Indian IT runs at two speeds.

As the Indian technology sector closes the fiscal year 2026, the market is splitting into two clear halves. The era of 'Generative AI experimentation' has ended. Companies are now pushing hard to run 'Agentic AI' (AI that carries out tasks on its own) at full scale, and to rebuild their data systems. The Nasscom Strategic Review 2026 projects the Indian technology industry to reach $315 billion in FY26. That is a steady 6.1% rise on the year before. AI work has grown fast and now adds a real $10 billion to $12 billion to total revenue.

The biggest change shows in the jobs data. The sector added only 135,000 net new jobs, for a total workforce of about 6 million. Revenue grew well while hiring barely moved. This confirms that the old rule, where revenue grew in step with hiring, is broken for good.

The Indian technology sector is tracked by benchmarks like the Nifty IT and BSE Information Technology indices. It can no longer be treated as one single block that lives on cheap offshore labour. A sharp but needed change in how the business earns money is reshaping it. Traditional IT services rely heavily on Application Maintenance and Support (AMS) and on running old infrastructure. Their revenue is now shrinking, and the cause is lasting. Generative AI tools and coding assistants cut the billable hours needed for routine work. So clients push hard for these productivity gains to be passed on to them. They demand price cuts right away when Time-and-Material (T&M) contracts come up for renewal.

At the same time, large sums are moving to the 'build and innovate' layers of technology. Most of the new value goes to firms with real engineering depth. That means Engineering Research & Development (ER&D), custom chip design (VLSI), and complex cloud-native systems. India's Global Capability Centers (GCCs) have also grown fast. They now number over 1,760 and employ roughly 1.9 million people. They were once seen only as rivals for talent. Now they are major buyers that shape the market. This changes the Total Addressable Market (TAM: the total spend vendors can compete for) for listed third-party vendors.

For institutional investors, traders, and analysts, the message is clear. In future, the market will value most highly the firms that earn their margin from their own intellectual property (IP). It will severely penalise firms stuck billing for effort.

IT matters now because the world's technology is being rebuilt on its biggest scale since the move to the cloud. GenAI is forcing companies to rebuild their data systems entirely. Indian IT firms do most of the hands-on work for this shift.

The sector also gives Indian investors a rare hedge against the home economy. The domestic market trades at premium valuations. But IT earnings are linked to the dollar, spread across the world, and completely immune to slowdowns in Indian spending or rural distress.

When the Nifty falls on domestic issues, IT often steadies the portfolio.

The meaning of 'IT Services' in the Indian stock market has split apart. Analysts now value each part of a company separately (a sum-of-the-parts approach). Over the past 12 to 18 months, mid-cap digital engineering and ER&D specialists have kept beating the large diversified giants. This shows the market draws a sharp line between old maintenance work and high-value product engineering.

In stock-market terms, the Indian technology sector has four distinct types of business. Each is paid from a different client budget and earns very different margins. The base layer is the large diversified IT services firms, which run company operations end to end around the world. They take most of the Chief Information Officer's (CIO) 'run' budget: IT infrastructure, Business Process Management (BPM), and application maintenance. This segment generates a lot of cash. But it is very exposed to AI cutting the amount of work, and to hard pressure on prices.

The high-growth edge of the sector covers Digital Engineering, ER&D, and Software Products. Digital engineering firms target the Chief Technology Officer's (CTO) 'build' budget. They break large, single-block company systems into small independent parts (microservices), so client data is ready for AI. ER&D specialists work where physical and digital engineering meet. They write embedded code (software that runs inside a device) for Software-Defined Vehicles (SDVs), aircraft systems (avionics), and medical devices. This work must meet strict rules, safety standards, and deep hardware needs. So it is the hardest field for newcomers to enter, and clients almost never switch.

In early April 2026, the Nifty IT index traded near the 30,441 level. It was recovering from a severe fall since the start of the year. Fears that AI would push prices down, and global economic worries such as high US-Iran tensions, caused that fall. The broader Indian market hit record highs at times in the year before. But the IT index moved sideways for a long time. This shows the market was carefully lowering the value it puts on technology stocks as AI disrupts the industry.

The Nifty IT index tracks the free-float market value of the 10 most traded Indian IT companies. It is the main benchmark that institutions use. At the end of March 2026, a few stocks made up most of the index. So it reflects how large-cap legacy work is doing, more than real innovation across the sector.

For a broader view, the BSE Information Technology index covers 80 listed technology firms. It includes the long tail of mid-cap and small-cap engineering specialists that the Nifty IT leaves out. The sector, led by large caps, now trades at about 17.3x to 21.6x its expected earnings for the next 12 months (forward P/E). It also pays a healthy dividend yield of roughly 2.4% to 2.8%. This is back to pre-pandemic averages. The extra premium the market paid during the 2021 digital boom has gone. But this average hides wide gaps underneath. Firms heavy in legacy work trade at big discounts. Fast-growing mid-caps with strong AI work in the pipeline trade at 30x to 40x or more.

Business model quality is measured by: (1) Revenue stickiness, (2) Margin protection from AI deflation, (3) Ability to grow revenue without growing headcount, (4) Client switching costs.

Generative AI coding assistants (e.g., GitHub Copilot, Anthropic's Claude) now make routine software building and testing 20% to 30% more productive. As a direct result, clients push hard for these savings to be passed on to them. This is breaking, for good, the effort-based billing model that built the Indian IT sector.

In the past, the sector grew in line with its people (a 'people-scale' model). It billed clients on a Time-and-Material (T&M) basis, charging by the hour for each Full-Time Equivalent (FTE) engineer deployed. In the Agentic AI era, this model is a serious financial weakness. If an Indian vendor works faster with its own AI tools on a T&M contract, it bills fewer hours. So its revenue falls directly. The client keeps all the savings, while the vendor's revenue shrinks.

To protect profits and keep value, Indian listed firms are moving fast to 'IP-scale' models, where revenue grows with what they own rather than with headcount. For many, this shift is a matter of survival.

In IT, the quality of revenue depends on four things: how long clients stay, how deep the industry focus is, how the work is billed, and whether demand lasts or comes and goes with the cycle. Not all revenue is equal. A dollar of ER&D revenue built into a carmaker's production line is worth 3x a dollar of generic application maintenance revenue that can be automated away.

Data from Q4 FY26 shows a sharp split in spending across industries. Optional spending in general retail and consumer-facing (B2C) BFSI is still weak. Geopolitical worries (such as US-Iran tensions) and delayed interest rate cuts weigh on it. In contrast, demand for Software-Defined Vehicles (SDVs), chip design, and core AI data systems has picked up. This shows a lasting shift from day-to-day running costs (OpEx) to long-term investment (CapEx).

Demand in the Indian technology sector comes from two separate client budgets. Which budget pays decides how good and how lasting the revenue is.

The story has shifted sharply. Firms used to announce 'AI Centers of Excellence' for publicity. Now institutional investors want Annual Contract Value (ACV) that can be checked. In Q3/Q4 FY26, tier-1 firms began to show scale. TCS reported AI services revenue running at $1.8 billion a year. HCLTech said its Advanced AI services grew 19.9% on the previous quarter to $146 million. But overall growth across the sector stayed weak. This suggests AI is eating into old revenue faster than new change projects can grow.

AI affects Indian listed companies in three distinct financial ways, and analysts must keep them strictly apart:

Traditional IT industries were weak in FY26. Even so, pure-play ER&D firms and the engineering arms of tier-1 IT majors kept growing, quarter after quarter. KPIT Technologies grew for its 22nd quarter in a row, with 20.6% EBITDA margins. LTTS won a landmark $100 million multi-year program from a US industrial equipment maker. Both show the lasting demand for work that joins physical and digital engineering.

Nasscom projects Engineering Research & Development (ER&D) to be a $63 billion segment in FY26. It is the best-protected business in Indian technology. In ER&D, a company engineers physical products for other companies, along with their digital copies (digital twins). Generic IT runs BPO and back-office systems. ER&D instead writes the embedded code that sits on chips inside connected products. These range from car ADAS and battery management systems to FDA-compliant medical imaging devices and aerospace turbines.

The Zinnov-Indiaspora GCC AI Opportunity Report (March 2026) found a stark fact. 55% of the work that GCCs in India do today (mainly routine, standard tasks) is at direct risk of being replaced by AI. To survive, GCCs are moving fast to higher-value work. They are bringing in-house high-margin 'frontier' skills in deep engineering, custom chips, and complex AI modelling (now 45% of the GCC mix). India ended 2025 with over 1,760 GCCs employing 1.9 million people. This makes them the dominant technology force inside the country.

Global Capability Centers are the in-house technology, engineering, and shared services hubs that global companies run in India. For listed Indian IT vendors, they cut both ways, more than anything else.

The Indian IT sector rises and falls with the health of US and European companies. When interest rates are high and economic visibility is low, clients delay discretionary tech spending. They focus on 'cost-optimization' deals rather than 'transformation' deals.

Vertical exposure defines a company's resilience. BFSI (Banking, Financial Services, and Insurance) is the largest vertical for almost all major IT firms. If US regional banks face a crisis, Indian IT feels the tremor immediately. Retail is highly sensitive to US consumer spending. Healthcare and Hi-Tech often move differently, sometimes against the cycle.

Relying on one region cuts both ways. A large US business gives scale, but ties the company fully to the US economy. Europe spreads the risk, but brings complex labour laws, language barriers, and swings between currencies.

The Q3 and Q4 FY26 results showed again how much cash the sector produces, even while revenue growth was weak. TCS reported Operating Cash Flow at 100.4% of Net Income. HCLTech kept free cash flow at 120% of net income (FCF/NI) over the trailing twelve months. Cyient's DET business turned an exceptional 157.6% of its profit after tax into free cash flow (FCF to PAT).

The defining financial strength of the Indian IT and ER&D sector is still how little capital it needs, and how strong its balance sheets are. The move to AI-led ways of working costs money. Even so, top firms earn exceptional Return on Equity (ROE) and Return on Capital Employed (ROCE). For example, TCS keeps its ROE above 50%, while Infosys and HCLTech sit safely in the 25% to 30% range.

However, investors must watch closely as mid-tier players need more capital. Firms like Coforge and Persistent Systems report rising contract assets and capex. They are paying for their own AI platforms and taking on more fixed-price change deals that favour the client.

In Q3 and Q4 FY26, operating margins moved in sharply different directions. It depended on how well firms ran and what kind of work they did. HCLTech faced near-term margin pressure (EBIT at 18.6% including restructuring costs). Persistent Systems held up with a 16.7% adjusted EBIT margin. It absorbed wage hikes through productivity gains from its AI platforms and lower subcontractor costs.

Operating margins in this sub-sector come from a constant balance of three things: pricing power, how much staff time is billed (utilisation), and how efficiently work is delivered.

Early 2026 brought a sharp fall, driven by fears of AI disruption, weak growth in legacy work, and worries over US tariffs. Nifty IT valuations have now reset. The sector's 1-year-forward P/E fell to about 17.3x to 21x. That is back to its 10-year historical average from before Covid.

The 2026 market has split in two. So it is a basic mistake to apply one 'IT Services' valuation multiple to the whole sector. Valuation now depends on two things: how far a firm's revenue can grow without adding staff, and how much of its work joins hardware and software.

In FY26 it became clear that the broad 'IT Services' label tells analysts very little. Analysts expected the same growth from a legacy infrastructure manager and a pure-play automotive software engineer. That led to serious mispricing.

To judge revenue visibility and execution risk correctly, Indian listed technology firms must be split up by how they actually earn their margins:

The gap between general IT system integrators and industry-specialist engineering firms grew much wider in Q4 FY26. Pure-play ER&D firms began to compete head-on with global tier-1 engineering consultants (such as AVL or Ricardo) to own core automotive products. In these bids, traditional Indian IT peers were not even in the race.

The listed Indian technology universe splits into four distinct groups. Each one needs to be judged in a different way:

Q3 and Q4 FY26 results confirmed a sharp gap in how well firms performed and how well they held up. Mid-tier engineering firms like Persistent Systems (4.0% QoQ CC growth) did better. Tier-1 majors reported weak quarter-on-quarter growth (-0.8% to +1.5% QoQ), as legacy maintenance pulled revenue down.

The market is now sorting the 'growth compounders' from the 'margin defenders'. TCS and Infosys protect margins through huge scale and internal efficiency. Mid-caps like Persistent and ER&D specialists like KPIT drive real, compounding revenue growth through niche industry skill.

AI is changing the sector so fast that a lead over rivals now lasts far less time. At contract renewal, clients now ask for 20% to 30% productivity savings to be passed on to them. This has turned AI from a far-off threat into a real, measurable hit to quarterly revenue across the sector.

Investors must weigh risks that overlap. Some are lasting, and some come and go with the cycle:

Management guidance for FY27 points to a cautious recovery that is split sharply in two. Large caps like Infosys and HCLTech are expected to guide for modest 3% to 6% CC (constant currency) growth, held back as new work eats into old revenue. In stark contrast, mid-tier digital engineers and ER&D specialists are guiding for lasting double-digit growth.

The sector has entered a messy period of change. Over the next 12 to 24 months, the stock market will ruthlessly separate the 'AI implementers' from the 'legacy maintainers.'

The old measure of the sector's health, total headcount added, no longer works. In FY26 the industry was projected at $315 billion, yet it added only 135,000 net new employees. This shows that AI-driven automation of delivery has broken the link between hiring and growth for good.

To judge business quality and valuations properly in 2026, analysts must track a new set of operating metrics:

Tracking IT requires interpreting management commentary closely. The small difference between 'cautious optimism' and 'deal pipeline strength' matters a great deal.

The Indian technology sector is going through a lasting upgrade that cannot be undone. For decades, the story was global labour arbitrage: supplying coding hours of acceptable quality at a fraction of Western wages. Generative AI and the growth of the GCC ecosystem have broken that story for good.

From here, the advantage lies entirely in selling skill (capability arbitrage). Value now comes from engineering depth, owned intellectual property, and the ability to join complex hardware and software systems. The fast rise of GCCs in India adds pressure. Listed vendors must move from simply carrying out work to becoming co-innovation partners that clients cannot do without.

For stock market participants, the order of valuations is clear. Premium multiples will go only to firms that run AI platforms well, work deep inside physical R&D cycles, and fully break the link between hiring more people and earning more revenue. The 'Run' layer is shrinking; the 'Build' layer is compounding. Investors must give company marketing hype very little weight. They should position their portfolios toward engineering depth that can be verified, and margins that grow without adding headcount.