Our Methodology
banking·saver uses publicly available Consumer Data Right (CDR) data and documented mathematical methods to calculate what each banking product actually costs or earns you. Here is exactly how every number is calculated.
1. Data Source
All banking product data is sourced from the Consumer Data Right (CDR) Product Reference Data API. Under Australian law, every Authorised Deposit-taking Institution (ADI) — banks, credit unions, and building societies — is required to publish their product data through this API. This means our database covers the entire regulated market.
We currently track products from 121+ institutions, including all four major banks (CBA, Westpac, NAB, ANZ), regional banks, credit unions, and digital banks.
Data is refreshed automatically every day at 03:00 UTC. When we detect a rate change, we record it in our rate history database, enabling the sparkline charts and rate change tracking you see throughout the site.
2. Savings Account Calculations
Australian savings accounts calculate interest daily on the closing balance and credit monthly. Our engine applies a progressive accuracy model:
Tier 0 — Market overview (zero input)
With no inputs, we display raw CDR data: base rates, bonus rates, total rates, and balance caps for every savings product. This gives you a comprehensive market snapshot but doesn’t account for your specific situation.
Tier 1 — Balance-aware (1 input)
When you enter your savings balance, we apply balance tier logic. Australian savings accounts use two tier models:
Whole-balance tiering (most common): one rate applies to the entire balance based on which tier it falls in. This creates potential “cliff effects” where depositing one extra dollar past a tier boundary can reduce total interest. We flag these.
Marginal tiering (less common): different rates apply to portions within each band, like tax brackets. The formula: annual interest = Σ(balance in tier × tier rate).
Tier 2 — Bonus condition assessment (3–5 inputs)
This is the largest accuracy improvement in our model. The ACCC found that 71% of bonus savings accounts fail to earn their bonus rate in any given month. We assess whether you can actually meet each product’s bonus conditions.
We parse bonus conditions from CDR data using regex pattern matching (Layer 1), covering the 10 most common condition types: minimum monthly deposit, no withdrawals, growing balance, balance cap, linked transaction account, minimum card transactions, age restriction, introductory period, salary deposit, and digital wallet requirements.
The calculation models three scenarios: all conditions met (best case), realistic conditions met (based on your inputs), and base rate only (worst case). The “realistic” estimate is displayed as the primary result.
3. Home Loan True Cost
The statutory comparison rate mandated by ASIC uses a $150,000 loan over 25 years — assumptions set decades ago that distort results for modern loan sizes. A $500 annual fee adds 0.33% to a $150K comparison rate but only 0.06% to an $800K loan.
Our “true comparison rate” recalculates using your actual loan amount via the same IRR equation (Newton-Raphson method), giving a more accurate picture of the true cost.
We also calculate: monthly repayments (standard PMT formula), total interest over the loan term, offset account annual savings (offset balance × loan rate), and estimated Lender’s Mortgage Insurance for LVR above 80%.
LMI estimates are approximate. Actual LMI varies by insurer (QBE, Helia), state, borrower profile, and loan type. Our estimates use mid-range industry benchmarks.
4. Credit Card Net Annual Cost
We calculate differently for transactors (pay in full) and revolvers (carry a balance):
Transactors: net annual cost = annual fee − rewards value − interest-free period value. Rewards are valued conservatively: Qantas and Velocity points at 1.5 cents per point, bank reward programs at 0.5 cpp, cashback at face value.
Revolvers: net annual cost = annual fee + interest charges − rewards value. The interest-free period is lost when carrying a balance. Interest is calculated on the average revolving balance at the card’s purchase rate.
5. Limitations & Known Constraints
CDR data quality: The ACCC has documented that 92% of CDR data holders have data reliability issues, with banks averaging 119 days to rectify problems. We display data as published by each institution; discrepancies between CDR data and bank websites may exist.
Condition parsing: Our Layer 1 regex engine handles approximately 80% of bonus conditions accurately. The remaining 20% — particularly bank-specific phrasing and complex multi-condition products — may be incompletely parsed. Layer 2 LLM-based extraction is planned but not yet live.
Assumptions: Savings calculations assume constant balance; home loan calculations assume fixed interest rates; credit card calculations assume consistent monthly spending. Tax implications are not included.
Update frequency: Product data is refreshed daily. Rate changes are typically reflected within 24 hours, but institutional delays in updating CDR feeds may cause temporary discrepancies.
6. What banking·saver Is Not
banking·saver is an information and calculation tool. It does not provide financial product advice, does not hold an Australian Financial Services Licence (AFSL), and does not receive commissions from any product provider. It does not recommend, rate, rank, or evaluate financial products.
All calculations are mathematical outputs based on inputs you supply and publicly available CDR data. Results show products that match your stated criteria — they are not recommendations. You should read the relevant Product Disclosure Statement (PDS) and Target Market Determination (TMD) and consider seeking independent financial advice before making financial decisions.