The AI panic and the case for doing the work

Summary: When AI headlines sparked a sharp sell-off in Indian IT stocks, fear ran ahead of facts. While the market reacted to a flashy demo, careful research told a far more nuanced story about what had actually changed and what hadn’t. It’s a reminder that in investing, doing the homework often matters more than following the noise.

I've been writing about AI (artificial intelligence) and investing for the better part of two years now, trying to navigate between breathless enthusiasm and reflexive dismissal. In a recent column, The Many Worlds of AI, I argued that we've been thinking about AI all wrong — treating it as a single phenomenon, a unified force that will sweep through the economy at roughly the same pace, when in reality it is many parallel stories moving at wildly different speeds. Some domains have achieved genuine escape velocity. Others remain stuck in what the industry calls ‘pilot purgatory’. And the distinction has almost nothing to do with AI's intelligence.

That column was published as a framework for thinking. What happened to Indian IT stocks in February was, in effect, a live demonstration of what happens when people don't use the framework — when they treat AI as a single undifferentiated wave and react to a story rather than examine the facts beneath it.

What the market heard and what it missed

In the first two weeks of February 2026, the Indian technology sector experienced a classic bout of fear. Roughly Rs 50,000 crore of market value disappeared from Indian IT stocks in a handful of trading sessions, with the Nifty IT index falling nearly a fifth from its early February peak. The immediate trigger was an AI product demonstration — a tool that could stitch together multi-step workflows across legal, finance, sales and analysis. Investors saw the demonstration and drew a straight line: if a piece of software can review contracts, build financial models and test code, why would clients keep paying Indian vendors to do the same?

The market heard “AI can do what Indian IT does” and sold. It was more like a panic — everything with ‘IT’ in the description was sold, no matter the quality. Foreign institutions, which for years had treated Indian IT as a sort of growing annuity, turned net sellers overnight.

But the market missed the framework that would have saved it from panic. As I wrote in The Many Worlds of AI, AI thrives where mistakes are cheap, private and correctable — in coding, data transformation and internal analytics. It struggles where mistakes are expensive, public and permanent — in customer-facing communication, compliance-heavy workflows and complex enterprise operations. The Indian IT services model doesn't neatly fit into either world. It spans both. These companies perform work in the fast world — coding, software development — but sell it as a service wrapped in domain knowledge, client relationships, legacy system understanding and institutional memory. That service wrapper involves exactly the ambiguity and human judgement that characterise the slow worlds of AI adoption.

Selling everything with ‘IT’ in the name because of one AI demonstration is like closing every restaurant in the city because someone invented a better microwave. The tool changed. The need for someone who understands your particular kitchen — your systems, your regulations, your customers, your history — did not.

The research that was ready when the panic wasn't

At Value Research, we were not selling. We were doing what we always do: reading balance sheets, studying earnings transcripts, tracking cash flows and asking the only question that matters to a long-term investor — has the business actually changed, or has the price simply got frightened?

The answer, for one company in particular, was unambiguous. Here was a business at the centre of the sell-off whose fundamentals told a story the market was too panicked to hear.

A return on equity (ROE) above 15 per cent, sustained year after year with zero debt. And the business the market said AI would destroy? Its AI services revenue was growing at a healthy pace. It has set up dedicated AI labs for clients across industries. It is not being disrupted by AI. It is leveraging AI.

The stock had fallen from a 52-week high to trade at more than 20 per cent discount to its own history. That is a valuation typically reserved for businesses with serious structural problems. This business has none. It has healthy margins, a strong balance sheet, and reliable cash generation.

The market was charging a price usually reserved for broken businesses. The cash flows were not broken. The narrative was.

We are making this company our newest Value Research Stock Advisor recommendation. The full investment case — the company name, financial analysis, valuation basis, risk assessment and buy rationale — will be made available to subscribers soon.

But I want to be honest about something. This recommendation, as timely as it is, is not the real reason I'm writing this piece. It is the proof, the most recent, most vivid proof, of something more important.

The real question: what were you reading on February 13?

On the morning of February 13, when the sell-off was at its most intense, most investors were reading headlines. The headlines said AI would destroy Indian IT. The headlines said ‘sell’. The headlines were loud and confident, and, as headlines usually are, they were one-dimensional.

We were reading something different. We laid out for ourselves, in plain language, what this company actually earns, how it generates cash, what it returns to shareholders, what it is doing with AI and what the current price implied about the market's expectations. We could see the numbers behind the noise. We made the decision — not based on fear, not based on tips, but based on the kind of thorough, patient research that most individual investors simply cannot do alone.

Not because they lack intelligence. Because they lack time.

This is what Value Research Stock Advisor actually is. And it is what I want to talk about for the rest of this piece — because understanding what happened in February is useful, but understanding how to be prepared for the next February is far more valuable.

What Stock Advisor is and what it is not

Stock Advisor is not a tip sheet. It does not send breathless alerts. It does not shout “BUY!” before you've had your morning coffee. It does not promise to double your money or reveal ‘hidden gems’.

Stock Advisor is a research desk. A quiet, thorough, slightly obsessive research desk that does one thing well: the work.

Every recommendation begins with the business — not the stock price. We study what the company makes, who it sells to, why customers stay, what its advantages are and whether those advantages are durable. We examine the financial history across cycles. We assess valuation against realistic expectations. We describe risks with the same care as opportunities.

The result is not a target price. It is an argument — a complete, documented argument you can read, question, and use to make your own informed decision.

What the subscription includes

• Full investment cases: The complete argument for every recommendation: business model, financial history, valuation, and risks.
• A monitored portfolio: Every recommendation tracked and updated as events unfold.
• Sell calls when needed: We tell you when to leave, not just when to arrive.
• Stock Advisor Live: Live sessions where we walk through active recommendations and answer subscriber questions. The first session is on March 14, 2026 at 12:30 pm.

Why this matters more than one good recommendation

A good stock recommendation is worth subscribing to. A reliable research process is worth staying for.

I want to draw a distinction here that I think most investors miss. A recommendation is a point in time: a company at a price with a thesis. It can be right or wrong. Over a long enough period, even the best research process will produce some recommendations that don't work out. That is inherent to investing in an uncertain world. What separates a good advisory service from a bad one is not a perfect batting average — no one has that — but the rigour and consistency of the process, the honesty about what went wrong when it does, and the discipline to act on new information rather than defend old positions.

At Value Research, the process has been the same since 1990. We read the annual reports. We track the cash flows. We talk to management when it is useful, but we trust the numbers more than the narrative. We ask uncomfortable questions: Is this company earning its cost of capital? Is the growth real, or is it funded by dilution? Is the management returning cash to owners, or building empires? Is the valuation pricing in a future that is genuinely likely, or one that requires everything to go right?

These are not glamorous questions. They do not make for exciting television or viral social media posts. But they are the questions that, over thirty-five years, have kept our readers from making expensive mistakes during exactly the kind of panics we saw in February.

We have studied Indian companies through liberalisation, through Harshad Mehta, through the dot-com collapse, through 2008, through demonetisation and through Covid. We have not always been right. But we have always done the reading, and we have never pretended that a headline is a substitute for a balance sheet.

The many worlds problem and why you need a research partner

Let me bring this back to the framework from The Many Worlds of AI, because it illustrates precisely why a service like Stock Advisor matters.

The central insight of that column was that AI is not one revolution — it is several parallel ones, moving at different speeds. It thrives in some domains and struggles in others, and the determining factors are structural: error economics, verifiability and reversibility. Executives, investors, and commentators who treat AI as a single, monolithic force will consistently overestimate its impact in the slow world and underestimate it in the fast one.

Investing in an era of AI requires the ability to make these distinctions. When the market panics about Indian IT, you need to be able to ask: Which part of this business is actually vulnerable, and which part sits in a world where AI adoption will be slow, messy and dependent on the kind of human judgement that cannot be automated? When a CEO proudly announces that AI is transforming customer service — as Salesforce did before its comprehensive reversal, or as Bajaj Finance did before becoming the butt of every joke on X — you need to know enough to be sceptical.

This is not easy work. It requires reading earnings transcripts, not just headlines. It requires understanding business models, not just stock charts. It requires patience to sit with ambiguity when the market demands certainty.

Most individual investors — even intelligent, financially literate ones — do not have the hours for this. They have jobs, families and lives that rightly take priority. What they need is someone who does have the hours, the access, and the institutional experience to do this work properly and present it in a form that a thoughtful non-professional can act on.

That is what Stock Advisor is. That is what it has been designed, from the beginning, to be.

Thirty-five years of not following the crowd

Value Research was founded in 1990, at a time when mutual fund information in India was virtually non-existent. We created the country's first fund ratings and scorecards. Over the decades, we built the most comprehensive database of Indian mutual funds and stocks available outside institutional terminals. Millions of Indian investors use our tools, our ratings and our publications to make informed decisions.

Stock Advisor grows from that same soil. It carries the same DNA: a bias toward facts over stories, a preference for long-term compounding over short-term excitement, a deep scepticism toward anything that sounds too good to be true and an absolute commitment to showing the work.

When we launched Stock Advisor, we made a deliberate choice. We would recommend fewer stocks, not more. We would write longer, more detailed research notes, not shorter ones. We would track every recommendation honestly — the losses alongside the gains. We would issue sell calls when the thesis no longer held, even if that meant admitting we were wrong. And we would never, under any circumstances, chase the market's mood.

In February, when the market's mood was in panic, we saw an opportunity. Not because we are contrarian for the sake of it, but because the work pointed unambiguously in that direction. The numbers pointed in that direction. The 35 years of experience that taught us to distinguish between a frightened price and a broken business — those pointed in that direction too.

What happens on March 14

On March 14 at 12:30 pm, we are holding the first session of Stock Advisor Live — a new format where we discuss our new recommendations with subscribers in real time. In this session, we will reveal the company behind this recommendation, walk through the complete investment case, explain the valuation analysis, discuss the risks and answer your questions live.

I will be there personally, because I believe this recommendation matters — not just as an investment, but as an example of what careful research looks like in practice. It is one thing to read a note. It is another to hear the argument, see the numbers, ask the questions and make up your own mind with full transparency.

Stock Advisor Live is open to all Stock Advisor subscribers. If you are already a subscriber, you will receive the details and the link in your inbox. If you are not yet a subscriber, this is a good time to join — you'll get immediate access to the full recommendation note, plus the live session on March 14

A note on what kind of investor this is for

I want to end with a note of honesty, because Ogilvy — whose advertising principles I admire — always said the worst thing you can do is trick someone into buying something that isn't for them.

Stock Advisor is not for everyone. It is not for people who want daily trading calls. It is not for people who measure success in days or weeks. It is not for people who want to be told what to do without understanding why. And it is not for people who believe there is a shortcut to building wealth in the stock market.

It is for people who believe that understanding what you own matters. That doing the work — or having someone trustworthy do it on your behalf — is not optional but essential. That the best investment decisions are made calmly, with good information, not reactively, in the grip of fear or greed. And that the rare, uncomfortable moments when the market offers a genuinely good business at a genuinely good price are worth being prepared for.

If that sounds like you, we would be glad to have you.

Subscribe to Stock Advisor today