Which MCP server compares an Indian stock against its peers?

What does the peer comparison tool return?

get_peer_comparison takes a company name, NSE symbol, ISIN or BSE code. It finds the peer group, then returns one row per company with market cap, P/E, P/B, trailing twelve month revenue and net income, margins, ROE, ROA, ROCE, interest coverage, debt to equity, EV/EBITDA, cash-flow ratios, book value per share, a Piotroski F-score and an Altman Z-score. It also returns peer_medians for the main ratios. Every value comes from a cache of filings and end-of-day prices.

Return and margin ratios are fractions, so ROE of 0.3259 is 32.59 percent. Banks, NBFCs and insurers return null for ROCE, interest coverage, EV and EV/EBITDA, and for both scores. The ratios page lists which ratios are general-only.

What does a real answer look like?

This is Infosys (INFY) with limit 4 on 29 October 2026. The peer group is "Computers - Software & Consulting". The table holds Infosys, TCS, HCL Technologies and Wipro, and the median P/E is 16.0425. The median matches the middle of all four rows, Infosys included. The values were cached on 28 October 2026 at 14:45 UTC.

{
  "query_name": "Infosys Limited",
  "grain": "isubgroup",
  "grain_value": "Computers - Software & Consulting",
  "grain_widened": false,
  "size_banded": true,
  "peer_count": 4,
  "peer_medians": { "pe": 16.0425, "pb": 4.79525, "roe": 0.28945, "roce": 0.39225, "debt_to_equity": 0.00625 },
  "peers": [
    { "nse_symbol": "TCS",     "pe": 17.2462, "roe": 0.511,  "roce": 0.6965, "debt_to_equity": 0.0 },
    { "nse_symbol": "INFY",    "pe": 14.8388, "roe": 0.3259, "roce": 0.4511, "debt_to_equity": 0.0, "is_query": true },
    { "nse_symbol": "HCLTECH", "pe": 20.6869, "roe": 0.253,  "roce": 0.3334, "debt_to_equity": 0.0125 },
    { "nse_symbol": "WIPRO",   "pe": 14.4706, "roe": 0.1546, "roce": 0.1877, "debt_to_equity": 0.2343 }
  ],
  "cached_at": "2026-10-28T14:45:02+00:00"
}

How is a bank compared?

Here is HDFC Bank with limit 3. The peer group is "Private Sector Bank", with ICICI Bank and Kotak Mahindra Bank. ROCE and debt to equity are null for all three, and the median for each is null too. The tool returned a ratios_caveat because Kotak Mahindra Bank carried the ratios_incomplete flag. In this response the Kotak row still counts in the P/E and P/B medians, so read the flag before you read the median.

{
  "query_name": "HDFC Bank Limited",
  "grain_value": "Private Sector Bank",
  "peer_medians": { "pe": 16.7063, "pb": 2.248, "roe": 0.1457, "roce": null, "debt_to_equity": null },
  "peers": [
    { "nse_symbol": "HDFCBANK",  "pe": 13.1712, "pb": 1.7509, "roe": 0.1457, "roce": null, "is_query": true },
    { "nse_symbol": "ICICIBANK", "pe": 16.7063, "pb": 2.4112, "roe": 0.1668, "roce": null },
    { "nse_symbol": "KOTAKBANK", "pe": 20.3798, "pb": 2.248,  "roe": 0.1269, "roce": null,
      "ratios_incomplete": true }
  ],
  "ratios_caveat": "One or more rows are flagged ratios_incomplete: a structural event (rights/demerger/spin-off/merger) in the last 52 weeks means that company's valuation ratios (P/E, P/B) may not be comparable across the event."
}

What are the two peer sets?

get_relevant_peers takes an optional purpose (the default is operating) and an optional event_context. When it has no reviewed tags for a company it says so in mapping_status and mapping_note.

{
  "query_name": "Infosys Limited",
  "purpose": "operating",
  "peer_count": 0,
  "selection_quality": "unavailable",
  "coverage": { "mapping_status": "missing_reviewed_mapping", "query_reviewed_tag_count": 0, "sebi_fallback_used": false },
  "peers": [],
  "mapping_note": "No active reviewed business tags exist for the query company.",
  "reason_code": "no_data"
}

What do the data-quality flags mean?

• valuation_suspect: a row with an implausible figure, such as a corrupted P/B or a market cap with a unit slip. The row is kept and left out of peer_medians.
• ratios_incomplete: a rights issue, demerger, spin-off or merger in the last 52 weeks. P/E and P/B may not compare across the event. We saw this flag on Kotak Mahindra Bank.
• non_recurring_items: margins dominated by a one-off exceptional or discontinued item. The row is left out of the medians.
• grain_widened: the peer set fell back to a broader SEBI industry grain. It was false for both companies we pulled.
• size_banded: the peer set was restricted to a market-cap band. It was true for both.

We did not see valuation_suspect or non_recurring_items in the two tables above. Their meaning comes from the tool documentation.

What are the limits?

• Industry groups are broad. A SEBI group can hold companies with different business models. The tool does not say a peer is comparable. It says the peer shares a classification.
• Business-tag peers are partial. Three of three companies returned no peers from get_relevant_peers on 29 October 2026. We cannot say how many companies have reviewed tags.
• Cached data. The peer tables were cached on 28 October 2026 at 14:45 UTC. Prices are end of day.
• Nulls for financials. Banks, NBFCs and insurers have null for several ratios by design.
• No verdict. The tool returns numbers and a median. It does not rank the peers or say which one is better.

How do I ask for it?

For a general guide to asking an assistant to compare companies, see Compare Indian stocks with AI. To set up the server, connect your AI, or follow the guide for Claude or ChatGPT, Cursor and other clients. The server is at https://mcp.dalalos.in/mcp. The public repo is DalalOS-Indian-Stock-Research on GitHub.

How we checked

On 29 October 2026 we called get_peer_comparison for Infosys and HDFC Bank and get_relevant_peers for Infosys, HDFC Bank and Mazagon Dock on the live server. Flag definitions for valuation_suspect and non_recurring_items come from the tool documentation. Figures change as companies file and caches refresh. This is research data, not investment advice, and DalalOS does not rank or rate stocks. Found an error? Email hello@dalalos.in.