roic.ai Alternative: Where the ROIC Number Comes From
By Chad Hartman
Published · Last updated
There is no line in any SEC filing labeled ROIC. No company reports it, no accounting standard defines it, and no regulator specifies its inputs. Every platform publishing a ROIC figure built that figure — from a numerator that has to be tax-adjusted and a denominator that has to be assembled out of three separate balance sheet decisions. So the question worth asking about any platform named after the metric is not whether its ROIC is accurate. It is what happened to the filing between the moment it was submitted and the moment the number appeared on screen.
What roic.ai Does Well
roic.ai is a well-built platform with a clear thesis, and several of its strengths have no equivalent here.
Coverage is the first and largest. Its data spans tens of thousands of companies across more than seventy global exchanges, which means a European industrial or an Asian semiconductor name gets the same treatment as a US large cap. GeminIQ covers SEC filers. For an investor holding anything outside the US reporting system, that difference settles the question before any other consideration enters.
History depth is the second. roic.ai carries decades of financial statements, well beyond what most retail-accessible platforms offer, and for long-cycle businesses or full-cycle analysis that depth is the point.
Programmatic access is the third. A REST API with a free tier, spreadsheet endpoints for Excel and Google Sheets, an MCP server, and a ChatGPT integration mean the data can move into a model, an agent, or a chat workflow rather than only being read on a page. Earnings call transcripts sit alongside the statements, and the single-page company overview is a clean piece of product design — everything about a business on one screen, which remains the fastest way to form a first impression of a company.
None of that is in dispute. The question this post is about is narrower.
Standardization Happens Before You See the Number
Both platforms draw from the same source. Every US filing either one reads is the same 10-K or 10-Q sitting on EDGAR, tagged in XBRL by the company that filed it.
What differs is the step immediately after ingestion. roic.ai standardizes and reformats filings into its own presentation templates — a consistent set of line items applied across every company, so a balance sheet from one filer lines up visually with a balance sheet from another. That is a deliberate design choice with a real benefit: comparison across companies gets easier when everyone's statements share a shape.
GeminIQ skips that step. Filings are parsed into structured data with the issuer's own line items and XBRL tags preserved, which means the balance sheet on screen has the same rows, in the same arrangement, as the balance sheet the company filed. Two companies looking different from each other counts as information rather than as a formatting problem to solve.
Neither approach is wrong. They optimize for different things, and the tradeoff only becomes visible when a calculated metric sits downstream of the choice.
Why ROIC Is the Metric Where It Shows
ROIC is the most standardization-sensitive metric in common use, and the sensitivity lives almost entirely in the denominator.
Invested Capital is not reported anywhere. It gets built from total equity plus total debt minus excess cash, and each of those three inputs carries a decision the filing never makes. Which borrowings count as debt — and specifically whether operating lease liabilities and the current portion of long-term debt belong in the total. How much cash is excess versus operationally necessary. Whether the figure feeding the ratio is a single period's snapshot or a two-period average.
Standardizing a balance sheet means mapping the issuer's line items onto template rows, and that mapping is where debt classification gets decided. A filer reporting five separate borrowing lines under its own labels has to collapse into a template's debt row somehow, and the collapse is a judgment call made once, at ingestion, invisibly to whoever reads the resulting ROIC six steps later.
The consequence is not error. It is that the number becomes difficult to reproduce. An investor who calculates invested capital by hand from the 10-K and gets a different answer than the platform shows has no way to locate the divergence, because the intermediate step — which filing line became which template row — is never exposed.
The Same Company, Two Defensible Figures
Two platforms reading the identical filing can publish ROIC figures a full percentage point or more apart without either one making a mistake.
Include the current portion of long-term debt in invested capital and the capital base grows, so ROIC falls. Exclude it and the base shrinks, so ROIC rises. Fold operating lease liabilities into debt and a retailer or airline looks materially more capital-intensive than it does under a convention that leaves leases out. Subtract all cash rather than only the excess above an operating threshold and a cash-rich business posts a dramatically higher return on a dramatically smaller base.
Every one of those is a defensible methodology. None of them is disclosed on the face of a ROIC figure. That is why a metric that looks like a fact is closer to a position — and why reconciling two platforms' ROIC figures is a methodology exercise rather than an arithmetic one.
What Traceability Actually Buys
GeminIQ's answer is not a claim that its convention is the correct one. It is that the convention stays visible, consistent, and checkable.
Every figure in the Financial Statements view carries its XBRL tag, so any number traces back to the specific tag in the specific filing that produced it. The metric library documents what each calculation includes: Return on Invested Capital states its denominator, Invested Capital states its debt definition and its excess cash rule, Net Debt-to-EBITDA states what enters total debt. One convention applies to every company in the database, which is the property that makes a cross-company screen mean anything at all.
The practical test is simple. Take a ROIC figure, open the filing it came from, and rebuild it by hand. If the platform's inputs are traceable, the rebuild either matches or the difference is locatable in a specific decision. If the statements were reformatted on the way in, the rebuild will disagree and the reason will not be recoverable from anything on screen.
Where Each One Fits
Choose roic.ai when coverage or access shape the requirement. Non-US holdings, decades of history, API or agent-based workflows, and earnings transcripts alongside the statements are all reasons it is the better tool, and no amount of traceability compensates for a company that isn't in the database.
Choose GeminIQ when the requirement is verification. As-filed line items with XBRL tags attached, calculated metrics whose conventions are documented rather than inferred, and a screener built on inputs that reconcile against the source filing are what the platform optimizes for — at the cost of covering SEC filers only.
The honest framing is that these tools disagree about what a financial statement is for. One treats it as data to be made comparable. The other treats it as a document that already said something specific, and holds that a metric derived from it should still be recoverable from the original. For most investors that distinction never surfaces. For anyone rebuilding a number and finding it doesn't match, it is the whole story.
Frequently Asked Questions
What is the best roic.ai alternative?
It depends on what roic.ai was being used for. For global coverage, deep history, or API and agent access, the alternatives are other broad-coverage data providers. For US filing data where every figure traces back to its XBRL tag in the source filing, the alternative is an as-filed platform — a different design, not a cheaper version of the same one.
Does roic.ai use SEC data?
Yes, for US companies. Both roic.ai and GeminIQ read the same 10-K and 10-Q filings from EDGAR. The difference is what each does after ingestion: roic.ai standardizes statements into consistent templates, while GeminIQ preserves the issuer's own line items and XBRL tags.
Why do ROIC figures differ between platforms?
Almost always because of the denominator. Invested capital is not reported in any filing and has to be assembled from equity, debt, and cash — and whether operating leases count as debt, whether the current portion of long-term debt is included, and how much cash counts as excess are three separate methodology choices. Each is defensible, and together they can move ROIC by a percentage point or more.
Which platform has better data?
Neither, on accuracy. Both read the same filings. They differ on what gets preserved: standardized templates make cross-company comparison visually easier, while as-filed structure makes any individual figure reproducible from the source document. Which one is better depends entirely on whether the work involves scanning many companies or verifying one.
Does GeminIQ cover non-US companies?
No. GeminIQ covers SEC filers, which means US-listed companies filing 10-K and 10-Q reports. Investors holding companies outside the SEC reporting system need a platform with international exchange coverage.
Research Faster. Invest Smarter.
Most financial websites rely on third-party aggregators that simplify or process data before you ever see it. We built GeminIQ because we believe you deserve a better fundamental analysis tool—one that goes beyond basic price charts and processed numbers. We extract our data directly from SEC 10-K and 10-Q filings to ensure that when you look at a balance sheet or a cash flow statement, you are seeing the numbers exactly how the company reported them. Our goal is to give you the tools to verify the narrative for yourself using clean, traceable data. Start researching now at GeminIQ.com.
Disclaimer: The content in this blog is for educational and entertainment purposes only and does not constitute financial, legal, or tax advice. Investing involves risk, including the loss of principal. The views expressed are my own and not intended as financial advice or a guarantee of future performance.