Loan Book Growth Forecasting
This use case explains how to build a driver based loan book growth model for financial services, lending and fintech in Model Reef.
You will:
Represent loan portfolios and products in the branch and variable structure.
Build originations, churn and amortisation drivers for each product.
Model interest income, funding cost and net interest margin over time.
Connect loan book growth to P&L, Balance Sheet, Cashflow and valuation.
Use scenarios to test growth, rate and funding conditions.
Model Reef is not a regulatory capital or full risk system. It models commercial loan book growth, yield and funding at planning level using structured drivers and time series.
When to use this pattern
Use this pattern when:
You run a lending or credit business (bank, non bank, fintech, B2B or B2C).
Loan book size and mix drive most of your economics.
You want transparent growth, yield and funding assumptions rather than opaque spreadsheets.
You need scenario based views of loan growth and net interest margin.
It is a foundation for:
Credit Loss and Provision Modelling
Unit Economics for Fintech Products
Multi Product Financial Model
Build a Capital Structure Model
Architecture overview
Step 1: Define portfolios, products and branches
Start by mapping your loan book into branches that reflect how you manage and report it, for example:
Lending Business
Portfolio - Home Loans
Product - Owner Occupier
Product - Investor
Portfolio - SME Loans
Portfolio - Asset Finance
Portfolio - Consumer BNPL
Each product branch will hold:
Asset variables for loan balances.
Revenue variables for interest income and fees.
Drivers for originations, churn and rates.
You can also use branches for channels (for example broker, direct, digital) if you want channel specific views.
Step 2: Build originations and churn drivers
In the Data Library, create drivers for each product such as:
New Originations per period (by value).
Average Ticket Size and number of accounts if you want account counts.
Churn Rate or Prepayment Rate.
Scheduled Amortisation Profile for term products.
For amortising products (for example term loans), you can approximate amortisation using:
Scheduled principal repayments as a percentage of opening balances, or
Average life based curves such as constant payment or annuity profiles.
For revolving products (for example credit cards or lines), use:
Net growth drivers that capture utilisation and credit limit changes.
These drivers define how balances move from period to period.
Step 3: Implement loan balance roll forward logic
For each product, create an Asset variable such as:
Assets - Loan Book - Home Loans - Owner Occupier.
Implement a balance roll forward conceptually as:
Closing Balance = Opening Balance + New Originations − Scheduled Repayments − Prepayments and Churn.
In Model Reef you do this with drivers and formulas in the variable, not cell based logic. For example:
Base Balance Driver with total gross balances per period, imported or calculated.
Additional drivers or variables for originations and exits if you want to make them explicit.
If you already have history, you can import historical balances and attach growth and runoff assumptions for the forecast horizon.
Step 4: Model interest income, fees and funding cost
Create Revenue variables for interest and fees per product, such as:
Revenue - Interest Income - Home Loans - Owner Occupier.
Revenue - Fee Income - Home Loans - Owner Occupier.
Attach rate drivers, for example:
Product Rate = Reference Rate plus Margin.
Effective Rate drivers per product.
Compute interest income conceptually as:
Interest Income = Average Balance × Effective Rate.
Balance Series can be approximated as opening, closing or mid period balances depending on how much precision you need.
For funding, create Liability and Interest Expense variables such as:
Liabilities - Wholesale Funding.
Interest Expense - Wholesale Funding.
Attach funding rate drivers and compute:
Funding Cost = Funding Balance × Funding Rate.
Net interest margin at product or portfolio level is then:
NIM = Interest Income − Funding Cost, or
NIM Percentage = (Interest Income − Funding Cost) ÷ Average Earning Assets.
Step 5: Connect to P&L, Balance Sheet, Cashflow and valuation
With loan and funding variables set up:
P&L will show interest income, interest expense and net interest outcome.
Balance Sheet will show loan asset balances and funding liabilities.
Cashflow Statement will show interest received and paid, plus net loan drawdowns and repayments.
Cash Waterfall will include loan cashflows in the operating and financing sections depending on how you structure the flows.
If you use the valuation engine, FCFF and FCFE will already incorporate loan book growth, interest and funding as part of overall business cashflows.
Step 6: Use scenarios for growth, rate and funding changes
Clone the base model into scenario models to test:
Faster or slower loan book growth.
Different originations by channel or geography.
Changes in product rates and spreads.
Funding mix changes between deposits, wholesale and equity.
Stress conditions such as rising rates or reduced funding availability.
In each scenario, adjust:
Origination, churn and amortisation drivers.
Rate and spread assumptions.
Funding cost and availability.
Any linked credit loss and provision drivers if you use that pattern.
Compare scenarios using:
Loan book growth and composition.
Net interest income and margin.
Cashflow and funding gaps.
Valuation metrics where relevant.
Check your work
Loan balances reconcile to historical data in the base period.
Growth, amortisation and churn assumptions are grounded in observed patterns.
Rates and spreads line up with current pricing and funding conditions.
Scenario results are intuitive to credit, treasury and strategy teams.
Troubleshooting
Related guides
Last updated