Loan Underwriting

The financial institution verifies your income, expenses, and property details in order to issue final approval on your mortgage application. The decision to pass the loan is done for this journey in two stages:

1. automatically: applicants must fulfill the requirements such as not being blacklisted, not having county court judgments, and credit history for all the applicants.
2. manually.

The journey is the following: the application reaches the automatic stage, and from there it can be either rejected or moved to the manual assessment.

Automatic Assessment

The automatic assessment filters all the applications for the blacklisted people. The assessment is done by searching for the person on the database of the bank and on the Lexis Nexis LexisNexis Legal & is a global provider of legal, regulatory and business information and analytics that help customers increase productivity, improve decision-making and outcomes, and advance the rule of law around the world. platform to get the confidence score. This connector is an integration we recommend, but other third-party systems can be integrated. The search is conducted using the full name of the applicant and/or Personal Identification Number.

The search within the bank's database is done using the following data:

• first name: mandatory for search
• last name: mandatory for search
• date of birth: optional for search
• nationality: optional for search
• country of residence: optional for search
• customer PIN: optional for search.

If by searching the database a result is found for the first name, last name, date of birth, and nationality, then the applicant fails the request and a rejection screen is displayed. If by searching the database a result is found for the customer PIN, then the applicant fails the request and a rejection screen is displayed, otherwise, the applicant is approved and the solution runs the Lexis Nexis assessment.

The search within the Lexis Nexis connector is done:

Field	Required	Type	Description
Entity unique ID	No	Text	A unique system-generated number that identifies the entity.
Entity score	No	Numeric	This field displays a confidence score between 0 and 100 that Lexis Nexis Bridger Insight XG generates to indicate how closely the list record matched the input entity information. A score of 90 indicates the match is closer than a potential match with a score of 80. When a valid score is not available, ‐1 is returned.
Entity name	No	Text	The main entity name from the list record that generated the potential match.
Best name score	No	Text	The score of the name that most closely matched the input information.
Best name	No	Text	The name that most closely matched the input information.
Best country	No	Text	The country that most closely matched the input information.
Reason listed	No	Text	The reason the country is on the list.

If by searching the database a result is found, and the <Entity Score> is >90%, then the applicant fails the request and a rejection screen is displayed, otherwise, the customer is approved and the solution runs the county court judgments assessment.

County Court Judgments

By using a connector, the solution incorporates an endpoint that calls the API to bring data from an external source. Such an external source can be the Experian connector. This connector queries the Experian database to find the information. However, the solution can accommodate any other third-party software, this is one such software integration developed by FintechOS. Depending on localization, other connectors produced by FintechOS may be used.

The data used is the status being settled/not settled:

• country court judgment date: the date when the country court Judgment was posted
• country court judgment value: the amount to pay.

The result can be:

• If the country court Judgment is not settled, then the applicant fails the request and a rejection screen is displayed
• If the country court Judgment is satisfied/If there is more than one country court judgments in the past 12 months/If there are two or more, then the system adds the total value of them and if the total value is over 1000, then the applicant fails the request and a rejection screen is displayed.

If the assessment is passed, then the Credit Bureau assessment is next.

Credit Bureau Interrogation

Also, at this stage, we will need to pull a credit report for all the applicants, no matter if they have information fetched from open banking - we will need this information in the scoring matrix.

The BC check will be mocked since we don't have a connector in this area but we will need to mock the following attributes:

For search, we will use the following attributes:

• First Name (mandatory)
• Last Name (mandatory)
• PIN (optional)
• Date of birth (mandatory)
• Current address, and postal code (mandatory).

To get the result, the values for the metadata below are mocked:

• Credit type
• Credit is active
• Currency
• Start date
• End date
• Loan amount
• Monthly instalment
• Principal due
• Max DPD in the past 3 months (can be from 0 to 999, or it can be no history)
• Max DPD in the past 6 months (can be from 0 to 999, or it can be no history)
• Max DPD in the past 12 months (can be from 0 to 999, or it can be no history)
• Max DPD in the past 24 months (can be from 0 to 999, or it can be no history).

Second Automatic Underwriting Process

The information provided in the full approval step of both applicants is validated here:

• Proof of ID: this validation is already done since the ID and selfie are confirmed
• Proof of address: Because the applicant uploads the documents, this validation is done manually by an underwriter.
• Proof of income: If all applicants chose to offer the income via open banking, the validation is done here.
  NOTE  The open banking integration registers the net income.
• Proof of expenses: If all applicants chose to offer the expenses via open banking, the validation is done here.

DTI

The DTI is calculated here:

• Predicted monthly income amount: used as income for all applicants
• Predicted monthly essential expenditure: used as expenses for all applicants.

Then, the system calculates the income weight for income and expenses based on the data from the open banking integration.

The system adds the incomes for both applicants to calculate the total income available for the current loan application and also adds the expenses. This results in a single value for the income at an application level and one value for the expanses.

Then, the system calculates the existing DTI at an application level.

application expenses / application income = current DTI

Then, the system runs the knock-out criteria:

• if the current DTI is bigger than 50% , the loan request is rejected
• If the applicant has income or expenses via open banking, the system looks within Credit Kudos's response to check the gambling expenses. If gambling expenditure ratio to income is more than 5%, then the loan application is rejected.

Scoring Matrix for Both Applicants

The system calculates the scoring matrix and the maximum DTI permitted for the applicant by combining FICO score and applicant score. To establish max DTI, the system uses the long-term DTI for the calculated performance category (A,B,C,D).

In case there are 2 applicants, the system uses the performance category graded the lowest between the 2 applicants.

Using the maximum permitted DTI by the matrix for the loan application, the system calculates the requested offer using the PMT formula to establish if the application is eligible for the requested loan. For the maximum offer, the system uses the maximum loan period available for he selected product.

In case there is only one applicant and they are not eligible for the requested loan (requested amount/requested loan period), the system calculates the maximum offer (maximum amount/maximum period) for the selected product in the DIP section using the PV formula.

The Result

Passed Automatic Underwriting

The loan application passes to the manual underwriting. If the application doesn't have a requester offer available, but has a maximum offer available in the refusal screen and they choose to accept that offer, the application continues to the manual underwriting stage.

Failed Automatic Underwriting

The application doesn't pass the automatic underwriting process because of KYC validations or the country court judgements validations, the applicant cannot continue the flow and .

If there is a single applicant and the application failed because of DTI or scoring criteria, the system displays the Unfortunately, you don't seem to qualify for the requested loan page that allows the applicant to add a co-applicant and try again. This screen will also show the maximum offer available.

If there are already multiple applicants and they are not eligible for the requester amount, the Unfortunately, we are not able to approve your request for a mortgage page is displayed.