Sterlite Technologies Ltd STLTECH

Summary

Sterlite Technologies (STL) is India's largest optical-fibre and optical-fibre-cable manufacturer, with ~8-9% of the global ex-China OFC market. Inside that business sits a genuinely exciting growth story: per CRISIL's rating rationale dated 1 September 2026, STL's data-centre segment grew to 21% of Q1 FY27 revenue, up from just 1% a year earlier — the highest disclosed data-centre revenue concentration of any company in this report's entire universe. That growth is backed by a confirmed USD 1.1 billion, three-year data-centre contract sitting inside a record ₹18,618 crore order book (as of 30 June 2026), a dedicated product architecture (the Neuralis brand spanning Data Center Interconnect and Data Center Whitespace, Hollow Core Fibre, and NFPA 262-compliant plenum-rated Fiber Trunk Assemblies launched 10 September 2026), and named senior leadership — a Chief Business Officer for Hyperscale Data Centers and a COO for Optical Interconnectivity & DC Terminated Solutions — run as a distinct business line, not a marketing label.

The balance sheet has also genuinely improved: a ₹1,500 crore Qualified Institutions Placement, completed 2 July 2026, took STL to net-debt-free status, and CRISIL responded by upgrading STL's long-term rating to CRISIL AA/Stable from AA-/Stable on 1 September 2026, with interest coverage expected above 9x medium-term versus 2.9x previously. The company has also announced a funded ~50% capacity expansion (₹3,000 crore of capex through FY29) and a "Lakshya" strategic plan targeting ₹20,000 crore of FY29 revenue at 27%+ EBITDA margin.

The SELL call, however, does not rest on doubt about that demand story — it rests on the reliability of the numbers underneath the stock's ~178x trailing P/E. For the identical Q1 FY27 quarter, three sources give three different profit-after-tax figures: screener.in shows ₹125 crore, STL's own 24 July 2026 press release claims a "record" ₹197 crore, and Groww shows a reported loss of ₹17 crore. This report does not resolve that conflict to a single number — the three-way disagreement on the company's own most recently reported, most-hyped quarter is itself the central reason for caution, and it sits on top of a recent loss history (FY24 PAT: ₹(57) crore; FY25 PAT: ₹(123) crore, screener.in consolidated) that makes FY26's small return to profit (₹56 crore) a thin base against a ₹42,091 crore market capitalisation. A live $96.5 million Prysmian trade-secrets jury verdict (2024, under appeal) and an unverified promoter/ownership structure (Wikipedia names STL a Vedanta Limited subsidiary; this could not be independently confirmed against a primary shareholding filing) add further, independent reasons for caution.

Net: a SELL. On the most conservative, most defensible earnings base this report can construct — FY26 PAT divided by the QIP-expanded current share count — fair value sits well below CMP under essentially any plausible target multiple. This is a valuation and data-reliability judgment, not a judgment on the underlying data-centre demand story, which is real and, by disclosed revenue concentration, the strongest in this report's universe.

*Screener.in figures shown; Groww separately shows P/E 174.28x (TTM) and book value ₹70.86/share — a ~52% gap on book value alone. See "What gives us pause" below; this report treats these as unresolved data-quality flags, not as figures to silently average.

• The highest disclosed data-centre revenue concentration of any company in this report's 20-company universe. 21% of Q1 FY27 revenue, up from just 1% a year earlier (CRISIL rationale, 1 September 2026) — the cleanest, most material, best-quantified data-centre revenue story this report finds anywhere.
• A confirmed USD 1.1 billion, three-year data-centre contract sits inside a record ₹18,618 crore order book (as of 30 June 2026) — a scale of booked, DC-specific revenue visibility this report does not find matched elsewhere in its universe.
• Purpose-built data-centre product architecture. The Neuralis brand (Data Center Interconnect + Data Center Whitespace), Hollow Core Fibre marketed as "India's first" for data-centre networks, Multiverse MCF, and NFPA 262-compliant plenum-rated Fiber Trunk Assemblies launched 10 September 2026 — genuine, differentiated engineering, not a relabelled generic cable line.
• Dedicated senior data-centre leadership — a named Chief Business Officer for Hyperscale Data Centers & Optical Interconnectivity and a COO for Optical Interconnectivity & DC Terminated Solutions — evidence this is run as a distinct business line at the C-suite level.
• A genuine credit upgrade. CRISIL raised STL's long-term rating to AA/Stable from AA-/Stable on 1 September 2026, citing the ₹1,500 crore QIP (completed 2 July 2026) that took the company to net-debt-free status, with interest coverage expected above 9x versus 2.9x previously.
• A funded, quantified growth plan. A ~50% capacity addition backed by ₹3,000 crore of capex through FY29, alongside the "Lakshya" plan targeting ₹20,000 crore of FY29 revenue at 27%+ EBITDA margin.

• A three-way conflict on STL's own most recent quarterly PAT — the central reason for this report's SELL call. For the identical Q1 FY27 quarter (ended June 2026), screener.in shows PAT of ₹125 crore, STL's own 24 July 2026 press release claims a "record" ₹197 crore, and Groww shows a reported loss of ₹17 crore. This report does not resolve this conflict to a single figure; the disagreement itself, on the company's most-hyped recent quarter, is the finding.
• This sits atop a genuinely weak recent earnings history. FY24 PAT was ₹(57) crore and FY25 PAT was ₹(123) crore (screener.in, consolidated); FY26's return to profit (₹56 crore) is still a very small base relative to STL's ₹42,091 crore market capitalisation.
• The disclosed ~178x trailing P/E is itself unstable. Using this report's more conservative FY26-PAT-divided-by-current-share-count construction (see Valuation), the implied EPS is ~₹1.09 against a disclosed-multiple-implied EPS of ~₹4.60 (CMP ÷ 178x) — over a 4x gap between two defensible ways of computing the same ratio, on the same stock, the same day.
• Book value per share is separately disputed: ₹46.5 (screener.in) versus ₹70.86 (Groww), a ~52% gap — a second, independent data-quality flag beyond the PAT conflict.
• Basic ownership and governance facts could not be confirmed. Wikipedia names STL a subsidiary of Vedanta Limited with Anil Agarwal (Vedanta Resources' founder-chairman) as STL's "Founder/Chairman," but this could not be corroborated against STL's own management-page listing or a primary shareholding filing in this research — an open question on who the company's actual promoter and Chairman are.
• A live $96.5 million jury verdict against STL in the Prysmian trade-secrets litigation (South Carolina, 2024) remains under appeal — a contingent liability roughly comparable in size to a full year of the company's already-small disclosed PAT, independently corroborated by both Wikipedia and CRISIL's own 1 September 2026 rating rationale.
• Price action tied to group-level, not company-specific, news. The stock hit a 5% lower circuit on 16 September 2026 amid a broader Vedanta-group selloff, per Groww's news aggregation — a reminder that STL's price is not always driven by its own fundamentals.
• Promoter holding fell from 44.52% to 42.29% in a single month (June to July 2026) — worth watching alongside the ownership-structure verification gap above.

• Reconciliation of the three conflicting Q1 FY27 PAT figures against the audited FY27 annual report — the single most important item this report needs closed before STL's earnings can be trusted.
• Confirmation of STL's sitting Chairman and the promoter/parent structure (Vedanta Limited versus an Agarwal-family holding vehicle) from a primary filing.
• The outcome and any provisioning of the Prysmian $96.5 million appeal.
• Delivery against the "Lakshya" FY29 targets and the pace of execution on the $1.1bn data-centre contract.

• Financial-reporting-reliability risk (dominant). The three-way PAT conflict on the same quarter is, on its own, sufficient reason for caution, independent of the underlying data-centre demand story.
• Valuation risk. On this report's more conservative earnings construction, essentially every plausible target multiple implies material downside from CMP.
• Litigation risk. The $96.5 million Prysmian verdict remains under appeal and unresolved.
• Governance-verification risk. Basic facts — Chairman, promoter identity, statutory auditor — could not be confirmed.
• Group/contagion risk. Price moves tied to broader Vedanta-group news rather than STL-specific fundamentals.

We use approach (a): FY26 PAT ÷ implied current share count (market cap ÷ CMP). We deliberately do not use a trailing-P/E approach here: STL's disclosed trailing P/E of ~178x is built on a TTM earnings base that necessarily blends in the disputed Q1 FY27 quarter — and depending on which of the three conflicting PAT figures for that quarter is used, the "correct" trailing multiple could plausibly range from materially above 178x (using Groww's reported loss, which would make a P/E undefined) to well below it (using the company's own higher ₹197 crore claim). We regard building a scenario table on that foundation as irresponsible. Instead, we anchor on FY26's full-year PAT of ₹56 crore (screener.in, consolidated, not itself in dispute) divided by an implied current share count of market cap ÷ CMP — ₹42,091 crore ÷ ₹819 = ~51.41 crore shares, reflecting the July 2026 QIP's expanded base — giving a conservative FY26 EPS of ~₹1.09. We hold this EPS constant across scenarios and vary only the target multiple. We want to be explicit about what this does and does not tell you: this is the most defensible single EPS figure we can construct without taking a side in the three-way PAT dispute, but it is also almost certainly conservative if STL's Q1 FY27 run-rate genuinely reflects a step-change in profitability, as the company's own press release claims. We are not confident enough in any single number to publish one P/E multiple without this caveat attached.

Bear applies a 100x multiple — still a rich, growth-stock multiple by any conventional measure — reflecting a scenario in which the market re-prices STL toward a more normal multiple once the PAT conflict is resolved unfavourably. Base applies 175x, deliberately close to STL's own disclosed trailing P/E of ~178x, to isolate the point this report is making: even if you accept the market's current multiple as reasonable in isolation, pairing it with the more defensible FY26-PAT/current-share EPS of ₹1.09 — rather than whichever higher, disputed EPS actually produces the headline 178x figure — implies a fair value of just ₹191, a (76.7)% gap from CMP. Bull applies 275x, assuming continued growth-multiple expansion in line with the market's evident enthusiasm (CLSA raised its target by 47% to ₹1,400 with an Outperform rating on ~13 September 2026, citing AI-driven expansion) — even then, on this report's earnings base, the implied target remains well below CMP. We do not know CLSA's own methodology or which of the three PAT figures it relies on, and we flag the gap between its target and all three of our scenarios rather than attempt to resolve it. Every scenario here implies downside from CMP, consistent with this report's SELL rating; the range communicates how unstable the valuation is at any multiple, not a false precision around one number.

Source: screener.in (20 Sep 2026 fetch) for the FY24-26 annual table; CRISIL rating rationale dated 1 September 2026 for the order book and data-centre revenue-mix figures; STL's own 24 July 2026 press release and Groww (fetched 20 Sep 2026) for the three conflicting Q1 FY27 PAT figures, presented here side-by-side rather than reconciled to one number.