Bank X-Ray is a decision-support solution that aggregates and analyses banking transactions to monitor cash activity, detect signs of economic distress and unearth suspicious behaviour.

How does Bank X-Ray work?

To generate a Bank X-Ray report, Digital Connect users need to send information about their clients’ banking transactions on the platform. This can be done by uploading their bank statements (in True PDF format). Learn more

We then use optical character recognition to analyse the content of the bank statements and classify transactions per type and date. That is the first step that enables us to reconstitute the cash of a company, assess its management and spot irregularities or economic issues.

All this information is gathered in one visual Bank X-Ray report about your client’s financial health and behaviour with relevant decision support indicators. Let’s deep dive into the report.

Overview of the Bank X-Ray report

Untitled

On each Bank X-Ray report, you will find:

The Bank X-Ray Score based on the aggregation of the score assigned to every data points analysed (debt stress, overdraft, payment behaviour…) by Bank X-Ray. The overall score is then translated to an alphanumeric system going from A to E.
The Analysis Details section, which shows the number of bank accounts analysed, the period taken into account and the extraction status for each bank statement.
The Balance section, which gives an overview of the banking movements over a pre-set period with its trend evolution.
The top Counterparties of the company in amount and number of transactions.
A series of charts to deep dive into your customer’s Payment Behaviour regarding salaries, social security, taxes and financial providers.
An Alerts section to highlight transactions containing pre-set keywords.
A Transactions tab with all the transactions extracted from the bank statements, sorted chronologically and the possibility to filter by category, name or amount.

How can you use Bank X-Ray?

You can use Bank X-Ray for three different purposes:

Risk analysis: complete your assessment with a real-time view of the company cash management and detect automatically signs of economical distress within the company.
Fraud detection: leverage the information contained in banking transactions to unveil inconsistencies that could reveal misbehaviours or fraud attempts.
Portfolio monitoring: follow the evolution of your existing clients over time by updating their Bank X-Ray report on a regular basis.

Bank X-Ray Scorecard

The first section of the Bank X-Ray report gives an overall assessment of the company’s cash health using a letter-based scoring system from A to E (from overall positive to overall negative).

If the data sanity, which refers to the quality of the extracted data, is lower than 70%, then no rating will be assigned to the Bank X-Ray report.

How are Bank X-Ray scores calculated?

The overall score is based on the aggregation of scores assigned to 9 indicators analyzed by Bank X-Ray. These scores range from A to E, with each indicator having its own weight that impacts the overall Bank X-Ray score. Below, you can find the indicators listed in descending order of weight on the overall score:

Indicators	Description	Scoring method	Scoring rules
Alerts frequency	Ratio of transactions with alerts to total transactions	Benchmark on cases from companies belonging to the same sector and country	• A: The indicator is better than 90% of benchmark
• B: The indicator is better than 70% and under 90% of benchmark
• C: The indicator is better than 50% and under 70% of benchmark
• D: The indicator is better than 20% and under 50% of benchmark
• E: The indicator is worse than 20% of benchmark
Alert-related amount	Significant amounts on transactions with alerts	Benchmark on cases from companies belonging to the same sector and country	• A: The indicator is better than 90% of benchmark
• B: The indicator is better than 70% and under 90% of benchmark
• C: The indicator is better than 50% and under 70% of benchmark
• D: The indicator is better than 20% and under 50% of benchmark
• E: The indicator is worse than 20% of benchmark
Overdraft	Number of days with negative balance	Benchmark on cases from companies with similar revenue levels
• A: The indicator is better than 90% of benchmark
• B: The indicator is better than 70% and under 90% of benchmark
• C: The indicator is better than 50% and under 70% of benchmark
• D: The indicator is better than 20% and under 50% of benchmark
• E: The indicator is worse than 20% of benchmark
Social security	Regularity of monthly social security payments	Hardcoded rule	• A: Social Security payments are made regularly
• C: Social Security payments are missing due dates less than 40% of cases.
• E: Social Security payments are missing due dates in more than 40% of cases and there is no payment during 3 months.
Tax payment	Regularity of tax payments	Hardcoded rule	• A: Tax payments are made regularly
• C: Tax payments are missing due dates less than 40% of cases.
• E: Tax payments are missing due dates in more than 40% of cases and there is no payment during 3 months.
Payroll Management	Regularity of monthly salary payments	Hardcoded rule	• A: Salaries are paid regularly
• C: Salary payments are missing in less than 40% of cases.
• E: Salary payments are missing in more than 40% of cases and there is no payment during 3 months.
Debt stress	Ratio of existing debt to cash generated	Benchmark on cases from companies belonging to the same sector and country	• A: The indicator is better than 90% of benchmark
• B: The indicator is better than 70% and under 90% of benchmark
• C: The indicator is better than 50% and under 70% of benchmark
• D: The indicator is better than 20% and under 50% of benchmark
• E: The indicator is worse than 20% of benchmark
Whole-number amounts	Ratio of transactions with whole-number amounts to total transactions	Benchmark on cases from companies belonging to the same sector and country	• A: The indicator is better than 90% of benchmark
• B: The indicator is better than 70% and under 90% of benchmark
• C: The indicator is better than 50% and under 70% of benchmark
• D: The indicator is better than 20% and under 50% of benchmark
• E: The indicator is worse than 20% of benchmark
Balance trend	Important deviation between opening balance and ending balance	Benchmark on cases from companies belonging to the same sector and country	• A: The indicator is better than 90% of benchmark
• B: The indicator is better than 70% and under 90% of benchmark
• C: The indicator is better than 50% and under 70% of benchmark
• D: The indicator is better than 20% and under 50% of benchmark
• E: The indicator is worse than 20% of benchmark

Bank X-Ray overall score is based on:

The score assigned to each indicator (C, D and E scores will generate alerts),
The weight of each indicator on the scoring model.

The Bank X-Ray overall score is therefore calculated by summing the weight of all indicators with alerts. In the table below, we share a general guideline to assess the score:

Bank X-Ray overall score	Rules
A	The sum is between 0 and 4
B	The sum is between 5 and 8
C	The sum is between 9 and 12
D	The sum is equal to 13
E	The sum is equal or above 14

How is the benchmark built?

The scores are computed by checking which percentile the borrower falls into. The population that makes up these percentiles is split according to country and industry (identified from NACE code) - so a company is only ever compared with companies with similar characteristics. The population used for the Italian benchmark consists of 1000+ companies after all filters are applied. We have selected only companies that gather these conditions:

Bank X-Ray duration over 30 days
Data Sanity above 70%
At least 70% of the uploaded statements were used to generate the Bank X-Ray

These filters ensure that the data used is healthy and well-representative of the population, and that we can use it as a benchmark.

Most of the values computed to compare with the benchmark are ratios, with the exception of Overdraft which measures the number of days with negative balance. In that sense, Overdraft is the only flag that also takes into account the revenue of the company, and therefore the company is benchmarked against companies of similar revenue. Revenue groups are not split by country nor industry, which would drastically reduce the number of companies within a certain benchmark and consequently reduce the reliability of the population. When computing the percentiles, we have observed that distinct revenue groups (from different countries and industries) behave similarly across most bank flags, and hence opted to only use revenue for the Overdraft criterion.

← Previous

Concepts: Doc X-Ray