WO2008035312A2

WO2008035312A2 - A system for calculating debt service capacity of a loan applicant and a method therefor

Info

Publication number: WO2008035312A2
Application number: PCT/IB2007/053841
Authority: WO
Inventors: Stephen Ankunam Quaye
Original assignee: Stephen Ankunam Quaye
Priority date: 2006-09-22
Filing date: 2007-09-21
Publication date: 2008-03-27
Also published as: ZA200901984B; WO2008035312A3

Abstract

A method and system of calculating debt service capacity of a loan applicant includes capturing, indexing, reclassifying and grouping transaction data and profile data of the loan applicant. Next, capturing a loan structure, risk factors, financial variables and reconciliation rules and finally using the aforementioned and loan applicants' data and cash flow projections to calculate loan applicants' debt service capacity with time value of money calculations.

Description

A SYSTEM FOR CALCULATING DEBT SERVICE CAPACITY OF A LOAN APPLICANT AND A METHOD THEREFOR

BACKGROUND OF THE INVENTION

The present application relates to a system and method of calculating debt service capacity of a loan applicant using inter alia cash flow projections.

SUMMARY OF THE INVENTION

According to one example embodiment, a method of calculating debt service capacity of a loan applicant includes:

capturing, indexing, reclassifying and grouping transaction data and profile data of the loan applicant;

capturing a loan structure, risk factors, financial variables and reconciliation rules; and

using the aforementioned and loan applicants' data and cash flow projections to calculate loan applicants' debt service capacity with time value of money calculations. According to another example embodiment, a system for calculating debt service capacity of a loan applicant includes:

a memory for storing data relating to:

a loan applicant's transaction data; a loan risk structure; loan risk factors; financial variables; projected loan applicants' data; and reconciliation rules; and

a processor in connection with the memory, the processor adapted to access data stored in the memory and to:

index, reclassify and group the loan applicant's transaction data and profile data into a loan risk structure; and

use the data to calculate loan applicants' debt service capacity with time value of money calculations.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a block diagram illustrating an example methodology;

Figure 2 illustrates how a loan applicant and their financial activities are grouped; Figure 3 illustrates a risk structure reflected into an accounting and financial format;

Figure 4 illustrates the reconciliation and enhancement of data;

Figure 5 illustrates the adjustment of extrapolations for age groups;

Figure 6 illustrates adjustment factors applied to the age-group interpolated curve for each financial line;

Figure 7 illustrates the expected time to revenue interruption (ETRI) mapped to loan tenor;

Figure 8 shows Expenditure Line Discounting for Financial Factors/Variables and Debt Service Coverage Ratio (DSCR)

Figure 9 shows a loan zone;

Figure 10 shows projected cash annuities within the loan zone;

Figure 11 illustrates an example of revenue projection refinement with profile matching; and

Figure 12 illustrates an example of a system to implement the methodologies described herein.

DESCRIPTION OF EMBODIMENTS

A method of calculating debt service capacity of a loan applicant is described with reference to the accompanying drawings.

It is envisaged that the described system and method will be used by a financial institution in which case the loan applicant will be a customer or potential customer of the financial institution. As such, the terms loan applicant and customer are used interchangeably in the remainder of the document.

The method and system uses the following inputs.

Input 1 : Customer Profile Data

All information in the loan applicant's application form, including inter alia:

name

ID age

• employer profession / trade

• position at employer temporary or permanent home address work address

Input 2: Customer Transaction Data

This data set includes information pertaining to the loan applicant's financial transaction history, for example:

vendor

• place of transaction date of transaction time of transaction

• transaction location vending hardware number

• card number lnput 3: Prudential Information

This input includes all data used to make the loan computation prudent for a particular user (financial institution) and compliant in terms of the user's legislative requirements.

Four main classes of prudential information as described below are used:

i) Customer risk information risk measures derived or set attributed to the individual customer. This includes inter alia ratios for financial standing, credit ratios, liquidity and solvency ratios; time and cash discount factors, for example.

ii) Customer group (or Group) risk information including

base interest rates inflation as commonly known and used in commerce and finance for the cost of various goods and services escalation typically refering to the increase in salaries of various customer groups.

• depreciation typically referring to the cost of supply to various customer groups.

iii) Internal compliance user's credit policy. The use may have a set of policies that distinct from legislative compliance. These may be the basis for applying prudential rules to their loan issues.

iv) Legislative requirements National Credit Act Bank's Act lnput 4: Loan Structures

There are three main types of loan structures with full range time-scales and their date and time feeds. The loan types are:

1. Time-based - where the user defines a loan by a fixed time period to repay for all customers. As an illustrative example: all customers, where prudent, will receive a loan with a repayment period of one week.

2. Amount-based - where the user defines a fixed amount that can be lent to all customers. As an illustrative example all customers will receive a loan of fixed amount of R100.00 at varying times.

3. Aggregated (for joint and several lending) projected cash flow or income statement loans. A number of loan applicants may decide to borrow in a borrowing club. In which case the system will add, aggregate or cluster loan applicants the cash flows and compute their borrowing capacity with any combination of the above methodologies.

4. Or any combination of the above

Input 5: Customer Repayment Data

These inputs are the inputs including the customer repayment data for further interest rate margin pricing and re-pricing updates. This is usually obtained from the financial institution where the loan applicant may already have existing loans.

Using the above inputs the method proceeds as follows. 0. Install, Update Risk Structure; Classify and Profile Customers

The system manages the risk and return profile of the customer and classified, ranked and consolidated customer group. Figure 2 illustrates how the customer and their financial activities are grouped.

The system orders the customers and customer's financial activity according the following ranking:

1. Customers' trade or profession (occupational or operational risk). This ranking reflects both the earnings (return) and risk associated with type of job;

2. Sub-classify and rank the above, into customer earnings or income type being irregular earnings or fixed earnings. The customer may have the same trade or profession but earn a fixed revenue (from for example a company) or irregular revenue (from offering services informally to a neighbourhood);

3. Sub-classify and rank customers with earnings type further into those who make essential versus non-essential expenditure.

This system of classification is further exemplified in the following Table.

The aforementioned risk structure is reflected into an accounting and financial format as illustrated in Figure 3.

The system also captures, classifies and stores customer profile information (all the information found in customer application form) for further computation steps. 1. Capture and Index Customer Transaction Data

The system captures any and all customer transactions from any combination of the following electronic sources:

• user data base;

» live data feeds, or;

• real-time from point of sale devices at vendors.

Each transaction item is indexed to enable accurate computation of customer's future earnings and expenditure trends per transaction items as one would normally account for with generally accepted account principles (GAAP) as exemplified in the table below.

An additional system of indexation relates to time-based indexation. The following indexes are used to tag earnings and expenditure items dynamic behaviour through time and projections:

• cyclical transactions - for example, expenses incurred on a daily basis to and from work; or food expenses during the morning, lunch and dinner.

• seasonal transactions - for example, expenses that are driven by seasons such as electricity or coal expenditure during winter months.

• incremental transactions - for example, transactions that are mainly influenced by customer's age. • adverse incidental transactions - for example, hospitalisation.

Profile data input as Input 1 above is used to qualify revenue and expenditure in further computational steps.

For example, revenue transactions are further qualified with customer profile data such as:

• years of employment

• attendance

• employer

• position in employer

and some expenditure line items maybe are further qualified with:

• place of employment β place of abode

• place of purchase β religion

• etc.

These user defined correlations between customer profile data and financial/account events are used to enhance the quality of predictions.

2. Group Risk Ranked Customers with Corresponding Indexed Transaction Data

The system maps the risk ranked customers according to step 0 with their corresponding transaction data as per the above indexation system in step 1. 3. Reconcile and Enhance Data

Next the data from step 2 is reconciled and enhanced.

With reference to Figure 4, the system uses the datasets of:

1. customer profile

2. customer transaction data, and;

3. reliable and public customer data

to reconcile and enhance the quality of data required to run further computational process.

The following examples indicate the principles applied and coded rules (intelligence) for data reconciliation and enhancement:

Example 1 :

Reconciliation and Enhancement of Trade/Profession Data Field Provided in Customer Profile

The customer trade or profession data field is reconciled and enhanced by cross-referring it with the indexed and classified revenue item from customer transaction data. The revenue data will typically itemise the employer and hence be used as a more reliable reflection of the customer's data field. The system overrides the customer input data in the customer profile with the trade deduced from the revenue transaction item and this logic.

Example 2:

Reconciliation and Enhancement of Customer Revenue Type (fixed or irregular) The system has coded logic that measures and analyses the periodicity between revenue events and cross-refers it with the source of revenue or customer employment. This analysis is used to re-classify the customer revenue type as either fixed (for example an permanent employer) or irregular (for example, as temporary employer).

Example 3:

There may be the requirement to accurately account and reconcile for the customer's liabilities and overall debt position, over and above customers representation in their application form. The system has coded logic to read expenditure items from known banks, lenders or financial service providers and correctly list customer's liabilities. This is used as the basis of a crosscheck for customer liability register.

Example 4:

The system has a coded logic to cross-reference and reconcile the customer's stated ID with names from other reliable and public sources of ID data.

4. Capture/Derive Risk Projections

The system captures the financial institutions views on various economic and financial factors and variables such as inter alia interest rate forward curves, inflation projections, money supply and salary escalation which will be used for further computation.

5. Select Loan Structure

The system accepts four main types of loan structures as the basis for loan computations.

The user may chose to retail: 1. A time-based loan structure (one which has set and fixed loan tenor period for a particular time-scale as the basis of loan computation);

2. A loan amount-based loan structure (one which has set and fixed loan amount as the basis of loan computations);

3. A clustered (or joint & several) loan structure (one which aggregates customers' financial projections as the basis for loan computations);

4. Islamic banking structures (which do not account for the charging of interest on capital, normally found in the Western world).

6. Convert Compliance and Legislative Requirements into Coded Prudential Rules

The system has user's internal compliance requirements and legislative requirements which are translated into code or logic which informs the issue of final loans to be disbursed to customer.

Computational Routine Paths 7A and 7B

The following computational paths 7A and 7B refer to time value of money (TVM) computations with prudential adjustments.

Path 7A computes the consolidated customer group view for debt service capacity and path 7B computes the individual customer view for debt service capacity.

Path 7A (computation for group debt service capacity)

7A.1 Construct Customer Group Financial Statement The system takes customer transactions and re-classifies from step 3 into their accounting format according to generally accepted accounting practices (GAAP). It automatically takes client transaction data and states them in the consolidated income statement (I/S) (on a realised cash basis) and the balance sheet statements (B/S). The group cash flow (ClF) statement per customer group is reconciled from the I/S and B/S according to GAAP.

7A.2 Project Financial Statements by Extrapolation

The system takes the customer group financial statements and projects them using a number of simple and reasonable curve-fitting or trend-line techniques. This extrapolation process is not driven by any underlying financial assumptions or theory. This is performed for the expected life time of the customer. Extrapolation is particularly important for the income and cash flow statements.

7A.3 Refine Projections from 7A.2 with Age-Group Interpolation

The customer population is used and classified into age groups. This trend-line is then used to refine financial extrapolations. This is illustrated in Figure 5.

7A.4 Refine Projections from 7A.3 with Group Factors and Variables

The forward views from step 4 are used and the adjustment factors are applied to the age-group interpolated curve for each financial line (earnings and expenditure etc). For example, the transport expenditure line item will be typically adjusted or weighted for fuel inflation. This is illustrated in Figure 6.

7A.5 Compute Group Prudent Cash Annuities Tenor or Group Loan Tenor The system measures the expected time to revenue interruption (ETRI) and maps it to loan tenor or projected cash annuity tenor via a formula. This formula makes use of a user defined loan tenor coverage ratio (LTCR).

The system measures the periodicity between revenue events and computes user defined statistical (average, mean, mode, standard deviation etc) time period for the customer to reach temporary ( as opposed to terminal) loss of revenue.

With reference to Figure 7 the ETRI becomes the time basis for lending computations.

ETRI is the time the customer group takes to get to no revenue, for a chosen time-scale.

LTCR is a user-defined prudential factor to be applied to ETRI

LT is calculated with the following formula:

LT = ETRI / LTCR (ETRI divided by LTCR)

There are other variants of ETRI used for other sub-groups.

The system measures other variants of cash flow interruption used in the appropriate computation of other customer sub-groups, for example:

1. Expected time to free cash flow interruption (ETFCFI) used to calculate LT in the irregular revenue sub-groups;

2. Expected time to net income interruption (ETNII) used to calculate LT in the irregular revenue sub-groups; or 3. Expected time to essential expenditure interruption (ETEEI) used to calculate LT in a customer group in very difficult financial situation.

7A.6 Compute Loan Zone & Prudent Cash Annuities

With reference to Figure 8 the system further discounts the expenditure lines after 7A.4 with a debt service coverage ratio factor (DSCR).

This DSCR-adjusted expenditure curve with the LT vertical line forms the boundaries of the loan zone. The loan zone is a maximum theoretical boundary or zone defined within a time and money envelope.

With reference to Figure 9 below, the loan zone which is the lending envelope or shape within A, B, C and D. This is the maximum amount of future cash annuities the customer group could potentially borrow.

Prudent cash annuities are the projected cash annuities that square-in the loan zone. The projected cash annuities (PCA) are driven by the selected loan structure and time-scale to issue the loan.

With reference to Figure 10 and depending on the selected loan structure and time-scale; the system searches to fit the largest rectangle (Y Rand high by 5 times X Days wide) within the loan zone ABCD read in conjunction with Figure 9.

This becomes the basis for steps used to calculate group debt service capacity (GDSC).

7A.7 Compute Group DSC (GDSC)

With reference to Figure 10 as an illustrative example, GDSC is calculated as the: • Present value of the group projected cash annuities (PCA₁, PCA₂, PCA₃ , PCA₄, PCA₅,) of Y Rand high;

• Over the period of 5 times X Days or 5Xdays wide; and

• User defined interest rate forward curve over the period 5X Days.

In financial or mathematical notation:

GDSC = PV (J(X_1-5), 5X, series PCA_1-5)

or

PV (J(X)₁ X_1-5, PCA_1-5) = (PCA₁Z(Hi₁)¹) + (PCA₂/(1+i₂)²) + (PCA₃/(1+i₃)³) + (PCV(I +i₄)⁴) + (PCA₅/(1+i₅)⁵)

7A.8 Reconcile GDSC with Customer Group Balance Sheet (Solvency) Prudential

The system reconciles GDSC with the customer group's balance sheet or solvency (debt to net worth) position.

The system reads financial institution's prudential settings for debt (liabilities) to equity (net worth or net asset value). This is derived from transactional data and profile data as the sum of customer groups assets and liabilities pro-rated down to meet this solvency restriction on the GDSC calculation.

If the GDSC exceeds the user-defined debt to equity ratio (D:E), the system pro-rates the GDSC down until it meets the use-set D:E.

Path 7B 1 -10 (computation for individual debt service capacity (DSC))

7B.1 Construct Customer Financial Statement The system takes customer transactions and re-classifies from step 3 into their accounting format according to generally accepted accounting practices (GAAP). It automatically takes client transaction data and states them in the consolidated income statement (I/S) (on a realised cash basis) and the balance sheet statements (B/S). The group cash flow (C/F) statement per customer group is reconciled from the I/S and B/Saccording to GAAP.

7B.2 Project Financial Statements by Extrapolation

The system takes the customer financial statements and projects them using number of simple and reasonable curve-fitting or trend-line techniques. This extrapolation process is not driven by any underlying financial assumptions or principle. The extrapolation is performed for the expected life time of the customer. Extrapolation is particularly important for the income and cash flow statements.

7B.3 Refine Projections with Age-Group Interpolation

The following step uses customer population, classifies it into age groups and uses this trend-line or curve to refine financial extrapolations in Figure 5.

7B.4 Apply Group Factors and Variables

The system takes the user's forward views of economic variables and financial factors from step 4 and applies the adjustment factors to the age- group interpolated curve for each financial line (earnings and expenditure etc). With reference to Figure 6., for example, the transport expenditure line item will be typically adjusted or weighted for fuel inflation.

7B.5 Refine Projection with Customer Peer Profile Matching The system uses a customer peer comparison from the customer profile data base to make more accurate predictions and projections.

For example, the system uses customer's position in employment to enhance revenue projections.

The system reads the applying customer's age, trade, employer and employment position searches for a customer peer with the same trade, employer and employment but a year older. The older customer peer becomes the future projection bench mark for the applying customers.

With reference to Figure 11 the system uses customers of same or similar profiles to project loan applicant envisaged revenue or expenditure.

7B.6 Compute Customer Prudent Cash Annuities Tenor or Loan Tenor

This step in the system measures the expected time to revenue interruption (ETRI) and maps it to loan tenor or projected cash annuity tenor via a formula. This formula has user-defined factor known as loan tenor coverage ratio (LTCR).

The system measures the periodicity between revenue events and computes user-defined statistical (average, mean, mode, standard deviation etc) time period for a customer to reach temporary (as opposed to terminal) loss of revenue or ETRI.

LTCR is a user-defined prudential factor to be applied to ETRI LT is calculated with the following formula:

LT = ETRI / LTCR (ETRI divided by LTCR)

There are other variants of ETRI used for other sub-groups.

The system also measures other variants of time to cash flow interruption for their respective customer sub-groups, for example:

2. expected time to net income interruption (ETNII) used to calculate LT in the irregular revenue sub-groups, or;

3. expected time to essential expenditure interruption (ETEEI) used to calculate LT in a customer group in very difficult financial situation.

With reference to Figure 8, this step further discounts the group factors and variables with the DSCR.

This double discount for group factor and variables and DSCR are to generate a loan (for example, transport expenditure to generate a car loan) to form a loan zone.

7B.7 & 8 Compute Customer Loan Zone & Prudent Cash Annuities

This step further discounts the expenditure lines after 7A.5 with the DSCR as in Figure 8.

The loan zone is formed by the boundary outlined by the line in Figure 8 with the vertical line making the ETRI the shape formed by points ABCD in Figure 9. This loan zone is a maximum theoretical boundary where one could prudently square-in a series of projected cash annuities as the basis for present value or discounted cash flow calculations for a given time-scale.

Prudent cash annuities are the projected cash annuities that could within this loan zone. The projected cash annuities (PCA) are driven by the loan structure selection and time-scale to make the loan.

With reference to Figures 9 & 10 and depending on the selected loan structure and time-scale; the system searches to fit the largest PCA rectangle (Y Rand high by 5 times X Days wide) within the loan zone ABCD in Figure 9.

This becomes the basis for steps used to calculate customer debt service capacity (CDSC) as illustrated in Figure 10.

7B.9 Compute Customer DSC (CDSC)

CDSC is calculated as the:

• present value of the group projected cash annuities (PCA₁, PCA₂, PCA₃ , PCA₄, PCA₅,) of Y Rand high;

• over the period of 5 times X Days or 5XDays; and

• using interest rate forward curve over the period 5X Days.

In financial or mathematical notation:

GDSC = PV (J(X_L5), 5X, series PCA_1-5)

or

PV (i(X), X_1-5, PCA_1-5) (PCA₁Z(H-J₁)¹) + (PCA₂/(1+i₂)²) + (PCA₃Z(I-H₃)³) + (PCA₄ZCRi₄)⁴) + (PCA₅/(1+i₅)⁵)

7B.8 Reconcile CDSC with Customer Balance Sheet (Solvency) Prudential

The system reconciles CDSC with user policy on individual solvency positions.

The system reads user's settings for debt (liabilities) to equity (net-worth or net asset value) which is derived from transactional data and profile date for the sum of customer groups assets and liabilities; and reconciles this restriction to the CDSC calculation.

If the CDSC exceeds the user defined debt to equity ratio (D:E), the system pro-rates the CDSC down until it meets the user set D: E.

8.1 Reconcile the Sum of CDSC with GDSC

This reconciliation step guarantees that the demand or issue of creditZloans does not exceed the prudent supply of creditZloan as computed at GDSC level.

If the ∑ CDSC > GDSC the system prorates ∑ CDSC down, so that:

∑ CDSC ≤ GDSC.

8.2 Apply Prudential Rules to CDSC

There are further user prudential requirements to lending. The system allows for the user to input prudential rules after translating them into into code. 9.1 Monitor and Calculate the Sum of Customer Arrears and Write-Offs

The system takes feeds from user on customer arrears and write-offs as the basis for full cost recovery of these losses via re-pricing these future loans.

9.2 Re-Price Interest Margin

The system takes the sum in step 9.1 and divides it by the GDSC computed base retail interest rate. This marginal interest rate is added the base retail interest and used to price future loans.

A system for implementing the above methodologies is schematically illustrated in Figure 12.

The system includes a memory 10 for storing data. The memory may be in the form of a database. The data stored in an example embodiment relates to: a loan applicant's transaction data; a loan risk structure; loan risk factors; financial variables; projected loan applicants' data; and reconciliation rules; and

A processor 12 is in connection with the memory 10. The processor 12 is adapted to access data stored in the memory 10 and to index, reclassify and group the loan applicant's transaction data and profile data into a loan risk structure and to use the data to calculate loan applicants' debt service capacity with time value of money calculations.

The processor may include modules which are implemented by a machine- readable medium embodying instructions which, when executed by a machine, cause the machine to perform any of the methods described above.

It will be appreciated that embodiments of the present invention are not limited to such architecture, and could equally well find application in a distributed, or peer-to-peer, architecture system. Thus the modules could be located on one or more servers operated by one or more institutions.

Claims

CLAIMS:

1. A method of calculating debt service capacity of a loan applicant includes:

using the aforementioned and loan applicants' data and cash flow projections to calculate loan applicants' debt service capacity with time value of money calculations.

2. A method according to claim 1 wherein the profile data includes at least some of:

name;

ID; age; employer; profession or trade; position at employer; temporary or permanent employment; home address; and work address.

3. A method according to claim 1 wherein the transaction data includes financial transaction history data.

4. A method according to claim 3 wherein the transaction history data includes at least some of: vendor; place of transaction; date of transaction; time of transaction; transaction location; vending hardware number; and card number.

5. A method according to claim 1 wherein the loan structure is one or more of a time-based loan structure, an amount-based loan structure and an aggregated projected cash flow or income statement loan structure.

6. A system for calculating debt service capacity of a loan applicant includes:

a memory for storing data relating to: