JP2009259196A - Extended assurance loss cost prediction device and prediction system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an extended assurance loss cost prediction device and an extended assurance loss cost prediction system, capable of enhancing dramatically prediction precision. <P>SOLUTION: This extended assurance loss cost prediction device is provided with a means for calculating a loss cost in each sales year and month and each lapse month number in an already lapse period of maker assurance, based on a sales number in each sales year and month in the already lapse period of the maker assurance as to an assurance-objective article, and based on a loss result data in each sales year and month and each lapse month number in the already lapse period of the maker assurance as to the article, a means for calculating a loss cost in each sales year and month and each lapse month number in an already lapse period of extended assurance, based on a contract number in each sales year and month in the already lapse period of the extended assurance as to the article, and based on a loss result data in each sales year and month and each lapse month number in the already lapse period of the extended assurance as to the article, and a means for calculating a loss cost in each sales year and month and each lapse month number in an unlapsed period by a chain ladder method, based on the loss cost in each sales year and month and each lapse month number in the already lapse period of the maker assurance, and based on the loss cost in each sales year and month and each lapse month number in the already lapse period of the extended assurance. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an extended warranty loss cost prediction apparatus and an extended warranty loss cost prediction system for predicting a loss cost of an extended warranty after a manufacturer warranty for an article with a manufacturer warranty.

  When an insurance company provides an extended warranty after the manufacturer's warranty for a high-priced item with a manufacturer's warranty, such as an automobile, this is equivalent to the total amount of repair costs, replacement costs, loss costs, etc. during the extended warranty period of the unit product. It is necessary to predict the extended warranty loss cost.

  The conventional method for predicting the extended warranty loss cost of high-priced goods, for example, automobiles, is the actual loss data such as the number of articles (number of cars) that have been actually extended in the past and the repair costs actually required during the extended warranty period. Was used to predict the average value of the past. Of course, it was assumed that the target article itself or the parts constituting it would not be greatly evolved in the future. In addition, there is an analytical prediction method that simply applies continuous conditions such as the Weibull distribution from the present, but this is effective in predicting the occurrence of failure itself, but repair requests are targeted for prediction. In the present invention, since there is no consideration of the time lag of repair request from the occurrence of a failure, that is, the uneven distribution of claims caused by the milestones and timings at which a person takes charge action, it has not been adopted by the present applicant.

  According to the investigation by the present applicant, there is no publicly known document about the prediction of the extended warranty loss cost for such a repair request.

  According to the applicant's conventional method of predicting the extended warranty loss cost from the number of articles and loss history data during the extended warranty period, whether or not the target article itself or the components that make up the product have evolved rapidly. Since the extended guaranteed loss cost is predicted without knowing the degree and without reflecting the influence at all, the prediction accuracy is limited.

  Accordingly, an object of the present invention is to provide an extended guarantee loss cost prediction apparatus and an extended guarantee loss cost prediction system capable of dramatically improving prediction accuracy.

In the present invention, the extended warranty loss cost prediction device is
A computer for loss cost prediction;
It is connected to a computer for loss cost prediction, and the manufacturer's warranty for the entire product is stored and stored in a matrix in the form of a sales date corresponding to the sales date for each manufacturer's warranty period In addition, the loss history data by sales year and the number of elapsed months in the lapse of the manufacturer warranty for each article is stored and saved in a matrix form corresponding to the sales year and the elapsed month. A data storage unit of
About the extended warranty that covers only a part of the contract of goods The number of contracts by sales month in the elapsed period of the extended warranty for every article to be guaranteed is stored and stored in a matrix form corresponding to the sales date , It is configured to store and save loss data by sales year and elapsed month for all items subject to warranty in a matrix form corresponding to the sales year and elapsed month. A second data storage unit;
A third data storage unit;
A fourth data storage unit;
A fifth data storage unit;
A sixth data storage unit;
A seventh data storage unit;
An eighth data storage unit;
A ninth data storage unit;
A tenth data storage unit.
The computer for loss cost prediction
From the first data storage unit, the number of sales by sales year and the number of months of sales in the past period corresponding to the sales period and the number of months of the goods to be predicted By reading different loss record data and dividing the read loss record data by the number of sales read, the loss cost is calculated for each sales year and elapsed month in the manufacturer warranty period, and the calculated loss cost is sold. A first calculation means for storing and storing in a matrix in the third data storage unit corresponding to the year and month and the number of elapsed months;
From the second data storage unit, the number of contracts by sales year and the number of contracts in the past period corresponding to the sales date and the number of months of the article to be predicted Read another loss record data and divide the read loss record data by the number of contracts read to calculate the loss cost by sales year and month in the elapsed period of the extended warranty, and sell the calculated loss cost Second calculation means for storing and storing in a matrix in the fourth data storage unit corresponding to the year and month,
From the third data storage unit, the loss cost by sales year and the number of elapsed months in the past period of the manufacturer warranty of the article to be predicted is read, and the extended warranty of the article to be predicted is already read from the fourth data storage unit. Read loss costs by sales month and number of months in the elapsed period, and unify the read loss costs related to the manufacturer's warranty with the read loss costs related to the extended warranty. The cumulative loss cost in the elapsed months direction by elapsed months is calculated, and the cumulative loss cost in the elapsed months by sales year and elapsed months in the calculated elapsed period is calculated according to the sales month and elapsed months. 3rd calculation means which memorize | stores and preserve | saves it at matrix form in the data storage part of 5;
From the fifth data storage unit, the cumulative loss cost in the elapsed months for each elapsed month and the number of elapsed months in the elapsed period is read, and the cumulative total in the sales month direction for each elapsed month and in the elapsed month Calculates the loss cost and stores the total loss cost in the elapsed month direction and the sales month direction by sales year and elapsed month in the elapsed period obtained in the sixth period. A fourth calculating means for storing and storing the data in a matrix in the unit;
From the sixth data storage unit, read the accumulated loss cost in the elapsed months and sales months by sales year and month in the past period, and bound the first month of the non-elapsed period by sales year Month, the loss development for only the boundary month by dividing the cumulative loss cost of the same number of months elapsed as the boundary month in the immediately preceding sales month by the cumulative loss cost of the elapsed months immediately preceding the boundary month in the immediately preceding sales month A fifth calculation means for calculating a factor, and storing and saving the loss development factor of only the boundary month obtained in the calculation in a matrix form in the seventh data storage unit corresponding to the sales year and the elapsed month number;
Reads the loss development factor for the boundary month from the seventh data storage unit, and sums the loss development factor for the boundary month to the power within the range of the sales month and the number of months that have not passed. A sixth calculation means for calculating all the loss development factors of and storing all the loss development factors of the total non-elapsed periods obtained in the calculation in a matrix form in the eighth data storage unit;
Read the total loss cost for the last month of the past elapsed period from the fifth data storage unit by sales year and month, and read all loss development factors for the total non-elapsed period from the eighth data storage unit. By month, the cumulative loss cost for the last month of the past period is multiplied by the loss development factor for each past month, and the past and total past periods in the past and the total past months are shown for each sales year and each past month. Calculate the accumulated loss cost in the elapsed months direction, read the accumulated loss cost in the elapsed months direction by sales year and elapsed months in the total elapsed period from the fifth data storage unit, and calculate the elapsed losses in the calculated elapsed period Cumulative loss cost in the direction of the number of months by sales year and the number of months elapsed, and the total number of sales and months in the past period read out And unifying a total loss cost of months another elapsed months direction, a seventh calculation means for storing saved in a matrix in the data storage unit of the ninth,
Read the accumulated loss cost by sales year for the maker's warranty and extended warranty for the goods to be predicted from the ninth data storage unit, and the lapse of the manufacturer's warranty and extended warranty Calculate the cumulative loss cost for each sales year in the past and the past of the extended warranty only by subtracting the cumulative loss cost of the manufacturer's warranty from the cumulative loss cost in the total and the total of the past. And an eighth calculating means for storing and saving the accumulated loss cost in the past and only past extended warranty periods in the tenth data storage unit corresponding to the sales year and month.

  Note that “loss cost” used in the present specification means “fault repair cost that occurs during one month of one article”. In other words, during the manufacturer warranty period, the failure repair costs related to the failure repair performed based on the manufacturer warranty contract are totaled for each sales year and every elapsed month, and this is already passed for every sales year and every elapsed month. Divided by the number of sales. In addition, during the extended warranty period, breakdown repair costs related to breakdown repairs made based on the extended warranty contract are totaled for each sales month and number of months passed, and this total has elapsed for each sales year and month elapsed. Divided by the number of contracts. In addition, due to contract differences between manufacturer warranty contracts and extended warranty contracts, the warranty coverage for breakdown repairs is naturally different, but it should be as homogeneous as possible. Therefore, when the warranty scope of the manufacturer warranty contract covers the entire scope, but the warranty scope of the extended warranty contract covers only a specific part (for example, the warranty scope includes the special warranty scope and the general warranty scope). If the manufacturer's warranty contract is 5 years for the special warranty scope and 3 years for the general warranty scope, and the extended warranty contract consists of the general warranty scope for 2 years after the manufacturer's warranty) Out of range is excluded from the repair cost during the manufacturer warranty period. Naturally, the reverse is also conceivable, but since the influence is small, it is assumed that a separate frame is calculated, and in the present application, it is assumed that the common guarantee range is met.

  It calculates the loss cost by sales month and elapsed month in the manufacturer warranty period, and calculates the loss cost by sales year and elapsed month in the extended warranty period. The accumulated loss cost in the direction of the number of months by sales year and by the number of months passed is calculated from the loss cost by the sales month and the number of months in the past period of the manufacturer warranty and extended warranty. Calculate the cumulative loss cost by month and by number of months. By subtracting the accumulated loss cost for the manufacturer warranty's past and unexpired periods from the calculated accumulated loss cost for the manufacturer's warranty and extended warranty's past and unexpired periods, the past and remaining periods for the extended warranty only Get the cumulative loss cost for the period. Thus, in the present invention, the loss cost is adopted as a common analysis index in order to align the bases of the manufacturer-guaranteed sales number and loss record data and the extended warranty contract number and loss record data. That is, the ordinary chain ladder method predicts the cumulative loss performance data based on the number of sales or the number of contracts. In the present invention, however, the manufacturer guarantee (of course, the loss performance data) covers all sales items. Is also very large) and the extended warranty (of course, the loss actual data is also small) that covers only some selected contracts of the goods sold can be analyzed at the same time. Alternatively, the total loss cost is predicted using the loss cost divided by the number of contracts. Furthermore, in order to enable detailed predictions, the chain ladder method, which is usually used on a yearly basis, is applied on a monthly basis. The manufacturer warranty, which has a track record prior to the extended warranty, can check the presence and extent of the impact of the evolution of the product itself and its components, so the product itself and its components that have not yet appeared in the extended warranty You can take into account the impact of evolution. As a result, the prediction accuracy can be dramatically improved. Further, it becomes possible to predict in advance the failure risk of a specific component that has not yet appeared in the extended warranty.

  The article to be guaranteed is an automobile, and the number of sales and loss record data stored and stored in the first data storage unit are classified according to the sales period in the lapse of the manufacturer warranty for every automobile to be guaranteed. The number of sales and the number of contracts and losses stored in the second data storage unit, which are loss data by sales year and by the number of months in the lapse of the manufacturer warranty for all vehicles subject to warranty The actual data shows the number of contracts by sales month for the extended warranty period for all vehicles subject to warranty and the sales month and month for the extended warranty period for all vehicles subject to warranty. Another loss record data is preferable. In this case, the manufacturer warranty may be 36 months from the date of sale, and the extended warranty may be 24 months from the end of the manufacturer warranty.

The third calculating means calculates the loss cost by sales year and elapsed month in the elapsed period of the manufacturer warranty of the article to be predicted read from the third data storage unit, C _ij (i is the sales date, j is elapsed months), the fourth data available from the storage unit in the earned period extended warranty of the article to be predicted readout years by and number of elapsed months another loss cost and c _{ij (i} sales years, j elapsed months Number), the cumulative loss cost R _{ij in the} direction of the number of months by sales year and the number of months elapsed in the total period of manufacturer warranty and extended warranty is expressed as R _ij = ΣC _ij + Σc _ij (j = 1 to final It is preferable to include means for calculating from the calculation of (elapsed months). 7a and 7b of the embodiment described later for the loss cost C _ij , FIGS. 8a and 8b of the embodiment described later for the loss cost c _ij , and FIG. 9a of the embodiment described later for the accumulated loss cost R _ij . Understanding is facilitated by referring to 9b and 9c.

The fourth calculation means calculates the total loss cost in the direction of elapsed months by sales year and elapsed month in the total elapsed period of the manufacturer warranty and extended warranty read from the fifth data storage unit, R _ij (i is Assuming that the sales date and j are the number of months elapsed), the cumulative loss cost T _ij in the elapsed month direction and sales month direction in the elapsed period and the sales month direction is represented by T _ij = ΣR _ij (i = first It is also preferable to include means for calculating from the calculation of (sales date to past last sales date). The cumulative loss cost R _ij can be easily understood by referring to FIGS. 9a, 9b and 9c of the embodiment described later, and the cumulative loss cost T _ij with reference to FIGS. 10a, 10b and 10c of the embodiment described later.

The fifth calculation means calculates the total loss cost in the elapsed month direction and the sales month direction by sales year and month in the past elapsed period read from the sixth data storage unit, and T _ij (i is the sales date) , J is the number of elapsed months), means for calculating the loss development factor F _ij for only the boundary month from the calculation of F _ij = T _{(i−1) j} / T _{(i−1) (j−1)} It is also preferable to include. For the accumulated loss cost T _ij , refer to FIGS. 10 a, 10 b and 10 c of the embodiment described later, and for the loss development factor F _ij for only the boundary month, it is easy to understand by referring to FIGS. 11 a, 11 b and 11 c of the embodiment described later. It becomes.

If the loss development factor of only the boundary month read from the seventh data storage unit is F _ij (where i is the sales month and j is the number of elapsed months), the sixth calculation means It is also preferable to include means for calculating the loss development factor LDF _ij from the calculation of LDF _ij = ΠF _kl (the power of all boundary month loss development factors F _kl satisfying k ≦ i and l ≦ j). Refer to FIGS. 11a, 11b, and 11c of the later-described embodiment for the loss development factor F _ij for only the boundary month, and refer to FIGS. 12a, 12b, and 12c of the later-described embodiment for all the loss development factors LDF _ij . Then, understanding becomes easy.

Calculating means 7, all losses Development factor the LDF ij of unexpired period total read out from the data storage unit of the 8 _(i sales years, j elapsed months), the fifth data storage means If the cumulative loss cost for the last month of the past period is SC _i (i is the year of sale) by sales year read out from, the past period in the non-elapsed period and the total sales year and the past month in the non-elapsed period It is also preferable to include means for calculating the cumulative loss cost MC _{ij in the} direction of the number of elapsed months according to the number from the calculation of MC _ij = SC _i × LDF _ij . The seventh calculating means calculates the total loss cost in the direction of the number of months by sales year and month in the total elapsed time read from the fifth data storage unit, R _ij (where i is the sales date and j is the time Number of months)) and the accumulated loss cost MC _{ij in the} direction of the number of months elapsed by sales year and total number of months elapsed in the calculated period not yet elapsed, It is more preferable to use the cumulative loss cost for each sales year in the elapsed and unexpired periods of the manufacturer warranty and extended warranty. For all loss development factors LDF _ij , refer to FIGS. 12 a, 12 b and 12 c of the embodiment described later, and for the accumulated loss cost obtained by integrating the accumulated loss cost R _ij and the accumulated loss cost MC _ij , FIG. 13 a of the embodiment described later. , 13b and 13c make it easy to understand.

In the present invention, the extended warranty loss cost prediction system further includes a client side terminal and a server side extended warranty loss cost prediction device connected to the client side terminal via a communication network.
Extended warranty loss cost prediction device
A computer for loss cost prediction;
It is connected to a computer for loss cost prediction, and the manufacturer's warranty for the entire product is stored and stored in a matrix in the form of a sales date corresponding to the sales date for each manufacturer's warranty period In addition, the loss history data by sales year and the number of elapsed months in the lapse of the manufacturer warranty for each article is stored and saved in a matrix form corresponding to the sales year and the elapsed month. A data storage unit of
About the extended warranty that covers only a part of the contract of goods The number of contracts by sales month in the elapsed period of the extended warranty for every article to be guaranteed is stored and stored in a matrix form corresponding to the sales date , It is configured to store and save loss data by sales year and elapsed month for all items subject to warranty in a matrix form corresponding to the sales year and elapsed month. A second data storage unit;
A third data storage unit;
A fourth data storage unit;
A fifth data storage unit;
A sixth data storage unit;
A seventh data storage unit;
An eighth data storage unit;
A ninth data storage unit;
A tenth data storage unit.
The computer for loss cost prediction
From the first data storage unit, the number of sales by sales year and the number of months of sales in the past period corresponding to the sales period and the number of months of the goods to be predicted By reading different loss record data and dividing the read loss record data by the number of sales read, the loss cost is calculated for each sales year and elapsed month in the manufacturer warranty period, and the calculated loss cost is sold. A first calculation means for storing and storing in a matrix in the third data storage unit corresponding to the year and month and the number of elapsed months;
From the second data storage unit, the number of contracts by sales year and the number of contracts in the past period corresponding to the sales date and the number of months of the article to be predicted Read another loss record data and divide the read loss record data by the number of contracts read to calculate the loss cost by sales year and month in the elapsed period of the extended warranty, and sell the calculated loss cost Second calculation means for storing and storing in a matrix in the fourth data storage unit corresponding to the year and month,
From the third data storage unit, the loss cost by sales year and the number of elapsed months in the past period of the manufacturer warranty of the article to be predicted is read, and the extended warranty of the article to be predicted is already read from the fourth data storage unit. Read loss costs by sales month and number of months in the elapsed period, and unify the read loss costs related to the manufacturer's warranty with the read loss costs related to the extended warranty. The cumulative loss cost in the elapsed months direction by elapsed months is calculated, and the cumulative loss cost in the elapsed months by sales year and elapsed months in the calculated elapsed period is calculated according to the sales month and elapsed months. 3rd calculation means which memorize | stores and preserve | saves it at matrix form in the data storage part of 5;
From the fifth data storage unit, the cumulative loss cost in the elapsed months for each elapsed month and the number of elapsed months in the elapsed period is read, and the cumulative total in the sales month direction for each elapsed month and in the elapsed month Calculates the loss cost and stores the total loss cost in the elapsed month direction and the sales month direction by sales year and elapsed month in the elapsed period obtained in the sixth period. A fourth calculating means for storing and storing the data in a matrix in the unit;
From the sixth data storage unit, read the accumulated loss cost in the elapsed months and sales months by sales year and month in the past period, and bound the first month of the non-elapsed period by sales year Month, the loss development for only the boundary month by dividing the cumulative loss cost of the same number of months elapsed as the boundary month in the immediately preceding sales month by the cumulative loss cost of the elapsed months immediately preceding the boundary month in the immediately preceding sales month A fifth calculation means for calculating a factor, and storing and saving the loss development factor of only the boundary month obtained in the calculation in a matrix form in the seventh data storage unit corresponding to the sales year and the elapsed month number;
Reads the loss development factor for the boundary month from the seventh data storage unit, and sums the loss development factor for the boundary month to the power within the range of the sales month and the number of months that have not passed. A sixth calculation means for calculating all the loss development factors of and storing all the loss development factors of the total non-elapsed periods obtained in the calculation in a matrix form in the eighth data storage unit;
Read the total loss cost for the last month of the past elapsed period from the fifth data storage unit by sales year and month, and read all loss development factors for the total non-elapsed period from the eighth data storage unit. By month, the cumulative loss cost for the last month of the past period is multiplied by the loss development factor for each past month, and the past and total past periods in the past and the total past months are shown for each sales year and each past month. Calculate the accumulated loss cost in the elapsed months direction, read the accumulated loss cost in the elapsed months direction by sales year and elapsed months in the total elapsed period from the fifth data storage unit, and calculate the elapsed losses in the calculated elapsed period Cumulative loss cost in the direction of the number of months by sales year and the number of months elapsed, and the total number of sales and months in the past period read out And unifying a total loss cost of months another elapsed months direction, a seventh calculation means for storing saved in a matrix in the data storage unit of the ninth,
Read the accumulated loss cost by sales year for the maker's warranty and extended warranty for the goods to be predicted from the ninth data storage unit, and the lapse of the manufacturer's warranty and extended warranty Calculate the cumulative loss cost for each sales year in the past and the past of the extended warranty only by subtracting the cumulative loss cost of the manufacturer's warranty from the cumulative loss cost in the total and the total of the past. And an eighth calculating means for storing and saving the accumulated loss cost in the past and only past extended warranty periods in the tenth data storage unit corresponding to the sales year and month.

  Note that “loss cost” used in the present specification means “fault repair cost that occurs during one month for one article”. In other words, during the manufacturer warranty period, the failure repair costs related to the failure repair performed based on the manufacturer warranty contract are totaled for each sales year and every elapsed month, and this is already passed for every sales year and every elapsed month. Divided by the number of sales. In addition, during the extended warranty period, breakdown repair costs related to breakdown repairs made based on the extended warranty contract are totaled for each sales month and number of months passed, and this total has elapsed for each sales year and month elapsed. Divided by the number of contracts. In addition, due to the contractual difference between the manufacturer warranty contract and the extended warranty contract, the warranty coverage for failure repair is naturally different, but it should be as homogeneous as possible. Therefore, when the warranty scope of the manufacturer warranty contract covers the entire scope, but the warranty scope of the extended warranty contract covers only a specific part (for example, the warranty scope includes the special warranty scope and the general warranty scope). If the manufacturer warranty contract is 5 years for the special warranty scope, 3 years for the general warranty scope, and the extended warranty contract consists of the general warranty scope for 2 years after the manufacturer warranty), the warranty of the extended warranty contract Out of range is excluded from the repair cost during the manufacturer warranty period. Naturally, the converse is also conceivable, but since the influence is small, it is assumed that a separate frame calculation is performed, and in the present application, it is premised on a common guarantee range.

  It calculates the loss cost by sales month and elapsed month in the manufacturer warranty period, and calculates the loss cost by sales year and elapsed month in the extended warranty period. The accumulated loss cost in the direction of the number of months by sales year and by the number of months passed is calculated from the loss cost by the sales month and the number of months in the past period of the manufacturer warranty and extended warranty. Calculate the cumulative loss cost by month and by number of months. By subtracting the accumulated loss cost for the manufacturer warranty's past and unexpired periods from the calculated accumulated loss cost for the manufacturer's warranty and extended warranty's past and unexpired periods, the past and remaining periods for the extended warranty only Get the cumulative loss cost for the period. Thus, in the present invention, the loss cost is adopted as a common analysis index in order to align the bases of the manufacturer-guaranteed sales number and loss record data and the extended warranty contract number and loss record data. That is, the ordinary chain ladder method predicts the cumulative loss performance data based on the number of sales or the number of contracts. However, in the present invention, the manufacturer guarantee (of course, the loss performance data) covers all sales items. Is also very large) and the extended warranty (of course, the loss actual data is also small) that covers only some selected contracts of the goods sold can be analyzed at the same time. Alternatively, the total loss cost is predicted using the loss cost divided by the number of contracts. Furthermore, in order to enable detailed predictions, the chain ladder method, which is usually used on a yearly basis, is applied on a monthly basis. The manufacturer's warranty, which has a track record prior to the extended warranty, allows you to confirm the presence and extent of the impact of the evolution of the product itself and its components, so the product itself and its components that have not yet appeared in the extended warranty You can take into account the impact of evolution. As a result, the prediction accuracy can be dramatically improved. Moreover, it becomes possible to foresee in advance the failure risk of specific components that have not yet appeared in the extended warranty.

  The article to be guaranteed is an automobile, and the number of sales and loss record data stored and stored in the first data storage unit are classified according to the sales period in the lapse of the manufacturer warranty for every automobile to be guaranteed. The number of sales and the number of contracts and losses stored in the second data storage unit, which are loss data by sales year and by the number of months in the lapse of the manufacturer warranty for all vehicles subject to warranty The actual data shows the number of contracts by sales month for the extended warranty period for all vehicles subject to warranty and the sales month and month for the extended warranty period for all vehicles subject to warranty. Another loss record data is preferable. In this case, the manufacturer warranty may be 36 months from the date of sale, and the extended warranty may be 24 months from the end of the manufacturer warranty.

The third calculating means calculates the loss cost by sales year and elapsed month in the elapsed period of the manufacturer warranty of the article to be predicted read from the third data storage unit, C _ij (i is the sales date, j is The number of months elapsed), and c _ij (i is the sales month, j is the elapsed month), and the loss cost by sales month and elapsed month in the past period of the extended warranty of the article to be predicted read from the fourth data storage unit Number), the cumulative loss cost R _{ij in the} direction of the number of months by sales year and the number of months elapsed in the total period of manufacturer warranty and extended warranty is expressed as R _ij = ΣC _ij + Σc _ij (j = 1 to final It is preferable to include means for calculating from the calculation of (elapsed months). Loss cost with respect to _{C ij} with reference to FIGS. 7a and 7b of the embodiments described below, referring to Figure 8a and 8b embodiments will be described later loss cost _{c ij,} Figure 9a embodiment with respect to total loss cost _{R ij} later, Understanding is facilitated by referring to 9b and 9c.

The fourth calculation means calculates the total loss cost in the direction of elapsed months by sales year and elapsed month in the total elapsed period of the manufacturer warranty and extended warranty read from the fifth data storage unit, R _ij (i is Assuming that the sales date and j are the number of months elapsed), the cumulative loss cost T _ij in the elapsed month direction and sales month direction in the elapsed period and the sales month direction is represented by T _ij = ΣR _ij (i = first It is also preferable to include means for calculating from the calculation of (sales date to past last sales date). The cumulative loss cost R _ij can be easily understood by referring to FIGS. 9a, 9b and 9c of the embodiment described later, and the cumulative loss cost T _ij with reference to FIGS. 10a, 10b and 10c of the embodiment described later.

The fifth calculation means calculates the total loss cost in the elapsed month direction and the sales month direction by sales year and month in the past elapsed period read from the sixth data storage unit, and T _ij (i is the sales date) , J is the number of elapsed months), means for calculating the loss development factor F _ij for only the boundary month from the calculation of F _ij = T _{(i−1) j} / T _{(i−1) (j−1)} It is also preferable to include. For the accumulated loss cost T _ij , refer to FIGS. 10 a, 10 b and 10 c of the embodiment described later, and for the loss development factor F _ij for only the boundary month, it is easy to understand by referring to FIGS. 11 a, 11 b and 11 c of the embodiment described later. It becomes.

If the loss development factor of only the boundary month read from the seventh data storage unit is F _ij (where i is the sales month and j is the number of elapsed months), the sixth calculation means it is also preferred to include means for calculating from the calculation of (the power of all the boundaries month Los Development factor _{F kl} satisfying k ≦ i and l ≦ j) the loss Development factor _{LDF ij LDF ij} = _{ΠF kl.} Refer to FIGS. 11a, 11b, and 11c of the later-described embodiment for the loss development factor F _ij for only the boundary month, and refer to FIGS. 12a, 12b, and 12c of the later-described embodiment for all the loss development factors LDF _ij . Then, understanding becomes easy.

The seventh calculating means uses all the loss development factors of the non-elapsed periods read out from the eighth data storage section as LDF _ij (where i is the sales year and month and j is the number of elapsed months), and the fifth data storage means. If the cumulative loss cost of the final month of the earned period when viewed in another reading was sold years and SC _{i (i} sales month and year) from the sale date by and elapsed month of earned period and unearned period total in the unexpired period It is also preferable to include means for calculating the cumulative loss cost MC _{ij in the} direction of the number of elapsed months according to the number from the calculation of MC _ij = SC _i × LDF _ij . The seventh calculating means calculates the total loss cost in the direction of the number of months by sales year and month in the total elapsed time read from the fifth data storage unit, R _ij (where i is the sales date and j is the time Number of months)) and the accumulated loss cost MC _{ij in the} direction of the number of months elapsed by sales year and total number of months elapsed in the calculated period not yet elapsed, It is more preferable to use the cumulative loss cost for each sales year in the elapsed and unexpired periods of the manufacturer warranty and extended warranty. For all loss development factors LDF _ij , refer to FIGS. 12 a, 12 b and 12 c of the embodiment described later, and for the accumulated loss cost obtained by integrating the accumulated loss cost R _ij and the accumulated loss cost MC _ij , FIG. 13 a of the embodiment described later. , 13b and 13c make it easy to understand.

  According to the present invention, the manufacturer's warranty, which has a track record prior to the extended warranty, can be confirmed whether there is an impact of the evolution of the article itself or its component parts, and the extent, and has not yet appeared in the extended warranty. The influence of the evolution of the article itself and its component parts can be taken into account in advance. As a result, the prediction accuracy can be dramatically improved. Further, it becomes possible to predict in advance the failure risk of a specific component that has not yet appeared in the extended warranty.

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

  FIG. 1 is a block diagram schematically showing the overall configuration of an extended guaranteed loss cost prediction system according to an embodiment of the present invention. However, this embodiment relates to an extended warranty loss cost prediction system for automobiles suitable for use in the automobile insurance or guarantee industry as a high-priced item with a manufacturer guarantee. In this embodiment, it is assumed that the manufacturer warranty is 3 years from the date of sale, and the extended warranty is 2 years from then.

  As shown in the figure, the extended warranty loss cost prediction device 10 on the server side is connected to a plurality of client-side terminals 12 via a communication network 11 such as a local area network (LAN), the Internet, or a dedicated network line. For example, the server may be a main personal computer (PC) and the client may be a terminal connected to the client via a LAN, or the server may be a head office server computer and the client may be a terminal of each branch. Of course, the extended guarantee loss cost prediction apparatus 10 may be operated independently without being connected to the network 11.

  The client-side terminal 12 includes a communication function that can be connected to the communication network 11 in addition to a computer, a keyboard, a mouse, a display, and the like for operation by the user. The client side terminal 12 may realize a user interface by a program such as a WEB browser when communicating with the server.

  The extended warranty loss cost prediction device 10 on the server side is provided with at least a communication control device 10a for controlling communication via a network, a loss cost prediction computer 10b, and a database 10c.

Although not shown, the loss cost prediction computer 10b includes a ROM that stores an operating system (OS), a CPU that executes various programs, a RAM that functions as a work area for various processes, and the like. or processes incoming and outgoing data to and from the 10a, the first data storage unit 10c ₁ in the database 10c, a second data storage unit 10c _2, the third data storage unit 10c _3, the fourth data storage unit 10c _4, the data storage unit 10c 6 of the fifth data storing unit 10c ₅ and 6 allowing it to read and write data stored stored _(see FIG. 2), executes a program stored in the ROM.

  FIG. 2 is a block diagram schematically showing a functional configuration of the loss cost prediction computer and database in the embodiment of FIG.

As can be seen, loss cost prediction computer 10b of this embodiment, sales date by the earned period of connected and manufacturer warranty to the first data storage unit 10c ₁ and the third data storage unit 10c ₃ and a first calculation means 10b ₁ for calculating the number of elapsed months (1-36 months) another loss cost, of connected and extended warranty to the second data storage unit 10c ₂ and the fourth data storage unit 10c ₄ A second calculation means 10b ₂ for calculating loss costs by sales year and month (37 to 60 months) in the elapsed period, a third data storage unit 10c ₃ , a fourth data storage unit 10c _4, and Third calculation means 10 connected to the fifth data storage unit 10c ₅ and for calculating the cumulative loss cost in the direction of elapsed months (horizontal) by sales month and by elapsed months (1 to 60 months). b ₃ , connected to the fifth data storage unit 10 c ₅ and the sixth data storage unit 10 c ₆ , in the sales year (longitudinal) direction according to the sales month and the elapsed months (1 to 60 months) It is connected to the fourth calculation means 10b ₄ for calculating the cumulative loss cost, and the sixth data storage unit 10c ₆ and the seventh data storage unit 10c ₇ , and the loss development factor of only the boundary month in the chain ladder method is calculated. and calculation means 10b ₅ of the fifth calculating is connected to a data storage unit 10c ₈ of the seventh data storage unit 10c ₇ and 8, the loss Development of all months of unexpired period total in the chain ladder method and calculating means 10b ₆ sixth calculating the factor, the fifth data storage unit 10c _5, connected to the data storage unit 10c ₈ and the ninth data storage unit 10c ₉ eighth In addition, the chain ladder method is used to calculate the cumulative loss cost by sales year and the number of months (1 to 60 months) in the unexpired period, and the cumulative loss cost for the manufacturer warranty and extended warranty for the past and unexpired periods. 7 and calculation means 10b ₇ of is connected to the data storage unit 10c ₁₀ of the data storage unit 10c ₉ and 10 of the ninth, cumulative in manufacturer's warranty, and earned period of extended warranty and unexpired period total loss cost to from by subtracting the total loss cost in earned period and unexpired period total of the manufacturer warranty, and a calculation unit 10b ₈ eighth to calculate the cumulative loss cost in earned period and unexpired period total extended warranty only Yes.

The first data storage unit 10c ₁ includes a part of the sales period and the number of months (e.g., part of the manufacturer's warranty, such as partly shown in FIG. 3a and all shown in FIG. 3b. 1 to 36 months) another sales number D _ij (the sales date is i and the elapsed month number is j), part of which is shown in FIG. 4a and all of which is shown in FIG. 4b Loss performance data L _ij by sales year and month (1 to 36 months) in the lapse of the manufacturer's warranty period (with the sales date as i and the elapsed month as j) for each vehicle Stored and stored in a matrix. However, in FIGS. 3 a, 3 b, 4 a and 4 b, the column (vertical) direction represents the sales year / month i, and the row (horizontal) direction represents the number of elapsed months j. The unit of sales number D _{ij is} a unit, and the unit of loss record data L _ij is 1,000 yen.

The second data storage unit 10c _2, partially as are its entirety shown in Figure 5b is illustrated, sales date by the number of subscribers in the earned period extended warranty d _ij in FIG. 5a (Where i is the sales date and j is the number of months), and the sales year in the past period of the extended warranty, some of which are shown in FIG. 6a and all of which are shown in FIG. 6b. Loss record data l _ij by month and by elapsed months (37 to 60 months) is stored and stored in a matrix for each vehicle. However, in FIGS. 5a, 5b, 6a and 6b, the column (vertical) direction represents the sales year / month i, and the row (horizontal) direction represents the number of elapsed months j. The unit of the contract number d _{ij is} a unit, and the unit of the loss record data l _ij is 1,000 yen.

The third data storage unit 10c _3, as in whole in Figure 7b there is shown a part in FIG. 7a is shown, manufacturer warranty earned period in sales years by and number of elapsed months ( 1 to 36 months) Another loss cost C _ij (the sales date is i and the elapsed month number is j) is stored and saved in a matrix for each vehicle. However, in FIGS. 7a and 7b, the column (vertical) direction represents the sales year / month i, and the row (horizontal) direction represents the number of elapsed months j. Note that the unit of the loss cost C _ij is 1,000 yen / unit.

The fourth data storage unit 10c _4, such as its entirety in Figure 8b there is shown a part in Figure 8a is shown, extended warranty earned period in sales years by and number of elapsed months ( 37 to 60 months) Another loss cost c _ij (the sales date is i and the elapsed month number is j) is stored and saved in a matrix. However, in FIGS. 8a and 8b, the column (vertical) direction represents the sales year / month i, and the row (horizontal) direction represents the number of elapsed months j. The unit of the loss cost c _ij is 1,000 yen / unit.

The fifth data storage unit 10c ₅ includes a total of manufacturer warranty and extended warranty, a part of which is shown in FIG. 9a and the whole of the first half and the latter half are shown in FIGS. 9b and 9c, respectively. The cumulative loss cost R _{ij in} the elapsed months (horizontal) direction for each sales year and elapsed months (1-60 months) in the total elapsed period of each (where i is the sales date and j is the elapsed month) Each vehicle is stored and saved in a matrix.

Further, the sixth data storage unit 10c ₆ of the first half and second half of the whole in Figure 10b and Figure 10c there is shown a portion in FIG. 10a is as shown respectively, elapsed months (lateral ) and cumulative further sales date (vertical) direction the cumulative loss cost of direction, sales date by and number of elapsed months (1-60 months in earned time total) another loss cost T _{ij (sales} years i, elapsed months The number j) is stored and saved in a matrix for each vehicle. However, in FIG. 9a, FIG. 9b and FIG. 9c, and FIG. 10a, FIG. 10b and FIG. 10c, the column (vertical) direction represents the sales year / month i and the row (horizontal) direction represents the number of elapsed months j. . The unit of the cumulative loss cost T _ij is 1,000 yen / unit.

The seventh data storage unit 10c ₇ further shows only a part of the boundary month as shown in FIG. 11a, part of which is shown in FIG. 11b and FIG. 11c, respectively. The loss development factor F _ij (where the sales date is i and the elapsed month number is j) is stored and saved in a matrix for each vehicle. However, in FIG. 11a, FIG. 11b, and FIG. 11c, the column (vertical) direction represents the sales year / month i, and the row (horizontal) direction represents the number of elapsed months j.

The eighth data storage unit 10c _8, further its has partially shown in Figure 12a, as the whole first and second halves are shown in FIGS. 12b and Fig. 12c, described later, non All loss development factors LDF _ij (where i is the sales date and j is the elapsed month) are stored and stored in a matrix for each vehicle. However, in FIG. 11a, FIG. 11b, and FIG. 11c, the column (vertical) direction represents the sales year / month i, and the row (horizontal) direction represents the number of elapsed months j.

The ninth data storage unit 10c ₉ also includes a manufacturer's warranty and extension, part of which is shown in FIG. 13a and all of the first half and second half are shown in FIGS. 13b and 13c, respectively. Cumulative loss costs by sales year and total number of months (1 to 60 months) in the elapsed and non-elapsed periods of the warranty are stored and stored in a matrix for each vehicle. However, in FIG. 13a, FIG. 13b and FIG. 13c, the column (vertical) direction represents the sales date, and the row (horizontal) direction represents the number of elapsed months. The unit of accumulated loss cost is 1,000 yen / unit.

The tenth data storage unit 10c ₁₀ stores and saves the accumulated loss cost by sales year and month and the number of elapsed months in the non-elapsed period of only the extended warranty as shown in FIG.

  FIGS. 15 to 18 are flowcharts schematically showing a part of the program of the loss cost prediction computer 10b. Hereinafter, processing contents of the loss cost prediction computer 10b will be described with reference to these drawings.

First, as shown in FIG. 15, the first calculation means 10b ₁ is stored and stored in a matrix in the first data storage unit 10c ₁ , and is shown in FIGS. 3a and 3b and FIGS. 4a and 4b. Data, the number of sales by sales year D _{ij in} the past period of the manufacturer's warranty for the subject car and the loss by sales month and elapsed month (1 to 36 months) in the manufacturer's warranty The actual data L _ij is read (step S1), and the division of (loss actual data L _ij ) / (sales number D _ij ) is performed for each read data, so that the maker warranty has passed the sales period and the elapsed time A loss cost C _ij for each month (1 to 36 months) is obtained (step S2). The obtained loss costs C _ij (shown in FIGS. 7a and 7b) in the lapse period of the manufacturer warranty are stored and stored in a matrix in the third storage unit 10c ₃ (step S3).

On the other hand, as shown in FIG. 16, the second calculation means 10b ₂ is stored in a matrix in the second data storage unit 10c ₂ and is shown in FIGS. 5a and 5b and FIGS. 6a and 6b. Data, the number of contracts d _ij by sales month for the extended warranty period of the subject automobile and the loss results by sales month and elapsed month (37 to 60 months) for the extended warranty period The data l _ij is read (step S11), and the division of (loss actual data l _ij ) / (contract number d _ij ) is performed on each read data, so that the sales are performed according to the sales year and the elapsed month in the extended warranty period. A loss cost c _ij is obtained for each number (37 to 60 months) (step S12). The obtained loss costs c _ij (shown in FIGS. 8a and 8b) in the elapsed period of the extended warranty are stored and stored in a matrix in the fourth storage unit 10c ₄ (step S13).

Further, as shown in FIG. 17, the third calculation means 10b ₃ is stored and stored in a matrix in the third data storage unit 10c ₃ , and among the data shown in FIGS. 7a and 7b, the manufacturer and sales date by and number of elapsed months (1-36 months) another loss cost C _ij in earned time guarantees, shown in the fourth data in the storage unit 10c ₄ are stored and saved in a matrix Figures 8a and 8b The loss cost c _ij by sales year and the number of months (37 to 60 months) of the extended warranty currently being read is read (step S21), and both of the read loss costs are unified into the manufacturer warranty and extension. The loss cost is classified according to the sales year and the number of months (1 to 60 months) in the elapsed warranty period (step S22), and then the loss cost is accumulated in the direction of the number of months. And allows to calculate the cumulative loss cost R _ij manufacturer warranty and sale date by and number of elapsed months in the earned period total extended warranty (1-60 months) by the elapsed months (horizontal) direction, the fifth data The data is stored and stored in a matrix in the storage unit 10c ₅ (step S23). That is, the third calculation means 10b ₃ calculates R _ij = ΣC _ij + Σc _ij (operation for accumulating j for each i from 1 to the maximum value 60 (last elapsed months), j = 1 to 60 )I do. 9a, 9b, and 9c show the accumulated loss cost R _ij in the elapsed months (horizontal) direction calculated and stored in the fifth data storage unit 10c ₅ in this way.

Next, the fourth calculation means 10b ₄ reads the accumulated loss cost R _ij in the elapsed month (horizontal) direction stored in the fifth data storage unit 10c ₅ in a matrix, and this elapsed month (horizontal) direction. The accumulated loss cost R _ij is accumulated in the sales month direction, and the elapsed month (horizontal) direction and sales month (vertical) by sales month and elapsed month (1-60 months) in the past period. ) calculates the direction of the total loss cost _{T ij,} a matrix in store stored in the data storage unit 10c ₆ of 6 (step S24). That is, the fourth calculating means 10b ₄ accumulates the calculation of T _ij = ΣR _ij (i for each j from the first sales date (2001/01 in this example) to the maximum elapsed past sales date that is the maximum value). (Calculation). 10a, 10b, and 10c are calculated in this way, and the accumulated loss cost T _{ij in the} elapsed month (horizontal) direction and sales date (vertical) direction stored in the sixth data storage unit 10c ₆ is calculated. Is shown.

Next, a cumulative loss cost MC _ij is calculated for each sales year and elapsed month in the non-elapsed period by the chain ladder method. The chain ladder method is a method for estimating the loss cost of the non-elapsed period on the assumption that the regularity is maintained in the future by paying attention to the regularity of the transition of the loss cost appearing in the actual statistics of the elapsed period. Hereinafter, the method will be schematically described with reference to FIG.

FIG. 19 shows the accumulated loss cost R _ij (the value obtained by accumulating the loss cost in the elapsed month direction) for each sales month and every elapsed month (1 to 60 months). However, in the figure, the column (vertical) direction represents the sales date i, and the row (horizontal) direction represents the number of elapsed months j (1 to 60 months). More specifically, in FIG. 19, ◯, A to Z, and a to y are data that have already passed and have actual values, and Δ and z (shaded portions) are non-elapsed periods and have no actual values. Is necessary data.

The sales date is 2007/12 and the z of the elapsed month of April is the boundary month. To calculate the cumulative loss cost, first, (cumulative value of cumulative loss cost data a to y) / (cumulative loss cost data) (A cumulative value of A to Y) is calculated to calculate the loss development factor F _ij only for the boundary month in z, and the calculated loss development factor is the cumulative loss cost Z (= SC _i) of the last month in the past period. ) To calculate the cumulative loss cost z.

Hereinafter, a method for calculating the cumulative loss cost MC _ij not only for the boundary month but also for the entire non-elapsed period, by sales year and by the number of elapsed months, will be described in detail using this chain ladder method.

First, calculation means 10b ₅ of the fifth, FIG 10a, as shown in Figure 10b and Figure 10c, sales date by and passed the earned time total which is stored in a matrix in the data storage unit 10c ₆ of the sixth months (1-60 months) by the elapsed months (horizontal) direction and sales date (vertical) reads the direction of total loss cost T _ij, boundaries month is the first month of the unexpired period seen by sales date In the example of FIG. 19, the loss development factor F _ij is calculated only for the month of z, for example, and is stored in the seventh data storage unit 10c ₇ in a matrix (step S25). In other words, the fifth calculating means 10b ₅ uses the same elapsed time as the boundary month in the immediately preceding sales month for the total loss cost in the elapsed month direction and the sales month direction by the sales year and the elapsed month in the total elapsed period. F _ij = T _{(i−1) j} / T _{(i−1) (j−)} , which is an operation for dividing the cumulative loss cost of months by the cumulative loss cost of the elapsed months immediately before the boundary month in the immediately preceding sales month. Perform the operation of ₁₎ . Figure 11a, Figure 11b and Figure 11c show the manner calculated, seventh Los Development factor _{F ij} of the data storage unit 10c ₇ only the boundary months stored in.

Next, the sixth calculation means 10b ₆ performs the loss development factor F _ij for only the boundary month stored in the seventh data storage unit 10c ₇ in a matrix as shown in FIGS. 11a, 11b and 11c. Are calculated, all loss development factors LDF _ij for the total of the non-elapsed periods are calculated, and stored in a matrix in the eighth data storage unit 10c ₈ (step S26). In other words, the sixth calculation means 10b ₆ calculates LDF _ij = ΠF _kl (k ≦ i), which is an operation to raise the loss development factor F _ij of the boundary month within the range of the sales month and the number of months that have not passed. And the power of all boundary month loss development factors F _kl satisfying l ≦ j. 12a, 12b, and 12c show all the loss development factors LDF _ij calculated in this way and stored in the eighth data storage unit 10c ₈ for the total of the non-elapsed periods.

Next, the seventh calculating means 10b ₇ calculates the sales by month and year in the total elapsed time stored in a matrix in the fifth data storage unit 10c ₅ as shown in FIGS. 9a, 9b and 9c. The cumulative loss cost SC _i of the last month (the last month of the elapsed period as viewed by sales year) is read from the cumulative loss cost R _{ij in} the elapsed months (horizontal) direction by number of months, and FIG. 12a, FIG. 12b and FIG. as shown in 12c, earned in the eighth all Los Development factor LDF _ij read out, unexpired period by multiplying each of both unexpired period total of the data storage unit 10c ₈ stored in a matrix of Cumulative loss cost MC _ij for the period and the non-elapsed period is calculated (the sales date is i and the elapsed month number is j) (step S27). That is, the seventh calculator 10b ₇ first calculates MC _ij = SC _i × LDF _ij .

Further, the seventh calculating means 10b ₇ calculates the past period and the total loss cost MC _{ij of} the non-elapsed period calculated using the chain ladder method as described above, and FIGS. 9a, 9b and 9c. As shown in the figure, a single cumulative loss cost R _{ij in} the elapsed (monthly) direction for each sales year and elapsed month in the total elapsed period stored in a matrix in the fifth data storage unit 10c ₅ To calculate the accumulated loss cost in the elapsed months (horizontal) direction by sales year and elapsed months (1 to 60 months) in the elapsed period and the total period not including the elapsed period, and the ninth data storage The data is stored in a matrix in the unit 10c ₉ (step S28). FIG. 13a shows the cumulative loss cost by sales month and elapsed month (1 to 60 months) in the total elapsed period including the non-elapsed period stored and saved in the _ninth data storage unit 10c ₉ in this way. A part of the whole is shown in FIGS. 13b and 13c, respectively.

As described above, the fifth calculating means 10b 5 _of the sixth calculation means 10b ₆ and seventh calculation means 10b by a calculation using the chain ladder method by _7, sales date by and number of elapsed months in the earned period of ( 1 to 60 months) From the accumulated loss cost in different elapsed months (horizontal) direction and sales date (vertical) direction, first, a part thereof is shown in FIG. 11a, and the first half and the latter half thereof are shown in FIGS. 11b and 11c. As shown respectively, the loss development factor for the boundary month only is calculated, then all of the parts are shown in FIG. 12a, part of which is shown in FIGS. 12b and 12c, respectively. The loss development factor is calculated. That is, the calculation of (cumulative value of cumulative loss cost data a to y) / (cumulative value of cumulative loss cost data A to Y) shown in FIG. 19 described above calculates the loss development factor of the first month that has not passed yet for each elapsed month. The loss development factor for the second and subsequent months that have not yet elapsed is obtained by raising the loss development factor for the first month that has not passed until the number of months that have elapsed. For example, the loss development factor 2.78 for the six months elapsed in the sales date 2007/11 shown in FIG. 12a is calculated as 2.78 = 1.49 × 1.29 × 1.22 × 1.19. The After that, by multiplying these loss development factors by the cumulative loss cost of the last month, the cumulative loss cost by the sales period and the number of elapsed months in the non-elapsed period is obtained. By integrating the cumulative loss cost by number of months, as shown in FIG. 13a to FIG. 13c, the elapsed month by sales month and elapsed month (1 to 60 months) in the total of past and non-elapsed periods as shown in FIGS. The cumulative loss cost in the number (lateral) direction is required. As described above, the chain ladder method handled by the fifth calculation unit 10b ₅ , the sixth calculation unit 10b ₆ and the seventh calculation unit 10b ₇ includes the fifth data storage unit 10c ₅ and the sixth data storage unit. 10c ₆ , seventh data storage unit 10c ₇ , eighth data storage unit 10c _8, and ninth data storage unit 10c ₉ prepare matrices as shown in FIGS. 9a to 13c and FIG. Or, create and repeat the processing calculation over and over the numerical values on these matrices to complete the final matrix, and predict the numerical values for the non-elapsed period, using computer hardware resources This is a specific implementation of processing by software.

As shown in FIG. 18, the eighth calculating means 10b ₈ is stored in the ninth data storage unit 10c ₉ , a part of which is shown in FIG. 13a, and the whole first half is shown in FIGS. 13b and 13c. In the accumulated loss cost data shown in the second half and the latter half, the accumulated loss costs in the past and unexpired period totals (1 to 60 months) of the manufacturer warranty and extended warranty are read (step S31), and the read manufacturer warranty and extension are read. By subtracting the accumulated loss cost for the manufacturer's warranty for the past and unexpired period (1 to 36 months) from the accumulated loss cost for the warranty's past and unexpired period, the elapsed and only periods for the extended warranty only The accumulated loss cost for the total period (37 to 60 months) is calculated (step S32). The accumulated loss cost (shown in FIG. 14) in the total elapsed time and the total elapsed time (37 to 60 months) of only the extended warranty obtained is stored and saved in the _tenth data storage unit 10c ₁₀ (step S33). .

  As described above, according to the present embodiment, the loss cost for each sales year and the number of elapsed months in the elapsed period of the manufacturer warranty is calculated, and the loss cost for each sales year and the elapsed month in the elapsed period of the extended warranty. Is calculated. The accumulated loss cost in the direction of the number of months by sales year and by the number of months elapsed is calculated from the loss cost by the sales month and the number of months in the past period of the manufacturer warranty and extended warranty. Calculate the cumulative loss cost by month and by number of months. By subtracting the accumulated loss cost for the manufacturer warranty's past and unexpired periods from the calculated accumulated loss cost for the manufacturer's warranty and extended warranty's past and unexpired periods, the past and remaining periods for the extended warranty only Get the cumulative loss cost for the period. In this way, the loss cost is adopted as a common analysis index in order to align the bases of the manufacturer-guaranteed sales number and loss record data and the extended warranty contract number and loss record data. That is, the ordinary chain ladder method predicts the cumulative loss performance data based on the number of sales or the number of contracts. However, in the present invention, the manufacturer guarantee (of course, the loss performance data) covers all sales items. Is also very large) and the extended warranty (of course, the loss actual data is also small) that covers only some selected contracts of the goods sold can be analyzed at the same time. Alternatively, the total loss cost is predicted using the loss cost divided by the number of contracts. Furthermore, in order to enable detailed predictions, the chain ladder method, which is usually used on a yearly basis, is applied on a monthly basis. The manufacturer's warranty, which has a track record prior to the extended warranty, allows you to confirm the presence and extent of the impact of the evolution of the product itself and its components, so the product itself and its components that have not yet appeared in the extended warranty You can take into account the impact of evolution. As a result, the prediction accuracy can be dramatically improved. Further, it becomes possible to predict in advance the failure risk of a specific component that has not yet appeared in the extended warranty.

  In the above-described embodiment, the article is an automobile and is used for calculating the insurance premium of the extended guarantee insurance related to the automobile. However, the present invention is also based on the extended guarantee following the manufacturer guarantee in the article with a high manufacturer guarantee. It can be widely used in fields related to failure risk prediction. For example, it is clear that goods can be used for the calculation of failure risk guarantee fees and insurance premiums for homes and high-priced appliances.

  All the embodiments described above are illustrative of the present invention and are not intended to be limiting, and the present invention can be implemented in other various modifications and changes. Therefore, the scope of the present invention is defined only by the claims and their equivalents.

It is a block diagram which shows roughly the whole structure of the extended guarantee loss cost prediction system in one Embodiment of this invention. It is a block diagram which shows roughly the functional structure of the computer for loss cost prediction in the embodiment of FIG. 1, and a database. It is a figure which shows a part of sales number data classified by sales year and the number of elapsed months in the lapse period of the manufacturer guarantee stored in the first data storage unit. It is a figure which shows all the sales number data classified by sales year and the elapsed month in the lapse period of the manufacturer guarantee stored in the first data storage unit. It is a figure which shows a part of loss performance data according to the sales month and the elapsed months in the lapse period of the manufacturer guarantee stored in the first data storage unit. It is a figure which shows all the loss performance data according to the sales month and the elapsed months in the elapsed period of the manufacturer guarantee memorize | stored in the 1st data storage part. It is a figure which shows a part of contract number data classified by sales year and elapsed month in the already elapsed period of the extended guarantee memorize | stored in the 2nd data storage part. It is a figure which shows all of the number-of-contracts data according to the sales month and the elapsed months in the elapsed period of the extended guarantee memorize | stored in the 2nd data storage part. It is a figure which shows a part of loss actual data according to the sales month according to the past elapsed period of the extended guarantee memorize | stored in the 2nd data storage part, and the number of elapsed months. It is a figure which shows all the loss performance data according to the sales month and the elapsed months in the already elapsed period of the extended guarantee memorize | stored in the 2nd data storage part. It is a figure which shows a part of loss cost data classified by sales year and the number of elapsed months in the lapse period of the manufacturer guarantee stored in the third data storage unit. It is a figure which shows all the loss cost data classified by sales year and elapsed months in the elapsed period of the manufacturer guarantee memorize | stored in the 3rd data storage part. It is a figure which shows a part of loss cost data classified by sales year and the number of elapsed months in the elapsed period of the extended guarantee memorize | stored in the 4th data storage part. It is a figure which shows all the loss cost data classified by sales year and the number of elapsed months in the elapsed period of the extended guarantee memorize | stored in the 4th data storage part. It is a figure which shows a part of total loss cost data of the direction of the number of months by sales year and month in the total of the elapsed time stored in the fifth data storage unit. It is a figure which shows the first half of all the accumulation loss cost data of the elapsed months direction according to the sales month and the elapsed months in the total elapsed period memorize | stored in the 5th data storage part. It is a figure which shows the second half of all the total loss cost data of the direction of elapsed months for every sales year and elapsed months in the total of the elapsed periods stored in the fifth data storage unit. It is a figure which shows a part of total loss cost data of the elapsed month direction according to the sales month and the elapsed month number in the 6th data memory | storage part, and the sales month direction memorize | stored. It is a figure which shows the first half of all the accumulated loss cost data of the elapsed month direction according to the sales month and the elapsed month number in the 6th data memory | storage part, and the elapsed month direction, and a sales year direction memorize | stored. It is a figure which shows the second half of all the total loss cost data of the elapsed month direction according to the sales month and the elapsed month number in the 6th data storage part, and the elapsed month direction according to the elapsed month. It is a figure which shows a part of loss development factor data of only the boundary month memorize | stored in the 7th data storage part. It is a figure which shows the first half of all the data of the loss development factor of only the boundary month memorize | stored in the 7th data storage part. It is a figure which shows the second half of all the data of the loss development factor only of the boundary month memorize | stored in the 7th data storage part. It is a figure which shows a part of data of all the loss development factors memorize | stored in the 8th data memory | storage part. It is a figure which shows the first half of all the data of all the loss development factors memorize | stored in the 8th data memory | storage part. It is a figure which shows the second half of all the data of all the loss development factors memorize | stored in the 8th data memory | storage part. It is a figure which shows a part of accumulated loss cost data according to the sales month and the elapsed month number in the past elapsed period and the total elapsed period stored in the ninth data storage unit. It is a figure which shows the first half of all the total loss cost data classified by sales year and the number of elapsed months in the total elapsed period and the total elapsed period stored in the ninth data storage unit. It is a figure which shows the second half of all the total loss cost data according to the sales month and the elapsed month number in the total elapsed period and the non-elapsed period stored in the ninth data storage unit. Stored in the tenth data storage unit, only the manufacturer warranty and non-elapsed period and the total elapsed period, manufacturer warranty and extended warranty non-elapsed period and total elapsed period, and extended warranty only non-elapsed period and It is a figure which shows the accumulated loss cost data according to the sales month in the total elapsed period. It is a flowchart which shows roughly a part of program of the computer for loss cost prediction in embodiment of FIG. It is a flowchart which shows roughly a part of program of the computer for loss cost prediction in embodiment of FIG. It is a flowchart which shows roughly a part of program of the computer for loss cost prediction in embodiment of FIG. It is a flowchart which shows roughly a part of program of the computer for loss cost prediction in embodiment of FIG. It is a figure for demonstrating the chain ladder method.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Extended guarantee loss cost prediction apparatus 10a Communication control apparatus 10b Loss cost prediction computer 10b ₁ 1st calculation means 10b ₂ 2nd calculation means 10b ₃ 3rd calculation means 10b ₄ 4th calculation means 10b ₅ 5th calculation means 10b ₆ 6th calculation means 10b ₇ 7th calculation means 10b ₈ 8th calculation means 10c Database 10c ₁ 1st data storage part 10c ₂ 2nd data storage part 10c ₃ 3rd data storage part 10c _{4 4th} 4 of the data storage unit 10c ₅ the fifth data storage unit 10c ₆ of the sixth data storage unit 10c ₇ seventh data storage unit 10c ₈ eighth data storage unit 10c ₉ ninth data storage unit 10c ₁₀ tenth of the Data storage unit 11 Communication network 12 Client-side terminal

Claims (18)

A computer for loss cost prediction;
Connected to the computer for loss cost prediction, and stores the number of sales by sales date in the manufacturer's warranty for all items for each manufacturer's warranty in the form of a matrix corresponding to the sales date. In addition to saving, the loss history data for each article by the sales year and the number of months in the lapse of the manufacturer warranty for every article is stored and saved in a matrix form corresponding to the sales year and the number of months. 1 data storage unit;
About the extended warranty that covers only a part of the contract of goods The number of contracts by sales month in the elapsed period of the extended warranty for every article to be guaranteed is stored and stored in a matrix form corresponding to the sales date , It is configured to store and save loss data by sales year and elapsed month for all items subject to warranty in a matrix form corresponding to the sales year and elapsed month. A second data storage unit;
A third data storage unit;
A fourth data storage unit;
A fifth data storage unit;
A sixth data storage unit;
A seventh data storage unit;
An eighth data storage unit;
A ninth data storage unit;
A tenth data storage unit,
The loss cost prediction computer is:
From the first data storage unit, the number of sales by sales date in the elapsed period corresponding to the sales date and the number of elapsed months of the article to be predicted, and the sales date and progress in the elapsed period of the article to be predicted By reading the loss actual data by number of months and dividing the read loss actual data by the number of sales read, the loss cost by sales year and month by the elapsed time of the manufacturer warranty is calculated, First calculation means for storing and saving the calculated loss cost in a matrix in the third data storage unit in correspondence with the sales year and the number of months elapsed;
From the second data storage unit, the number of contracts by sales month in the elapsed period corresponding to the sales date and the number of elapsed months of the article to be predicted, and the sales year and month in the elapsed period of the article to be predicted and By reading the loss actual data by the number of elapsed months and dividing the read loss actual data by the number of contracts read out, the loss cost by sales year and elapsed month in the elapsed period of the extended warranty is calculated, Second calculation means for storing and saving the calculated loss cost in a matrix in the fourth data storage unit in correspondence with the sales date and the number of months elapsed;
From the third data storage unit, the loss cost for each sales year and the number of elapsed months in the elapsed period of the manufacturer warranty of the article to be predicted is read and the item to be predicted is read from the fourth data storage unit. Read the loss cost by sales month and number of months in the past period of the extended warranty, unify the loss cost related to the read manufacturer warranty and the lost cost related to the extended warranty, and add up the total of the manufacturer warranty and extended warranty. Calculate the cumulative loss cost in the elapsed months direction by sales month and elapsed months in the elapsed period, and calculate the cumulative loss cost in the elapsed months direction by sales month and elapsed months in the calculated elapsed period. And a third calculation means for storing and storing in a matrix in the fifth data storage unit corresponding to the number of months elapsed,
From the fifth data storage unit, the cumulative loss cost in the elapsed months by sales year and the elapsed months in the elapsed period is read, and in the sales month direction by the sales month and the elapsed months in the elapsed period. The cumulative loss cost is calculated, and the cumulative loss cost in the elapsed month direction and the sales month direction by sales year and elapsed month in the elapsed period obtained by the calculation is calculated according to the sales month and elapsed month. 4th calculation means which memorize | stores and preserve | saves it at matrix form at the data storage part of these,
From the sixth data storage unit, the cumulative loss cost in the elapsed months and the sales month direction by the sales month and the elapsed months in the elapsed period is read, and the first month of the non-elapsed period is determined by sales year and month. As a boundary month, the loss development method for only the boundary month is calculated by dividing the cumulative loss cost of the same number of elapsed months as the boundary month in the previous sales month by the cumulative loss cost of the elapsed months immediately before the boundary month in the previous sales month. Fifth calculation means for calculating a loss factor and calculating and storing the loss development factor for only the boundary month in a matrix in the seventh data storage unit corresponding to the sales year and the elapsed month number; ,
Read out the loss development factor of the boundary month only from the seventh data storage unit, and increase the power of the loss development factor of the boundary month within the range of the sales month and the number of months that have not passed. Sixth calculation means for calculating all the loss development factors for the total period and storing all the loss development factors for the non-elapsed periods obtained in the calculation in a matrix form in the eighth data storage unit. When,
Reading the cumulative loss cost of the last month of the elapsed period from the fifth data storage unit by sales year and month, and reading all the loss development factors for the total non-elapsed period from the eighth data storage unit, Multiply the accumulated loss cost for the last month of the elapsed period by the sales month and the loss development factor for each elapsed month, respectively. Calculate the accumulated loss cost in the direction of elapsed months by number of months, read out the accumulated loss cost in the direction of elapsed months by sales year and month in the total elapsed period from the fifth data storage unit, and calculate The accumulated loss cost in the direction of the number of months elapsed and the number of months elapsed since the sales period and the number of months elapsed in the elapsed period and the period not yet elapsed And unifying a total loss cost of kicking sold years by and number of elapsed months another elapsed months direction, a seventh calculation means for storing saved in a matrix in the data storage unit of the ninth,
From the ninth data storage unit, the accumulated loss cost for each sales year in the elapsed period of the manufacturer's warranty and extended warranty for the article to be predicted and the total period of unexpired period is read. By subtracting the accumulated loss cost of the manufacturer's warranty from the accumulated loss cost for the elapsed period and the accumulated loss cost for the unexpired period, the accumulated loss cost for the extended warranty only, and the total loss cost for each sales year for the accumulated period, And an eighth calculation means for storing and storing the calculated total loss cost in the past and only past extended warranty periods in the tenth data storage unit in correspondence with sales year and month. Extended warranty loss cost prediction device.   The article to be guaranteed is a car, and the sales number and loss record data stored and stored in the first data storage unit are sales years in the lapse of manufacturer warranty for all cars to be guaranteed Contracts stored in the second data storage unit, which are loss data by sales year and by number of months in the past period of the manufacturer warranty for all automobiles subject to warranty and the number of sales by month The number and loss results data are based on the number of contracts by sales month for the extended warranty period for all vehicles covered by the warranty and the sales year for the extended warranty period for all vehicles covered by the warranty and The extended guaranteed loss cost prediction apparatus according to claim 1, wherein the data is loss result data for each number of elapsed months.   3. The extended warranty loss cost prediction apparatus according to claim 2, wherein the manufacturer warranty is 36 months from the date of sale, and the extended warranty is 24 months from the end of the manufacturer warranty. The third calculation means calculates a loss cost for each sales year and every elapsed month in the manufacturer warranty of the article to be predicted read out from the third data storage unit by C _ij (i is the sales date, j is the number of months elapsed), and c _ij (i is the year of sales, j) the loss cost by sales month and month of the extended warranty period of the article to be predicted read from the fourth data storage unit. When the number of elapsed months), manufacturer's warranty, and the total loss cost _{R ij} sales years by and number of elapsed months another elapsed months direction in earned period total extended warranty _{R ij = ΣC ij + Σc ij} (j = 1 The extended guaranteed loss cost prediction apparatus according to claim 1, further comprising: a calculation unit that calculates from a calculation of (last elapsed month number). The fourth calculation means calculates the total loss cost in the direction of elapsed months by sales year and elapsed month in the elapsed period of the total manufacturer warranty and extended warranty read from the fifth data storage unit, R _ij ( If i is the sales month and j is the number of months elapsed), the cumulative loss cost T _ij in the elapsed month direction and the sales month direction in the elapsed period and the sales month direction is represented by T _ij = ΣR _ij (ii The extended warranty loss cost prediction apparatus according to claim 1, further comprising means for calculating from a calculation of (= first sales date to past last sales date). The fifth calculating means calculates the total loss cost in the elapsed month direction and the sales month direction by sales year and month in the elapsed period read from the sixth data storage unit, and T _ij (i is the sales (Year, month, j is the number of elapsed months), the loss development factor F _ij for only the boundary month is calculated from the calculation of F _ij = T _{(i−1) j} / T _{(i−1) (j−1).} The extended guarantee loss cost prediction apparatus according to claim 1, further comprising: means. When the loss development factor of only the boundary month read from the seventh data storage unit is F _ij (where i is the sales month and j is the number of elapsed months), the sixth calculation means And means for calculating all loss development factors LDF _ij from the calculation of LDF _ij = ΠF _kl (a power of all boundary month loss development factors F _kl satisfying k ≦ i and l ≦ j). The extended guarantee loss cost prediction apparatus according to any one of claims 1 to 6. The seventh calculation means sets all the loss development factors read from the eighth data storage unit to the total undeveloped period as LDF _ij (where i is the sales month and j is the number of elapsed months), the fifth When viewed from the data storage means by the read selling date cumulative loss cost of the final month of earned time and SC i _(i is sold years), earned period and unexpired period total sales date by the unexpired period The unit according to any one of claims 1 to 7, further comprising means for calculating a cumulative loss cost MC _{ij in the} direction of elapsed months for each elapsed month from the calculation of MC _ij = SC _i × LDF _ij . Extended warranty loss cost prediction device. Said seventh calculation means, said fifth cumulative loss cost R _{ij (i} sales years by and number of elapsed months another elapsed months direction of earned time total read out from the data storage unit of the sales date, j And the accumulated loss cost MC _{ij in} the elapsed month direction by sales year and elapsed month in the calculated non-elapsed period. 9. The extended warranty loss cost prediction apparatus according to claim 8, wherein the estimated loss cost is a cumulative loss cost for each sales year in the total and non-elapsed periods of the manufacturer's warranty and the extended warranty. A client-side terminal, and a server-side extended guaranteed loss cost prediction device connected to the client-side terminal via a communication network,
The extended warranty loss cost prediction device
A computer for loss cost prediction;
Connected to the computer for loss cost prediction, and stores the number of sales by sales date in the manufacturer's warranty for all items for each manufacturer's warranty in the form of a matrix corresponding to the sales date. In addition to saving, the loss history data for each article by the sales year and the number of months in the lapse of the manufacturer warranty for every article is stored and saved in a matrix form corresponding to the sales year and the number of months. 1 data storage unit;
About the extended warranty that covers only a part of the contract of goods The number of contracts by sales month in the elapsed period of the extended warranty for every article to be guaranteed is stored and stored in a matrix form corresponding to the sales date , It is configured to store and save loss data by sales year and elapsed month for all items subject to warranty in a matrix form corresponding to the sales year and elapsed month. A second data storage unit;
A third data storage unit;
A fourth data storage unit;
A fifth data storage unit;
A sixth data storage unit;
A seventh data storage unit;
An eighth data storage unit;
A ninth data storage unit;
A tenth data storage unit,
The loss cost prediction computer is:
From the first data storage unit, the number of sales by sales date in the elapsed period corresponding to the sales date and the number of elapsed months of the article to be predicted, and the sales date and progress in the elapsed period of the article to be predicted By reading the loss actual data by number of months and dividing the read loss actual data by the number of sales read, the loss cost by sales year and month by the elapsed time of the manufacturer warranty is calculated, First calculation means for storing and saving the calculated loss cost in a matrix in the third data storage unit in correspondence with the sales year and the number of months elapsed;
From the second data storage unit, the number of contracts by sales month in the elapsed period corresponding to the sales date and the number of elapsed months of the article to be predicted, and the sales year and month in the elapsed period of the article to be predicted and By reading the loss actual data by the number of elapsed months and dividing the read loss actual data by the number of contracts read out, the loss cost by sales year and elapsed month in the elapsed period of the extended warranty is calculated, Second calculation means for storing and saving the calculated loss cost in a matrix in the fourth data storage unit in correspondence with the sales date and the number of months elapsed;
From the third data storage unit, the loss cost for each sales year and the number of elapsed months in the elapsed period of the manufacturer warranty of the article to be predicted is read and the item to be predicted is read from the fourth data storage unit. Read the loss cost by sales month and number of months in the past period of the extended warranty, unify the loss cost related to the read manufacturer warranty and the lost cost related to the extended warranty, and add up the total of the manufacturer warranty and extended warranty. Calculate the cumulative loss cost in the elapsed months direction by sales month and elapsed months in the elapsed period, and calculate the cumulative loss cost in the elapsed months direction by sales month and elapsed months in the calculated elapsed period. And a third calculation means for storing and storing in a matrix in the fifth data storage unit corresponding to the number of months elapsed,
From the fifth data storage unit, the cumulative loss cost in the elapsed months by sales year and the elapsed months in the elapsed period is read, and in the sales month direction by the sales month and the elapsed months in the elapsed period. The cumulative loss cost is calculated, and the cumulative loss cost in the elapsed month direction and the sales month direction by sales year and elapsed month in the elapsed period obtained by the calculation is calculated according to the sales month and elapsed month. 4th calculation means which memorize | stores and preserve | saves it at matrix form at the data storage part of these,
From the sixth data storage unit, the cumulative loss cost in the elapsed months and the sales month direction by the sales month and the elapsed months in the elapsed period is read, and the first month of the non-elapsed period is determined by sales year and month. As a boundary month, the loss development method for only the boundary month is calculated by dividing the cumulative loss cost of the same number of elapsed months as the boundary month in the previous sales month by the cumulative loss cost of the elapsed months immediately before the boundary month in the previous sales month. Fifth calculation means for calculating a loss factor and calculating and storing the loss development factor for only the boundary month in a matrix in the seventh data storage unit corresponding to the sales year and the elapsed month number; ,
Read out the loss development factor of the boundary month only from the seventh data storage unit, and increase the power of the loss development factor of the boundary month within the range of the sales month and the number of months that have not passed. Sixth calculation means for calculating all the loss development factors for the total period and storing all the loss development factors for the non-elapsed periods obtained in the calculation in a matrix form in the eighth data storage unit. When,
Reading the cumulative loss cost of the last month of the elapsed period from the fifth data storage unit by sales year and month, and reading all the loss development factors for the total non-elapsed period from the eighth data storage unit, Multiply the accumulated loss cost for the last month of the elapsed period by the sales month and the loss development factor for each elapsed month, respectively. Calculate the accumulated loss cost in the direction of elapsed months by number of months, read out the accumulated loss cost in the direction of elapsed months by sales year and month in the total elapsed period from the fifth data storage unit, and calculate The accumulated loss cost in the direction of the number of months elapsed and the number of months elapsed since the sales period and the number of months elapsed in the elapsed period and the period not yet elapsed And unifying a total loss cost of kicking sold years by and number of elapsed months another elapsed months direction, a seventh calculation means for storing saved in a matrix in the data storage unit of the ninth,
From the ninth data storage unit, the accumulated loss cost for each sales year in the elapsed period of the manufacturer's warranty and extended warranty for the article to be predicted and the total period of unexpired period is read. By subtracting the accumulated loss cost of the manufacturer's warranty from the accumulated loss cost for the elapsed period and the accumulated loss cost for the unexpired period, the accumulated loss cost for the extended warranty only, and the total loss cost for each sales year for the accumulated period, And an eighth calculation means for storing and storing the calculated total loss cost in the past and only past extended warranty periods in the tenth data storage unit in correspondence with sales year and month. Extended warranty loss cost prediction system.   The article to be guaranteed is a car, and the sales number and loss record data stored and stored in the first data storage unit are sales years in the lapse of manufacturer warranty for all cars to be guaranteed Contracts stored in the second data storage unit, which are loss data by sales year and by number of months in the past period of the manufacturer warranty for all automobiles subject to warranty and the number of sales by month The number and loss results data are based on the number of contracts by sales month for the extended warranty period for all vehicles covered by the warranty and the sales year for the extended warranty period for all vehicles covered by the warranty and The extended guaranteed loss cost prediction system according to claim 10, which is loss actual data by elapsed months.   12. The extended warranty loss cost prediction system according to claim 11, wherein the manufacturer warranty is 36 months from the date of sale, and the extended warranty is 24 months from the end of the manufacturer warranty. The third calculation means calculates a loss cost for each sales year and every elapsed month in the manufacturer warranty of the article to be predicted read out from the third data storage unit by C _ij (i is the sales date, j elapsed months), the fourth data sales date by and the number of elapsed months another loss cost c _{ij (i} in earned period extended warranty storage unit article to be predicted read from the sale date, j Is the elapsed month), the cumulative loss cost R _ij in the elapsed months for each sales year and elapsed month in the total period of the manufacturer warranty and extended warranty is calculated as R _ij = ΣC _ij + Σc _ij (j = 1 The extended guaranteed loss cost prediction system according to any one of claims 10 to 12, further comprising means for calculating from a calculation of (... last elapsed months). The fourth calculation means calculates the total loss cost in the direction of elapsed months by sales year and elapsed month in the elapsed period of the total manufacturer warranty and extended warranty read from the fifth data storage unit, R _ij ( If i is the sales month and j is the number of months elapsed), the cumulative loss cost T _ij in the elapsed month direction and the sales month direction in the elapsed period and the sales month direction is represented by T _ij = ΣR _ij (ii The extended warranty loss cost prediction system according to any one of claims 10 to 13, further comprising means for calculating from a calculation of (= first sales date to past last sales date). The fifth calculating means calculates the total loss cost in the elapsed month direction and the sales month direction by sales year and month in the elapsed period read from the sixth data storage unit, and T _ij (i is the sales years, j is when the number of elapsed months), and calculates the loss Development factor _{F ij} of only the boundary month from the calculation of _{F ij = T (i-1} ) j / T (i-1) (j-1) 15. The extended guaranteed loss cost prediction system according to claim 10, further comprising: means. When the loss development factor of only the boundary month read from the seventh data storage unit is F _ij (where i is the sales month and j is the number of elapsed months), the sixth calculation means And means for calculating all loss development factors LDF _ij from the calculation of LDF _ij = ΠF _kl (a power of all boundary month loss development factors F _kl satisfying k ≦ i and l ≦ j). The extended guarantee loss cost prediction system according to any one of claims 10 to 15. The seventh calculation means sets all the loss development factors read from the eighth data storage unit to the total undeveloped period as LDF _ij (where i is the sales month and j is the number of elapsed months), the fifth If the cumulative loss cost for the last month of the past period is SC _i (i is the year of sale) by the sales year read from the data storage means, the past period in the non-elapsed period and the total sales year / month in the non-elapsed period The unit according to any one of claims 10 to 16, further comprising means for calculating a cumulative loss cost MC _ij in the elapsed month direction for each elapsed month from the calculation of MC _ij = SC _i × LDF _ij . Extended warranty loss cost prediction system. The seventh calculation means calculates the total loss cost in the direction of the number of months of sales and the number of months elapsed in the total elapsed time read from the fifth data storage unit, R _ij (i is the year of sales, j And the accumulated loss cost MC _{ij in} the elapsed month direction by sales year and elapsed month in the calculated non-elapsed period. The extended warranty loss cost prediction system according to claim 17, wherein the loss warranty cost is a cumulative loss cost for each sales year in the past and non-elapsed periods of the manufacturer warranty and the extended warranty for the goods to be processed.

Images