and long-term temperature forecast data over one or more months. The system includes a weather reference module for calculating the reference weather data, wherein the weather reference module applies a stochastic time series model to the past temperature data to provide a second cumulative distribution function of the temperature data. Configured to ask for
The value prediction module is configured to calculate the predicted value by applying a structure parameter of the weather-based structured financial product to a predicted weather index obtained from the first cumulative distribution function, and the reference module is , A reference value is calculated by applying a structure parameter of the weather-based structured financial product to a reference weather index obtained from the second cumulative distribution function, and the quality indicator module includes the first cumulative distribution A first ranking probability score is calculated based on the function, a second ranking probability score is calculated based on the second cumulative distribution function, and the first ranking probability score and the second rank are calculated. A computer configured to calculate the quality indicator as a ranking probability skill score from the ranking probability score System. When executed
A function in which the value prediction module calculates a predicted value based on predicted weather data for a predetermined period and a predetermined region;
A function in which a reference module calculates reference weather data from past weather data of the predetermined period and the predetermined area;
A function for a quality index module to calculate a quality index indicating a predicted quality associated with the predicted weather data based on the predicted weather data and the reference weather data;
The value prediction module causes a computer to execute a function of calculating a predicted value of a weather-based structured financial product based on the predicted value weighted by the quality indicator and a reference value calculated from the reference weather data; Having computer readable program code embodied in a computer readable storage medium;
The predicted value is a put option value V _{forwarded-put} defined by V _{forwarded-put} = E [min (max ((P _strike −I _forwarded ) · P _tick , 0), P _limit )],
P _strike is the strike price value of the weather-based structured financial product,
P _tick is the tick price value of the weather-based structured financial product,
P _limit is the limit price value of the weather-based structured financial product,
I _forecasted is a computer readable storage medium that is a random variable that follows the distribution of the predicted weather index. A computer readable storage medium according to claim 21, comprising:
When executed
A computer-readable storage medium further comprising program code for causing the computer to execute a function for the reference module to calculate a reference value based on the reference weather data. A computer readable storage medium according to claim 21, comprising:
When executed
The value prediction module calculates the predicted value by applying a structure parameter of the weather-based structured financial product to a predicted weather index obtained from the predicted weather data; and the reference module includes the reference weather Computer readable storage further comprising program code for causing the computer to execute the function of calculating the reference value by applying the mechanism parameter of the weather-based structured financial product to a reference weather index determined from data Medium.   24. The computer-readable storage medium according to claim 23, wherein the predicted weather data, the reference weather data, and the past weather data include temperature data, and the predicted weather index and the reference weather index are the predetermined values. A computer readable storage medium comprising one of a period and an average temperature, cumulative temperature, heating degree days, and cooling degree days of the predetermined area. A computer readable storage medium according to claim 21, comprising:
When executed
A function for calculating a ranking probability score for the predicted weather data; a function for calculating a ranking probability score for the reference weather data; and a function for calculating the ranking probability score for the reference weather data. A computer readable storage further comprising program code for causing the computer to execute a function of calculating the quality index as a ranking probability skill score from the ranking probability score and the ranking probability score of the reference weather data. Medium. A computer readable storage medium according to claim 21, comprising:
When executed
A computer-readable storage medium further comprising program code for causing the computer to execute a function of calculating the forecast weather data from past weather data of a plurality of years and long-term weather forecast data over one or more months. A computer readable storage medium according to claim 26, comprising:
When executed
Calculating a function of weather prediction module asks for the first predetermined period of time the cumulative distribution function of the historical weather data, a cumulative value which the weather forecast module is included in the third position of the first cumulative distribution function And the weather forecasting module uses the quantile level obtained from the long-term weather forecast data for the cumulative value included in the third quantile of the first cumulative distribution function. A computer-readable storage medium further comprising program code for causing the computer to execute a function of obtaining a second cumulative distribution function of the predicted weather data for the predetermined period by scaling the cumulative distribution function downward. A computer readable storage medium according to claim 26, comprising:
When executed
A function of determining for said predetermined period of time reference climatology information weather prediction module comprises a deterministic component from said historical weather data, the weather forecast module the reference climatology information, and ensemble prediction of the predetermined time period A computer-readable storage medium further comprising program code for causing the computer to execute the function of calculating the predicted weather data from the computer. The computer-readable storage medium according to claim 21, wherein the database has a function of storing a plurality of sets of predicted weather data, and each of the plurality of sets of predicted weather data follows the predetermined period. further comprising a function assigned to the period, the program code for executing weather forecast module from the plurality of predictive weather data set and a function for calculating the forecasted weather data the computer to the computer, forecasted weather data Each set is weighted with a weighting factor having a value that increases between one period and a later period. The computer-readable storage medium according to claim 21, wherein the predicted weather data is a first of temperature data obtained from temperature data of a plurality of past years and long-term temperature prediction data over one or more months. wherein the cumulative distribution function, weather reference module, by applying a stochastic time series model the historical temperature data, a function of obtaining a second cumulative distribution function of the temperature data, the value prediction module, said A function of calculating the predicted value by applying a structure parameter of the weather-based structured financial product to a predicted weather index obtained from a first cumulative distribution function; and the reference module includes the second cumulative distribution function A function for calculating the reference value by applying the structure parameter of the structured financial product to the reference weather index obtained from A function for the quality index module to calculate a first ranking probability score based on the first cumulative distribution function; and a second rank for the quality index module to be calculated by the computer based on the second cumulative distribution function. A function for calculating a ranking probability score, and a function for the quality index module to calculate the quality index as a ranking probability skill score from the first ranking probability score and the second ranking probability score by the computer. A computer-readable storage medium further comprising program code for causing the computer to execute the program. The method of claim 1, comprising:
The reference weather data is obtained from deterministic data applicable to the predetermined period and from probabilistic data obtained for the predetermined period via autoregression. The method of claim 1, comprising:
The method wherein the reference weather data is simulated by applying a probabilistic time series model to the past weather data. The method of claim 1, comprising:
The predicted weather data is not used for obtaining the reference weather data. 32. The method of claim 31, wherein
The method wherein the deterministic data includes historical trend data and seasonal pattern data.

Images