Infosys:
The AI Hype Train That’s Actually
Executing. ₹45,479 Cr Revenue.
5.5% AI Revenue. Margins Hold.
The stock fell 19% in three months. The company just posted its strongest quarterly revenue in near-history and doubled down on AI strategy at an investor day. Welcome to the gap between stock price and business reality.
When Wall Street Hates What You’re Selling, But You Keep Selling It Anyway
- 52-Week High / Low₹1,728 / ₹1,264
- Q3 FY26 Revenue₹45,479 Cr
- Q3 FY26 PAT₹6,666 Cr
- Q3 EPS (₹)16.41
- Annualised EPS (Q3×4)₹65.64
- Book Value₹205
- Price to Book6.39x
- Dividend Yield3.29%
- Debt / Equity0.11x
- 3-Month Return-19.0%
India’s Second-Largest IT Company Just Went Full AI. Here’s Why No One’s Paying Attention.
Infosys is the second-largest listed IT services company in India (by revenue) after TCS, with a 50-year track record of executing boring consulting gigs, managing legacy systems, and occasionally winning a big cloud migration deal. It’s not Tesla. There is no Elon Musk moment waiting. There is no “platform shift” narrative you can pitch to a 20-something at a music festival. It’s software engineering services to 1,869 active clients across 50+ countries, with 323,578 employees, of which 317,240 are actual software people. And it’s profitable every single quarter without exception.
Then, on February 17, 2026, management held an Investor AI Day. CEO Nandan Nilekali and team redefined Infosys as an “enterprise AI execution partner”—not building foundation models (that’s OpenAI, Anthropic, Google), but taking enterprises’ data, legacy systems, AI models, and transforming their entire operating model. The Topaz Fabric platform is now live with 600+ agents. AI projects span 90% of top-200 clients. And the company is hiring 20,000 college grads this year to build out the AI-first delivery model.
The street heard “margin compression from AI pricing” and sold first. They asked questions much later. The stock is down 24% in a year. Meanwhile, cash surplus has hit ₹51,027 crore, an ₹18,000 crore buyback just concluded, and management expects expansion > compression in AI deal value. Q3 results—the strongest revenue quarter in 13 years—landed with barely a market flinch. Here’s the breakdown of what’s actually happening, minus the Netflix documentary tone.
They’re Not Building ChatGPT. They’re Building the Bridge Between ChatGPT and Your ERP System.
Infosys generates revenue via digital services and traditional consulting. Digital services (higher margin, higher growth) include AI, cloud migration, modernization, IoT, engineering. Traditional services include application management, support, testing, infrastructure. Revenue is split roughly 35–40% digital / 60–65% traditional (though the gap is narrowing). Clients pay by the hour, by the project, by the outcome (still rare, growing), or via platform SaaS models (Finacle, McCamish, Edgeverve). Geographically, 56% comes from North America, 33% from Europe, 9% from rest of world, and 3% from India—a typical global offshore services company structure.
Q3 FY26 revenue was ₹45,479 crore (+8.9% QoQ), the strongest in over a decade. It came from a mix of: (1) a slight macro recovery after a 2024 slowdown, (2) digital and AI work finally penetrating into deals beyond the “innovation lab” phase, and (3) large deals from prior years ramping into delivery. Management’s guidance is 0–3% constant-currency growth for FY26 overall (a full year starting April 2025), implying caution on macro but confidence in execution.
Q3 FY26: The Numbers That Should’ve Moved the Needle
Result type: Quarterly Results | Q3 FY26 EPS: ₹16.41 | Annualised EPS (Q3×4): ₹65.64 | Full-year (TTM) EPS: ₹67.73
Source table
| Metric (₹ Cr) | Q3 FY26 Dec 2025 |
Q3 FY25 Dec 2024 |
Q2 FY26 Sep 2025 |
YoY % | QoQ % |
|---|---|---|---|---|---|
| Revenue | 45,479 | 41,764 | 44,490 | +8.9% | +2.2% |
| Operating Profit | 10,634 | 10,115 | 10,535 | +5.1% | +0.9% |
| OPM % | 23.4% | 24.2% | 23.7% | -80 bps | -30 bps |
| PAT | 6,666 | 6,822 | 7,375 | -2.3% | -9.6% |
| EPS (₹) | 16.41 | 16.39 | 17.73 | 0.1% | -7.4% |
What’s This Company Actually Worth in an AI World?
Method 1: P/E Based
TTM (Trailing Twelve Months) EPS = ₹67.73. Industry median P/E = 21.0x. Infosys’s justified multiple given (1) lower growth vs TCS (2) higher execution risk in AI transition (3) but strong cash flow and buyback support): 16x–22x.
Range: ₹1,084 – ₹1,490
Method 2: EV/EBITDA Based
TTM EBITDA ≈ ₹45,000 Cr (approx). Current EV ≈ ₹5,19,854 Cr → EV/EBITDA = 11.5x. Quality software services trade at 10x–15x. Infosys at 11.5x is middle-of-pack. Fair range 10x–13x reflects execution uncertainty post-AI transition.
EV range (10x–13x): ₹4,50,000 Cr – ₹5,85,000 Cr → Per share:
Range: ₹1,108 – ₹1,439
Method 3: DCF Based (Simplified)
Operating CF: ₹35,694 Cr (FY25). Growth: 4–6% (cautious, given 0–3% guidance). Terminal growth: 2.5%. WACC: 9.5%.
→ Terminal Value (2.5% growth / 7% cap rate): ~₹4,80,000 Cr
→ Total EV: ~₹6,40,000 Cr (less net debt ~₹(8,000) Cr, ie no net debt issue)
Range: ₹1,200 – ₹1,500
The Plot Twists in the Longest-Running IT Services Show
🔴 The Big One: Infosys’s “Enterprise AI Rewrite” Strategy (Feb 17, 2026 Investor Day)
Management redefined the company not as a “cloud player” or “digital services vendor,” but as an enterprise AI execution partner. The Topaz Fabric platform (launched earlier, matured now) abstracts away model choice—enterprises can use OpenAI, Anthropic (notably a highlighted partner), Gemini, or open-source models. 600+ pre-built agents handle modernization, process automation, customer experience, operations. AI work is now 5.5% of revenue at run-rate; target: meaningfully higher without cannibalizing traditional services. Management says “expansion number looks larger than compression number” but won’t quantify (translation: too early). Large-deal cycles have not compressed from 12–18 months to weeks. But the mix is shifting: AI-first, AI-augmented, industry-specific blueprints.
✅ Hiring Momentum & Workforce Reconfig
- • 20,000 college hires this year; 20,000 planned for next FY
- • Net headcount +13,000 in 9 months despite macro headwinds
- • New Y-shaped career track: generalist AI-enabled + specialist
- • 90% of organisation enabled on AI tools; forward-deployed engineers as context translators
- • Voluntary attrition fell to 12.3% (Q3) from 14.3% (Q2) — talent stickiness improving
⚠️ Margin Headwinds & Pricing Model Shift
- • OPM down 80 bps YoY, 30 bps QoQ due to employee compensation & depreciation
- • Pricing models shifting: hourly → outcome-based, platform SaaS, hybrid
- • CFO expects S&M & AI investments to absorb ~50 bps; offset by margin expansion programs
- • Risk: if AI “model-driven pricing” (clients’ expectations) runs ahead of “enterprise deployment maturity,” compression pressure rises fast
They’ve Got More Cash Than Some Countries Have GDP
Source table
| Item (₹ Cr) | Mar 2025 | Sep 2025 | TTM (Annualised) |
|---|---|---|---|
| Total Assets | 147,795 | 160,380 | — |
| Net Worth (Equity + Reserves) | 95,818 | 103,330 | — |
| Borrowings (Lease debt) | 8,227 | 8,755 | — |
| Cash & Equivalents* | 51,027 | ~52,000 est. | — |
| Net Debt | +43,000 (Net Cash) | +43,000+ est. | — |
For context, that’s 10.8% of their market cap sitting in the bank. Most growth companies are burning cash. Infosys is printing it.
Borrowings are entirely lease-debt (Ind AS 116 treatment). Interest coverage at 98x. Default risk is not on the periodic table of elements.
Management’s policy: return 85% of free operating CF to shareholders (dividends + buyback) over 5 years. That’s ₹30,000+ crore over the period. Supported by net worth growth from retained earnings.
Sab Number Game Hai, But This Game Favours Infosys
Source table
| Cash Flow (₹ Cr) | FY23 | FY24 | FY25 | TTM (Annualised) |
|---|---|---|---|---|
| Operating CF | +22,467 | +25,210 | +35,694 | +35,694+ |
| Investing CF | -1,071 | -5,093 | -1,864 | -2,000 est. |
| Financing CF | -26,695 | -17,504 | -24,161 | -25,000+ est. |
| Net Cash Flow | -5,299 | +2,613 | +9,669 | +8,000+ est. |
Where Infosys Shines, and Where It’s Walking a Tightrope
Revenue Growth +10.2% CAGR. But Have Profits Kept Up?
Source table
| Metric (₹ Cr) | FY22 | FY23 | FY24 | FY25 | TTM |
|---|---|---|---|---|---|
| Revenue | 121,641 | 146,767 | 153,670 | 162,990 | 173,173 |
| Operating Profit | 31,491 | 35,130 | 36,425 | 39,236 | 40,986 |
| OPM % | 26% | 24% | 24% | 24% | 24% |
| PAT | 22,146 | 24,108 | 26,248 | 26,750 | 28,003 |
| EPS (₹) | 52.56 | 58.08 | 63.20 | 64.32 | 67.73 |
The good news: revenue growing double-digit. The caveat: profit growth (5.5%) is half that. Why? Mix shift to higher-volume, lower-margin services (traditional IT), tax volatility, and now employee cost inflation. If AI work really does “expand” as management claims, profit growth should re-accelerate. If profit growth stays <6% with revenue at 10%, that's a derating story. We're watching FY26 closely.
TCS Dominates. Infosys Stabilises. Everyone Else Prays.
Source table
| Company | Q Revenue (₹ Cr) | Q PAT (₹ Cr) | P/E | ROCE % | Revenue Growth % |
|---|---|---|---|---|---|
| Infosys | 45,479 | 6,666 | 18.4x | 37.5% | +8.9% |
| TCS | 67,087 | 10,720 | 18.1x | 64.6% | +4.9% |
| HCL Tech | 33,872 | 4,082 | 21.5x | 31.6% | +13.3% |
| Wipro | 23,556 | 3,145 | 15.5x | 19.5% | +5.5% |
| Tech Mahindra | 14,393 | 1,119 | 27.1x | 18.6% | +8.3% |
TCS is the ROCE leader (64.6%). Infosys is the growth leader (8.9% QoQ, 2.2% QoQ). HCL is the growth story (13.3% YoY). Wipro is the value play (15.5x P/E). Infosys at 18.4x is fair given execution but not premium enough to justify “AI leader premium” thesis—which is exactly why the stock has de-rated from ₹1,600+ to ₹1,308.
Who Owns This Giant, and Are They Betting on AI?
- Promoters14.5%
- FIIs30.3%
- DIIs (incl. LIC 11.3%)41.1%
- Public13.7%
Pledge: 0.00%. Shareholders: 26.3 lakh (Dec 2025) vs 28 lakh in Mar 2023. Slight retail reduction—likely due to stock de-rating. LIC is the largest DII holder at 11.28%.
Promoters: The Murty & Nilekani Families
Sudha Gopalakrishnan (2.57%), Rohan Murty (1.64%), Nandan Nilekani (1.10%), Akshata Murty (1.05%). Collective ~14.5%, down from 15%+ in prior years due to secondary market sales and buyback. CEO Nandan Nilekali is not the founder (that’s N.R. Narayana Murthy, ~0.41%); he’s the professional manager. This is a professionally-run, low-founder-holding company. Alignment via stock, not control.
FIIs at 30.3%, Withdrawing Gradually
Singapore Government Pension Fund (exited 2024). Vanguard funds combined 2.6%+ Still invested, but less conviction. DIIs (41%) now larger than FIIs—institutional India is anchoring. The ₹18,000 crore buyback at ₹1,800 was partly de-risking for promoters/insiders who likely participated, leaving external investors to absorb the stock decline post-announcement. Smart money move.
Auditors, Boards, Pledges, and AI Day Optics
✅ The Institutional Strength
- ✓ CRISIL AAA/Stable rating (reaffirmed Jun 2025) — solvency is not a variable
- ✓ Clean audit history — no material qualifications
- ✓ Board 78% independent; 22% women; fully independent committees (audit, risk, remuneration)
- ✓ 323,578 employees; 39.5% women; attrition improving (12.3% Q3 vs 14.3% Q2)
- ✓ 1,869 active clients; 41 of 100+ million USD clients (sticky enterprise customer base)
- ✓ Zero material data breaches in FY25 (vs 1 in FY24) — ops are tightening
⚠️ Watch List
- ⚠ Promoter stake declining (14.5%) — standard in mature, listed companies, but watch for future capital needs
- ⚠ Significant FII ownership (30.3%); if FIIs exit en masse due to “US tech downcycle fears,” it could trigger a re-rating downward
- ⚠ AI strategy credibility depends entirely on execution. If 40,000+ AI hires don’t translate to 10%+ incremental revenue growth in 3–4 years, margin + growth thesis breaks
- ⚠ Analyst estimates assume “modest expansion >compression.” If compression hits 2–3% of revenue, FY26–27 EPS will disappoint badly
The Great Paradox: AI Will Obsolete Code. So Why Are Infosys & TCS Hiring 60,000+ People?
Global IT services market is ~$1.3 trillion annually. Of that, application development + support ~40%, infrastructure management ~20%, consulting + digital ~25%, BPO/support ~15%. The consensus narrative in 2024–25 was “AI will compress margins on coding by 40–50%.” It’s not wrong. A coder working with Cursor + Claude can write 2–3x more code in the same time. But here’s the industry insight: enterprises don’t have a “code shortage.” They have an “execution shortage.” Legacy systems are 10–30 years old, written in languages (COBOL, Perl) that nobody under 40 knows. They’re sitting on mountains of tech debt, siloed data, fragmented vendors. A CIO’s existential question is not “how do I save on code-writing costs?” It’s “how do I stay relevant if I don’t modernize 10 million lines of legacy systems while AI is reshaping every business process?” That’s a 3–5 year program, costs hundreds of millions, and requires change management, data stewardship, process redesign, vendor orchestration. And the only companies with the scale, global delivery capability, vendor relationships, and customer trust to do this at enterprise grade are TCS, Infosys, Accenture, Cognizant, IBM—the 50-year-old consultancies. So yes, coding gets cheaper. But “enterprise AI modernization delivered at scale” gets more expensive, more mission-critical, and more valuable. Infosys is betting on this thesis. If true, AI is not a headwind for Infosys. It’s a multi-billion-dollar tailwind that requires hiring 40,000+ people to meet the demand.
💡 Infosys’s Positioning: From “IT Services Vendor” to “Enterprise AI Execution Partner”
The rebranding (Feb 2026 Investor Day) was not marketing optics. It reflects a real shift in how large deals are structured. Instead of “build me an e-commerce platform” (commodity, high price sensitivity), it’s now “re-architect my supply chain, CRM, and HR using AI agents; integrate 7 existing vendors; retrain 500 people.” This is high-complexity, high-margin, high-stickiness work. And only a handful of firms can execute it. Management’s “Topaz Fabric + 600 agents” narrative is credible because 4,600+ AI projects are live with clients. Not vapourware. Not PowerPoint slide solutions. Live, paying customers deploying agentic AI into production.
⚠️ The Execution Risk: “AI Hype vs AI Delivery”
Every tech vendor is saying “AI this, AI that.” Infosys is one of many. The street’s skepticism is partly warranted: if AI truly disrupts IT services margins by 30–40%, no amount of hiring and re-skilling can offset that. Management says “expansion > compression.” They haven’t quantified it. If expansion is +200 bps and compression is -150 bps = net +50 bps, that’s not “explosive” growth. It’s margin stability at best. The stock is priced for that. But if compression is -300 bps, margins tank regardless. Timing matters. I’d expect FY26–27 clarity by end of FY26 Q4 (March 2026). If Q4 shows OPM stabilising and AI revenue >6%, the market will re-rate upward. If OPM dips to 22% and AI revenue stalls at 5.5%, downward re-rating is on the table.
🌐 Macro Backdrop: Macro-Lite, Not Macro-Boom
Management guidance of 0–3% constant-currency growth for FY26 reflects caution. US (56% of revenue) faces uncertain regulatory environment, rate cycles, and potential tech spending pullback if recession fears escalate. Europe (33%) is slower-growth. Upside: if US macro stabilises and CIOs unlock budgets for “AI modernization as urgency,” growth could surprise upward. Downside: if US recession materialises, AI projects get deferred, and FY26 ends at 0% growth (the lower end of guidance). The range is wide. Infosys’s execution inside that range will matter more than macro.
🚀 The Data Centre Bet: Long, But Real
AI infrastructure (GPUs, memory, cooling) is becoming a major capex theme. Hyperscalers (AWS, Azure, GCP) need 3–5 year partnerships with system integrators to plan, deploy, and maintain data centre infra. Infosys has entered this space via partnerships with Cognition, pilot programs with Indian hyperscalers. Revenue ramp is 2–3 years out. But if successful, it’s a new beachhead in a growing market. Management isn’t over-hyping it, which is a good sign.
The AI Paradox: Pricing for Disruption, But Data Says Execution
Infosys is a₹5.3 lakh crore IT services company trading at a de-rated valuation (18.4x P/E, below peer median 21x) despite executing the strongest quarterly revenue (₹45,479 Cr) in 13+ years, maintaining 37.5% ROCE, holding OPM at 23%+, and announcing a strategic pivot to enterprise AI execution that could unlock significant upside. The stock fell 19% in 3 months. The business fundamentals improved. That gap is either an opportunity or a trap.
Q3 FY26 Execution: Revenue +8.9% QoQ, the strongest growth in quarters. AI revenue reached 5.5% (₹2,500+ Cr annualised run-rate). OPM at 23.4%, down 80 bps YoY but within guidance. Operating cash flow at ₹35,694 Cr (FY25), up 41% in 3 years. EPS flat YoY due to tax & depreciation timing, not fundamental weakness. The execution is solid.
The AI Transition (Feb 2026 Investor AI Day): Management repositioned Infosys as an “enterprise AI execution partner,” not a foundation-model builder. Topaz Fabric platform abstracts model choice, 600+ agents live, 4,600+ AI projects in delivery, 90% of top-200 clients engaged. 40,000+ hires planned over 2 years to build out AI-first delivery capability. Revenue expansion from AI (new modernization + process automation deals) expected to exceed compression from AI-driven coding efficiency. No quantified net figure disclosed, but management tone is confident.
Historical Context: Stock delivered 0% CAGR over 5 years; +8% over 10 years. But total returns (dividends + buyback participation) were higher. The company is a “cash machine for shareholders,” not a “capital appreciation play.” It’s been trading at 17–20x P/E for a decade. Current 18.4x is mid-range. The de-rating from 22–24x is justified if margins compress 200+ bps permanently. It’s not justified if margins hold at 23–24% and AI revenue grows to 8–10% of total.
Base Case: Margins Hold, AI Penetration Rises
OPM stabilises at 23–24%. AI work grows to 7–8% of revenue by FY27 (from 5.5% today). Large-deal sales cycles remain 12–18 months (not shortened). Revenue growth at 5–7% CAGR. EPS CAGR 6–8%. Fair value range ₹1,084–1,500 holds. Stock re-rates to 19.5–20x in 12–18 months as execution clears (higher confidence, lower uncertainty premium).
Bull Case: AI Expansion Exceeds Compression
AI work scales to 10%+ of revenue by FY27. Modernization deals (higher margin, mission-critical) expand faster than traditional services shrink. OPM improves to 24–25% by FY27. Revenue growth 7–9% CAGR. EPS CAGR 9–11%. Fair value range ₹1,400–1,700. Stock re-rates as market recognizes AI-driven leverage.
Bear Case: Compression > Expansion
AI coding efficiency collapses margins on traditional services by 300+ bps. Large deal values fall as “AI commodification” spreads. Revenue growth stalls at 2–3%. Margin compression is permanent at 21–22%. EPS growth negative. Fair value range ₹800–1,050. Stock re-rates downward as consensus downgrades multiply.
✓ Strengths
- Second-largest IT services company in India; 323,578 employees; 1,869 clients
- 37.5% ROCE — best-in-class capital efficiency in traditional services
- ₹51,027 Cr cash surplus; net cash positive; zero solvency risk
- ₹35,694 Cr operating cash flow; 41% CAGR in 3 years
- Topaz Fabric platform with 600+ agents; 4,600+ AI projects in delivery
- 90% of top-200 clients engaged on AI initiatives; 5.5% revenue at run-rate
- 65–70% dividend payout; ₹18,000 Cr buyback completed (management conviction)
✗ Weaknesses
- OPM down 80 bps YoY; margin compression narrative gaining traction despite guidance
- Revenue growth (5–6%) slower than historical 10%+ due to traditional services maturity
- EPS growth (5–6%) not matching revenue growth — profit leverage has eroded
- Promoter stake declining (14.5%); low founder alignment in AI transition narrative
- FII stake shrinking (30.3%, down from 35%+); concentration shifting to DIIs
- AI revenue (5.5%) still small base; expansion thesis unproven at scale
→ Opportunities
- Enterprise AI modernization market (₹300–400 Bn by 2030 per external estimates)
- Legacy system upgrades (COBOL/Perl migration) remain urgent for Global 2000 companies
- Data centre orchestration as AI infrastructure scales (2–3 year timeline)
- Higher-margin consulting/outcomes-based pricing as AI work matures
- Attrition reduction (12.3%) improving delivery efficiency and utilisation
- Large-deal TCV (₹11.6 Bn in FY25, ₹9.5 Bn in FY26) still robust despite macro caution
⚡ Threats
- AI-driven coding efficiency permanently compressing traditional services margins
- Generalist GenAI tools (ChatGPT, Claude) enabling clients to “self-serve” modernization
- US tech spending slowdown if recession fears escalate (56% of revenue at risk)
- Talent competition for AI engineers — wage inflation could persist 2+ years
- FII exit flow as India growth narrative fades in favour of US mega-cap tech
- Analyst estimates for FY26–27 assume margin stability — any miss triggers multiple compression
Infosys is a high-quality business with a temporary valuation gap. The stock is priced for permanent margin compression and structural revenue decline. The company’s execution (Q3 results, AI penetration, cash generation) suggests neither is happening yet.
For income-seeking investors or those with a 3–5 year horizon, the dividend (3.29% yield) + potential buyback participation offer asymmetric upside if management’s AI thesis plays out. For short-term traders betting on immediate margin recovery, the 18.4x P/E is fair but not compelling. For those who believe AI kills IT services, stay away. For those who believe enterprise AI execution is a $300 Bn opportunity and Infosys is the primary beneficiary, this is a rare de-rated entry point. The market in 12 months will have more clarity. Until then, every quarter’s OPM and AI revenue growth rate is a data point that matters.