Skip to content
StockSummarizer
Menu
How to analyze stocks

How to Triangulate Filings, Sell-Side Notes, and Your Own Model

How to reconcile filings, sell-side research and your own model line by line, explain each gap and treat analyst notes as context rather than conclusions. Educational, not advice.

By Published Updated

White concrete triangulation pillar on a grassy hilltop with a wide valley view
Triangulation pillar photograph by Anthony O'Neil, CC BY-SA 2.0, via Wikimedia Commons. Cropped and resized.

You have three views of the same company. The filings say what happened. Sell-side research notes (from brokerage analysts) say what the market expects and why. Your own model says what you think. Triangulating means comparing the three on purpose, line by line, and writing down where they disagree. Those disagreements are either your edge or your error, and you need to know which.

Educational only, not financial advice. This is a workflow for reconciling sources. It applies the source hierarchy framework to numbers. For triangulating valuation methods (multiples vs DCF vs peers), see triangulating multiples, DCF and peer checks.

Direct answer

Build your model from the filings first, so your starting numbers are audited facts. Then compare it line by line with whatever sell-side or consensus information you can access: revenue growth, margins, capex, share count, and FCF. For each meaningful gap, write one sentence on why you differ. If you can’t explain a gap, assume it’s your error until you can. Treat sell-side notes as context. Read the analyst’s conflict disclosures, and never use a price target in place of your own reasoning.

FilingsWhat happened(audited / filed)Sell-side notesWhat consensusexpects + whyYour modelWhat you thinkand whyReconcileWhere do thethree disagree?Log the gapVariant view oryour error?
The useful output is the list of disagreements and whether you can explain each one.

What each source is good and bad at

SourceGood forWeak atCheck
Filings (FS, MD&A, AIF)Facts, definitions, segment data, risksForward view (limited to forward-looking info)Reconcile non-GAAP to GAAP
Sell-side notesIndustry context, model structure, consensus expectationsIndependence; tendency to clusterConflict disclosures; date of estimate
Your modelExplicit assumptions you understandBlind spots; overconfidenceCompare to history and to peers
Consensus aggregatesQuick read of expectationsStale or mixed estimates; hides dispersionNumber of estimates; range
Audited filings & MD&AFacts and management's required discussionYour model (built on filings)Your assumptions, explicitly writtenSell-side research notesUseful context; check the conflict disclosuresConsensus aggregatesAverages hide the spread and stale estimatesHeadlines & price targetsLowest signal for decisions
This is a variant of the source hierarchy framework, applied to reconciling numbers rather than reading documents.

The reconciliation table

Line itemFilings (last FY)Your model (next FY)Consensus (next FY)Gap explained?
Revenue growth5.0%6.0%4.0%Yes: contract start in Q2 per MD&A
EBITDA margin24.0%23.5%25.0%Partly: I assume wage inflation; need source
Capex / revenue9.0%12.0%9.0%Yes: capex plan in MD&A (capex post)
Diluted shares100.0M101.0M100.5MYes: RSU vesting (share-count post)

All figures are hypothetical. Gaps marked “Partly” go into your notes as open questions, with a date to resolve them.

Revenue growth gap (pts)2EBITDA margin gap (pts)1.5Capex / revenue gap (pts)3Diluted share count gap (%)0.5
Hypothetical. The capex gap is the biggest, so that's where to look first: is it your assumption or a timing difference?

Access and independence notes

  • Many brokerages give clients some research access. Libraries may offer financial databases too (check with your local library).
  • Research reports usually include disclosures about the dealer’s relationships with the issuer. Read them before weighing the conclusion.
  • Price targets and ratings are outputs. The assumptions in the model are what you should actually compare.
  • For peer comparisons, use peer tables without cherry-picking.

Sources

  • NI 51-102 Part 4A (forward-looking information in filings) and NI 52-112 (non-GAAP reconciliation). Confirm the current text on the CSA website.
  • CIRO (Canadian Investment Regulatory Organization) Investment Dealer and Partially Consolidated Rules on research reports and analyst conflict disclosure. Confirm the current rule text on ciro.ca.
  • Michaely & Womack (1999), “Conflict of interest and the credibility of underwriter analyst recommendations,” Review of Financial Studies.

Media credits

  • Hero: “Triangulation pillar, top of Wavering Down - geograph.org.uk - 1855262.jpg” by Anthony O’Neil, CC BY-SA 2.0, via Wikimedia Commons. Cropped and resized to 1280x720. Real photograph, not AI-generated.

By Andrew · stocksummarizer.com

Guides · coming soon

Want the playbook in one place?

We are compiling free literacy hubs into paid ebook playbooks (Research, Valuation without fake precision, Dividend safety). Soft launch only for now — keep using these free guides, and browse ebooks when a guide ships.

Preview ebook shelf

Educational only — not financial advice. Do your own research. See our disclosures.