How to Research Indian Midcaps Alongside US Stocks

The financial table can look familiar when you move from a US company to an Indian midcap: revenue, margins, cash flow, valuation multiples. The harder transfer is knowing which disclosures support the next year's assumptions and who controls the decisions behind them.

A global terminal can organise both markets. It cannot make two companies equally well covered, make their accounting definitions identical, or turn an old management target into a verified outcome without the relevant evidence.

The useful cross-market workflow therefore keeps the financial comparisons and local source work together. It does not assume US companies are uniformly widely held, that Indian companies all behave alike, or that one platform has every answer.

What global terminals already do

Koyfin's data documentation describes global financials, valuation ratios and consensus estimates, with EPS, revenue and EBITDA estimates supplied by Capital IQ. TIKR's published plan information includes global financials, transcript history and, on applicable plans, management guidance. It would be inaccurate to describe either simply as a US ratio table or to claim that global platforms have no guidance information.

The practical check is company-specific. Does the ticker have estimates? How many contributors are represented? When were they updated? Does the transcript history cover the relevant target and its deadline? A blank field is missing coverage, not a zero estimate or proof that nobody researches the company.

The same discipline applies to filings. The SEC's guide to reading a 10-K explains the annual report's business, risk and financial sections. That does not make forward guidance a universally standardised SEC table. For Indian post-earnings calls within the rule's scope, SEBI LODR Regulation 46(2)(oa) requires website transcripts and simultaneous exchange submission within five working days. Neither filing system removes the need to reconcile an old target with later results.

Check ownership without turning it into a quality score

The archived Inve two-market research ledger, dated September 2026, reports a 59.0% median promoter holding across its 1,476-company Indian shareholding sample, using filings through August 2026. It reports majority promoter ownership above 50% at 71.3% of those 1,476 companies. This is a dated sample of companies in the guidance corpus, not a current census of Indian midcaps or a matched comparison with US ownership.

Promoter ownership also does not necessarily mean a single family owns the whole stake. The practical task is to identify the controlling holders and examine related-party transactions, capital allocation and minority-shareholder treatment. A stake close to the sample median is not an investment threshold.

Start with promoter shareholding disclosures, then read the actual ownership and pledge filings. In the US, use the ownership disclosures relevant to that company rather than assuming a hired CEO and dispersed shareholders. The question is the same in both markets: whose incentives shape the decisions that your valuation relies on?

What the two-market backtest actually compared

Inve's archived study compared a simple trailing-growth signal with a signal based on prior guidance outcomes. Its eligible observations comprised 2,978 Indian call-dates across 1,152 companies, from 11 March 2024 to 28 August 2026, and 2,793 US call-dates across 801 companies, from 10 January 2025 to 25 August 2026.

The growth measure used trailing-twelve-month sales and net-profit growth against the previous twelve months. It was not the production five-horizon ranking. The outcome measure required at least three resolved commitments. That is a study eligibility rule, not a recommendation to judge a management team from a small count.

For the crossing test, entry was at or above the 85th percentile and exit below the 75th, subject to the study's maximum-hold and end-of-data rules. The separate fixed-horizon test used a 180-day hold. The figures below are the archived ledger's cleaned, equal-weighted mean returns per trade, not annualised returns or returns on a funded portfolio.

Source: Inve's September 2026 two-market verification ledger, checked rows 1–8. Means are shown to two decimals so the differences are interpretable; these are historical study outputs, not a new October backtest.

The crossing-test difference between the last two pure signals was 5.39 percentage points per trade; the 180-day difference was 8.44 points. The archive records one-sided bootstrap tail fractions around 0.010 and 0.0044, respectively, at higher resample counts. These are resampling outputs under that analysis, not the probability that the conclusion is true or false. Trades are not necessarily independent, and multiple thresholds and blends were explored.

The growth and guidance-only crossing sets shared 30 stock-and-entry-date trades, out of 250 and 229 respectively. That establishes limited overlap between trade sets, not almost disjoint companies: the archive separately records 53 shared stocks among 248 growth-selected and 226 guidance-selected stocks.

The US result limits the broader claim. Growth-only crossing trades averaged 30.72% across 217 trades, with 59.9% of those 217 profitable. The five largest stock-level contributions accounted for 61.9% of the sum of individual trade returns. That is not an investable portfolio attribution.

After excluding the study's specified 12 memory/AI-semiconductor names, growth-only averaged 14.14% across 208 trades, with a median of 3.97% and 58.2% of those 208 profitable. Guidance-only averaged 8.95% across 217 trades, with a median of 4.41% and 64.5% of those 217 profitable. The higher win rate did not produce the higher mean. The exclusion is a sensitivity analysis, not a reason to discard inconvenient winners.

Why these results are exploratory

The study applied a 45-day lag to financial periods, but that does not make it an out-of-sample experiment or eliminate all future-information risk. The code constructs within-month percentile ranks using the month's observations and assigns the final recorded verdict from the date of the first resolving history step. Both choices need further point-in-time validation.

The archived review also identifies US fallback-date leakage. The universe selects companies with parsed calls, the verdicts were extracted retrospectively, prices were unadjusted, large daily jumps were filtered, and transaction costs were omitted. The India window is about 29.6 months and the US window 19.5 months; neither is a full business cycle.

The result weakens under alternative settings too. At entry 80, India's crossing gap was around 2.2 points with a bootstrap tail fraction near 0.15. At a 90-day horizon, the gaps ranged from about 0.9 to 3.2 points. This does not establish a universal trading signal, an optimal 20%–40% blend, or a causal claim that credibility produces returns.

The useful conclusion is narrower: guidance evidence and trailing growth asked different questions in this sample. A research workflow should inspect both.

KPIT: read the whole call, including the correction

KPIT Technologies illustrates why comparable financial tables still need a careful source reading. In Q3 FY25 it raised its EBITDA-margin outlook from above 20.5% to above 21%. Its FY25 results release reports 21.1% for Q4 FY25 and 21.0% for the full year, with FY25 constant-currency revenue growth of 18.7%. The quarter and year are different observations.

The later history was less smooth. The Q3 FY26 transcript describes organic growth as negative by less than 1%, following guidance for flattish-to-positive organic constant-currency growth. Calling the outcome exactly minus 1% would add precision the source does not supply.

The Q4 FY26 call reported FY26 EBITDA margin of 20.8%, against the earlier 21% objective, and guided FY27 EBITDA margin at 20.5%–21.2%. Q1 FY27 then reported 17.2% EBITDA margin and 0.1% year-on-year constant-currency revenue growth. The KPIT concall archive provides the source trail. One weak quarter puts pressure on an annual objective; it does not by itself prove the full-year outcome.

There is an even more instructive example in the Q4 FY26 transcript. Early wallet-share commentary can be read as a move toward 15%–20% of client spending. Later, on printed page 29, management explicitly corrected the statement: the immediate objective was a 20% relative increase from a roughly 10% wallet share, not a move from 10% to 20% share. On that illustrative base, the former implies about 12% share.

The local extracted guidance row retains the earlier interpretation. Copying its 15%–20% target would therefore miss the correction inside the very same source. This is a useful reason to read the whole exchange, not a reason to conclude that a global terminal or an automated tracker has already settled the investment case.

A favourable record needs the same scrutiny

Coforge discussed a 14% FY26 reported EBIT-margin objective on its Q1 FY26 call. It reported 14.0% in Q2 FY26, then 13.4% in Q3 while describing a plan for 15% in Q4. Its Q4 FY26 transcript reported 16.6% quarterly EBIT margin and 14.4% for the full year.

The same transcript explains that realised cash-flow hedge gains and losses moved from revenue to other income/expense and that comparatives were restated. The final margin therefore needs that accounting bridge before it is compared with the earlier target. A favourable ledger does not remove the need to reconcile definitions, and it does not tell you to hold through a share-price decline.

KPIT and Coforge illustrate disclosure checks, not investment views. Historical backtest results do not guarantee future performance.

How Inve fits a two-market research workflow

Keep the workflow organised around evidence:

1. Compare financials on consistent definitions. Use a domestic or global financial-data tool, then check the source where scope or adjustments differ.
2. Check ownership locally. Identify controlling holders, related-party transactions and relevant pledge or insider disclosures. Do not use a sample ownership median as a buy threshold.
3. Reconcile guidance across documents. For India, Concall AI organises quarterly commentary and the Promise Tracker supplies target histories and recurring-question records. Detailed Indian company ledgers are Pro-gated. The US guidance-tracking guide explains the separate US workflow.
4. Return to valuation. Test whether the price requires stronger delivery than the evidence supports. The India DCF calculator can help make those assumptions explicit where the model is suitable.

Our post-screener guide develops the second-stage checks. The four-layer India research stack places them beside operating metrics and annual-report notes. Inve has an incentive to emphasise its own guidance tools; the examples here show why their output still needs source verification.

Frequently asked questions

Can I use Koyfin or TIKR for Indian midcap research?

Yes, their published documentation includes global financial coverage and other research features. Check the actual ticker, plan, transcript depth and estimate freshness. Do not assume coverage is identical across companies or that a missing field means a negative business fact.

Does management credibility matter more in India than in the US?

The archived study found different return patterns in different samples and windows. It did not isolate ownership or disclosure rules as the cause, and it does not establish a universal ranking of the two markets. Examine incentives and delivery evidence in both.

What did the two-market test find?

In the archived Indian crossing test, guidance-only trades averaged 9.58% across 229 trades versus 4.19% across 250 growth-only trades. In the US sensitivity sample excluding the specified 12 names, guidance-only had a higher win rate and median but a lower mean. These exploratory, gross per-trade results carry the timing and selection limitations described above.

How many quarters do I need?

Four to six consecutive quarters can help reconstruct an annual guidance cycle if coverage is complete. That is a starting point for reading, not enough evidence to settle management quality. Check earlier cycles and other filings where available.

Why pair a summary with a guidance history?

The summary-versus-tracking comparison shows the difference: one organises a document, the other reconciles documents. Neither should override a correction in the original source.

A transferable research process is one that asks the same hard question in either market: does the source evidence still support the assumption you are paying for?