Resources
How lenders can verify income using open banking transaction data
Use Fiskil’s banking and Income APIs for income assessment, with a worked transaction example, provider questions and clear lending responsibilities.
An Australian lender can evaluate Fiskil’s Banking API for customer-authorised transaction evidence and its Income API for structured insights derived from enriched transactions. Assess those outputs against the lender’s own evidence policy, exception handling and credit process.
That distinction matters when choosing an API. A provider offering bank transactions is not necessarily offering a complete income-verification report, and a recurring credit is not automatically salary.
Compare three layers of the income workflow
| Layer | What the lender should establish | Question for the evaluation |
|---|---|---|
| Bank transaction evidence | Which authorised accounts and history are available | Can the relevant credits be traced to the returned source records? |
| Provider income analysis | Which inflows become identified sources or recurring streams | How are transfers, irregular pay and insufficient history handled? |
| Lender assessment | How the evidence supports the particular application | Which cases need another source or human review? |
The Income guide describes an analysis based on enriched transactions after synchronisation. Ask for a representative income profile and the meaning of each amount, frequency and history field. Do not treat a bank credit as a verified gross salary or infer employment status from a recurring pattern alone.
Use the same synthetic cases across providers: two employers, irregular pay, a transfer between the applicant's own accounts, a loan credit and a short history. Record the expected interpretation and the evidence that would justify it. The purpose is to establish whether the proposed service supports the lender's workflow, not to manufacture an accuracy percentage.
Which API calls should a lender evaluate?
Start with the output your assessor needs. If your team applies its own income rules, test whether it can retrieve the credits behind an application. If an assessor needs a summary of recurring income, test whether they can trace a figure in that summary back to those credits.
| What to test with Fiskil | Documented operation | What to inspect |
|---|---|---|
| Retrieve the bank evidence | GET /v1/banking/transactions |
Select the end user, account and date range. Follow the pages and check the posting status of the credits you intend to count. |
| Retrieve the income analysis | GET /v1/banking/income |
After transaction synchronisation, request the same end user and accounts for the same dates. Inspect income_sources for the receiving account, category, frequency and transaction count. Inspect income_summary for totals by currency. |
The transaction reference describes date-time filters. The income reference uses inclusive calendar dates and lets you select accounts and income categories. Set the period explicitly for both requests. The Income guide describes a default analysis of up to six months of salary-category data. Record the categories you requested before interpreting a missing source.
For the worked example below, ask an assessor to trace the $5,000 payroll deposit and explain what happened to the $1,200 transfer. If the summary includes the transfer, record how the assessor spotted it and what they would exclude. If the account or the period is missing, record the evidence still needed. A deposit can support an assessment of money received. Establishing gross pay or continuing employment may require a payslip or other evidence under your policy.
A packaged report or an assessment inside loan origination calls for a different demonstration. Basiq’s Consumer Affordability Report defines measures such as median monthly salary. NextGen’s ApplyOnline launch describes checking declared PAYG income inside the application. Give each provider the same case and ask to see exactly what reaches the assessor when a credit needs explanation.
Work through the credits before calculating income
Consider the following synthetic example. The labels are an illustrative assessment method, not Fiskil categories or an actual borrower outcome.
| Transaction | Amount | Proposed treatment | Why it needs that treatment |
|---|---|---|---|
| Monthly payroll deposit | $5,000 | Candidate employment income | Recurrence and description support investigation; corroborate as your policy requires |
| Transfer from the customer's savings account | $1,200 | Internal transfer | Moving money between accounts does not create another source of income |
| Retail purchase refund | $240 | Refund | Returning prior expenditure should not inflate salary |
| One-off equipment sale | $800 | Non-recurring credit | A single receipt does not establish ongoing earnings |
| Freelance receipts | $900 | Separate variable-income review | Business receipts may involve expenses and different evidence requirements |
In this one-month example, summing all credits gives $8,140. Treating every credit as regular income would hide the distinction between $5,000 of candidate payroll income and the other receipts. The freelance amount needs its own assessment; it is not automatically eligible net income.
Repeat the analysis across the policy-defined period. Identify missing weeks or accounts before turning an incomplete history into a monthly average. Explain periods of leave, changed employers and irregular payment schedules through a review process rather than silently smoothing them away.
A practical assessment workflow
- Define the evidence needed. Record the applicant, accounts, relevant period and purpose of collection.
- Obtain appropriate access. Use the approved consent and participation arrangement for the service.
- Retrieve and validate the data. Complete pagination and check coverage, dates, status and freshness.
- Classify candidate receipts. Distinguish employment, business, transfers, refunds and other credits using documented rules.
- Resolve ambiguity. Request supporting evidence or human review where the source or sustainability is unclear.
- Create an explainable result. Keep the calculation, source references, exclusions and reviewer outcome available under the agreed retention controls.
This is a proposed implementation pattern. It is not a credit policy, automated approval rule or statement that every listed receipt is acceptable for a particular loan.
Handle common exceptions explicitly
A customer may split salary between accounts, receive reimbursements from an employer or use a business account for several purposes. Transfers between connected accounts can appear on both sides of the dataset. A reversed payment should not count as two independent events.
Pending transactions and date fields also need care. Fiskil's banking guide explains the preference for posted datetime with an execution-datetime fallback where applicable. Align the analysis window with the actual fields returned and document transactions that cannot be placed reliably.
Do not infer that an empty result means the applicant has no income. It could reflect the selected accounts, consent scope, a retrieval problem or an incomplete period.
Compare providers using an actual assessment case
| Buying question | Evidence to request |
|---|---|
| Is this data access or a finished income product? | A documented response and demonstration of the proposed service |
| Which customer accounts can be connected? | Tests covering the applicant types and banking products in scope |
| Can an analyst explain the result? | Source references, classification rules and exception handling |
| What happens with irregular income? | A worked variable-income example without invented certainty |
| How is unavailable data handled? | A visible incomplete-evidence path and a safe retry process |
| What does the lender still operate? | Written allocation of assessment, review, privacy and support responsibilities |
Keep the lender's responsibilities visible
Where responsible lending obligations apply, access to a banking API does not replace the lender's assessment. ASIC's responsible lending guidance is the starting point for determining the applicable inquiries, verification and suitability requirements.
Use only the data necessary for the stated service. Set retention, access and deletion controls with the responsible privacy and compliance teams. A richer dataset is useful when it answers the assessment question, not simply because it is available.
Fiskil provides the banking-data connection and a documented Income API for this workflow. Talk to Fiskil with the account types, evidence window and analysis outputs your lending team requires. Confirm which capabilities are supplied by Fiskil and which are implemented in your assessment system before committing to a production design.


