Source Hierarchy: Ranking Evidence From Filing to Commentary
A source hierarchy ranks evidence about a company by how close it sits to the underlying fact and who is accountable for it. Regulatory filings sit at the top, audited and legally the company's own statement. Call transcripts follow, then investor presentations, then broker research, then news and commentary. Lower rungs are faster and more opinionated, higher rungs slower and checkable. Every claim is traced up to the rung on which it can be verified.
Information degrades as it moves away from the fact and away from anyone who answers for it. The pattern is familiar from ordinary life. At a wedding, the story that the groom's business is in trouble arrives fourth-hand: a cousin heard it from a neighbour who heard it from someone who once supplied the shop. Nobody in that chain saw a ledger, and nobody in it will lose anything if the story is wrong. The same degradation happens to every claim about a listed company, only faster and at scale. A source hierarchyA ranking of the places a fact can come from, ordered by how reliable each one is for checking that fact. is a ladder that makes the degradation visible. A claim's weight can then be read off the rung it stands on, and a rumour can be climbed from a social post to an audited note, or watched dying on the way up.
What is a source hierarchy, and what does it rank by?
Start with something as ordinary as an electricity bill. The bill itself, printed by the utility with the meter reading on it, is the fact. A neighbour saying over the wall that bills are up this month is a report of the fact. A message forwarded on the residents' group saying that everyone's bill has doubled is commentary on the report. The amount owed is settled by the bill, and if the neighbour and the group disagree with it, the bill wins. Nobody would argue about that. Yet the same person will read a forwarded message about a company they hold shares in and act on it without ever opening the document the message is, at best, a distant echo of.
A source hierarchy ranks evidence by two properties at once: proximity, how close the source sits to the fact, and accountability, who answers if the source is wrong. Proximity is the distance the information has travelled. The audited note in Varnika Ceramics' annual report is the fact, or the nearest thing an outsider can get to it. A transcript is the company describing the fact and being questioned on it. A broker noteA research report on a listed company written by an analyst at a brokerage or investment firm, usually for that firm's clients. is an outsider interpreting the company's description. A news report is a summary of the interpretation, and a social post is a reaction to the summary. AccountabilityThe property of a source that somebody identifiable bears a cost, legal or professional, if what the source says turns out to be false. runs down the same slope. Directors sign a filing and can be prosecuted for a false one; a statutory auditor signs an opinion on the accounts; a registered analyst signs a note and must disclose interests; an editor stands behind a report; an anonymous account stands behind nothing. Down the ladder, distance from the fact grows and accountability thins, and the picture below sets the six rungs out with both properties written on each.
Two things about the ladder are worth sitting with before it is used. First, it ranks reliability for checking a fact, not usefulness in general. A good broker note can be more useful than a filing for deciding what to look at, which is a separate question from which of the two settles a fact. Second, the rungs are not interchangeable in the way a lower rung sometimes pretends. The post might have quoted the filing wrongly, quoted it out of context, or invented it. So a social post that quotes a filing is still a social post until the filing itself has been opened. The ladder does not care where a claim says it comes from; it cares where the claim has actually been traced to. In the language people use loosely, the top rung is the primary sourceA first-hand record of the fact, made by the party responsible for it, as against a second-hand account of that record. and everything beneath it is a secondary sourceA source that reports, summarises or interprets a first-hand record rather than being the record itself.. The ladder spreads that second group out. How far each source has travelled then becomes visible.
| Rung | Source | Distance from the fact | Who answers for it | Audited? |
|---|---|---|---|---|
| 6 | Annual report note, exchange filing | The fact, as an outsider can see it | Directors who sign, the statutory auditor, the exchange that receives it | Yes |
| 5 | Earnings call transcript | The company describing the fact, under questions | Management, on the record and searchable | No |
| 4 | Investor presentation | The company describing the fact, unquestioned | Management, framing chosen by them | No |
| 3 | Broker note | An outsider interpreting the company | A registered analyst with disclosed interests | No |
| 2 | News report | A summary of the interpretation | An editor, at a distance | No |
| 1 | Social post, commentary | A reaction to the summary | Nobody identifiable | No |
Which two properties does a source hierarchy rank evidence by?
A social post quotes a line from Varnika Ceramics' annual report, naming the printed side it sits on. Which rung is that claim on right now?
Why do filings sit at the top?
Think about the difference between a promise made over tea and a registered sale deed. Both may describe the same plot of land, and the promise may even be sincere. Only one of them was signed in front of a registrar, carries a stamp, sits in a public record, and exposes the person who signed it to consequences if what it says is untrue. When two accounts of the same plot disagree, nobody sensible weighs them equally. A filing is the sale deed of company information.
Filings sit at the top because they are required, audited, signed and public, so every one of the four ways a claim can go wrong has somebody attached to it. Required: a listed company does not choose whether to publish its annual accounts or to disclose a material event; the listing rules oblige it, and the exchange receives the document with a timestamp. Audited: the annual financial statements carry the statutory auditor's opinion. An independent firm tested the numbers before they were published, and the audit committee approved related dealings such as Varnika Ceramics' Rs 44,00,00,000 of freight through Talwar Logistics. Signed: directors put their names to the report and the Chief Financial Officer, Meera Iyengar, certifies the statements. Public: the document is the same for every reader, so nobody can be shown a different version. A social post has none of these four. The ladder therefore puts an annual report of 240 printed sides above a post of 240 characters even when the post is more vivid, more recent and more confident.
One caution. The top rung is not perfect. Audited does not mean true; it means tested by an independent party to a materiality standard, and the notes are where the judgements live. Filings are also slow. The annual report arrives months after the year it describes. The lateness is exactly why the lower rungs exist and why people read them. The top rung is where a claim gets settled, not where it first appears.
Varnika Ceramics' annual report is audited. Which of these does the audit opinion actually mean?
Where do transcripts and presentations sit, and why in that order?
Here is a question students often get backwards. The presentation is polished, designed, and published by the company itself; the transcript is a raw record of an hour of talking. The polished one might seem to rank higher. The polished document does not rank higher, and the reason behind that is the most useful idea the ladder carries. Two accounts of a school's annual day make the point. One is the brochure the school printed. The other is a recording of the parents' meeting where a father stood up and asked why the science lab has been shut since March. Both come from the same building. Only one contains a question the school did not choose.
The transcript ranks above the presentation because both are unaudited company statements, but the transcript contains outsiders' questions and the record of what was declined, and the deck contains only what the company chose to show. Both documents sit at the same accountability level: management is on the record for what it says in each. Where they differ is proximity, and they differ because of who holds the pen. On Varnika Ceramics' results call, the sell-side analyst Priyanka Bhat asked what sell-through was, given that dealer inventory rose 35 per cent while sales rose 4 per cent, and Meera Iyengar declined to disclose it. The refusal is recorded on printed side 16 of the transcript. Slide 7 of the presentation for the same quarter says revenue up 4 per cent, 60 dealers added, healthy channel build, and says nothing at all about sell-through. A slide cannot record an absence, and nobody puts one there. So the transcript sits closer to the fact, not because management is more honest on a call, but because somebody else got to ask. The map below places all six sources on the two axes: the transcript sits level with the deck for accountability and to its right for proximity.
Predict before reading on. Transcript or presentation: which ranks higher on the ladder, and why?
What is a broker note, and why is it below the company's own documents?
The property broker is a familiar figure. He knows the locality in a way no buyer ever manages: which builder delivers late, which lane floods, what the last three flats actually sold for. He is also paid when the flat is bought. Neither fact cancels the other. A buyer would be foolish to ignore what the broker knows and foolish to treat his conclusion as neutral. A broker note is that same figure in print.
A broker note is expert, interested and unaudited. The mix puts a note below the company's own documents for checking a fact, and often above them for finding the right question. The analyst who writes it, someone like Priyanka Bhat, has followed Varnika Ceramics for years, listens to every call, talks to dealers, and builds a model. Her firm is registered with the regulator as a research analyst and must disclose its interests. Registration and disclosure are the accountability the rung carries. But her firm may also earn from trading in the shares, from banking for the company, or from being seen to be first with a view, and the note has never been audited by anyone. So the note that says dealer inventory looks high is informed, and it is the fastest route to learning that dealer inventory is the thing to look at. The note cannot settle whether the claim is true. Settling it takes a climb: the note cites the call, so the transcript is opened next; the transcript shows the question declined, so the annual report follows. A good analyst writes the note expecting exactly that climb, and supplies the references that make it easy. The comparison below sets what the note adds against what the transcript adds: they are different jobs rather than better and worse versions of one job.
A broker note on Varnika Ceramics says dealer inventory looks high and cites printed side 16 of the call transcript. What is the note most useful for?
Two brokerages publish notes on Varnika Ceramics the same week, and both say dealer inventory looks high. Does the second note move the claim up a rung?
Where do news and commentary sit, and when are they useful?
The residents' group makes the point again. The forwarded message about doubled electricity bills is almost never the place to learn what is owed, and it is very often the place where the existence of something to check first surfaces. Speed and distance travel together. The message arrived before the bill did precisely because it did not wait for the bill.
News and commentary sit at the bottom because they are furthest from the fact and thinnest on accountability, and they are useful for exactly one thing: signalling what to go and check. A news report on Varnika Ceramics is written in hours by someone covering forty companies, drawing on a broker note, a press release, and perhaps a phone call. An editor stands behind the report, and editorial accountability is real. The reporter has still not sat with the receivables note. CommentaryOpinion about a fact rather than a report of it: a column, a post, a forum thread, a forwarded message., the social post, the forum thread, the forwarded message, has less still: no editor, no name, and no cost for being wrong. Watch the attributionThe words a report uses to say where its information came from: the company said in a filing, against sources said, or nothing at all. in each. A line reading, the company said in its results filing, points straight up the ladder to a document that can be opened. A line reading, sources close to the matter said, points nowhere that can be followed, and a post that opens with everyone knows points at the floor. The best use of the bottom rungs is as a pointer: the news report is read, the claim is noted, the stated origin is noted, and the climb begins. If both drew on the same broker note, two agreeing reports corroborateTo confirm a claim with a second, independent source that did not draw on the first one. nothing. Two low rungs in agreement are one source counted twice. The timeline below shows why the bottom rungs are always first to speak and last to be checkable.
A broker note and a news report both say Varnika Ceramics' dealer channel is overstocked. Is the claim verified?
How does the hierarchy resolve the dealer-inventory claim about Varnika Ceramics?
Now the whole climb, with the dealer-inventory thread. A social post appears one evening: Varnika Ceramics is stuffing its channel. No number, no name, no source. On the ladder that is rung 1, and its weight is close to nothing. The post does carry a claim specific enough to check, so the climb begins.
The claim survives every rung and lands on an audited note. Landing on an audited note is what turns gossip into evidence. Rung 2: a business daily reports that dealer stock at Varnika Ceramics looks elevated, citing a broker note. The report has added a named intermediate source, nothing more. Rung 3: the broker note says dealer inventory rose about 35 per cent in the quarter against sales up 4 per cent, cites the earnings call, mentions dealer conversations and carries the firm's interest disclosures. Now the claim has numbers and a trail. Rung 4: slide 7 of the results presentation says revenue up 4 per cent, 60 dealers added, healthy channel build. The company confirms channel build in its own framing and gives no sell-through. Rung 5: printed side 16 of the transcript, where Priyanka Bhat asks for sell-through twice and Meera Iyengar declines. The number exists, was asked for and was withheld, on the record. Rung 6: the annual report's receivables note shows receivables up 30 per cent against 4 per cent sales growth, and the provisions note shows a first-time returns provision. Audited, signed, filed. Nothing on rung 6 says the word stuffing, and it does not need to. The audited note now supports what the post asserted, with a number nobody chose to show.
| Rung | What is opened | What it adds to the claim | Weight after this rung |
|---|---|---|---|
| 1 | The social post | A claim specific enough to check; no number, no name | Close to nothing |
| 2 | The news report | Names a broker note as its source | A pointer |
| 3 | The broker note | Dealer inventory up about 35 per cent, sales up 4 per cent; cites the call | Informed, interested |
| 4 | Slide 7 of the presentation | Revenue up 4 per cent, 60 dealers added, healthy channel build; no sell-through | Company confirms build, on its terms |
| 5 | Printed side 16 of the transcript | Sell-through asked for twice, declined on the record | Withheld number, now documented |
| 6 | Receivables and provisions notes | Receivables up 30 per cent against 4 per cent sales; first-time returns provision | Evidence, audited |
A prediction before the simulation runs. With the simulation switched to the false rumour, that the Gujarat plant was shut for a month, on which rung does its weight collapse, and why?
Climb the ladder with a claim and watch its weight change.
Pick one of two claims about Varnika Ceramics, then climb: use the buttons, the slider, or click any rung in the picture. The evidence meter and the panel of what this rung adds redraw at every step. The default is the true dealer-inventory thread at rung 1, the social post, where its weight is low. Then switch to the invented rumour and watch it die on the transcript rung.
How do lenders, analysts and households actually use the ladder?
A lender uses the ladder to decide what counts as notice. When Varnika Ceramics' bank hears a rumour that the channel is stuffed, the relationship manager does not call the company about a social post. He climbs to the quarterly filing, checks receivables days against the covenant pack the company sends under the loan agreement, and if the two disagree he has something to raise, on the record. A lender's evidence must be able to survive a dispute, so a lender lives on the top two rungs and uses the rest as an early warning.
An analyst lives on rung 3 by profession and climbs for a living. Priyanka Bhat writes the broker note, but the note is only as strong as the rungs it cites. A careful analyst footnotes the printed side of the transcript and the annual report note behind every claim, and readers trust her because they can follow her up. Practitioners do not ask whether a source is good or bad; they ask which rung it is on, what that rung can and cannot settle, and how far up the claim has actually been traced. A household investor holding a few hundred shares in Varnika Ceramics, a company with revenue of Rs 6,20,00,00,000 and 1,400 dealers, uses the ladder in the simplest way: a forwarded message is a reason to open the results filing, never a reason to sell, and the difference between those two reactions is most of what protects a small holding from a rumour.
The error that gets made, and what it costs
The reader who stops at the broker note because it agrees with the social post, and acts. The post said stuffing; the note said dealer inventory looks high; the two match, and the match feels like confirmation. Sometimes the broker is right. Sometimes the broker note was written the morning after the same social post, from the same social post, and the agreement is one source wearing two coats. Either way the reader has never opened the transcript, where the declined question lives, or the receivables note, where the audited number lives. The evidence trail stopped at the interested rung, and the checkable rungs above were never visited.
The cost is a decision whose evidence was never traced above the rung where interest and speed live. The cost is the same whether the decision turns out well or badly. The reader cannot tell the difference between having been right and having been lucky.
The post, the news report and the broker note have all been read, and they agree. What is the one action the ladder calls for before deciding anything?
References
| Source | Document | Where |
|---|---|---|
| Securities and Exchange Board of India (SEBI) | SEBI (Research Analysts) Regulations, 2014, the registration and disclosure frame for broker research | sebi.gov.in |
| SEBI | SEBI (Listing Obligations and Disclosure Requirements) Regulations, 2015, the frame that makes filings required | sebi.gov.in |
Varnika Ceramics Limited, Talwar Logistics, Rohan Talwar, Meera Iyengar and Priyanka Bhat are invented.
Educational material. Not advice on any investment, tax, budget or market position.
