TWI702556B - A system and a method for evaluating probability-weighted probability of default under multiple economic scenarios - Google Patents

Abstract

A system for evaluating probability-weighted probability of default (PD) under multiple economic scenarios is provided. The system has an input unit, a storage unit and a processing unit. The input unit receives a plurality of index information which are corresponding to an economic index. The storage unit stores the index information. The processing unit performs a classification procedure so as to obtain an index change type and an environment type. The processing unit counts a number of index information which belongs to the same index change type and the environment type, and obtains a weighting corresponding to occurrence probability of each of the environment type in each of index change type, and obtains the probability-weighted PD according to the weighting and a probability of default which are corresponds to the index change type and the environment type.

Description

System and method for evaluating probability-weighted default probability under multiple economic situations

This disclosure relates to a digital risk control technology, and in particular, relates to a system and method for evaluating probability-weighted default probability under multiple economic scenarios.

According to IFRS 9 (International Financial Reporting Standard 9; IFRS 9), it proposes that the expected credit loss should be determined by evaluating various possible outcomes to determine the unbiased and probability-weighted average amount. However, IFRS 9 does not explain how to evaluate various possible outcomes to obtain an unbiased and probability-weighted average amount.

In order to comply with my country's compliance with International Financial Reporting Standards No. 9 on January 1, 107, how to measure expected credit losses should be determined by assessing various possible results and determining the unbiased and probability-weighted average amount is a problem faced by those skilled in the art.

The present invention provides a system and method for evaluating probability-weighted default probabilities under multiple economic scenarios, and analyzes the probabilities of various possible future results based on the overall economic indicator information in the past history, thereby obtaining unbiased and probability-weighted default probabilities, and matching default losses Rate (Loss Given Default, LGD) and exposure at default (Exposure at Default, EAD) to calculate the amount of default loss.

According to an embodiment of the present disclosure, a system for evaluating probability-weighted default probability under multiple economic scenarios has an input unit, a storage unit, and a processing unit. The input unit receives multiple index information corresponding to the economic index. The storage unit stores indicator information. The processing unit is connected to the input unit and the storage unit. The processing unit performs a classification procedure on each indicator information to obtain the index change type and the situation type corresponding to each indicator information. The processing unit also calculates the number of the same index change type and situation type in the indicator information according to the index change type and the situation type. The processing unit also calculates the number of each situation type in the same index change type according to the index change type. The probability is used as the weight, and the probability weighted default probability is obtained according to the weight and the default probability corresponding to each index change type and situation type.

The method for evaluating probability-weighted default probability under multiple economic scenarios in an embodiment of the present disclosure has the steps of: receiving multiple indicator information corresponding to economic indicators; performing a classification process on each of the indicator information to obtain each of the indicator information. The corresponding index change type and situation type; according to the index change type and the situation type, the number of the same index change type and the same situation type in the index information is calculated, so as to obtain the same index change according to the number The probability of occurrence of the situation type in each of the types is used as the weight; and the type according to each index change Type and the corresponding weight and default probability of the situation type, obtain the probability weighted default probability.

Based on the above, this disclosure provides a system and method for evaluating probability-weighted default probability under multiple economic scenarios, so as to analyze the probabilities that various general economic scenarios may occur in the future based on historical overall economic indicator information, so as to use this probability and corresponding The default probabilities of different overall economic scenarios are then obtained to obtain an unbiased and probability-weighted default probability, combined with Loss Given Default (LGD) and Exposure at Default (EAD) to calculate the expected credit loss amount.

In order to make the above-mentioned features and advantages of the present invention more comprehensible, the following specific embodiments are described in detail in conjunction with the accompanying drawings.

100: A system for evaluating probability-weighted probability of default in a variety of economic scenarios

110: Input unit

120: storage unit

130: processing unit

S210~S250, S310~S360: steps

FIG. 1 shows a schematic diagram of a system for evaluating probability-weighted default probability in various economic scenarios according to an embodiment of the present disclosure.

FIG. 2 shows a flowchart of a method for evaluating a probability-weighted default probability executed by a system for evaluating a probability-weighted default probability in a variety of economic scenarios according to an embodiment of the present disclosure.

FIG. 3 shows a flowchart of a classification procedure executed by a processing unit according to an embodiment of the disclosure.

This disclosure provides a system and method for evaluating probability-weighted default probabilities in a variety of economic scenarios. It is used to look at the general economic situation in the future from a historical perspective through historical general economic indicators and their corresponding time points. This to present Some historical overall economic indicators are quantitatively analyzed to generate the probabilities that various general economic scenarios may occur in the future, and based on this, to obtain unbiased and probability-weighted risk scenarios in various scenarios.

Please refer to FIG. 1. FIG. 1 is a schematic diagram of a system for evaluating probability-weighted default probability under multiple economic scenarios according to an embodiment of the present disclosure. The system 100 for evaluating probability-weighted default probability under multiple economic scenarios has an input unit 110, a storage unit 120, and a processing unit 130.

The input unit 110 is used to input various types of data to the system 100 for evaluating probability-weighted default probabilities in a variety of economic scenarios. In particular, the user can input multiple indicator information corresponding to economic indicators through the input unit 110. For example, an economic indicator can be Gross Domestic Product (GDP), and multiple indicator information is the actual value of GDP per quarter or year. In other embodiments of the present disclosure, the economic indicator may be, for example, a consumer price index, and the multiple indicator information is a quarterly or annual consumer price index. This disclosure is not limited to the types of economic indicators.

In an embodiment of the present disclosure, if the system 100 for evaluating probability-weighted default probability under multiple economic scenarios is suitable for an environment with a network architecture, the input unit 110 is implemented with a communication chip, which can support global Mobile communication (Global System for Mobile communication, GSM), Personal Handy-phone System (PHS), Code Division Multiple Access (CDMA) system, Wideband Code Division Multiple Access (Wideband Code Division) Multiple Access, WCDMA) system, Long Term Evolution (Long Term Evolution, LTE) system, Worldwide interoperability for Microwave Access (WiMAX) system, Wireless Fidelity (Wi-Fi) system or Bluetooth signal transmission components. If the system 100 for evaluating probability-weighted default probability under multiple economic scenarios runs as a single device, the input unit 110 can be a keyboard, mouse, touch screen, touch pad and other hardware devices that can be used for input.

The storage unit 120 is used to store necessary software, data, and various kinds of codes for operating and establishing the system 100 for evaluating the probability-weighted default probability under various economic scenarios. In particular, the storage unit 120 stores index information input by the user through the input unit 110. The storage unit 120 may be any type of fixed or removable random access memory (Random Access Memory, RAM), read-only memory (Read-Only Memory, ROM), flash memory (flash memory), hard disk (Hard Disk Drive, HDD), Solid State Drive (SSD) or similar components or a combination of the above components.

The processing unit 130 is connected to the input unit 110 and the storage unit 120. The processing unit 130 may be a central processing unit (CPU), or other programmable general-purpose or special-purpose microprocessor (Microprocessor), digital signal processor (DSP), or programmable The present disclosure is not limited to the integrated controller, Application Specific Integrated Circuit (ASIC) or other similar components or a combination of the above components.

In order to allow the bank to assess all possible defaults Different scenarios and results are comprehensively and simultaneously evaluated. In the embodiments of this disclosure, the changes in historical economic indicators and market trends are evaluated to evaluate the probabilities of different scenarios, based on the probabilities of various scenarios and The probability of default assesses credit risk.

FIG. 2 shows a flowchart of a method for evaluating a probability-weighted default probability executed by a system for evaluating a probability-weighted default probability in a variety of economic scenarios according to an embodiment of the present disclosure. Hereinafter, the details of the operation of the probability-weighted default probability system and method for evaluating the probability-weighted probability of default in various economic scenarios of the present disclosure will be described with reference to FIGS. 1 and 2.

In step S210, the processing unit 130 obtains multiple index information corresponding to the economic index. It should be noted that in this embodiment, the economic indicator uses the rate of change of Gross Domestic Product (GDP) as the main indicator, and the indicator information is the rate of change of the GDP per quarter. For example, in the first quarter to the fourth quarter of the Republic of China, the rate of change of GDP was 1.54%, 2.02%, 1.94%, and 3.57% respectively. Each piece of data is the indicator information. In other words, there will be a time series relationship between these indicator information. The processing unit 130 receives the indicator information input by the user via the input unit 110, or obtains corresponding indicator information from the storage unit 120.

In step S220, the processing unit 130 performs a classification procedure on each index information to obtain the index change type and the situation type corresponding to each index information.

In the following description, the index information that the processing unit 130 is performing the classification process is called current index information. It should be noted that, when the classification procedure is executed, although the processing unit 130 classifies the current index information, the actual Meaning, it is to classify the overall situation of past indicator information that affects current indicator information and future indicator information that is affected by current indicator information. How the processing unit 130 executes the classification procedure will be detailed below.

Please refer to FIG. 3. FIG. 3 is a flowchart of a classification procedure executed by the processing unit according to an embodiment of the disclosure. In the actual economic environment, because the economic situation at different time points will have continuity due to changes in the overall environment, in this embodiment, the economic indicators of the past two consecutive quarters will be classified, and then the economic indicators at the time Analyze the future situation types, and calculate the weights of different economic situation types according to different economic situation types.

Specifically, in step S310, the processing unit 130 obtains the previous period indicator information and the previous two period indicator information of the current indicator information. Continuing from the example from the first quarter to the fourth quarter of the Republic of China, if the current indicator information is the third quarter of the Republic of China (1.94%), the previous period of indicator information is the indicator information of the second quarter of the Republic of China (2.02%), while the indicator information for the first two periods is the indicator information for the first quarter of 2004 (1.54%).

In step S320, the processing unit 130 obtains the first economic indicator change and the second economic indicator change. The first economic indicator change is the change between the current indicator information and the previous period of indicator information, and the second economic indicator change is the change between the previous period of indicator information and the previous two period of indicator information. Continuing from the above, the first economic indicator change in this embodiment is the change between the indicator information in the third quarter of 2004 and the indicator information in the second quarter of 2004, that is, 1.94%-2.02%=-0.08%. The second economic indicator change is the change between the indicator information in the second quarter of the Republic of China and the indicator information in the first quarter of the Republic of China, namely: 2.02%-1.54%=0.48%.

In step S330, the processing unit 130 judges the index change type corresponding to the current index information according to the first economic indicator change and the second economic indicator change. In an embodiment of the present disclosure, the index change types are divided into three types: the first economic index change and the second economic index change are both greater than 0, the first economic index change and the second economic index change are opposite, and Both the first economic indicator change and the second economic indicator change are less than zero. Following the previous example, for the current indicator information in the third quarter of 2004, the index change type is the opposite of the first economic indicator change and the second economic indicator change.

In step S340, the processing unit 130 also obtains the index information of the next period following the current index information to obtain the third economic index change. Following the previous example, the next period of indicator information of the current indicator information is the indicator information of the fourth quarter of 2004, which is 3.57%. The third economic indicator change is the change between the current indicator information and the next period of indicator information, that is, 3.57%-1.94%=1.63%.

In step S350, the processing unit 130 also obtains the average economic indicator change in the preset time interval according to the current indicator information and the preset time interval. In the embodiment of the present disclosure, the preset time interval is the 12th to the 24th month after the calculation of the time corresponding to the current indicator information. In other words, what the processing unit 130 obtains is the average economic index change for the entire year of the next year from the time point where the current economic index is located. Following the previous example, when the current indicator information is the third quarter of 2004, the processing unit 130 obtains the average economic indicator changes from the third quarter of the Republic of China to the second quarter of the Republic of China. For example, if the rate of change of indicator parameters between the third quarter of 2005 and the second quarter of 2006 are 2.92%, 2.71%, 1.99%, and 3.25%, The average economic index change was 2.72%.

In step S360, the processing unit 130 also judges the context type to which the current indicator information belongs based on the third economic indicator change and the average economic indicator change.

依據上述表一，當第三經濟指標變化(即：短期)屬於擴張階段，平均經濟指標變化(即：長期)也屬於擴張階段時，則判斷當前指標資訊所屬的情境類型為情境1；當第三經濟指標變化(即：短期)屬於收縮階段，平均經濟指標變化屬於持平階段時，則判斷當前指標資訊所屬的情境類型為情境8，以此類推。如何分辨第三經濟指標變化、平均經濟指標變化屬於何種階段將稍後再進行闡述。 Specifically, the third economic indicator change is the next quarter economic indicator change after the current indicator information (ie: short-term change), and the average economic indicator change is the average economic indicator change for the next year after the current indicator information (ie: mid- to long-term change) ). In other words, the processing unit 130 looks at the short-term and medium- and long-term economic situation types of the future market through the time point of the current indicator information. Take the foregoing example to illustrate, that is, under the situation where the change in the rate of change of GDP for two consecutive periods is reversed, what kind of situation will the future market economy be in the short and medium-term. In this embodiment, the situation types are divided into nine different situation types based on short-term and medium- and long-term. Short-term situation types have short-term expansion, short-term flatness, and short-term contraction. Long-term situation types have long-term expansion, long-term flatness, and long-term contraction. Therefore, please refer to the following table 1 for the nine different situation types obtained based on the short-term situation type and the long-term situation type:

According to the above table 1, when the third economic indicator change (i.e. short-term) belongs to the expansion stage and the average economic indicator change (i.e. long-term) also belongs to the expansion stage, the situation type to which the current indicator information belongs is judged as situation 1; 3. Changes in economic indicators (ie, short-term) are in the contraction phase, and when the average economic indicator changes are in a flat phase, the situation type to which the current indicator information belongs is judged as situation 8, and so on. How to distinguish the change of the third economic indicator and the stage of the change of the average economic indicator will be explained later.

In an embodiment of the present disclosure, the processing unit 130 sequentially performs a classification process on all index information to find the index change type and the context type to which each index information belongs. In other embodiments of the present disclosure, the processing unit 130 may also perform the classification procedure only based on the indicator information in a specific time interval, and the present disclosure is not limited to this.

Please return to FIG. 2, after the processing unit 130 has performed the classification procedure, in step S230, the processing unit 130 calculates the number of the same index change type and situation type in the index information according to the index change type and the situation type. Please refer to Table 2 below. Table 2 is the combined results of all index change types and situation types.

在表二中，所有指數變化類型以及情境類型的組合共有27種。處理單元130會計算屬於相同指數變化類型以及相同情境類型的指標資訊(例如：同屬於第一經濟指標變化與第二經濟指標變化皆大於0以及情境1的指標資訊)共有幾個。舉例來說，在本實施例中是以136季的指數資訊作為樣本，並且經由處理單元130進行分類與計算後，分屬於相同指數變化類型以及相同情境類型的指標資訊的數量如表二的內容。

In Table 2, there are 27 combinations of all index change types and situation types. The processing unit 130 calculates several index information that belong to the same index change type and the same situation type (for example, the index information that both belong to the first economic index change and the second economic index change are greater than 0 and the situation 1). For example, in this embodiment, the index information of 136 seasons is used as a sample, and after classification and calculation by the processing unit 130, the number of index information belonging to the same index change type and the same context type is shown in Table 2. .

In step S240, the processing unit 130 calculates the probability of occurrence of each situation type in the same index change type according to the number belonging to the same index change type as a weight. As mentioned above, the index change type is based on the index data of the past two quarters. The result of the classification of changes in information, the probability of the situation type is the result of looking into the future at the time of the current indicator information. Therefore, for the known indicator information, the sum of the different types of scenarios that may occur in the future should be 100%. Based on this, the processing unit converts the amount of indicator information of each situation type into a corresponding probability according to the index change type, and uses the probability of each situation type in the history as the weight of different situation types that may occur in the future. Please refer to Table 3. Table 3 shows the weights of all index change types and situation types.

所有指數變化類型及情境類型的權重是通過過往已知的指標資訊 進行分析，藉此以獲取在不同種類已發生的指數變化類型中，對應到未來發生不同情境類型的機率。

The weights of all index change types and context types are analyzed based on previously known indicator information to obtain the probability that different types of index changes have occurred in the future corresponding to different context types.

In step S250, the processing unit 130 obtains the probability-weighted default probability according to the weight and the default probability corresponding to each index change type and context type. In detail, different types of overall economic situations and different periods will correspond to different preset default probabilities. Therefore, under the known index information, the weight of each situation type in the index change type corresponding to the known index information can be further passed, and the weight is combined with the corresponding overall economic environment and the default probability of default in different periods. Multiplying the product, the processing unit 130 may further obtain the probability-weighted default probability.

For example, if the current reading of this example is in the fourth quarter of 106, and the indicator information in the second to fourth quarters of 106 is 3.25%, 2.48%, and 2.28%, respectively, and the processing unit 130 is in When judging the probability-weighted probability of default in the next year, it will first determine that the changes in the indicator information in the past two quarters are negative based on the indicator information, so the changes in the first economic indicator and the second economic indicator in Table 3 are both less than 0 The weighted weight in this column. Then, the processing unit 130 will further obtain the corresponding default probabilities in different context types, and obtain the probability-weighted default probability based on the product of the weighted weight and the default probability.

Not only that, the processing unit 130 can also receive the Loss Given Default (LGD) and Exposure at Default (EAD) through the input unit 110, and the processing unit 130 weights the default probability, LGD and EAD according to the probability. The default exposure generates expected credit losses (ECL). Based on this, through this system of evaluating the probability of weighting the probability of default, it is possible to evaluate each This kind of possible outcome produces an unbiased and probability-weighted amount of default loss.

The following will explain how to distinguish the change of the third economic indicator and the change of the average economic indicator belonging to the process of expansion, flatness and contraction.

In an embodiment of the present disclosure, the processing unit 130 performs statistical analysis based on changes in all historical indicator information to obtain a confidence interval with a confidence level of 95% in the changes in indicator information. If the change of the judged indicator information falls within this confidence interval, it means that the economy is flat, if it is higher than this confidence interval, it means that the economy is expanding, and if it is below this confidence interval, it means that the economy is shrinking.

In the embodiment of the present disclosure, for the short-term, that is, the third economic indicator change, the confidence interval is [-0.4%, 0.38%]. In other words, if the change of the third economic indicator falls within the range of -0.4% to 0.38%, it means that the economy is flat in the short term, and if the change of the third economic indicator is greater than 0.38%, it means that the economy is expanding in the short term. If the change of the third economic indicator is less than -0.4%, it means that the economy is contracting in the short term.

For the long-term, that is, the average economic indicator change, the confidence interval is [-0.29%, 0.18%]. In other words, if the average economic indicator change falls within the range of -0.29% to 0.18%, it means that the economy is flat in the long-term. If the average economic indicator change is greater than 0.18%, it means that the economy is expanding in the long-term. The change in economic indicators is less than -0.29%, which means that the economy is contracting in the long term.

In summary, in the system and method for assessing probability-weighted default probability provided in this disclosure, the processing unit will be used to classify historical overall economic indicators, and look forward to the future from the perspective of historical overall economic indicators Overall economic indicators. In this way, through the evaluation of the relationship between historical overall economic indicators and future overall economic indicators Estimate the probability of various general economic scenarios in the future and set corresponding weights. In addition, the processing unit also obtains the probability-weighted default probability under different scenarios and their corresponding time points based on the probabilities that may occur in various general economic scenarios in the future and their corresponding default probabilities, and looks at the overall future from a historical perspective. Economic context, to obtain the weights (probabilities) that may occur in various general economic scenarios in the future under the existing historical overall economic indicators, and use this weight and the corresponding default probability under various general economic scenarios to obtain various Unbiased and probability-weighted risk situations in the context.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention. Anyone with ordinary knowledge in the technical field can make some changes and modifications without departing from the spirit and scope of the present invention. The scope of protection of the present invention shall be determined by the scope of the attached patent application.

100‧‧‧A system for evaluating probability-weighted default probability in multiple economic scenarios

110‧‧‧Input Unit

120‧‧‧Storage Unit

130‧‧‧Processing unit

Claims (6)

A system for evaluating probability-weighted default probabilities under multiple economic scenarios, including: an input unit, receiving multiple indicator information corresponding to economic indicators, a storage unit, storing the indicator information; and a processing unit connected to the input unit and the storage The processing unit performs a classification procedure on each of the index information to obtain the index change type and the context type corresponding to each of the index information. The processing unit also calculates the index change type and the context type respectively according to the index change type and the context type. The number of indicators that belong to the same index change type and the same situation type is weighted based on the probability of each situation type occurring in the same index change type according to the number, and is based on each index change Type and the weight and default probability corresponding to the situation type to obtain the probability-weighted default probability. The processing unit also receives the loss given default (LGD) through the input unit according to the weighted default probability and the The default loss rate generates an expected credit loss rate (Expected Credit Loss, ECL). When the processing unit executes the classification procedure, the processing unit also obtains the previous period indicator information and the previous two period indicator information of the current indicator information to obtain The first economic indicator change and the second economic indicator change, the processing unit also judges the index change type corresponding to the current indicator information based on the first economic indicator change and the second economic indicator change, and the processing unit also obtains the subsequent current indicator Index information for the next period of the news, with To obtain the third economic indicator change, the processing unit also obtains the average economic indicator change in the preset time interval according to the current indicator information and the preset time interval, and the processing unit also obtains the average economic indicator change in the preset time interval according to the third economic indicator change and the average economic indicator. The indicator change determines the type of situation to which the current indicator information belongs. As described in item 1 of the scope of patent application, the system for evaluating probability-weighted default probabilities in multiple economic scenarios, where the index change types include two consecutive periods of economic indicator changes greater than zero, two consecutive periods of economic indicator changes being reversed, and two consecutive periods The economic indicator is less than at least one of zero; and the situation type includes at least one of short-term expansion, flatness, and contraction, and a combination of at least one of long-term expansion, flatness, and contraction. For example, the system for evaluating probability-weighted default probability under multiple economic scenarios as described in item 1 of the scope of patent application, wherein the preset time interval is the 12th to 24th month after the time corresponding to the current indicator information. A method for evaluating probability-weighted default probabilities under multiple economic scenarios is suitable for systems that evaluate probability-weighted default probabilities under multiple economic scenarios. The system includes a processing unit, and the method includes: receiving multiple economic indicators by the processing unit Index information; the processing unit executes a classification process on each of the index information to obtain the index change type and the situation type corresponding to each of the index information; the processing unit depends on the index change type and the situation Types are calculated for the index information that belong to the same index change type and the situation type The quantity is weighted based on the probability of each situation type in the same index change type obtained according to the quantity; the weight and default probability corresponding to each of the index change type and the situation type by the processing unit , Obtain the probability weighted default probability; and wherein, when the processing unit receives a Loss Given Default (LGD), the expected credit loss (ECL) is generated according to the weighted default probability and the LGD ; The classification procedure includes: obtaining the previous period indicator information and the previous two period indicator information of the current indicator information by the processing unit to obtain the first economic indicator change and the second economic indicator change; by the processing unit according to the The first economic indicator change and the second economic indicator change determine the type of index change corresponding to the current indicator information; the processing unit obtains the index information of the next period following the current indicator information to obtain the third economic indicator change; The processing unit obtains the average economic indicator change in the preset time interval based on the current indicator information and the preset time interval; and the processing unit determines the current economic indicator change based on the third economic indicator change and the average economic indicator change The situation type to which the indicator information belongs. As described in item 4 of the scope of patent application, the method of evaluating the probability-weighted default probability under multiple economic scenarios, where the index change types include two consecutive periods of economic indicator changes greater than zero, two consecutive periods of economic indicator changes being reversed, and two consecutive periods The economic indicator is less than at least one of zero; and The situation type includes at least one of short-term expansion, flatness, and contraction, and a combination of at least one of long-term expansion, flatness, and contraction. As described in item 4 of the scope of patent application, the method for evaluating the probability-weighted default probability under various economic scenarios, wherein the preset time interval is the 12th to 24th month after the time corresponding to the current indicator information.

Images