On June 8, 2026, Sterlite Technologies Limited (STL) was locked at its 5% lower circuit limit at ₹588.30 on the BSE — the second consecutive session in which the stock hit its daily downside boundary. The immediate context was a global AI reversal trade that pulled down technology and infrastructure stocks broadly, reducing demand for the momentum positions that had accumulated in STL over the preceding weeks.
Those two sessions of 5% declines are notable primarily for what they reveal about the stock’s underlying character in the current moment: they followed a period in which STL hit back-to-back 5% upper circuit limits on multiple consecutive sessions, was locked at its daily upside boundary for four straight sessions after the $1.11 billion deal announcement, and accumulated a one-month gain of 56%, a six-month gain of 513.90%, and a year-to-date gain of 474% by June 8. A stock that travels in 5% daily increments — up and down — is not a stock being priced by fundamental analysis. It is a stock being traded by momentum participants responding to a narrative re-rating event. The 5% lower circuit days are the momentum unwinding; the 5% upper circuit days were the momentum accumulating.
Understanding what changed to produce 474% year-to-date appreciation — and what has not changed — requires understanding the order that triggered everything.
The $1.11 Billion Order That Changed the Story
In late May 2026, Sterlite Technologies disclosed that one of its subsidiaries had received a Product Award Letter from an unnamed hyperscale partner — a company in the category of Amazon, Google, Microsoft, or Meta — for the multi-year supply of optical connectivity products valued at approximately $1.11 billion (approximately ₹10,000 crore) over FY27 to FY29. The products involved are optical connectivity solutions for AI data centre infrastructure build-outs in the United States.
The deal was disclosed following a BSE exchange filing, and the market’s response was immediate and extreme. STL hit 5% upper circuit on the day of disclosure and continued to lock in the upper circuit for four consecutive sessions. The logic was straightforward: STL’s entire FY26 revenue was ₹4,745 crore. A single order worth ₹10,000 crore to be executed over three years implied that the hyperscaler contract alone could add more revenue in FY27–FY29 than the company generated in all of FY26 — assuming execution at the contracted rate. The company’s order book, which was already up 67% year-on-year to ₹7,309 crore by the end of FY26, was now supplemented by this mega-order.
CLSA initiated coverage on the stock on May 26, 2026 — immediately following the deal announcement — with an Outperform rating and a price target raised from ₹405 to ₹655, a 60% increase. The brokerage’s note said: “STL’s order book in FY26 was already up 67% to ₹7,300 crore, which has been followed by this mega win. We build this US$1.1bn order into our estimates, resulting in forecast increases of 25–64% over FY27–29CL.” CLSA’s EBITDA CAGR estimate for STL over FY26–FY29 was 49% — a figure that justifies a significant re-rating when applied to a company that had just returned to profitability from a loss-making position.
The deal did something more than add contract value: it reframed what kind of company STL is. From a telecom optical fibre manufacturer that serves domestic and international carrier networks, the market began treating it as an AI data centre infrastructure play — a company in the same category as Marvell and Corning on the US side, positioned to benefit from the buildout of AI clusters that demand dramatically more fibre per rack than conventional cloud infrastructure. STL’s Celesta IBR cable, launched in Q2 FY26 and featuring up to 6,912 fibres in a high-density format designed specifically for AI and cloud data hall environments, became the product symbol of that repositioning.
MD Ankit Agarwal’s statement on the deal was explicit: “With its optical solutions, STL will support building AI data centre infrastructure in the US for this hyperscaler.”
The Financial Baseline That the Narrative Sits On
The narrative re-rating was not built on an imaginary foundation. STL’s FY26 financial results showed genuine operational recovery.
For the full year FY26, STL reported revenue of ₹4,745 crore — up 18.8% year-on-year from ₹3,992 crore. The company returned to profitability with a net profit of ₹56 crore, a meaningful reversal from the net loss of ₹72 crore recorded in FY25. EBITDA margin expanded to 13.2%, up from approximately 9% in the prior year, supported by higher capacity utilisation and a shift toward higher-value products. Q4 FY26 was the strongest quarterly showing: revenue rose 37% year-on-year to ₹1,441 crore, EBITDA grew to ₹218 crore from ₹146 crore, and net profit reached ₹59 crore — which included a one-time gain of ₹31 crore but was still a dramatic improvement from ₹5 crore in Q4 FY25. Order book at year-end stood at ₹7,309 crore. Enterprise and data centre revenue contributed approximately 19–20% of FY26 revenues, with management guiding for that share to reach 30% over the next 12–18 months.
The product and technology story is also real. STL’s AI data centre cable portfolio includes the Celesta IBR series with up to 864 fibres in a sub-12mm diameter optimised for micro-duct jetting inside data halls — specifically engineered for the GPU-dense computing environments that AI training requires, where the density and low-latency requirements of intra-facility interconnect demand solutions that conventional telecom cables cannot provide. The company’s vertical integration from optical fibre manufacturing through to finished connectivity solutions creates a cost and customisation advantage that pure-play systems integrators or component resellers cannot replicate easily. In collaboration with C-DOT, STL also demonstrated India’s first 100-km quantum key distribution over 4-core multi-core fibre — a capability that positions it at the frontier of quantum-secured network infrastructure, relevant to governments and defence clients beyond the commercial hyperscaler market.
The Valuation Question the Rally Created
The issue with a 474% year-to-date return is that it tends to resolve the valuation question in the direction of expensive, regardless of the underlying story’s quality.
At the pre-lower-circuit price of approximately ₹617 before the June 7 decline, STL was trading at approximately 634 times its FY26 net profit of ₹56 crore — a P/E multiple that, as Rakesh Jhunjhunwala analysts noted, “leaves limited room for disappointment.” CLSA’s ₹655 price target was derived from forward estimates built into FY27–FY29 EBITDA projections that incorporate the hyperscaler order. At ₹588 (the June 8 lower circuit level), the stock remains significantly above the pre-deal price but has given back approximately 10% from its peak. It is still 474% higher for the year.
Two structural risks accompany the re-rating at any price level. First, the order is a Product Award Letter — a committed allocation with purchase orders to be released periodically during the contract period. The total contract value of $1.11 billion represents maximum potential revenue; actual revenue will depend on purchase order releases and execution cadence, which are subject to the hyperscaler’s own data centre build timeline and capital allocation decisions. If the hyperscaler reduces or delays its AI infrastructure capex — the same risk dynamic that created Broadcom’s post-earnings decline — STL’s delivery schedule is affected.
Second, executing a $1.11 billion supply contract requires capacity. STL’s current installed manufacturing capacity and the capital allocation required to fulfil annual delivery commitments from FY27 to FY29 will determine whether the order translates into the margin expansion CLSA projects or is absorbed partly by the capex required to build out that capacity. STL’s debt profile — which has historically been a concern for analysts given the capital intensity of its manufacturing operations — will be the financial variable to watch as the company ramps toward the hyperscaler contract’s delivery requirements. The company’s FY25 net debt was not negligible, and funding capacity expansion without materially increasing leverage or diluting equity will require operating cash flows to improve substantially ahead of deliveries.
The AI Reversal Trade and What It Signals
The June 7–8 two-day 10% decline in STL came in the specific context of what BusinessToday described as “markets globally declining on an AI reversal trade” — a broad unwinding of AI-themed momentum positions following the post-Broadcom-earnings selloff, gold’s correction after the May jobs report, and a general reduction in risk appetite across high-multiple growth stocks globally.
In that context, STL’s lower circuit is not specific to the company. It is a function of the stock’s position in a high-conviction, high-concentration trade. Stocks that gain 56% in a month on a single catalyst attract momentum participants who hold no fundamental conviction in the business — they are trading the narrative, not the earnings. When global risk appetite retreats, those participants sell first, and they sell into lower circuit conditions because the exit is compressed into a narrow daily band.
The underlying thesis — that AI data centre buildout is a durable, multi-year demand driver for high-density optical connectivity, and that STL’s vertical integration and hyperscaler relationship position it to capture a meaningful share of that demand — has not changed across these two sessions. The valuation at which that thesis is priced has corrected. Whether the correction is sufficient to make the stock attractive at a fundamental level depends entirely on how accurately CLSA’s FY27–FY29 EBITDA estimates translate into actual delivered contracts, and whether STL can execute the hyperscaler supply commitment without the debt or dilution that would erode the shareholder economics of the re-rating.
Disclaimer: Investments in securities markets are subject to market risks. Read all the related documents carefully before investing. The securities quoted are exemplary and are not recommendatory.
