Jul 11, 2023
Credit scoring has always been the foundation of lending. From the humble beginnings of using credit bureau data to assess the credit worthiness of an individual or a business, modern practices involve using sophisticated technology, alternative data providers as well as machine learning models.
When it comes to online lending, consumers expect a fast and easy process, and fintechs are meeting that demand by investing in the best practices available. No wonder then, that online lending was estimated to be “USD 10.7 billion in 2021 and is projected to reach USD 20.5 billion by 2026”.
In this guide we’ll explore the core concepts of credit scoring, how different models are calculated, as well as showing you how to combine different approaches like alternative data to offer great experiences for your applicants.
What is credit scoring?
Credit scoring is the process of evaluating the creditworthiness of an individual or a business. Credit scoring models use various factors, such as credit history, payment behavior, income, and other financial information to determine a credit score.
The credit score is a numerical representation of the likelihood that a borrower will repay a loan.
So, who uses credit scoring models? Credit scoring models are used by:
Other financial institutions
Credit scoring models are used to assess the risk of lending money to an individual or a business. The credit score is an important factor in determining the interest rate, loan amount, and other terms of a loan.
Credit scores are based on data from the three major credit bureaus:
What do these credit bureaus do?
Credit bureaus collect individual's credit use information. Most credit scores fall within a range from 300 to 850, with the higher score translating to less risk for the lending entity.
In terms of scoring models, one can differentiate between traditional and alternative models. The traditional scoring models are mostly FICO & VantageScore, while alternative methods rely on alternative data to assess credit worthiness.
The idea is that today there is no one size fits all approach. Due to intense competition and stark differences in worldwide markets, fintechs have to innovate on how they approach credit scoring. This way, they can not only offer better credit products, but can up their approval rates while hedging against fraud / and default risk.
Let’s dive in on how to approach this.
Combining traditional scoring models with alternative data
Modern fintechs rely on combining traditional credit scoring models with alternative data, incorporating several other systems in the credit underwriting process to maximize approval rates while minimizing risk. This is normally done through what is called a credit decisioning engine, which allows lenders to combine multiple approaches under a single hood.
Here are some potential scoring strategies to consider:
Alternative Data Sources
In addition to traditional credit reports, tap into alternative sources of data such as cash flow, accounting information, and loan payment history.
This data can provide additional information on an individual's creditworthiness beyond what is captured in traditional credit reports. It can also be used to verify the applicant’s employment history, education, and address.
To assess risk and determine creditworthiness quickly and efficiently, leverage machine learning algorithms to analyze the data generated by traditional and alternative data sources. These algorithms could process large amounts of data to make faster and more accurate lending decisions.
Real-Time Access to Cash Flow Data
Leveraging real-time access to an applicant’s cash flow data would help determine an applicant's ability to repay a loan. This minimizes any risk of default which comes from underwriting based on past trends. This is important for businesses that are very seasonal.
To prevent fraud and identity theft, use biometric authentication during the application process. This would verify an applicant's identity using facial recognition, thumbprint scanning, or other biometric markers.
Compliance Check Systems
To ensure that lending regulations are being followed, use compliance check systems to identify potential compliance issues. The compliance system would be able to catch any violations before they turn into larger issues.
Credit Simulation Models
It's always better to play a few moves ahead, so implement credit simulation models that allow us to run "what if" scenarios on different credit terms. This would help us decide if a loan is worth the risk based on the loan term and interest rate agreed upon.
By combining traditional credit scoring systems, alternative data sources, machine learning, cash flow data, biometric authentication, compliance check systems, and credit simulation models, an online lender can make more informed and efficient lending decisions while preventing fraud and minimizing risk.
This approach would allow the lender to approve more loans without increasing the risk and reduce the amount of human intervention needed.
Oscilar allows you to do exactly that. Using our solution, you can combine your credit scoring models with your KYC flows while tapping into alternative data sources during the credit underwriting process for instant decisioning.
Of course, the fundamentals of credit scoring in the US are still reliant on bureau data, as well as the dominant players - FICO and VantageScore. In the next section, we’ll go over the basics and how they differ from one another.
FICO Scoring Model
The FICO scoring model is one of the most widely used in the world.
FICO scores are calculated based on five factors, which are weighted differently:
Factor / Weight in %
1. Payment history – 35%
2. Amounts owed – 30%
3. Length of credit history – 15%
4. New credit – 10%
5. Types of credit used – 10%
FICO scores range from 300 to 850, and scores in the 670 - 739 range are considered good. This system has been around for over 30 years, and is being utilized by 90% of lenders in the US.
The VantageScore model is another credit scoring model that is gaining popularity, as it was created as an alternative to FICO.
The VantageScore ranges from 300 to 850, with a higher score indicating a lower risk of default.
Factor / Weight in
Payment history – 40%
Age and type of credit – 21%
Percentage of credit limit used – 20%
Total balances and debt – 11%
Recent credit behavior – 5%
Available credit – 3%
VantageScore used trended data, and weighs in the applicant’s credit utilization date, making it increasingly popular among lenders - the company reported a 30% year on year increase in adoption in 2021.
Differences between FICO and VantageScore
While both FICO and VantageScore are scoring models, there are some differences between the two. The four key differences include:
Number of factors: The FICO score is based on five factors, while the VantageScore is based on six factors.
Payment history: Payment history is the most important factor in both models, but payment history accounts for 35% of the FICO score and 40% of the VantageScore.
Credit timeline: The FICO score considers the length of credit history, while the VantageScore considers the age and type of credit.
Credit balance: The FICO score considers the credit card balances and types of credit used, while the VantageScore considers the total balances and debt.
Differences aside, it’s a good idea to use both services together. By utilizing both FICO and VantageScore, you can compare the credit scores provided by each model. This comparison helps validate the consistency and accuracy of the credit information.
If the scores from FICO and VantageScore align closely, it provides more confidence in the applicant's creditworthiness. On the other hand, if there is a significant difference between the scores, it prompts us to further investigate and evaluate the factors influencing the variations.
FICO and VantageScore Credit Scores for Small Businesses
Although both scoring models were originally designed to assess the creditworthiness of individuals, both FICO and VantageScore have also developed credit scoring systems specifically for small businesses.
FICO Credit Scores for Small Businesses
For small businesses, FICO offers three different credit scores to assess credit risk in different situations:
The Small Business Credit Score for term loans and lines of credit
The LiquidCredit Small Business Scoring Service for credit cards and working capital loans
The Small Business Scoring Service for Commercial Credit
These credit scores evaluate data from business and personal credit reports, available from credit bureaus such as Experian, Equifax, and Dun & Bradstreet.
VantageScore Credit Scores for Small Businesses
Like FICO, VantageScore credit scores use a variety of factors to determine creditworthiness, including payment history, credit utilization, length of credit history, new credit, and credit mix.
Recently, VantageScore introduced a credit scoring system specifically for small businesses: the VantageScore Small Business credit score.
The VantageScore Small Business credit score uses both business and personal credit data, along with additional financial information such as cash flow, assets, and revenue, to assess credit risk.
This takes into account the financial history of the business and the business owners. VantageScore estimates that approximately 27 million small businesses in the United States have a credit file, and potentially around 7 million of these files could be scored using the VantageScore Small Business credit score.
Differences Between FICO and VantageScore Credit Scores for Small Businesses
While both FICO and VantageScore consider a variety of factors such as payment history and credit utilization, there are some differences between the two scoring systems:
Use of alternative data: VantageScore Small Business credit score incorporates alternative data sources such as financial statements from account software providers. FICO on the other hand only uses data from credit bureaus.
Credit score ranges: FICO credit scores for small businesses can range from 0 to 300, while VantageScore Small Business credit scores can range from 1 to 100.
Lender adoption: FICO credit scores are more widely used than VantageScore credit scores by banks and other lenders.
Credit scoring model practice changes
Scoring models are constantly evolving to better predict credit risk. In recent years, there have been some changes to credit scoring model practices.
One change is the use of alternative data. Alternative data includes things like utility bills, rent payments, and cell phone bills. By using alternative data, scoring models can better assess the creditworthiness of individuals who may not have a traditional full credit report or history.
Another change is the use of trended data. Trended data looks at a borrower's credit history over time, rather than just a snapshot at a specific point in time. By using trended data, credit scoring models can better predict credit risk.
Other credit score models
In addition to FICO and VantageScore, there are other scoring models that are used by lenders and financial institutions. Some of these models include the TransUnion CreditVision Risk Score, the Experian PLUS Score, and the Equifax Credit Score.
What are the methods of credit scoring?
There are two main methods of credit scoring: traditional credit scoring and alternative credit scoring.
Traditional credit scoring uses credit history and other financial information to determine a credit score, while alternative credit scoring uses alternative data, such as utility bills and rent payments, to assess credit risk.
In conclusion, the credit scoring system is an important tool used by lenders and financial institutions to assess credit risk. FICO and VantageScore are the most widely used scoring models, but there are other models available as well.
Credit scoring models are constantly evolving to better predict credit risk, and credit scoring tools are used by lenders and financial institutions to make informed decisions about lending money.
Beyond the data provided by the major credit bureaus and the different credit scoring models, online lenders increasingly rely on alternative data as well as machine learning for their credit underwriting process.
If you are interested in an out-of-the-box solution, consider booking a demo with us today.