In today’s Finshots, we break down the Indian IT selloff and what this means for you as an investor.

The Story

Over the past 12 months, the Nifty IT index has fallen 21%. And almost 95% of that fall happened in just the last one month. In fact, that’s its worst drop in 23 years.

To give you a sense of scale, companies in the NIFTY IT index together lost over ₹6 lakh crore in market value in a single month. And unless you’ve been living under a rock, you probably know why.

One word. Anthropic.

In early February 2026, it launched a set of “agentic” AI tools powered by its Claude model. These are tools that not just help you write a few lines of code but can also handle entire engineering tasks on their own — writing code, testing it, fixing errors, deploying it, and even monitoring performance after launch.

One good example is Claude Code. It can work with old programming languages like COBOL (Common Business Oriented Language), a 60-year-old system still used in banking. It can scan massive codebases, understand how everything connects, document it, and even convert it into modern languages like Java or Python.

Put simply, instead of telling a developer “Hey, code this feature”, you tell the AI, “Build, test, deploy and maintain this entire system”, and it does most of the work with minimal supervision.

And that kind of scared investors because these are exactly the kinds of services Indian IT firms have traditionally billed its global clients for. For decades, these clients came to Indian companies because we had large teams of engineers who could customise software, test it, maintain it, fix bugs, and run support operations — all at lower costs.

Besides, application maintenance and small upgrades such as routine fixes and updates, have historically contributed close to half of revenues for many firms. And these are precisely the tasks AI now claims it can automate. What once required years and “armies of consultants” can now happen in months.

And naturally, investors panicked. If US clients, who form the bulk of Indian IT revenues, start thinking, “Why pay for a 50-member team when AI can do this cheaper and faster?”, that hits the core business model.

And that explains why Indian IT stocks experienced such a sharp sell-off.

But is this panic fully justified, you ask?

Well, first things first. It’s true that agentic AI can hurt Indian IT companies. For context, earlier, large contracts signed by Indian IT firms usually lasted around five years. When clients renewed them, companies would offer discounts of about 15–20%. That was normal. But now, with AI promising higher automation and lower costs, clients are asking for steeper discounts — closer to 20–30%. That puts pressure on both revenues and margins, which means that in the short term, revenues could slow to around 6–7% annual growth.

But that’s just the short-term picture.

When you’re looking at the sector as an investor, the bigger question is what happens in the long run.

And here’s where things get interesting. The actual ground reality looks quite different from the panic in the markets. Investors seem to be reacting as if AI will completely wipe out Indian IT companies in a few years.

But is that realistic?

Manish Sonthalia, Director and Chief Investment Officer at Emkay Investment Managers, shared a simple example with The Economic Times. A few years ago, when electric vehicles (EVs) became the next big thing, markets almost wrote off traditional fuel-based vehicles or ICE (internal combustion engine) cars. The assumption was that they would soon become obsolete. But that didn’t happen. In fact, stocks of companies linked to ICE vehicles, including auto manufacturers and component makers, went on to rally three to four times.

This simply tells you how markets sometimes overreact to disruption. And reality often turns out to be more balanced than the initial fear. The AI effect on Indian IT could be very similar.

Just because agentic AI has arrived, it doesn’t automatically mean Indian IT or SaaS (Software as a Service) companies will become obsolete. If someone believes that, they probably haven’t spent much time working inside enterprise IT. And that’s not us saying it. That’s Bernard Golden, CEO of Navica, a US-based tech consulting and investment firm.

According to him, sure, AI makes coding easier. But writing code is only a small part of building real enterprise software. Business software isn’t just lines of code. It involves industry knowledge, regulatory compliance, security checks, legal contracts, customisation for large clients, integrations with dozens of other systems, ongoing upgrades, and long-term maintenance. There are sales teams, support desks, and consultants who help clients across countries.

Along with software, large companies also buy reliability and accountability. This means that they need vendors who can operate in 20 countries, provide local support, handle complex contracts, and take legal responsibility if something goes wrong. An internal IT team can’t simply replicate all of that just because AI can generate code faster.

Golden also points out something really interesting. In the past, many companies tried building their own systems thinking it would be cheaper or more efficient. But most underestimated the complexity and failed. And eventually, they went back to established vendors.

So writing obituaries for Indian IT companies might seem a bit premature.

And that could give you a good reason to look at the brighter side of this in the long term.

See, currently nearly 80% of enterprise IT spending goes towards maintenance and only a small portion goes towards real innovation. But if AI reduces maintenance costs, companies won’t suddenly slash their overall tech budgets. That’s because in large organisations, once you cut a budget, it’s very hard to get it increased again. So most firms prefer to keep the budget intact.

That means the same budget, but more room for innovation.

In fact, AI could even reduce the risks that come with innovation. Take a bank that still runs on decades-old code written by engineers who’ve long retired. Moving that system to modern software is risky and expensive. But if AI can understand old code, translate it, test it and reduce errors, the risk falls sharply. That’s not a threat. That’s an opportunity, no?

So how should you, as an investor, read into all this?

Well, one option is to follow the herd. Sell now, and worry about AI later. Many foreign investors have done exactly that by shifting money into “AI infrastructure” themes like chips, power and domestic cyclical stocks.

But for long-term investors, that may not be the smartest move.

Now, it’s true that the old model — billing clients for large teams and long hours, could see pressure. Lower-end, people-heavy work may face slower growth over the next 3–5 years. But Indian IT companies have adapted before. They moved from Y2K fixes to client-server, then from on-premise to cloud. Now the shift could be from selling hours to selling outcomes and AI-powered platforms.

Interestingly, this selloff has also cooled valuations. Indian IT stocks once traded at a 30–40% premium to US peers. Now many trade at around 21 times earnings (P/E), narrowing the gap to about 15–20%. That means they’re closer to fair value. For investors who stayed away because they looked expensive, this reset isn’t necessarily bad.

Because in the end, this isn’t just about Anthropic or one AI demo. Remember how back in 2021, markets overreacted with extreme optimism around SaaS and AI? Anything with “SaaS” or “AI” in its name traded like it would grow at 50% forever.

But today, we’re seeing the exact opposite or a pessimistic overreaction. Profitable, cash-rich IT companies are suddenly being treated as if they’re outdated, all because a slick demo video made AI look unstoppable.

And if history is any guide, neither extreme usually lasts.

Until next time…

Don’t forget to share this story with a friend or family member who’s been biting their nails over the Indian IT selloff, or even with curious strangers on WhatsApp, LinkedIn and X.

Myth Alert: I'm Too Young to Buy Life Insurance!

The other day, one of our founders was chatting with a friend who thought life insurance was something you buy in your 40s. He was shocked that it was still a widely held belief.

Fact: Life Insurance acts as a safety net for your family. The younger you are, the cheaper it is. And the best part? Once you buy it, the premium remains unchanged no matter how old you get.

Unsure where to begin or need help picking the right plan? Book a FREE consultation with Ditto's IRDAI-Certified advisors today.