Methodology v1.3
How we rate financial products
Every rate verified against the issuer's page and stamped with the date. One stated ranking rule per roundup, applied the same way to every product in it. A public graveyard for the products that failed. This is the version you can hold us to.
The short version
- 1.Every published number (APY, APR, annual fee, minimum) is verified against the issuer's own page and stamped with an "as of" date.
- 2.Each roundup is ordered by one stated rule: the verified rate for deposit accounts, the lowest APR for personal loans, and an editorial score for the other products. The rule is printed on the list and is the same for every product in it.
- 3.Closed, failed, and renamed products go to a public graveyard instead of being silently deleted.
- 4.Affiliate commissions never move a score or a ranking. The full list of what does not count is in section 4.
1. Rate verification and "as of" dates
The single biggest failure of financial comparison content is stale numbers presented as current. Our rule: no number ships without a source and a date. Before a rate appears on Financeradar, we open the issuer's own page (not a press release, not another comparison site), record the figure, and stamp it with the verification date. That date is shown to readers next to the number.
APY, minimum balance, monthly fees, withdrawal limits, and whether the rate is promotional. Source: the bank's rate page and account disclosures.
APR range, annual fee, rewards rates, bonus terms and spend requirements. Source: the issuer's pricing and terms document.
Fees, minimum investment, accreditation requirements, and target-return language. Source: the platform's offering pages and regulatory filings.
Rates are re-verified on a schedule, and any number flagged as stale (by us or by a reader) is re-checked before it appears in a ranking again. When a correction lands, the "as of" stamp updates with it.
2. How each list is ordered
Every roundup prints its ranking rule above the list. There are three kinds of rule, and none of them reads whether a product's links pay a commission.
High-yield savings, CDs and money market accounts are sorted by the APY we read on the bank's own page, highest first. Checking accounts are sorted by monthly fee, then by the interest they pay. Business bank accounts are grouped by the job they do (operating checking, savings, money market), then sorted by fee and APY. In each of these lists (within each job for business accounts), a figure we last checked more than 14 days ago stays listed with its date but is placed after the current ones.
Lenders are sorted by their lowest published APR, then by the lower APR ceiling, read on each lender's own page and dated. The floor is what the best-qualified borrowers get, so the order is a starting point, not a verdict.
Credit cards, brokers, IRAs, robo-advisors, private credit platforms, mortgage and HELOC lenders, payment processors, POS systems, small-business finance and tax software are sorted by an editorial score, described below. Products without a finished score are listed after the scored ones. Ties, and unscored products, go by the lowest annual fee for cards and the lowest minimum for private credit, otherwise by name.
The editorial score
The editor reads each product's published terms for the four things below and records the result as one number per product. It is a judgment written down as a number, not a formula: there are no sub-scores and no fixed weights between the four. The same four questions apply to every product in a category.
Value
Fees vs benefitsWhat you pay against what you get. For a card: annual fee against realistic rewards, not best-case spend patterns. For savings: APY against fee drag and minimums. For private credit: platform fees against the return profile after fees. For a broker: trading commissions, options and margin costs and the rate on idle cash, against account minimums. For a robo-advisor: the advisory fee in dollars on real balances, plus the fund costs on top, against what the service does for you.
Transparency
Fine print honestyHow honestly the product describes itself. Promotional rates with buried expiry dates, teaser APYs on capped balances, "target returns" with no track record behind them, and fee schedules that take four clicks to find all cost points here.
Track record
History and conductYears operating, regulatory actions, enforcement history, and failures. A platform two years old with no downturn behind it scores differently from an institution that has operated through several cycles. Regulator filings and public enforcement records are the source.
Accessibility
Who can actually use itMinimum deposits, credit score requirements, accreditation rules, and geographic restrictions. A product that is excellent for accredited investors with $25,000 minimums is scored in that context and labeled accordingly, not sold as an option for everyone.
How the "best for" picks are chosen
Not by the score alone. Each pick names its rule (for example "highest APY with no minimum") and goes to the qualifying product with the best figure under that rule, read from the product's own numbers (for a rate, among the current prints when there is one). A rule without a figure goes to the qualifying product placed highest in the list. A product is picked once; when it also wins a later rule, that rule is listed on its pick. The score is an editorial judgment, not a promise about performance.
3. The graveyard: closed, failed, and renamed products
Most comparison sites quietly delete a product when it shuts down, which erases exactly the information a careful reader needs. Financeradar keeps closed, failed, and renamed products listed with their history.
Knowing that PeerStreet went bankrupt, or that Yieldstreet rebranded after investor losses, changes how you read the category. Sector failure rates are part of the risk picture, and hiding them makes every surviving product look safer than it is.
What the product was, when and why it closed or changed (bankruptcy, wind-down, acquisition, rebrand), what happened to customer funds where documented, and links to primary sources: filings, regulator notices, official announcements.
A rebrand does not wipe a track record. When a product renames, the old name, the date, and the reason stay on the page, so the history follows the product instead of disappearing with the old domain.
If we published a fee, minimum, or promo as a fact and the issuer's page says otherwise, we change the listing. A rate that moved is an update, not a confession. Closed platforms stay in the failures tracker.
4. What does not influence a rating
- ×Affiliate commissions. Some product links pay us when a reader opens an account. Those links are disclosed, and the commission has zero effect on the score, the ranking, or the cons list. Full breakdown at /how-we-make-money.
- ×Provider payments of any kind. No position in any ranking is paid for, and no issuer or platform can pay for a score change, a review edit, or the removal of a con. There is no price for any of these.
- ×Provider outreach. Emails asking us to "take another look" are treated as input for the next scheduled review, nothing more. Disputes with a primary source attached get re-verified.
- ×Marketing volume. Heavy ad spend and viral promotion do not move a score. Documented changes to the product (a fee cut, a rate change, a regulatory action) do.
- ×Editor sympathy. The criteria are the criteria. Personal opinion goes in the written take, where readers can see it and argue with it.
5. How often things get refreshed
Re-verified on a rolling schedule, and immediately when a reader or provider flags a change with a source. The "as of" date tells you exactly how fresh each number is.
Re-scored when the product materially changes: a repricing, a new fee, a regulatory action, or a track-record event.
This page is reviewed quarterly. Material changes get a version bump at the top and an entry in the changelog below.
6. Methodology changelog
Financeradar launches with date-stamped rate verification, the four-criteria rubric (value, transparency, track record, accessibility), and the product graveyard. Initial coverage: credit cards, high-yield savings, private credit.
Added online brokers, robo-advisors and IRA roundups. Fees, minimums and cash rates are read on each provider's own page and dated. Each page lists regulator actions found in SEC and FINRA records, and the dollar cost examples are computed from the verified figures.
Corrections after a review of how the site reads to search assistants. Removed: a row of publication names that implied press coverage we do not have, app-store star ratings shown as if they rated bank accounts and cards, and the word "reviewed" where we compare published terms without using the product. Changed: every average now states how many accounts it covers and the dates they were read, the statistics pages print each figure's own read date, and audience and catalog sizes come from one source and are rounded down. The guides, roundups and comparison articles that quote a rate now take it from the product page when the page is shown, next to the date it was read, so a sentence cannot drift from the product it describes; dated snapshot reports keep their figures and say so. Re-read on the providers' own pages and updated where they had moved: deposit, card, private-credit and app products (each change is logged on the product page under "Recent changes to these figures"); a page we could not re-read keeps its old date.
Section 2 now describes how each list is actually ordered: deposit accounts by the verified rate, personal loans by the lowest APR, the other products by one editorial score each. Removed: the claim that the four criteria are scored one by one and weighted equally (the score is a single editorial number with no sub-scores or weights), and the claim that the score decides the "best for" picks (each pick follows its own stated rule). The category guide template lost a four-part rubric with fixed percentages that no code computed.
7. Questions we get
How does Financeradar verify rates?
Every rate we publish (APY, APR, annual fee, minimum deposit) is checked against the issuer's own page before publication and stamped with the verification date. Readers see an 'as of' date next to every number. Rates are re-verified on a schedule, and stale numbers are re-checked before they appear in any ranking.
How does Financeradar rank financial products?
By a rule that depends on the product, printed on every roundup. Savings accounts, CDs and money market accounts are sorted by the APY we verified on the provider's own page, current prints first; checking accounts by monthly fee, then interest; personal loans by the lowest published APR. Credit cards, brokers, IRAs, robo-advisors, private credit platforms, mortgage and HELOC lenders, payment processors, POS systems, small-business finance and tax software are sorted by an editorial score: one number the editor sets for each product after reading its terms for value, transparency, track record and accessibility. There are no sub-scores and no fixed weights.
Do affiliate commissions affect Financeradar ratings?
No. Some product links pay affiliate commissions, which are disclosed on the page. Commissions never affect scores, rankings, or which pros and cons appear. No position in any ranking is paid for, and no provider can pay to edit a review or remove a con.
Why does Financeradar list closed and failed products?
Because a product's history is decision-relevant. Knowing that PeerStreet went bankrupt or that Yieldstreet rebranded after losses matters when comparing platforms. Instead of silently dropping dead products, Financeradar keeps a public graveyard of closed, failed, and renamed products with their history.
Think a rating is wrong? Tell us why.
Disputes go to louis@dupple.com. Bring the URL plus a primary source: the issuer's page, a filing, or a regulator notice. We re-verify and document the outcome, and corrections update the "as of" stamp.