Credit Scoring

Summary

 

CROWDESTOR scoring is a mathematical Model designed to automatically evaluate the creditworthiness of a potential borrower applying for a funding on a CROWDESTOR platform. The Model is designed for business operating at least 2 years, minimum revenue 120’000 EUR and assets 20’000 EUR. The loan term is 3-18 months and amount from 10’000 until 200’000 EUR.

 

Model analysis is based on financial and non-financial aspects of the business. The financial evaluation takes into consideration financial health of the company, looking at its most important metrics – profitability, liquidity, leverage, etc. The second part of analysis focuses on non-financial factors such as management’s experience, personal credit history, macroeconomic and industry situation and trends.

 

Information for the assessment purposes is retrieved on-demand from 500 databases that covers more than 150 countries, 100 million legal entities and 300 million natural persons.

 

After evaluation, the Model automatically outputs one of following results - offer or rejection. Together with the result, each borrower receives an individual, tailor-made credit report. The report is automatically generated and is uniquely made for the exact borrower, on the exact date. Credit report gives an overview of borrower credit score, rating, probability of default and other important metrics.

 

The report metrics are presented by the Model to show borrower’s ability to repay the loan.  In case of a successful project placement on the platform, apart from other information regarding the project, investors can see output of the scoring Model. This gives possibility to make a balanced decision about the project before the investment.

 

Below you can find a detailed description of the Model.

 

Key facts about the Model

 

  • Both financial and non-financial factors are analysed, having equal weights in the Model (see section “Model” for further information),
  • Assessment of creditworthiness is done fully on-line, without a need to physically meet a client,
  • Model has a list of eligibility criteria for potential borrowers (see section “borrower’s eligibility criteria” below for further information),
  • Model is specifically designed for a loan secured against personal guarantee of borrower’s shareholder/board member,
  • Documents used in the analysis are bank statements, annual reports, credit reports, macroeconomic and other data,
  • Main outputs of the Model are probability of default, loss given default, expected loss, maximum loan amount, interest rate, loan term, credit score, credit rating.

 

The product

 

Borrower: legal entity

Type: business loan 

Amount: 10’000 - 200’000 EUR

Term: 3-18 months

Interest rate for investors: 12-32% p.a. (per annum)

Platform commission: starting from 3%

Interest rate for borrowers: from 15% p.a.

Payment schedule: instalment loan

Guarantee: personal guarantee

 

 

The product is targeting companies which are already operating for several years,  are at the growth stage, have stable income stream, yet are not at the level where they could get much cheaper public funding (i.e. bond listing on stock exchange)

 

 

 

Borrower’s eligibility criteria

 

  • Currently active company, i.e. not in the process of insolvency or having any other negative encumbrances,
  • Published at least 2 annual reports to the local IRS,
  • Have fixed ownership for a minimum of one year,
  • Published annual accounts for 2018,
  • Revenue in the last submitted annual report exceeds 120’000 EUR,
  • Total assets value in the last submitted annual report exceeds 20’000 EUR.

 

Data sources

 

Sources of information

 

In total, 500 databases are used to perform the scoring. Covering more than 150 countries, we have access to 100 million legal entities, 300+ million natural persons, 150+ million ultimate beneficial owners (UBOs), 450+ million ownership links, 200 thousand banks. We access and analyse potential borrower’s financials; UBOs, owners and management; company bank statements; personal credit history of company’s key personnel.

Specifically for SME scoring, we use several score providers, which specialize in different scoring approaches - qualitative, quantitative, reputational, compliance, environment; and specifically, on predicting probability of default of small-to-medium sized companies. For instance, one of such score providers is ESMA approved credit rating agency.

Some of the databases we use are strongly venture capital backed, while others are one of The Big Three credit rating agencies.

The databases we use, are also used by thousands of banks, credit facilitators, lenders, insurers, payment services and other industry players. Out of many database customers, here are few notable:  Raiffeisen Bank, Erste Sparkasse, Societe Generale, FCA, local country IRS (internal revenue services), DNB bank, London Stock Exchange, and many more.

 

Information provided by customer

 

  • Applicant’s ID/passport, that is verified by a third-party KYC solution service,
  • Filled application form, including information on borrower (for instance registration number) and possible guarantor,
  • Filled KYC and AML questionnaire,
  • Borrowers bank statement for the past 12 months either in PDF format, or via secure internet-bank connection (based on PSD2 directive),
  • General project description, purpose of the loan, minimum-maximum amounts, funds collection deadlines, project pictures, etc.,
  • Payment for the risk-assessment.

 

 

The SCORING Model

 

Analysed factors

 

Financial factors: 50% weight

 

Credit risk scoring Model analyses borrowers annual accounts for profitability, leverage, liquidity, coverage and activity aspects. For example, metrics include gross margin, debt/equity ratio, current ratio, collection days, and many more. These metrics are used as indicators of applicant’s financial health and are showcased in the borrower’s credit report and the credit report for investors. Improvements in these metrics will lead to improved credit scores, larger credit limit and lower interest rates for borrowers on CROWDESTOR platform. These factors account for 40% weight in the Model.

 

Moreover, credit risk scoring Model enriches the automatic evaluation of borrowers with bank account statement information. This is analysed to see the full picture of applicant company’s current financial position and cash-flow behaviour. Information from bank account statements is more recent than that in annual accounts and therefore, credit risk assessment is more appropriate, when analysis of these pieces of information are carried out in tandem. These aspects show potential borrower’s current ability to repay loans. These factors account for 10% weight in our model.

 

Moreover, the Model analyses bank account statement information to see the full picture of company’s current financial position and cash-flow behaviour. These factors account for 10% weight in the Model.

 

 

Non-financial factors: 50% weight

 

The credit risk scoring Model supplements previously described analysis with several non-financial factors, which are significant in accurately assessing company’s credit risk. First of such factors is market and industry outlook. We take into account the latest macroeconomic conditions and trends as well as current industry specifics. These factors account for 5% weight in our Model.

 

Another factor we analyse is personal credit history of potential borrower’s management representative who fill in the application for a loan. We take into account representative’s payment history and culture on loans, utilities and other financial obligations. It is well established that personal payment culture of company’s key management representative is a close proxy of company’s overall payment culture and attitude towards financial obligations. These factors account for 35% weight in the Model.

 

Finally, we take into account the company management’s experience in operating businesses, as this is one of the key factors in achieving success in business, thus, shows potential borrower’s ability to meet its financial obligations. These factors account for 10% weight in our Model.

 

 

Credit score Output

 

For each applicant company, CROWDESTOR credit risk scoring model assigns the credit score, the corresponding credit rating and the probability of default. CROWDESTOR credit score ranges from 0 to 1000. The system assigns credit rating based on the calculated credit score for an applicant. Credit rating groups have a corresponding minimum and maximum probability of default. Based on the credit score, credit rating and probability of default, we provide an interpretation of an applicant company credit scoring result. Only applicants with credit scores of 551 and larger, or risk rating of C and above are fundable on the platform. CROWDESTOR credit rating is modelled after credit rating scales developed by the largest and most credible credit rating agencies. Our credit ratings are similar due to similar methodology and data inputs used for analysis. Moreover, our credit ratings are tailor-made for evaluating SMEs and therefore, are supplemented by bank account and non-financial factors. The possible CROWDESTOR credit risk assessment results are summarized in the table below.

 

 

Credit Rating Scales

Score range

Credit rating

Probability of default

Interpretation

CROWDESTOR action

from

to

from

to

951

1000

AAA

0.0%

1.0%

Very high credit quality

fundable

901

950

AA

1.0%

2.0%

High credit quality

851

900

A

2.0%

3.0%

High credit quality

801

850

BBB

3.0%

4.0%

Medium-high quality

751

800

BB

4.0%

6.0%

Medium quality

701

750

B

6.0%

8.0%

Medium-low quality

651

700

CCC

8.0%

10.0%

Moderate-high credit quality

601

650

CC

10.0%

12.0%

Moderate credit quality

551

600

C

12.0%

16.0%

Moderate-low credit quality

501

550

DDD

16.0%

21.0%

Bad credit conditions and subject to high credit risk

non-fundable

451

500

DD

21.0%

26.0%

301

400

D

26.0%

31.0%

201

300

EEE

31.0%

46.0%

Very close to default or in default

101

200

EE

46.0%

66.0%

0

100

E

66.0%

100.0%

 

Model output

 

After scoring is done, the Model yields the following metrics:

  • Company rating,
  • Company score,
  • Personal score,
  • Probability of default,
  • Loss given default,
  • Expected loss,
  • List of rejection reasons (available only to borrowers),
  • List of things to improve,
  • Credit limit,
  • Loan term,
  • Interest rate offer.

 

The Model performs a credit risk analysis according to the Model scorecard and generates internal preliminary results. CROWDESTOR financial analysts manually verify all the information and Model outputs, before the result is finalized. After the loan has been manually reviewed and approved, the system automatically generates an offer or rejection and a credit assessment report for the borrower, and in the case of an offer, a credit assessment report for the investors.

 

So, the answer for borrowers is provided within 1 working day. Once it is ready, the potential borrower will receive a notification with a link to a credit report available in borrower’s personal cabinet on the platform. The link will be sent to an email address indicated in the application form. In the report, apart from other metrics, the potential borrower can see its financial analysis and indication on what to improve in the business. Suggested improvements are likely to increase potential borrower’s chances to obtain financing from CROWDESTOR or other sources.

 

In case of the actual project placement on CROWDESTOR’s platform, credit assessment report with key metrics will be available also for potential investors in order to provide them with a standardised assessment benchmark for all SME projects on our platform and more quality information to make decisions.

Thus, it gives possibility for investors to make more balanced decisions based on a transparent risk grade of each and every project, understandable risk-evaluation methodology and guidelines, a clear understanding of our decision-making process and debt collection procedure.

 

Glossary & terminology

 

Annual accounts – Financial reports prepared annually by an active company and submitted to the registration country’s corporate registry. Financial reports consist of Profit and Loss statement, Balance Sheet, and Cash Flow statement; these statements contain information on company’s assets, liabilities, equity, income and expenses, investments by or distributions to shareholders and cash flows.

 

Credit limit – A maximum loan amount in euros calculated by our system and offered to an applicant for placement on our platform. An appropriate credit limit for an applicant is calculated taking into account applicant’s company age, company score, revenue and total asset value.

 

Credit report - is a detailed breakdown of the borrower’s creditworthiness. It is a result of automated scoring model developed by CROWDESTOR, which retrieves and analyses the information about the potential borrower from 500 databases around the world. Main outputs of the analysis are credit decision (loan amount, interest rate, loan term) and metrics on borrower’s ability to repay the loan (probability of default, loss given default, company and personal scores, etc.). Moreover, report contains elaboration on what is good and what should be improved by borrower to receive better funding terms.

 

ESMA – European Securities and Markets Authority is a financial regulation agency in the European Union and European Supervisory Authority, located in Paris, France. ESMA is safeguarding the stability of the European Union’s financial system, and is directly supervising credit rating agencies and trade repositories. You can find more information about ESMA here: https://www.esma.europa.eu/about-esma/who-we-are

 

Expected Loss – Expected loss shows what percentage of the total loan is likely to be fully written off or not recovered. Expected loss is obtained by multiplying probability of default with loss given default. For example, expected loss of 1% states that for every euro invested in the loan, investors should expect to lose 1 cent.

 

FCA – Financial Conduct Authority is a financial regulatory body in the United Kingdom, which regulates financial services firms and financial markets in the UK. FCA is regulating more than 59’000 businesses, including banks, building societies, credit unions, insurers and major investment firms. You can find more information about FCA here: https://www.fca.org.uk/about/the-fca

 

KYC – Abbreviation of “Know Your Customer”. Usually referred to processes undertaken by the company to verify the identity, suitability and risks involved with maintaining or establishing a business relationship with its current or potential client. Today it is a significant element in the fight against financial crime and money laundering.

 

Loss given default – Loss given default shows what percentage of the loan amount is likely to be written off or lost after a borrower encounters interest or principal payment problems (i.e. defaults on the payment) and enters debt recovery. Companies with more fixed assets are likely to have lower loss given default.

 

Model – The SME credit risk scoring Model developed by CROWDESTOR, which is used to assess the credit risk and creditworthiness of a potential borrower.

 

Probability of default – Probability of default is the industry’s standard for measuring the likelihood that a borrower will not be able to make scheduled repayments over a specified period of time (in our case, consecutive 90 days, during the next 12 months). Simply put, it is a quantified probability of company delaying payments. Investors use probability of default to differentiate between safe and risky investments. Probability means that company may be late of one of its payments, but necessarily will be late. Also, it may occur any time during 12-month period. It is important to distinguish between probability of default and that of going bankrupt. Probability of default should be analysed together with loss-given default, to make more balanced decision.

 

PSD2 – The Revised Payment Services Directive, which is an EU Directive on regulate payment services and payment service providers throughout the EU and European Economic Area. The PSD2 directive replaces the PSD directive and it’s purpose is to create a more integrated European payments market, making payments safer and more secure and protecting consumer rights. You can find more information about PSD2 here: https://ec.europa.eu/info/law/payment-services-psd-2-directive-eu-2015-2366_en

 

SME – Small-to-medium sized enterprises that employ up to 250 people and generate annual turnover of up to EUR 50 million. These companies are the backbone of the European and other world regions’ economies.

 

UBO – Ultimate beneficial owner of the entity is the natural person who ultimately owns or controls a legal entity. Usually and particularly in case of SMEs, those are individuals who owns at least 25% of company’s share capital and has the right to exercise significant control over the company. Identification of the UBO is a crucial part of Anti Money Laundering (AML) and Know Your Customer (KYC) processes, which is mandatory for specific industries, such as banks, solicitors, accountants, etc.

 

 

LEGAL DISCLAIMER

 

The Model uses proprietary intellectual property and know-how created and developed by CROWDESTOR. Consequently, CROWDESTOR is the sole owner and holder of all intellectual property rights (whether registered or unregistered) and know-how associated with the Model. We constantly improve and develop the Model; therefore, we reserve rights to implement changes in the Model as part of the development process. The Model serves as a general guideline in making business decisions and we reserve rights to acquire additional information about the borrower and adapt application of the Model for the specific borrower to better understand the specific borrower’s individual situation. In preparing the output, the Model uses a variety available data from numerous databases, some of which have been generally indicated above, the exact number and information on databases used to acquire data for the Model is a proprietary know-how of CROWDESTOR.