What to Check After Running a Stock Screener in India

A clean stock screen leaves you with a harder question: which of these businesses deserves more reading? The ratios have narrowed the field. They have not checked whether the expansion, cash conversion or margin improvement in your valuation still matches what management is saying.

Imagine a screen for high returns on capital, modest debt and growing profits that leaves twenty-five candidates. That shortlist is an illustration, not a measured success rate. The useful next step is to put each candidate's financial record beside its guidance history, keeping the original metric and deadline visible.

There is a reason to do this. Inve's production snapshot of 4 September 2026 tracked 24,328 management commitments across 1,648 Indian companies. Of the 16,842 commitments with a verdict other than new, 2,271 were classified ghosted: 13.5% of those 16,842 eligible commitments, or 9.3% of all 24,328 tracked commitments.

Another 7,486 were still new. Here, eligible means there was a later call on which a commitment could be assessed; it does not mean every target had reached its deadline or a final outcome.

These are model-assigned classifications of the parsed record. A missing update can also be in an investor presentation, results release or annual report. The guidance ledger is an index to those questions, not proof of misconduct.

Company examples below illustrate a research method. They are not recommendations to buy, sell or hold a security.

Stage 1 narrows the field; Stage 2 checks the assumptions

Think of the first screen as reading a candidate's CV and the next stage as checking the work behind it. A good CV matters. So does finding out what happened when a project ran late.

Screener.in's screening guide describes custom queries and columns for comparing companies. Those are useful for a financial shortlist. This article compares research jobs, not the entire feature sets of competing platforms: a site that provides ratio tables may also provide transcripts, presentations and other research tools.

For example, compare return on capital employed with return on equity before interpreting a high ROE as business quality. Then read the original source behind any unusually attractive ratio. Annual audited accounts and interim results are different evidence; a quarterly number should not automatically be described as audited.

The three-filter screening workflow helps form the shortlist. The next stage should test the assumptions that survived it.

Kaynes: growing profit did not settle the cash question

Kaynes Technology shows why a revenue-and-profit screen needs a cash-flow follow-up. Summing the four Apr–Mar quarterly financial rows gives consolidated revenue of ₹1,126 crore in FY23 and ₹3,626 crore in FY26, about 3.2 times as much. Net profit rose from ₹95 crore in FY23 to ₹183 crore in FY24, ₹293 crore in FY25 and ₹364 crore in FY26 (Inve data; stored quarterly figures are rounded).

But consolidated operating cash flow was negative ₹42 crore in FY23, positive ₹88 crore in FY24, negative ₹82 crore in FY25 and negative ₹600 crore in FY26. Borrowings rose from ₹155 crore at March 2023 to ₹913 crore at March 2026, about 5.9 times as much. The cash-flow series is from Screener.in's consolidated Kaynes statements; the borrowing endpoints also match Inve's balance-sheet rows.

That is a reason to investigate profit versus operating cash flow, not an automatic rejection. A growing business can absorb cash in receivables and inventory. Capital expenditure is a separate investing cash-flow item, however: spending on a new factory does not by itself explain negative operating cash flow.

Now compare the cash results with management's own timetable:

Source: Inve guidance histories for Kaynes, FY25–FY26; Kaynes concall archive. Working-capital days here use the call's metric, not Screener.in's separately calculated ratio.

A qualification on the original working-capital target matters. In the Q1 FY26 transcript, the CFO described a 70-day objective excluding an extraordinary ₹350 crore acquired receivable. He also expected financing or collection arrangements to address that item. Comparing the target with subsequent reported days requires retaining that exclusion; presenting the two figures as automatically like-for-like would overstate the case.

The narrower conclusion is still useful. The positive-OCF targets were not met within the stated financial years. The reader should ask what changed in the business model and what evidence now supports the replacement timetable. That question is more precise than either dismissing the company for cash consumption or accepting the next assurance because revenue grew.

Three checks after the screen

1. Keep the original horizon beside the new one

Eternal, formerly Zomato, provides a good example of why the horizon belongs next to the percentage. On its Q1 FY25 call, management said food-delivery GOV growth above 20% could continue in the near term, with growth then running at 27%–28%. On the Q4 FY25 call, after quarterly growth slowed to 16%, management emphasised a long-term 4–5-year CAGR interpretation.

The qualification cuts both ways. Management also said on that later call that it had used the long-term framing previously. The record supports a change in near-term emphasis; it does not establish that the longer horizon was invented that day. The Q4 FY25 transcript lets you read that qualification in context.

Our Eternal guidance discussion explores the wider history. For this workflow, the lesson is simple: a long-term compound-growth aspiration cannot be substituted silently into a valuation that assumed the same growth next year. Also keep GOV and net order value distinct; their growth rates are not interchangeable labels.

2. Separate an explained revision from an unexplained omission

Neogen Chemicals initially guided FY26 battery-chemicals revenue at ₹400–500 crore in Q2 FY25. Its guidance history then narrowed to ₹300–500 crore in Q3 FY25, closer to ₹300 crore in Q4 FY25, and ₹30–40 crore in Q2 FY26, citing delays in customer gigafactory ramp-ups. The Q4 FY26 ledger update records ₹36 crore of revenue from Neogen Ionics (Inve data, commitment 2447).

Meeting the revised ₹30–40 crore range does not mean meeting the original ₹400–500 crore expectation. Nor does the large cut alone establish dishonesty. The useful evidence is that the lower expectation was stated, its cause named, and an outcome recorded against it.

That distinction is why the direction of guidance revisions needs to be read with the business circumstances. A candid cut can improve the information available to shareholders while the economics deteriorate.

3. Read repeated capital-allocation questions in context

Finolex Industries reported about ₹2,430 crore of net cash at 31 December 2025. Inve's Q1–Q3 FY26 financial rows total ₹177 crore of other income, ₹347 crore of operating profit and ₹430 crore of pre-tax profit. Other income was therefore about 41.2% of pre-tax profit. That accounting line should not all be relabelled interest earned on the cash pile without checking its components.

Across the five calls from Q3 FY25 through Q3 FY26, analysts repeatedly asked about surplus-cash deployment. On the last call, CFO Chandan Verma said: “We are yet to conclude upon what to do with this cash over the period in time.” The February 2026 transcript also records his references to planned capex and a previously declared dividend.

So the concern is an unresolved timetable for the remaining surplus, not that management never invested or returned any cash. Read repeated Q&A evasions as questions to pursue, with competitive sensitivity and board processes considered, rather than as an automatic finding of bad faith.

A positive example still needs matching accounting definitions

Coforge set out a path to 14% reported EBIT margin for FY26 on its Q1 FY26 call, when reported quarterly EBIT margin was 13.2%. The Q2 FY26 call reported 14.0%. In Q3 FY26 it reported 13.4% and described a plan for 15% in Q4 to reach the full-year objective.

The Q4 FY26 transcript reported 16.6% quarterly EBIT margin and 14.4% for FY26. It also disclosed a change that moved realised cash-flow hedge gains and losses from revenue to other income/expense, with comparatives restated. The final figures therefore cannot be used as a clean, unchanged-basis margin walk from the earlier calls.

This is a stronger research habit than simply applauding an achieved label: check the outcome, and check whether its definition changed. EBIT and EBITDA are different measures; an EBITDA result cannot by itself prove an EBIT target was met.

How Inve supports the post-screen review

Use Concall AI to find the quarterly operating figures, extracted guidance and the source quotes behind Q&A response classifications. The Promise Tracker adds the original target and its quarter-by-quarter history, together with persistent-question records. Detailed Indian company ledgers require Pro access.

Those tools help locate evidence. They do not replace the filings, cash-flow analysis or a valuation. The four-layer research stack shows where each document belongs. This is also a comparison of our own product with adjacent tools, so verify the source examples rather than treating our preferred workflow as independent proof of Inve's value.

For each shortlisted company, write down one assumption that matters to the valuation, its original deadline, the latest outcome and the next document that could resolve an uncertainty. Leave the row unresolved if the disclosure does not answer it.

Where this approach can mislead you

The roughly two-year research window discussed in these examples is not two years of complete coverage for every company. The dated production snapshot averaged 4.18 distinct commitment birth quarters per company. It cannot establish how a management team behaves across a full business cycle.

Coverage also selects companies that talk. A company without numerical guidance is not automatically inferior to one with a busy ledger. Read the annual report and capital-allocation record before penalising silence.

Finally, the corpus combines two model generations: gemini-2.5-pro across 1,476 company runs and gemini-3.1-pro-preview across 179. They were not tested on a common holdout. Extraction error and differences between models can affect classifications. The point of the tracker is to direct your reading, not to replace your judgement with a score.

Frequently asked questions

Does Inve replace a stock screener?

No. Financial screening narrows the candidates; guidance review tests the assumptions behind them. Platforms can cover more than one job, so compare the evidence you actually need rather than treating product names as exclusive categories.

Can I just read the latest concall summary?

Read it, but also check the earlier target it updates. A current summary can contain historical context when management supplies it; an omitted target requires prior documents. A quiet current call does not establish that all prior targets were met.

Which guidance deserves extra checking?

In the 4 September 2026 corpus snapshot, 71 of 279 eligible working-capital commitments were classified ghosted (25.4% of those 279), compared with 147 of 752 eligible capex commitments (19.5% of those 752). The baseline was 2,271 of 16,842 eligible commitments (13.5% of those 16,842). These are fuzzy, lower-bound categories matched on free-text metrics, not all guidance of either kind. Ghosting is not the same as a proven operational miss.

How long should the review take?

Use a time budget to organise an initial pass, not to certify a company as safe. A contradictory cash-flow series or changed accounting definition can require much longer than a quick ledger check.

The screen has done its job when it saves you from reading every company. Before turning a surviving row into a portfolio weight, ask what evidence would show that the assumptions behind it no longer hold.