WO2018074620A1 - Mortgage value prediction assessment system - Google Patents

Abstract

The present invention relates to a mortgage value prediction assessment system. The mortgage value prediction assessment system comprises: a data collection module for collecting market price fluctuation data of collateral; a historical volatility calculation module for calculating a remaining period from a current point of time to a maturity point of time to set the remaining period as a fluctuation calculation period, and then, by using the market price fluctuation data of the collateral collected by the data collection module, calculating historical volatility of the market price fluctuation data during a period that is retroactive by the fluctuation calculation period from the current point of time; and a mortgage value predictive value calculation module for calculating a mortgage value predictive value using the historical volatility calculated by the historical volatility calculation module and current value of the collateral.

Description

Collateral Value Prediction Evaluation System

The present invention relates to a collateral value prediction evaluation system, comprising: a data collection module (120) for collecting quote data of collateral; and using the quote data of the collateral collected by the data collection module (120), After calculating the remaining period from the current point to the expiration point T0 and setting it as the variation calculation period P (S1), the historical volatility of the current price variation data during the period retroactively by the variation calculation period P at the present time ( historical volatility calculation module 130 for calculating v (t)), the historical volatility v (t) calculated in the historical volatility calculation module 130 and the present value of the collateral p ( a collateral value prediction value calculating module 140 for calculating a collateral value prediction value CATV (t) using t)) (S3); It relates to the collateral value prediction evaluation system 100 configured to include.

There are many collaterals derived from various contractual relationships between individuals or between financial entities such as individuals and banks, but the most widely used common method is to set up collateral with real estate as an object. A common example is when a bank establishes a mortgage on a real estate in order to get a big loan, in which case the bank, in principle, fails to repay the loan in the time frame agreed upon, as a mortgage lender or mortgage lender. The real estate will be auctioned off and the winnings will be used to cover the bonds.

However, even if the creditor acquires a security right, such as a mortgage, the price of the real estate secured before the debt payment falls below the amount of the bond, or special circumstances arise such that the creditor does not know it. You can wear

Therefore, the secured creditor should be sensitive to changes in the factors that may affect the value of the security property, and it is necessary to anticipate the changes in the value of the security property due to such factors and prepare for injustice. For the prediction of the security value of such real estate, there are a number of prior inventions, including Patent Literature 1, "System and Method for Risk Prediction and Notification of Real Estate Secured Assets" (Republic of Korea Patent Publication No. 10-2006-0033976). It was.

On the other hand, with the recent development of network and IT technology, various financial transactions, which were previously performed offline, are being implemented online. As a result, various real estate (manufacturing machines, precious metals, etc.) in addition to the real estate mortgages that have been widely used have been implemented. Mortgage lending by financial institutions or peer-to-peer is also active.

In the case of such a property, as shown in FIG. 1, the variation in the value of the collateral value due to the change in the market price (or value) may be small, as shown in FIG. 1, and as shown in FIG. 2. In some cases, the value of collateral value may fluctuate due to changes in market prices (or values). In addition, as shown in FIG. 3, depreciation may be applied to gradually lower the value of collateral in the passage of time. This change in collateral value is particularly pronounced in the property, but in the case of real estate, it can be the same or similar.

As such, the change in the value of the collateral value due to the change in the collateral value of the real estate or real estate becomes an important variable in estimating the value of the collateral value at maturity (T0) when the loan period expires and the loan is to be recovered. That is, when the value of the collateral value is small as shown in FIG. 1, when the collateral value is set (t), when the value of the collateral (H1) is relatively high and the loan amount is determined accordingly, the collateral value (L1) at maturity (T0) May occur in a relatively low state. In this case, as shown in FIG. 1, when the value of the security value is small, the risk due to the decrease in the security value may be relatively low.

However, when the value of the collateral value is large as shown in FIG. 2, when the collateral is set (t) and the value of the collateral (H2) is relatively high, the loan amount is determined accordingly and the collateral value (L2) at maturity (T0). When is a relatively low state, unlike in the case of Figure 1 may be a greater risk due to the decrease in collateral value. Therefore, lenders need a system that can predict the value of mortgage at maturity as soon as possible and utilize it to determine loan amount as a part of investor protection measures, including volatility of collateral value according to characteristics of collateral. do.

The present invention solves the above-mentioned problems of the present invention, calculates historical volatility based on past market volatility of the set mortgage, and then provides a collateral value prediction evaluation system capable of estimating collateral value at maturity based on this. The task is to

In order to achieve the above object, the collateral value prediction evaluation system of the present invention includes a data collection module 120 for collecting data on price changes of collateral; and the market price variation of the collateral collected by the data collection module 120. Using the data, the remaining period from the present time to the expiration time T0 is calculated and set as the variation calculation period P (S1), and then the current price change during the period retroactively by the variation calculation period P at the present time. A historical volatility calculating module 130 for calculating historical volatility v (t) of data (S2); and the historical volatility v (t) calculated by the historical volatility calculating module 130 and the collateral A collateral value prediction value calculating module 140 for calculating a collateral value prediction value CATV (t) using the present value p (t) (S3); Characterized in that comprises a.

In addition, the present value of the collateral (p (t)) is characterized in that it is determined by selecting a smaller value of the current price of the collateral or the depreciation reflected value after the initial collateral setting of the collateral.

In addition, the historical variability v (t) is a retrospective of the standard deviation of the daily log return rate of the price change data during the period retroactive to the change calculation period P at the present time by the change calculation period P at the present time. It is characterized by being calculated by dividing by the root of the number of days in a period.

Further, the historical volatility v (t) is for each day of the period retroactively by the variation calculation period P, and the market price variation during the period retroactively by the variation calculation period P in the respective days. The standard deviation of the daily log return of the data is divided by the root of the number of days of the variation calculation period P to calculate historical volatility for each day, and then calculated as an average value of the historical volatility for each day.

On the other hand, if there is no price change data corresponding to the collateral, the historical volatility calculation module 130 calculates the historical volatility (v '(t)) by using the data of the price changes of similar collateral goods similar to the collateral In addition, the collateral value prediction module 140 may calculate the historical volatility v '(t) and the current value p' (t) calculated in the historical volatility calculation module 130. By using the degree of relevance (s) by using the following equation (2) to calculate the predicted value (CATV (t)) is characterized in that the calculation (S3).

[Equation 2]

CATV (t) = p '(t) * (1-v' (t)) * s

Here, the relevance (s) is a value that is determined according to the similarity between the collateral and the similar collateral goods, and is a value that is determined between 0 and 1.

On the other hand, the web operation module 110 is connected to the quote data server 200 and the financial institution server 300 through the communication network (N); It is configured to include more

 The web operation module 110 collects the price change data of the collateral through the quote data server 200 and transmits the data to the data collection module 120, and the collateral value prediction value CATV (t) is a loan value. If it is determined to be smaller (S4), if the financial institution server 300 is requested (S4a) additional collateral for the shortage by the shortage, and if the collateral value prediction value (CATV (t)) is determined to be greater than the loan value (S5) In addition, the financial institution server 300 is characterized in that the request to release the additional collateral setting (S5a).

According to the present invention, it is to provide a security value prediction evaluation system capable of calculating historical volatility based on past market volatility of a set collateral, and then predicting the collateral value at maturity corresponding to the characteristics of the collateral based on this. The advantage is that it is possible.

Figure 1: Graph showing the case where the collateral value fluctuation is small

Figure 2: Graph showing the case of large fluctuations in collateral value

Figure 3: Graph showing variation in collateral value when depreciation is applied.

4 is a graph for explaining a collateral value determination time of the collateral value prediction evaluation system according to an embodiment of the present invention.

Figure 5 is a schematic diagram of the overall configuration of the security value prediction evaluation system according to an embodiment of the present invention.

6 is a block diagram showing a configuration of a security value prediction evaluation system according to an embodiment of the present invention.

7 is a flow chart showing the operation of the security value prediction evaluation system according to an embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings, it will be described in detail the security value prediction evaluation system according to an embodiment of the present invention. First, in the drawings, the same components or parts are to be noted that as indicated by the same reference numerals as possible. In describing the present invention, detailed descriptions of related well-known functions or configurations are omitted in order not to obscure the subject matter of the present invention.

The collateral value prediction evaluation system of the present invention includes a data collection module 120 for collecting quote data of collateral as shown in FIG. 6, and the quote data of the collateral collected by the data collection module 120. Using the historical volatility calculation module 130 for calculating historical volatility v (t) (S2), and using the historical volatility v (t) and the present value of the collateral p (t). And a collateral value prediction value calculation module 140 for calculating the collateral value prediction value CATV (t) (S3). In addition, as shown in FIG. 5, it is possible to further include a web operation module 110 connected to the quote data server 200 and the financial institution server 300 through the communication network (N).

First, the data collection module 120 will be described. The data collection module 120 has a function of collecting quote data of collateral. In this case, the price change data of the collateral may be input to the data collection module 120 by a user, and the web operation module 110 may change the price of the collateral through the price data server 200. It is also possible to collect data to be delivered to the data collection module 120.

Next, the historical volatility calculation module 130 will be described. The historical volatility calculating module 130 is configured to calculate the historical volatility v (t) of the market price variation data.

In this case, it is preferable that the historical variability v (t) is calculated differently according to the period from the current time t to the expiration time T0. In other words, when the period from the current time t to the expiration time T0 is relatively short, it is more preferable to consider the retrospective changes in the past price of the relatively short period, and at the current time t, the expiration time T0. In the case where the period up to is relatively long, further consideration of retrospective past price changes can reflect the characteristics of the price changes. Therefore, as shown in FIGS. 4 and 7, the historical volatility calculation module 130 calculates the remaining period from the present point to the expiration point T0 and sets it as the variation calculation period P (S1). It is preferable to calculate (S2) the historical variability v (t) of the quote change data during the period retroactively by the change calculation period P.

This will be described in more detail with reference to FIG. 4. First, when the historical volatility v (t) is calculated at the first time point T1 in which a relatively long period (for example, three months) remains until expiration, the expiration time T0 at the first time point T1 Period P1 (in this case, three months) is set as the variation calculation period P (which is also three months), and the variation calculation period (which is three months) at the first time point T1. The historical volatility is calculated from the retrospective point. If the historical volatility v (t) is calculated at a second time point T2 where an intermediate period (e.g. two months) remains until expiration, then from the second time point T2 to maturity T0 The period P2 (in this case, two months) is set as the variation calculation period P (also two months), and retroactively for the variation calculation period (two months) at the second time point T2. Calculate historical volatility from one point. The historical volatility is calculated by setting the period P1 from the first time point T1 to the expiration time T0 as the variation calculation period P. When the historical variability v (t) is calculated at a third time point T3 where a relatively short period of time (e.g. one month) remains until expiration, from the third time point T3 to expiration time T0 The period P3 (in this case, one month) is set to the variation calculation period P (which is one month), and from the point of time retroactive to the variation calculation period (one month) at the third time point T3, Calculate historical volatility.

Meanwhile, the historical variability v (t) may be calculated in various ways that can indicate the variability of the market price variation data. In this case, the historical volatility v (t) of the daily log return rate of the quoted variation data during the period retroactively by the variation calculation period P at this time so that it can be normalized to a value between 0 and 1. The standard deviation is preferably calculated by dividing the standard deviation by the root value of the number of days (ie, the number of days of the variation calculation period P) retroactively by the variation calculation period P at the present time.

Here, the daily log return rate is the natural log ratio (ln (day price / previous price)) of the current price divided by the previous day's price for each day of the period retrospectively from the present time. it means.

The historical volatility v (t) may be calculated in another way to allow for more consideration of the average tendency of the volatility of the market value. In this case, the historical volatility v (t) is the price of the period of the period retroactively by the variation calculation period P in the respective days, for each day of the period retroactively by the variation calculation period P. It is preferable to calculate the historical volatility of each day by dividing the standard deviation of the daily log return rate of the fluctuation data by the root of the number of days of the fluctuation calculation period P, and then calculating the average value of the historical volatility of each day.

Next, the security value prediction value calculation module 140 will be described. As shown in FIGS. 6 and 7, the collateral value prediction module 140 may include the historical volatility v (t) calculated by the historical volatility calculating module 130 and the present value of the collateral p (t (t). ) To calculate the collateral value prediction value CATV (t) according to Equation 1 below (S3).

[Equation 1]

CATV (t) = p (t) * (1-v (t))

That is, the collateral value predicted value CATV (t) is calculated by multiplying the present value p (t) by the value obtained by subtracting the historical variability v (t) from 1. Therefore, even in the case of the same current price, when the historical volatility v (t) is high and the risk is high (that is, when the price of the collateral is largely changed), the security value prediction value CATV (t) is reflected. When the historical variability (v (t)) is low and the risk is low (that is, when the price of the collateral is low), the estimated value of the collateral value (CATV (t)) reflects the current price. It will have a value close to.

Also, the present value p (t) of the collateral is a smaller value of the current price of the collateral or the depreciation reflected value after the initial collateral setting of the collateral to reflect the application of depreciation to the collateral. It is preferable to determine by selecting.

On the other hand, there may be a case where there is no market price change data corresponding to the collateral according to the characteristics of the collateral (for example, a new item entered into the market). For this case, the historical volatility calculation module 130 is similar to the collateral similar collateral goods (for example, the LCD panel cutting machine having similar functions and characteristics when the collateral is a new LED panel cutting machine entering the market) The historical volatility v '(t) is calculated using the market price fluctuation data of, and the security value prediction value calculating module 140 calculates the historical volatility v' (t) calculated by the historical volatility calculating module 130. )) And calculating the collateral value prediction value CATV (t) according to Equation 2 below by applying the relevance s using the present value p '(t) of the similar collateral product (S3). desirable.

[Equation 2]

CATV (t) = p '(t) * (1-v' (t)) * s

Here, the relevance s is a value that is determined according to the similarity between the collateral and the similar collateral goods, and is a value that is determined between 0 and 1.

Meanwhile, the collateral value prediction evaluation system 100 according to an embodiment of the present invention may enable the security value prediction value CATV (t) calculated in this way to be actually reflected in the collateral operation. N) further comprises a web management module 110 connected to the quote data server 200 and the financial institution server 300 through, the web management module 110 is the collateral value as shown in FIG. When the estimated value CATV (t) is determined to be smaller than the loan amount (S4), the financial institution server 300 requests the additional collateral for a shortage by a shortage (S4a), and the collateral value predicted value CATV (t) is the loan value. If it is determined to be larger (S5), it is preferable to request (S5a) to release the additional collateral setting to the financial institution server 300.

Hereinafter, the operation of the collateral value prediction evaluation system 100 according to an embodiment of the present invention will be described, for example, when the collateral is gold.

First, when the price per 1g of gold is given as shown in Table 1 below, the daily log return for each day is displayed together.

Year Month Day	quote	Daily log yield
2016/01/04	41,080	0.010030665
2016/01/05	41,200	0.002916871
2016/01/06	41,350	0.003634165
2016/01/07	41,640	0.006988822
2016/01/08	41,900	0.006224583
2016/01/11	42,200	0.007134394
2016/01/12	42,410	0.004963962
2016/01/13	42,000	-0.009714565
2016/01/14	42,430	0.010186041
2016/01/15	42,300	-0.003068573
2016/01/18	42,540	0.005657724
2016/01/19	42,310	-0.005421345
2016/01/20	42,450	0.003303448
2016/01/21	42,530	0.001882797
2016/01/22	42,600	0.001644544
2016/01/25	42,600	0
2016/01/26	42,710	0.002578832
2016/01/27	42,900	0.004438741
2016/01/28	43,100	0.004651171
2016/01/29	43,170	0.001622812
2016/02/01	43,020	-0.003480686
2016/02/02	43,310	0.006718431
2016/02/03	43,600	0.006673595
2016/02/04	43,800	0.004576667
2016/02/05	44,140	0.007732583
2016/02/11	46,030	0.041926957
2016/02/12	48,000	0.041907653
2016/02/15	46,990	-0.021266198
2016/02/16	46,070	-0.019772834
2016/02/17	46,730	0.014224378
2016/02/18	47,020	0.006186686
2016/02/19	47,950	0.019585758
2016/02/22	47,880	-0.001460921
2016/02/23	48,000	0.00250313
2016/02/24	48,140	0.002912421
2016/02/25	48,810	0.013821778
2016/02/26	49,040	0.004701082
2016/02/29	48,700	-0.006957262
2016/03/02	48,400	-0.006179216
2016/03/03	48,630	0.004740811
2016/03/04	48,660	0.000616713
2016/03/07	48,650	-0.000205529
2016/03/08	48,870	0.004511903
2016/03/09	49,050	0.003676475
2016/03/10	48,740	-0.006340138
2016/03/11	49,100	0.007358987
2016/03/14	48,000	-0.022658024
2016/03/15	47,100	-0.01892801
2016/03/16	47,450	0.007403524
2016/03/17	47,700	0.005254873
2016/03/18	47,550	-0.003149609
2016/03/21	47,100	-0.009508788
2016/03/22	46,460	-0.013681274
2016/03/23	46,000	-0.009950331
2016/03/24	45,700	-0.006543099
2016/03/25	46,490	0.017138938
2016/03/28	45,840	-0.014080163
2016/03/29	45,800	-0.000872981
2016/03/30	46,070	0.005877888
2016/03/31	45,340	-0.015972334
2016/04/01	45,900	0.012275472
2016/04/04	45,500	-0.008752791
2016/04/05	45,700	0.004385972
2016/04/06	45,800	0.002185793
2016/04/07	45,950	0.003269758
2016/04/08	46,190	0.005209476
2016/04/11	46,360	0.003673694
2016/04/12	46,530	0.003660247
2016/04/14	46,000	-0.011455869
2016/04/15	45,850	-0.003266198
2016/04/18	46,050	0.004352564
2016/04/19	45,450	-0.013114942
2016/04/20	45,880	0.009416472
2016/04/21	45,930	0.001089206
2016/04/22	46,000	0.001522898
2016/04/25	46,000	0
2016/04/26	45,990	-0.000217415
2016/04/27	46,190	0.004339343
2016/04/28	46,180	-0.000216521
2016/04/29	46,620	0.00948283
2016/05/02	47,260	0.013634638
2016/05/03	47,030	-0.004878576
2016/05/04	47,200	0.003608197
2016/05/09	47,850	0.013677225
2016/05/10	47,400	-0.009448889
2016/05/11	47,440	0.000843526
2016/05/12	47,530	0.001895336
2016/05/13	47,690	0.003360642
2016/05/16	48,200	0.010637289
2016/05/17	48,090	-0.002284766
2016/05/18	48,230	0.002906979
2016/05/19	48,000	-0.004780223
2016/05/20	47,950	-0.00104221
2016/05/23	47,890	-0.001252087
2016/05/24	47,750	-0.002927647
2016/05/25	47,000	-0.015831465
2016/05/26	47,030	0.000638094
2016/05/27	46,630	-0.008541585
2016/05/30	46,420	-0.00451371
2016/05/31	46,620	0.004299233
2016/06/01	46,810	0.004067222

According to the above table, the historical variability (v (t)) and the collateral value prediction value (CATV (t)) at each time point shown in Figure 4 are calculated as shown in Table 2 below.

Maturity 3 months	T1	T2	T3	T0
Viewpoint	2016/03/02	2016/04/01	2016/05/02	2016/06/01
price	48,400	45,900	47,260	46,810
Period to Expiration = Variable Calculation Period (P)	3 months	2 months	1 month	0 months
Historical volatility v (t)	6.18%	7.58%	4.16%	0.00%
Relevance	100.00%	100.00%	100.00%	100.00%
depreciation	0	0	0	0
Final CATV	45,409	42,423	45,293	46,810
Loan	45,000	45,000	45,000	45,000
Difference from loan price	409	-2,577	293	1,810
Request for additional security		2,577
Extra collateral release			2,577

In this case, since the collateral value prediction value CATV (t) at the first time point T1 is determined to be greater than the loan price, it may be determined that the requested loan may be executed. On the other hand, when the collateral value prediction value CATV (t) is determined to be smaller than the loan price (in this case, -2,577 won) as in the second time point T2, the additional collateral corresponding thereto (in this case) 2,577 won) (S4a), and when it is determined that the collateral value prediction value CATV (t) is greater than the loan price (in this case, 233 won) as in the third time point T3 (S5), The request to release the additional collateral set value (in this case 2,577 won) (S5a).

In the foregoing description, optimal embodiments have been disclosed in the drawings and the specification. Although specific terms have been used herein, they are used only for the purpose of describing the present invention and are not used to limit the scope of the present invention as defined in the meaning or claims. Therefore, those skilled in the art will understand that various modifications and equivalent other embodiments are possible. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

Claims (6)

1. A data collection module 120 collecting data on price changes of the collateral;

   After calculating the remaining period from the present point to the expiration point T0 by using the quote change data of the collateral collected by the data collection module 120 and setting it as the change calculation period P (S1), the present point A historical volatility calculation module 130 for calculating a historical volatility v (t) of the market price variation data during the period retrospectively by the variation calculation period P;

   Using the historical volatility v (t) calculated by the historical volatility calculation module 130 and the present value of the collateral p (t), the collateral value prediction value CATV (t) A collateral value prediction value calculating module 140 for calculating S3; The collateral value prediction evaluation system 100 is configured to include.

   [Equation 1]

   CATV (t) = p (t) * (1-v (t))
2. The method according to claim 1,

   The present value of the collateral p (t) is

   The collateral value prediction evaluation system (100), characterized in that determined by selecting a smaller value of the current price of the collateral or the depreciation reflected value after the initial collateral setting of the collateral.
3. The method according to claim 2,

   The historical volatility v (t) is a period in which the standard deviation of the daily log return rate of the price change data is retroactively calculated by the variation calculation period P at the present time. Collateral value prediction evaluation system 100, characterized in that calculated by dividing by the root of the number of days.
4. The method according to claim 2,

   The historical volatility v (t) is the value of the quoted variation data during the periods retroactively calculated by the variation calculation period P in each of the days retroactively by the variation calculation period P. Mortgage value prediction evaluation system characterized in that by calculating the daily historical volatility by dividing the standard deviation of the daily log return by the root value of the number of days of the variation calculation period (P), 100).
5. The method according to claim 1,

   If there is no price change data corresponding to the collateral,

   The historical volatility calculation module 130 calculates historical volatility v '(t) using the data of the price variation of similar collateral goods similar to the collateral,

   The collateral value prediction module 140 may use the historical volatility v '(t) calculated by the historical volatility calculation module 130 and the present value of the similar collateral goods p' (t). Mortgage value prediction evaluation system 100, characterized in that by applying the degree of relevance (s) (S3) to calculate the collateral value prediction value (CATV (t)) according to the following equation (2).

   [Equation 2]

   CATV (t) = p '(t) * (1-v' (t)) * s

   Here, the relevance (s) is a value that is determined according to the similarity between the collateral and the similar collateral goods, and is a value that is determined between 0 and 1.
6. The method according to any one of claims 1 to 5,

   A web operation module 110 connected to the quote data server 200 and the financial institution server 300 through the communication network N; It is configured to include more

   The web operation module 110,

   Collect the price change data of the collateral through the quote data server 200 and transmit the data to the data collection module 120.

   When it is determined that the collateral value prediction value CATV (t) is smaller than the loan price (S4), the financial institution server 300 requests an additional collateral for the shortage by a shortage (S4a),

   When the collateral value prediction value CATV (t) is determined to be greater than the loan price (S5), the financial institution server 300 requests the financial institution server 300 to release the additional collateral set value (S5a). System 100.

Images