SOLUTIONS

Cashflow Score

A fast, explainable measure of credit risk based solely on account and transaction data.
Unlock more approvals
Assess thin-file and no-file applicants using current cash flow behavior, not only traditional credit history.
Make decisions faster
Use a 1–100 transaction-based risk score with clear drivers that help teams interpret borrower risk quickly.
Deploy quickly and easily
Use Cashflow Score as a standalone signal or alongside your existing bureau scores, rules, and risk models. No PII required.

Get a clearer view of creditworthiness – powered by cash flow behavior

Cashflow Score uses transaction data to provide a current, explainable view of borrower risk.
Traditional credit data
Credit utilization
Employment
Credit score
Cash flow behavior
Income and spending velocity
Liquidity and cash buffers
Income and payment stability
Debt obligation pressure
Balance recovery and resilience
High-risk outflows
Account balance volatility
and more...
Carrington Labs' Cashflow Score returns a 1–100 score, with 100 representing the highest credit quality, and organizes the behaviors behind the score into five categories of credit risk behavior: Velocity, Liquidity, Stability, Leverage, and Resilience.

Unlock your potential uplift

30%

more accurate in scoring high-risk customers

2.5x

more accurate in scoring low-risk, high-value customers

14%

higher margins with integrated limit-setting

Potential uplift our solutions can deliver based on a sample set of anonymized data. Actual outcomes vary by lender, product, portfolio, and implementation approach.

How it works

Provide transaction data
Send us bank transaction data — categorized or raw — via API or flat file.
We apply our transaction-based model
Our model processes the data, identifies borrower risk behaviors, and calculates a Cashflow Score from 1–100.
Get your results
Receive a score, risk segment, and explainable drivers organized around key cash flow behaviors.
Combine with your own logic
Use the score as-is or layer it with your existing bureau scores, policy rules, and credit risk models.

Compliance Ready

Cashflow Score is designed to support transparent credit risk assessment, internal governance, and lender-controlled decisioning.

No PII required
The model uses de-identified transaction data.
ECOA-compliant
Protected class information is not required or used in scoring.
Explainable outputs
Score drivers can support internal review, governance, and adverse action reason workflows where applicable.

See other solutions

FAQs

What is Cashflow Score?
Cashflow Score is a measure of credit risk based on how money moves into and out of a bank account over time. Instead of relying only on bureau data, it uses transaction-derived signals to assess borrower behavior, financial stability, repayment capacity, and risk. Carrington Labs' Cashflow Score is built from de-identified transaction data and returns a 1–100 score, risk segment, and explainable drivers to support decision transparency and governance.
How is Cashflow Score calculated?
Cashflow Score is calculated by analyzing transaction-derived signals that reflect how an applicant manages money day to day. These signals can include income patterns, spending behavior, liquidity, recurring obligations, balance volatility, and financial recovery. Carrington Labs' Cashflow Score organizes borrower risk signals into five behavioral categories: Velocity, Liquidity, Stability, Leverage, and Resilience. These categories help lenders understand the behaviors influencing the score.
How is Cashflow Score different from a credit score?
A traditional credit score is usually based on bureau data and past credit history. Cashflow Score adds a different lens by looking at current transaction behavior, including income, spending, liquidity, obligations, and balance recovery. That makes it useful as an additional underwriting signal, especially where bureau data is thin, limited, or less predictive for certain segments. Lenders can use Cashflow Score alongside bureau scores, policy rules, and existing risk models to make more informed decisions.
Can Cashflow Score help assess thin-file borrowers?
Yes. Cashflow Score can be especially useful for thin-file or no-file borrowers, or where bureau coverage is limited. Because it is based on transaction data rather than only traditional credit history, it can give lenders a broader view of borrower behavior and repayment capacity. Cashflow Score helps lenders identify applicants who may be stronger than their credit file suggests, while still supporting risk controls and explainability.
Why do lenders use Cashflow Score?
Lenders use Cashflow Score to improve risk separation, reduce false declines, and make faster, more consistent credit decisions. It adds a current cash flow lens to underwriting by helping lenders understand borrower behavior across income, spending, obligations, liquidity, and resilience. It is designed to support lender judgment, not replace it. Lenders remain in control of policy, cutoffs, approvals, declines, referrals, and customer treatment.
What data does Cashflow Score use?
It's built from de-identified bank account and transaction data, and works with customer-permissioned (open banking) data. No personally identifiable information is required.
Can we use it alongside our bureau score and existing models?
Yes. It's designed to complement bureau scores, policy rules, and existing risk models — you keep full control of cutoffs and decisions.
How is Cashflow Score delivered?
Via API or batch, returning a 1–100 score, a risk segment, and explainable drivers that drop into your existing decisioning.
Is it explainable enough for governance and adverse action?
Yes. Drivers are organized into clear behavioral categories, so credit teams can understand, govern, and communicate what's driving a score.
Can we get a model personalized to our portfolio?
Cashflow Score is a pre-tuned, ready-to-use score. If you want a model personalized to your own portfolio, products, and objectives, that's the Credit Risk Model — it can be built on transaction data, or on bureau, application, and internal performance data when transaction data isn't available.