Editorial Policy

Who writes FinSage

Every page is written and maintained by pasindu.nbuss, an Electrical Engineering graduate of the University of Moratuwa who works as a Software Engineer and has experience in software development and financial-market data. He is not a licensed financial adviser, investment adviser, insurance adviser, accountant, attorney or financial planner.

There is no external review board and no professional reviewer. FinSage does not claim that its content has been reviewed by a licensed financial professional, because it has not been. What it does instead is show the arithmetic and cite the sources, so that any claim can be checked independently rather than taken on authority.

How topics are selected

A topic makes it onto the site if it meets all four tests:

• It answers one clear question that a real person would type or ask — “how is EMI calculated”, not “everything about loans”.
• It can be settled with arithmetic or a citable source. Predictions, market calls and “best product” rankings are excluded by design; FinSage has no basis for them.
• It is not already covered. Existing pages are expanded rather than duplicated with a new URL, because thin near-identical pages help nobody.
• It can be written honestly at this level of expertise — concepts, calculations and how things work. Topics requiring a licence to advise on, such as which specific product suits an individual, are out of scope.

Article ideas come mostly from questions the calculators raise: if a result surprises people, the explanation deserves a page.

How calculations are checked

This is the part that matters most on a site full of numbers.

1. Standard formulas only. Calculators use the conventional finance formulas — amortisation for loans, the ordinary-annuity future value for regular contributions, compound interest for lump sums. The formulas are printed on the relevant article pages so you can check them rather than trust them.
2. Formulas and code must agree. Every published figure is generated by, or reconciled against, the same function the on-site calculator runs. When an article quotes “$400 a month at 8% becomes $596,144 after 30 years”, that figure comes from the growth calculator's own loop, not from a separate estimate.
3. Edge cases are tested. Each calculator is checked at its boundaries — zero interest rate, zero contribution, a one-year term, the minimum and maximum of every slider — to confirm it degrades sensibly rather than producing a nonsensical or infinite result.
4. Independent cross-checks. Results are compared against an independent implementation (a separate script) and, where available, against a regulator's own public calculator such as the US SEC's compound-interest tool.
5. Assumptions are published, not buried. Contribution timing (start or end of period), compounding frequency, and whether fees, taxes and inflation are included are stated next to every result. Two calculators that disagree usually differ in assumptions, not arithmetic.
6. Rounding is disclosed. Displayed values are rounded for readability; the underlying calculation is not, so small differences from hand-checks are expected.

How sources are selected

Sources are ranked, and the highest available tier is used:

1. Primary official sources — financial regulators, central banks, tax authorities, statistical agencies, deposit-insurance schemes, and legislation itself.
2. Original research — peer-reviewed papers and the original studies behind widely repeated rules of thumb, cited by author, year and journal so they can be located.
3. Recognised benchmark and index providers for market data definitions.
4. Official product documentation — a fund's own factsheet or key information document, an issuer's own attestation, an insurer's own policy wording.

What is deliberately not used as evidence: personal-finance blogs, content-marketing pages published by companies selling the product being discussed, aggregator sites that cite each other, social media, and AI-generated summaries. Where an industry body is cited because it holds the best data — for example the World Gold Council on gold — the commercial interest is stated in the citation itself so you can weigh it.

Every article ends with a visible “Sources and further reading” list. If a claim on FinSage carries no formula and no source, it should be read as opinion, and you are entitled to treat it that way.

How statistics and figures are handled

• No fabricated figures. No statistic, survey result, price or return is invented, estimated-and-presented-as-measured, or attributed to a source that does not contain it.
• Windows are stated. A historical return without a start and end date is meaningless, so FinSage either gives the window or declines to publish a single headline number — as it does for long-run gold returns.
• Illustrative prices are labelled illustrative, along with the market they describe.
• Uncertainty is stated rather than smoothed away. Where researchers disagree, the page says so instead of picking the most quotable figure.

Country and jurisdiction

FinSage has an international audience, and financial rules are national. Any page whose substance depends on one country's law, tax code, products or scoring systems carries a visible scope badge — for example “Primarily US-focused examples” — and separates the universal mechanics from the country-specific detail. Where a topic differs materially between markets, the differences are described rather than ignored.

This site does not attempt to give jurisdiction-specific legal, tax or regulatory guidance for any country, and none of its content should be relied on as a statement of the law where you live.

Corrections

Errors are inevitable; hiding them is a choice. The process:

1. Report an error to pasindu.nbuss@gmail.com with the page URL, the exact text or figure, and a source if you have one.
2. Reports are checked against the primary source or by re-running the calculation.
3. Substantive errors — a wrong number, a wrong formula, a misstated rule, a claim attributed to a source that does not support it — are corrected as a priority, and the page's “last reviewed” date is updated.
4. Material corrections are disclosed on the page, with a short note saying what was wrong, what it now says, and the date. The original error is described rather than silently deleted.
5. Typographical and formatting fixes are made without a note.
6. If a claim turns out to be unsupportable and cannot be fixed, it is removed rather than softened.

Updates and review dates

• Every article shows a published date and a last reviewed date. “Last reviewed” means a human re-read the page and re-checked its figures and links on that date — not that a typo was fixed.
• Articles are reviewed at least annually, and sooner when a source updates its figures, a rule changes, or a reader reports a problem.
• Where a change alters the substance of a page, the correction note above applies.
• Live market data is a different thing entirely: it comes from third-party APIs, is cached (about 90 seconds for crypto and the tokenised gold proxy, up to 6 hours for exchange rates), and carries its own timestamp on the page. It is not editorially reviewed and should never be used to transact — see how to verify live market data.

Advertising and commercial relationships

Current status: no advertising runs on this site, and no advertising account has been approved. There are no ad units, no ad placeholders and no advertising scripts loading anywhere.

The intention is to fund hosting through display advertising in future. The rules that will apply:

• Advertising will never influence editorial content. No advertiser will be able to commission, review, approve, amend or veto an article, and no article will be written to suit one.
• Ads will be clearly distinguishable from editorial content and labelled as advertising.
• No sponsored posts, paid links, paid guest articles or affiliate links are accepted. If that ever changes, existing pages will not be retro-fitted with commercial links.
• Any commercial relationship will be disclosed prominently on the page where it applies, in plain language, at the point a reader encounters it — not only in a policy page.
• No product recommendations are for sale. FinSage does not rank or recommend specific financial products at all, which removes the most common conflict in this field.
• Advertising, when introduced, will be described in the privacy policy before it goes live, including what cookies it sets and how to opt out of personalised advertising.

Editorial content is not personalised advice

Use of AI tools

Software tools, including AI assistants, are used for drafting support, code and proofreading, in the same way a spellchecker or an IDE is. They are not used as a source of facts. Every figure is either computed from a formula shown on the site or verified against a cited primary source by a human before publication, and every page is read end to end by pasindu.nbuss, who is responsible for its accuracy.

Questions about this policy

Email pasindu.nbuss@gmail.com, or see the contact page.