Why Stocks Fall After Good Results: The Concall Effect

The results look great on paper — PAT up 18%, revenue beats estimates by a margin. You wake up the next morning expecting green. The stock is down 6%.

This is one of the most common shocks in results season, and it has a precise explanation: the share price does not react to last quarter's numbers. It reacts to the next four quarters' numbers, as implied by what management just said on the concall. Across 24,328 management commitments tracked on Inve across 1,648 listed Indian companies, only about 57.8% end up delivered as stated — that is pooled across the 16,842 commitments old enough to carry a verdict, leaving out the ones still too new to judge (Inve data, 4 September 2026). The market knows this. When a concall suggests the next cycle of guidance will be weaker — or that management is no longer giving hard commitments at all — the stock falls, regardless of the headline.

Understanding this mechanism is the difference between reading a results season and actually using it.

The market prices the future, not the past

Here is the core idea, stated plainly. A quarterly result is history. A share price is a discounted estimate of all future earnings. When results land, analysts and institutional investors are not asking "was Q3 good?" — they already knew the broad shape of the quarter three days ago from channel-check data, sector trackers, and advance guidance. What they do not know is what management just said about Q4, FY27, margins, deal wins, and the order book.

The concall is the primary mechanism through which that forward information enters the market. And it is the only regular, public, unscripted forum where management faces live questions — from analysts who are paid specifically to catch inconsistencies and press on risks.

This is why a stock with strong PAT growth can fall 8% on results day while a stock that missed revenue estimates rises 4%: the first company's management just guided softer margins for the next two quarters; the second's guidance came in better than feared.

If you are reading only the results press release and ignoring the transcript, you are reading the data the market already priced — and missing the signal that is re-pricing the stock in front of you.

What "guidance" actually means (and when to discount it)

Guidance is management's forward-looking commitment: revenue growth, margin bands, capex spend, credit growth, order intake. Not all guidance is equal.

A hard numeric target — "we expect EBITDA margins of 18–19% in H2 FY27" — is a commitment that analysts will hold management to, quarter by quarter. Inve's guidance specificity score (0–100) tracks exactly this: how many of management's forward statements are pinned to a number, a period, and a named executive rather than floating as vague directional optimism.

Directional hand-waves — "we remain confident of healthy growth," "the business is tracking well" — contain no actionable information and carry no accountability. Companies that consistently issue only directional guidance tend to score low on Inve's guidance specificity score. The market prices this uncertainty as risk.

The most dangerous move in results season is treating directional language as a hard commitment and then being surprised when the next quarter's numbers come in below your implied expectation. "Healthy growth" is not a guidance range. It is the absence of one.

The Infosys FY26 guidance walk: reading four calls, not one

The clearest Indian large-cap illustration of how guidance evolves across a results cycle — and why tracking it call by call matters — is Infosys's FY26 constant-currency revenue growth guidance.

This is an example of how Inve's Concall AI parses and sequences guidance data. It is not a recommendation on Infosys shares.

A sequence of serial upward revisions like this is significant information — not just the individual quarterly number, but the pattern. A management team that guides conservatively and revises up is behaving differently from one that front-loads optimism and revises down. Both sequences can produce the same reported revenue, but one builds credibility and the other destroys it.

Inve's Concall AI reads every Infosys earnings call and investor day the same way: each forward commitment is logged, carried across quarters and closed out with a verdict — achieved, on track, missed or ghosted — which rolls up into a kept rate, a transparency trend and an overall trust grade. Those roll-ups move with every new call, so read them live on the ledger rather than from a number frozen into an article.

One logged deflection is a useful illustration. On a recent Infosys call, a Moneycontrol journalist asked about AI-services revenue specifically. CEO Salil Parekh's response, graded evasive by Concall AI: "It is much more growth but we are not giving the number, but it is growing very nicely here. … Is it 10% or 50%, we are not sharing the number."

That exchange does not appear in the investor presentation. It does not appear in the press release. It appears only in the transcript — and it is the kind of deflection that a trust score and an evasion log are designed to surface.

You can explore Infosys's full commitment record at inve.money/promise-tracker/INFY.

Management tone: the signal before the signal

The Q&A section of a concall is the most information-dense 30–45 minutes in a company's public communication calendar. The prepared remarks are marketing; the Q&A is examination. Management cannot see the questions in advance. Their answers — or their evasions — are unscripted.

Inve's Concall AI scores four dimensions of management communication per quarter:

• Confidence score (0–100) — based on language certainty, with the strongest and weakest moments quoted verbatim.
• Transparency score (0–100) — how directly management answered analyst questions; counts of direct versus deflected answers; areas of evasion listed explicitly.
• Data richness score (0–100) — did management back claims with actual figures, or use vague language like "significant" and "meaningful"?
• Guidance specificity score (0–100) — proportion of forward-looking statements that carry a number, a period, and attribution to a named executive.

These four scores combine into an overall tone assessment — Very Bullish through Defensive — and feed into the trust grade on the Promise Tracker.

The pattern to watch is tone drift across quarters. A management team that scores 78 on transparency in Q1, then 61 in Q2, then 54 in Q3 is showing you a deteriorating willingness to answer direct questions. This kind of drift typically precedes a guidance cut or a miss — it is the signal before the signal.

At the KPI Screener, you can compare quantitative operational metrics across companies in the same sector, which adds context to whether a tone shift is company-specific or sector-wide.

Guidance that quietly went silent: the silence that costs you money

One of the most underrated risks in results season is not what management says — it is what they stop saying.

Across the 24,328 commitments Inve tracks, 2,271 were simply ghosted — management made a public commitment, then never mentioned it again on any subsequent earnings call in our parsed corpus (Inve data, 4 September 2026). That is 9.3% of everything we track, a figure that includes commitments still too new to resolve; measured only against the 16,842 commitments old enough to carry a verdict, it is 13.5% — closer to one in seven that disappears from the record entirely. And 60.7% of the 1,648 companies tracked — 1,000 of them — have at least one piece of guidance that quietly went silent on their ledger.

This is not a failure mode that appears in a single quarter's results. It only becomes visible when you track what was previously committed to against what is currently being discussed. A diversified retail portfolio of 15–25 stocks means reading 15–25 transcripts per quarter — roughly 300–600 pages of dense financial communication every 90 days. Nobody does this manually at scale. Working from a structured earnings-call summary is how that reading gets done across a portfolio.

The Promise Tracker is built specifically to hold this multi-quarter memory. Every commitment is logged with its birth quarter, original wording, current verdict — Achieved, Missed, On Track, or Ghosted — and a quarter-by-quarter status trail. The friction ledger shows which topics management has persistently avoided, how many consecutive quarters they have dodged it, and the latest deflection quote.

When a stock "inexplicably" falls after a good quarter, the reason is often visible in the commitment record: a key commitment from four quarters ago has been quietly dropped, and enough institutional investors noticed.

How a results-season workflow actually looks

Most retail investors in India spend results season reading summarised press releases and watching YouTube reactions. Both are heavily filtered. Here is a more useful sequence:

Step 1 — Check the headline numbers, then move on. Beat or miss is already priced within minutes of the NSE filing. It is table stakes.

Step 2 — Go to the guidance table. On Inve's concall page for each stock, every forward-looking commitment from the call is extracted, pinned to the speaker, and tagged with a status versus last quarter — Revised Up, Revised Down, Achieved, Missed. Has any metric guidance been quietly cut? Has a previously prominent commitment disappeared?

Step 3 — Read the Q&A highlights. The toughest analyst questions are surfaced with management's exact response and a response quality grade: Direct, Partial, or Evasive. One evasive answer on margins is yellow; three consecutive quarters of deflection on the same topic is red.

Step 4 — Check the tone scores against last quarter. Transparency score falling while the headline numbers look fine is precisely the "good results, bad concall" pattern that precedes disappointment.

Step 5 — Cross-reference the commitment record. On the Promise Tracker, check whether commitments from prior quarters are progressing or going silent. A management team with a 35% kept rate and four ghosted commitments deserves a higher risk premium than one with a 75% kept rate — irrespective of last quarter's PAT.

The Concalls page gives you access to parsed summaries drawn from 8,220 earnings calls across 1,654 companies. You do not need to read the full transcript first — but you should know where to go when the summary raises a flag.

Why 609 dodges are a data point, not an opinion

A criticism of qualitative analysis is that it is subjective. "Management sounded evasive" is an opinion. But Inve's approach is counting, not characterising. Across all tracked calls: 609 persistent dodges — topics where an analyst asked, management deflected, pivoted or gave a non-answer, and then did it again on a later call (Inve data, 4 September 2026). Each is tagged to the topic, the quarter, and the transcript segment. These are extracted from the transcripts by our own pipeline and have no external check, so read the count as indicative rather than audited.

This is the distinction between secondary-source analysis ("experts say management tone matters") and first-party evidence ("we counted 1,857 repeatedly deflected topics across 452 companies"). The first is an editorial opinion you cannot verify. The second is a data point you can drill into.

When you see a company with an evasion streak of five consecutive quarters on a specific topic — say, receivables quality, or client concentration — that streak is a quantified signal, not a vibe. It is the kind of signal that the quarterly EPS line does not carry.

Frequently asked questions