Trust

Editorial Policy and Corrections

Personal finance is full of confidently wrong arithmetic. This page describes how CompoundCalc decides what to publish, how every printed number is verified before it goes live, and what happens when something turns out to be wrong anyway.

Who publishes this site

CompoundCalc is an independently operated site with no parent company, no investor and no commercial relationship with any bank, broker, platform or fund. It is not a content farm buying articles by the word: pages are written against a specific reader question, and pages that answer nothing get deleted rather than padded.

The site has no affiliate links. It earns money from display advertising only, which means there is no product it benefits from steering you toward. If that ever changes, the disclosure will appear on this page and on the affected pages before the first link goes live.

The rule that governs every number

No figure is published until it has been reproduced independently of the code that produced it.

Every worked example, comparison table and “about $X” claim on this site is recomputed outside the site's own JavaScript and the two results are compared before publication. If they disagree, nothing is published until the disagreement is understood.

In practice that means three checks:

  • Against the closed form. Lump-sum growth is checked against FV = P(1 + r/n)nt and contribution growth against the ordinary-annuity formula FV = PMT × [((1 + r/n)nt − 1) / (r/n)].
  • Against a period-by-period simulation. The site's engine steps through every compounding period rather than applying a single annual multiplier, which is where most published compound interest tables quietly go wrong. The simulation and the closed form must agree.
  • Against a known identity. Where one exists, it is used as a trap. Beginning-of-period contributions must come out to exactly 1 + r/n times end-of-period contributions, for example — a result that is off by even a fraction of a percent means the timing logic is broken.

This process has caught real errors before publication, including a worked example that overstated interest earned by a factor of two and a rate-comparison claim that overstated the benefit of moving from 7% to 8% by roughly $2,400. Both were wrong in the direction that flatters the site. That is exactly why the check exists.

The formulas, conventions and rounding rules are set out in full on the methodology page.

What gets written, and what does not

  • Tool first. If a page cannot help you calculate, compare or verify something concrete, it should not exist.
  • Assumptions stated on the page. Rate, compounding frequency, contribution timing and whether tax and fees are excluded are named where the number appears, not buried in a footnote.
  • No invented precision. Long-run projections are rounded to a sensible number of significant figures. Quoting a 30-year projection to the cent implies an accuracy that does not exist.
  • No fear or hype. No “secret the banks don't want you to know”, no guaranteed returns, no urgency. Compound interest is remarkable enough without dressing it up.
  • Ranges, not single answers. Where a rate assumption drives the result, the page shows more than one rate so the sensitivity is visible.

Review cycle

Calculator pages are re-tested whenever the shared calculation engine changes, and the engine's outputs are re-verified against the reference formulas at the same time. Guides are reviewed at least annually, and sooner when a rule they reference changes — contribution limits and tax wrappers being the usual triggers. The Updated date shown on a page reflects the last substantive review of that page's content, not an automated timestamp bump.

Corrections policy

Mistakes are fixed rather than quietly deleted.

  • Report an error to contact@compoundcalc.com. Naming the page and the figure you think is wrong gets it resolved fastest.
  • Reports of a numerical error are checked against the reference calculation, normally within a few days.
  • If the number was wrong, it is corrected and the page's Updated date is changed.
  • Where a correction materially changes the conclusion a reader would draw, a short note is left on the page saying what changed and when.
  • Typos and clarifications are fixed without a note.

Advertising and editorial independence

Advertising pays for the site, under three standing rules:

  • Ads are visually separated from content and carry an Advertisement label.
  • Advertisers have no input into what any guide says, and no page is written to please one.
  • Methodology, legal and policy pages carry no advertising at all, because those are the pages a reader consults precisely when they are deciding whether to trust the site.

Ad placement is also constrained by the reading experience: a calculator is never pushed below an advertisement, because the calculator is the reason you came.

Sources

The mathematics on this site is standard and not in dispute; it needs derivation, not citation, and each formula is shown so you can check it yourself. Where a page states a fact that is not mathematics — a contribution limit, a tax rule, a historical average — the source is named on that page and the date it was checked is given, because those figures go stale.

Use of automated tools

Software is used for drafting, code generation and bulk checks, as it is at most publishers now. It is not used to decide what is true. No figure reaches a published page on the strength of a language model having produced it: it reaches the page only after the independent verification described above. The site's position is simple — on a page about money, an unverified number is worse than no number.