HCL Technologies — The Year It Chose Growth Over Margin

The State of Play

HCLTech is India’s third-largest IT services company, a ₹3.1 lakh crore business that writes software, runs other companies’ technology, and engineers their products, for a roster of the world’s largest firms across 60 countries. It enters the middle of 2026 in a strange shape: the top line is still growing and the order book is the fattest it has been in years, yet annual profit actually slipped, because management spent the year deliberately trading margin for growth and absorbing a string of one-off costs. The stock sits near its 52-week low, foreign investors have been quietly leaving, and the company keeps paying out almost everything it earns as dividend. This is the story of how a single financial year — the four earnings calls from April 2025 to January 2026 — turned on one decision: keep the growth promise, break the margin one.

The Company

HCLTech sells three things. The biggest, about 74% of revenue, is IT & Business Services — running applications and data centres, moving clients to the cloud, automating their back-office processes. The second, ~16%, is Engineering & R&D Services: designing the actual products — chips, embedded software, cars, medical devices — for technology and industrial companies. The third and smallest, ~10%, is HCLSoftware, a portfolio of the company’s own software products, lower on revenue but the highest-margin of the three. In FY25 those segments earned a combined ₹1,17,055 crore.

The economics are unusually good for a business growing this slowly. Return on capital sits above 30%, the balance sheet is essentially debt-free (₹5,215 crore of borrowings against ₹74,622 crore of reserves), and the company throws off cash — free cash flow ran around ₹18,000 crore. Most of that cash goes straight back to shareholders: the dividend payout has hovered between 88% and 94% of profit, giving the stock a 4.65% yield that looks more like a utility than a tech company. The flip side, which screener flags bluntly, is a five-year sales growth of only 11.5%. This is a capital-return story wearing a technology company’s clothes.

Ownership is stable at the top and rotating underneath. The promoter — the Shiv Nadar family, through holding entities, with Roshni Nadar Malhotra as non-executive chairperson — holds a steady 60.8%, while the chief executive C. Vijayakumar (universally “CVK” on the calls) runs the company and is unrelated to the family. Beneath them, the live story is a handover: foreign institutions have trimmed their stake from roughly 19.6% to 15.5% over the past year, and Indian institutions have absorbed almost exactly what they sold, climbing from ~15.4% to ~19.0%.

The Story So Far

The backdrop: three years of slowing down

Before the four calls, the annual reports set the scene with a simple deceleration. FY23 grew revenue 18.5%, a genuine double-digit year led by engineering (+23.9%) and core IT services (+19.9%). FY24 cooled to 8.3%. FY25 cooled again to 6.5% — and in dollars, stripped of currency swings (the industry’s preferred “constant currency” yardstick, which removes the noise of the rupee and euro moving around), that was just 4.7%. The reports never sounded worried about it; each one was confident, even promotional, and each leaned harder on the same idea — artificial intelligence as the way to grow revenue without growing headcount in lock-step. By FY25 the CEO’s letter was organised entirely around AI, with a stated ambition to “decouple revenue from traditional models.” The growth engine was slowing; the pitch was that AI would re-engineer it.

April 2025 — bracing for a recession that didn’t come

The April call closed out FY25 and set the year ahead. The full year had landed at $13.84 billion, up 4.7% in constant currency, at an operating margin of 18.3%. The headline number, though, was the order book: $3 billion of net-new deals signed in a single quarter — total contract value, the dollar worth of fresh business — with fully half of it booked in March alone. For a quarter supposedly clouded by Trump-era tariff fear, clients were signing with what management called “a sense of urgency.”

Then came the guidance, and this is the promise to remember. For FY26, HCLTech guided to revenue growth of 2% to 5% and an operating margin of 18% to 19%. CVK was unusually candid about the low end:

“We are reasonably certain the environment will deteriorate… a recession is the most likely outcome” — and that recession, he said, was baked into the 2% floor.

So the setup was: a bumper order book in hand, a deliberately cautious revenue range because management genuinely expected a downturn, and a margin band — 18% to 19% — held roughly where the company had always operated. Within three months, two of those three things would change.

July 2025 — the stumble, and the trade

The June quarter is always HCLTech’s seasonally weakest, but this one was worse than seasonal. Revenue of $3,545 million grew a respectable 3.7% year-on-year, but the operating margin collapsed to 16.3% — well below plan — and the order book thinned to $1.8 billion as two large deals slipped out of the quarter on what management called “purely procedural delays.”

The margin drop came from three things stacking up at once: a slump in utilisation (the company had hired specialised staff in March who weren’t yet billable, an automotive program was winding down, and released employees couldn’t all be redeployed), a one-off hit from a client going bankrupt, and AI investments management chose to pull forward. On top of that, they announced a restructuring program to clean up facilities inherited from past acquisitions.

Faced with this, management made the defining move of the year. They raised the revenue guidance — to 3%–5% — because the feared deterioration simply hadn’t shown up (“the environment did not deteriorate as feared”). And they cut the margin guidance, from 18%–19% to 17%–18%, absorbing the one-offs and the restructuring rather than protecting the number. The recession CVK had called “most likely” in April had not arrived; the margin promise made in April was gone by July. The company was explicit about the logic — it had “traded short-term pain in profitability to secure long-term growth leadership.” CVK insisted the structural ambition was untouched:

“We are not setting, structurally, the margin bar lower… it will continue to be 19%–20%.”

He also promised the recovery would be quick: utilisation would “significantly moderate in Q2,” and “in Q3 we should be in line with the expected margins.” That promise is the one to check.

October 2025 — the recovery, on schedule

It checked out. The September quarter brought revenue of $3,644 million, up 4.6% in constant currency, and the operating margin climbed back to 17.5% — a 116 basis-point jump in a single quarter, landing exactly inside the recovery management had described in July. The order book also came back with a statement: $2.6 billion of net-new deals, the first time the company had crossed $2.5 billion without a single “mega deal” to flatter the number.

Two milestones sat underneath the recovery. The two deals that had slipped in June were now signed. And what HCLTech calls “Advanced AI” revenue — money from agentic systems (software that doesn’t just answer questions but takes actions), AI factories, and AI engineering, as distinct from ordinary work merely done with AI — crossed $100 million for the first time, about 3% of revenue. CVK framed it as the business moving “from the AI pilot stage to the AI monetization phase.”

The guidance moved again, in the same direction as July: services growth was raised to 4%–5%, though the company-wide range was held at 3%–5%, dragged by softness in the software products business. The margin band stayed at 17%–18%, with the restructuring now running “slightly on the higher side” of earlier estimates and the annual wage hike due to bite in the December quarter.

January 2026 — vindication, and a record book

The December quarter is where the year’s trade paid off. Revenue reached $3,793 million, up 4.8% year-on-year, pushing the company past a $15 billion annualised run-rate for the first time. The reported operating margin was 18.6%, but the more telling figure is the adjusted one: strip out the restructuring charge and the margin was 19.4% — within a whisker of the 19.5% it earned a year earlier, and right back inside the “structural” 19%–20% band CVK had defended through the whole downturn.

And the order book set a record: $3.0 billion of net-new deals, up 43% year-on-year, with the highest annual contract value in four years. Every one of the company’s four largest wins — including a five-year, $473 million mega-deal with a global apparel retailer — was won on the back of AI Force, HCLTech’s agentic-AI delivery platform, which CVK had taken to calling the company’s “killer app.”

The guidance was raised a third time: company-wide growth to 4.0%–4.5%, services to 4.75%–5.25%, the margin band still 17%–18% (it now also carried a one-time charge from India’s new labour code). Over three quarters the revenue promise had walked steadily up the range, from a cautious 2%–5% to a confident 4.0%–4.5%, while the order book went from $1.8 billion to $3.0 billion.

Where Things Stand

The reconciliation across the year is clean. The revenue promise was not just kept but repeatedly improved. The margin promise was broken early — and management was open about choosing to break it, sacrificing roughly a full point of margin to fund AI investment, restructuring, and growth, on the bet that it buys a stronger company. By the December quarter the underlying margin was effectively back to normal, which suggests the bet is being managed rather than lost.

The cost shows up where you’d expect: in the profit line. Revenue grew all year, but the most recently reported full year (ended March 2026) shows net profit of about ₹16,652 crore — slightly below the prior year’s ₹17,399 crore. A growing top line and a shrinking bottom line in the same year is exactly what “trading margin for growth,” plus a one-off labour-code charge and a weak quarter for non-operating income, looks like in the accounts.

A note on the very latest quarter: the company has since reported its March 2026 numbers — revenue of ₹33,981 crore, the highest quarterly figure on record, with net profit of ₹4,490 crore — but the transcript of that earnings call was not available through the public source at the time of writing, so it is not part of the quarter-by-quarter narrative above. The four calls covered here run through the December 2025 quarter.

The forward picture management has drawn is consistent: demand described as “constrained” but with “intact” structural appetite for transformation; spending shifting away from old discretionary projects toward AI-enablement work; restructuring meant to finish by the March 2026 quarter so the next year starts “on a clean slate”; and a firm refusal — repeated at every call — to give margin guidance for FY27 until April. The thing left unsaid sits in the shareholding and the share price: a debt-free, cash-rich, high-payout business growing in the mid-single digits, with foreign investors stepping back and domestic ones stepping in, and a market that has marked the stock down to the bottom of its yearly range while it waits to see whether the AI bet turns slowing growth back into fast growth.

The Four Checks

1. Quality and moat. A genuinely good business with a real but contestable moat. The advantage is switching costs and scale: once HCLTech runs a client’s applications, data centres and engineered products across multi-year contracts, ripping it out is expensive and risky — the kind of stickiness that lets a $15-billion run-rate business earn 30%+ on capital. The engineering and R&D arm (~16% of revenue, designing chips, cars and medical devices) is harder to replicate than commodity outsourcing, and HCLSoftware adds a small owned-IP layer. But the moat is shared with half a dozen rivals of equal or greater scale — TCS, Infosys, Accenture — who can and do contest every renewal, and the AI shift the company is betting on could just as easily compress industry pricing as expand it. Call it a sticky, well-run franchise in a structurally competitive industry, not a fortress.

2. Returns on incremental capital and runway. The returns are excellent; the runway is the problem. ROCE sits at 30.6% and ROE at 24%, and the trend is stable-to-rising (28% → 30% → 32% → 31% over FY23–26) — this is a capital-light machine where growth consumes almost nothing, so what little a rupee of growth requires earns a great deal. But the growth itself is modest: five-year sales growth of 11.5% (flattered by currency), FY26 guided and delivered at 4–5% in constant currency, and management’s own answer to the deceleration — paying out roughly 90% of profit — is an admission that there is nowhere inside the business to put the money. High returns on a narrow channel.

3. Capital allocation for the stage. Largely textbook for a mature cash machine, judged by deeds. The company keeps the balance sheet essentially debt-free (₹5,215 crore of borrowings against ₹74,622 crore of reserves), returns 88–94% of profit as dividend year after year, and confines acquisitions to small bolt-ons (ASAP in e-mobility engineering; HPE Telco Solutions, Jaspersoft and Wobby announced in FY26) rather than empire-building. The year’s deliberate margin-for-growth trade — absorbing restructuring and AI investment inside a cut guidance band rather than massaging the number — reads as honest spending, and the underlying margin was back to ~19.4% by the December quarter. The one quibble the data permits: no buyback is visible anywhere in the record, even with the stock at the bottom of its yearly range, where repurchases would beat a heavily taxed 4.7% dividend.

4. Price. Fair to mildly cheap, as of the June 2026 snapshot. The stock trades at ₹1,146 — just above its 52-week low of ₹1,103, against a high of ₹1,780 — at 17.9 times earnings with a 4.73% dividend yield. That is a modest multiple for a debt-free business earning 30% on capital, and the yield alone does much of the work while you wait. What the price is not paying for is growth: earnings actually slipped in FY26 (₹16,652 crore against ₹17,399 crore), and the market is priced as though mid-single-digit revenue growth is permanent. If the AI order book ($3 billion of net-new deals in the December quarter) converts into faster revenue, the multiple is undemanding; if it doesn’t, you own a high-yield utility at a sensible price.

Sources

This digest was built entirely from primary documents fetched from screener.in on 2026-06-05 (a logged-out session; the financial tables and the last three annual reports were current, the concall list reached back four quarters):

• Earnings call transcripts (BSE filings): Q4 FY25 / full-year FY25 + FY26 guidance (call 22 Apr 2025); Q1 FY26 (14 Jul 2025); Q2 FY26 (13 Oct 2025); Q3 FY26 (12 Jan 2026). The most recent call (Q4 FY26, ~Apr 2026) was unavailable as a PDF and is not covered.
• Annual reports (BSE): FY23, FY24, FY25 (high-signal sections — CEO/MD message, MD&A, risk framework, segment notes).
• Screener snapshot: company profile, key ratios, pros/cons, and the quarterly / 10-year P&L / shareholding tables (figures through the March 2026 quarter).

Full audit trail — the per-quarter and per-report research digests, the raw transcripts and reports, and the snapshot — lives in vault/Sources/Earnings/HCL Technologies Ltd/.