All Posts

Finance3 min read

Dynamic Pricing Engine for Banks: How to Price Loans, Deposits and Fees in Real Time

Dynamic Pricing Engine for Banks: How to Price Loans, Deposits and Fees in Real Time

Key Takeaways

  • A dynamic pricing engine calculates the rate, fee or spread for each customer and product at the moment of the offer, from the bank’s current pricing policy.

  • In banking, “dynamic” should mean faster policy changes and consistent application, not unsupervised price setting. People set the policy and limits; the engine applies them the same way in every channel.

  • Most bank prices follow a waterfall: base rate, risk adjustment, relationship adjustment, channel or promotion, caps and floors, and approved exceptions.

  • Pricing exceptions are where fair lending risk concentrates. A rules engine makes them visible, limited and logged.

  • US requirements that touch pricing include ECOA and Regulation B, FCRA risk-based pricing notices, Truth in Lending, and new state rules on algorithmic and automated decisions.

What Is a Dynamic Pricing Engine in Banking?

A dynamic pricing engine is software that calculates the price of a financial product for a specific customer and moment: the interest rate on a loan, the rate on a certificate of deposit, a fee or a foreign exchange spread. It combines market inputs such as benchmark rates, product rules, the customer’s risk and relationship, and the channel, and returns a price with the reasons behind it.

The opposite is static pricing: a rate sheet updated weekly or monthly, copied into each channel, with loan officers and branch staff applying exceptions by hand. Static pricing reacts slowly when rates move and makes it hard to show why two similar customers paid different prices. For the fundamentals of rules engines, see What Is a Rules Engine?.

What Changes When a Bank Moves from Rate Sheets to a Pricing Engine?

  Rate sheets and manual exceptions Pricing engine
Speed of change New sheet distributed to each channel One policy change published to every channel at once
Consistency Each channel and branch may apply rules differently Same inputs give the same price everywhere
Exceptions Negotiated case by case, often by email Limits by role, product and amount; every exception recorded
Personalization Few dimensions, mostly by product and term Risk, relationship, balances, channel and promotions combined
Testing Hard to see impact before launch New pricing run against past applications and accounts first
Audit Reconstructed from emails and files Inputs, rule version and result logged for every price

Which Bank Products Benefit from Dynamic Pricing?

Consumer and small business loans. After credit decisioning assigns a risk grade, pricing rules set the rate and fees by grade, term, collateral, loan-to-value and channel, and apply relationship discounts.

Deposits. Rates on savings, money market accounts and CDs change with the rate environment and funding needs. A pricing engine applies tiered rates by balance, term and relationship, runs promotional rates for new money only, and keeps exception rates for large depositors within limits set by treasury or the asset-liability committee (ALCO).

Fees and waivers. Account fees, fee waivers and relationship packages follow rules such as minimum balances, product bundles and customer segments. Moving them into rules makes waivers consistent instead of discretionary.

FX and treasury services for business clients. Spreads and margins vary by client tier, volume and currency pair, within limits set by the treasury desk.

Cross-sell offers. Pricing for a second product, such as a rate discount on a loan for customers with direct deposit, depends on rules about which relationships qualify.

How Does a Pricing Waterfall Work?

Most bank prices are built in steps. A simplified waterfall for an unsecured personal loan:

Step Rule Example
1. Base rate Product base rate, linked to the bank’s cost of funds or a benchmark 9.50%
2. Risk adjustment Add-on by risk grade and term Grade B, 48 months: +1.75%
3. Relationship adjustment Discount for existing deposits or direct deposit −0.25%
4. Channel or promotion Digital channel promotion, valid until a set date −0.25%
5. Caps and floors Product minimum and maximum, state rate caps, Military Lending Act limit for covered borrowers Within limits
6. Exception Loan officer exception within authority; above that, approval required None
Final price Logged with all inputs and rule versions 10.75%

Rates are illustrative. Each step is a table with an owner: treasury or ALCO owns the base rate, credit owns the risk adjustment, product owns relationship and promotion rules, compliance owns caps, and lending leadership owns exception authority.

When benchmark rates move, the bank changes step 1, tests the result and publishes. Everything else stays the same, and every channel quotes the new price at once.

Why Are Pricing Exceptions a Fair Lending Risk?

Because discretion is where similar customers can end up with different prices for reasons the bank cannot explain. Fair lending reviews look closely at discretionary pricing and exceptions, since they can produce disparities on a prohibited basis even when the rate sheet itself is neutral.

A rules engine reduces that risk in three ways:

  • Limits. Exception authority is set by role, product and amount, and anything beyond it goes to an approver.

  • Reasons. Every exception requires a reason code from an approved list, such as a competitor offer or a relationship retention case.

  • Monitoring. Because every exception is logged with its inputs, compliance can analyze exception patterns by branch, officer and customer segment.

This turns exceptions from a blind spot into data. For the wider compliance picture, see Business Rules Engines in Financial Compliance and Risk Management.

Which US Regulations Affect Dynamic Pricing for Banks?

Equal Credit Opportunity Act and Regulation B. Lenders may not discriminate on a prohibited basis in any aspect of a credit transaction, including pricing. When a lender takes adverse action, it must give the specific principal reasons. In May 2025, the CFPB withdrew its circulars on adverse action notices for complex algorithms (Federal Register, May 12, 2025); the requirement itself remains.

Fair Credit Reporting Act: risk-based pricing. When a lender uses a consumer report to offer credit on terms materially less favorable than the most favorable terms available to a substantial proportion of its consumers, it must give a risk-based pricing notice or use an exception such as the credit score disclosure (12 CFR Part 1022, Subpart H). Pricing rules need to know when that notice is triggered.

Truth in Lending (Regulation Z). Advertised and quoted terms must include required disclosures such as the APR. See Business Rules Engines in Loan Calculators for how this affects online quotes.

Algorithmic pricing disclosure. New York’s Algorithmic Pricing Disclosure Act, in effect since November 10, 2025, requires businesses that set prices with an algorithm using consumers’ personal data to show the statement “THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA.” Financial institutions regulated under the Gramm-Leach-Bliley Act or Section 801(f) of the New York Financial Services Law are exempt, as are entities regulated under New York insurance law (Consumer Finance Insights, December 2025). Nonbank lenders and fintech partners should check with counsel whether the exemption applies to them.

State rules on automated decisions. Colorado’s SB 26-189 applies from January 1, 2027 to automated decision-making technology used in consequential decisions, including lending, with notice, explanation, human review and record-keeping requirements (Seyfarth Shaw, May 2026).

The common thread: a bank must be able to show how a price was set and why it differs between customers.

Where Does Machine Learning Fit in Bank Pricing?

Machine learning can estimate things that inform pricing: probability of default, the likelihood that a customer accepts an offer, or the rate sensitivity of deposit balances. It should inform pricing, not set prices on its own.

The working pattern: the model estimates, the rules price, and both are logged. A model may estimate that a deposit customer is likely to move balances if the rate falls below a level. Rules decide whether that customer qualifies for an exception rate, within the limits ALCO has approved, and record the model version, the rule version and the outcome. This keeps pricing explainable and keeps prohibited characteristics and their proxies under compliance review.

How Do You Implement a Pricing Engine Without Replacing Core Systems?

  1. Start with one product. Personal loans, auto loans or CDs are common first choices.
  2. Document the current pricing. Rate sheets, adjustment tables, exception policies and the logic hidden in the loan origination or core system.
  3. Build the waterfall as decision tables. One table per step, each with an owner and an approver.
  4. Test on history. Run last quarter’s applications or accounts through the new rules and compare prices with what customers actually received. Explain every difference.
  5. Connect the channels. The loan origination system, online banking, the branch and partners call the engine through an API instead of keeping their own rate copies.
  6. Monitor exceptions. Review exception volumes and reasons monthly with compliance.
  7. Plan the timeline realistically. Platform go-live with Higson typically takes 3–6 months. After that, adding a new product takes weeks, and a single rule change, such as a new base rate, takes hours.

How Does Higson Support Dynamic Pricing for Banks?

Higson is a business rules engine built by Decerto for financial services companies and insurers. In a pricing setup, it holds the pricing waterfall and runs it in real time for every channel.

  • Decision tables that pricing teams maintain. Base rates, risk adjustments, relationship discounts, promotions, caps and exception limits live in decision tables that product, treasury and credit teams edit and test in Higson Studio. Rate sheets prepared in Excel are imported through CSV.

  • Testing before production. The tester and mass tester run a pricing change against past applications or accounts, so teams see how prices move before they publish.

  • Versioning, audit trail and granular permissions. Every change is versioned with rollback. Higson 4.3 LTS adds role-based permissions at the level of business domains, sessions and operations, with a full audit trail, so treasury can change base rates while compliance owns caps.

  • Real-time performance. Rule execution takes 0.23 ms per decision, with throughput of 9,000 requests/second, enough to reprice a quote as a customer changes the amount or term.

  • Models inside the rules. Higson’s ONNX runtime (since version 4.2) runs your own machine learning models inside rule execution, so a model estimate and the pricing policy around it are evaluated and logged together.

  • Integration and deployment. REST API and Java SDK; cloud (AWS, Azure, GCP), on-premise Kubernetes or hybrid. The Runtime API supports JWT, SSL and basic authentication; Higson Studio supports SSO through Active Directory, SAML and CAS.

This setup removes backlog from the IT queue: pricing, treasury and credit teams own the rules, and IT gets more time for architecture and integrations.

Notus Finance, a Polish mortgage brokerage that compares offers from multiple banks, runs bank offers, pricing and cross-sell conditions in Higson, integrated with its CRM, and business users update offers when banks change them (Notus Finance case study). In insurance, where rating tables change on a similar rhythm, TUW actuaries publish a new motor rate in 2 hours instead of 2 weeks.

See Your Pricing Waterfall in Higson

Bring one product and its rate sheet, and we will show you how its pricing waterfall and exception limits would work in Higson, tested on your own sample applications.

Book a 30-minute call

Download the Business Rules Engine Comparison Guide 2026

Take Full Control of Your Product Logic

We provide fee Proof Of Concept, so you can see how Higson can work with your individual business logic.