AI is Deflating IT Margins but Fuelling This Opportunity

Five years.

That's all it took for Anthropic to command a valuation that surpassed the combined market worth of TCS, Wipro and Infosys.

Let that sink in.

For decades, India's middle-class story has ridden on IT escalator. With an endless supply of English-speaking engineers billable at rates that offers huge cost arbitrage to the world, IT has shaped many cities in India.

And now, a five-year-old AI company is derailing that. No wonder there's an existential discomfort hanging over the sector.

At its core, the Indian IT model is simple: sell human hours in dollars. The more people you deploy, the more revenue you book.

But with AI writing code, test software and smarten itself with every use at a fraction of the cost, that cost arbitrage has shrunk and could even turn negative.

Even if our large IT firms successfully integrate AI (and many are trying hard), the uncomfortable truth remains: AI tends to be deflationary. It reduces the number of human hours required.

This means the very revenue engine that powered growth could start slowing structurally.

And it's not just IT. This will have repercussions for real estate, consumption, financial services sector and far reaching second order effects.

But this is not a doomsday note. When value disappears from one pocket, it doesn't vanish. It migrates. And the real opportunity for us as investors is to ask: Where is it migrating?

My bet is that a meaningful part of it is shifting toward AI hardware.

While the world debates software disruption, an estimated US$ 7 trillion is being funnelled into AI infrastructure and data centers globally. This is not a quarterly fad. It has the feel of a structural shift.

Chips, memory, storage and high-performance computing hardware...

Hyperscalers and data centers are absorbing supply at a pace that is squeezing the rest of the market. When the giants hoard components, the downstream effect is inevitable - shortages and price spikes.

That is precisely why PC and laptop prices have risen sharply - in some cases 20-30%...and it's not done yet.

If your budget is Rs 37,000-40,000 for a decent PC today, the options are shrinking fast. And with the dollar strengthening, it's getting worse.

When new products become expensive and scarce, what thrives?

The secondary market.

This supply-constrained environment is quietly turning into a tailwind for the refurbished PC segment. With affordability and availability, it offers sustainability, making the value proposition stronger.

This is where GNG Electronics (Electronics Bazaar brand), enters the frame.

The company refurbishes desktops and laptops at scale and sells them across 44 countries. Nearly 60% of its revenue comes from international markets - largely the US and Europe, with a meaningful presence in the Middle East.

The management claims it can deliver a device that matches the performance and aesthetics of a new machine at roughly one-third the price. What makes that claim commercially powerful is the warranty structure - up to three years in India and one year internationally.

Importantly, the cost of these warranties is already embedded in its financial model and managed through in-house quality control and after-sales systems.

Warranties, in this business, are not just comfort - they are entry barriers.

On the distribution side, the company has built over 4,700 customer touchpoints spanning enterprises, institutions, distributors and channel partners. It also has partnerships with global names like Microsoft, HP and Lenovo, and is a premium pro seller on Amazon.

Recently, the management nudged its revenue growth guidance with improving profitability expectations.

Of course, this isn't a straight-line story.

Technology can become obsolete quickly. Working capital discipline will be crucial, especially as the company extends credit to distributors while carrying high inventory in a growth phase. And the refurbished space does have unorganised competition.

But here's the bigger picture...

If AI truly reduces the need for incremental human hours, it will not reduce the need for machines. In fact, it will increase it. Every model trained, every application deployed, every AI-powered workflow ultimately runs on hardware.

And when new hardware becomes expensive, someone who can intelligently recycle, refurbish, and redistribute computing power sits at an interesting junction of affordability and necessity.

The Indian IT story made millionaires by exporting human intelligence. The next phase of wealth creation may well be about enabling AI.

Warm regards,


Richa Agarwal
Editor and Research Analyst, Hidden Treasure
Quantum Information Services Private Limited (Research Analyst)