India's artificial intelligence infrastructure boom has turned several companies into multibaggers, but beneath the surface lies a troubling disconnect between stock price appreciation and earnings fundamentals. As global warnings about an AI bubble intensify, investors need to separate genuine long-term opportunities from near-term valuation excess.
The company trades at a staggering 118 times earnings, despite revenue growth of just 29%. While its AI systems segment grew 460% year-on-year and now contributes 43% of revenue, the deteriorating capital efficiency suggests investments are still in deployment phase with measurable returns yet to materialize. Transcripts +1
Sterlite Technologies presents an even more extreme case. The stock has rocketed 474% year-to-date in 2026, yet its return on equity has collapsed from 40% to under 9% over six years. The company trades at 512 times earnings despite reporting negative net profit of ₹57 crore in FY25. Its price-to-book ratio of 12.7x values the company at 262% above its book value—a premium that demands flawless execution for a decade.
E2E Networks tells perhaps the most concerning story. While the stock has tripled over two years, return on equity has crashed from 36% to under 6%, and the company is reporting negative net profit growth across all quarters of FY26. The stock trades at 5.3 times book value despite deteriorating profitability, with negative earnings per share rendering traditional valuation metrics meaningless.
Anant Raj stands apart with relatively stronger fundamentals. The stock has doubled over two years, but return on equity has actually improved from 0.34% to 10.9%. Trading at 35 times earnings with a price-to-book of 3.3x, its valuation premium appears more reasonable given improving capital efficiency and strong domestic data centre demand. InvestorPresentations +1
The vulnerability of these companies became starkly apparent when examining their exposure to global hyperscaler spending patterns. HFCL Limited secured a landmark $1.1 billion contract with a global hyperscaler, representing 48% of its total order book of ₹21,200 crore. This creates significant concentration risk—if global AI capital expenditure recalibrates, HFCL's order book could decline by 40-50%. Transcripts
Sterlite Technologies derives 78% of its revenue from the Americas and Europe, making it highly sensitive to global AI infrastructure investment cycles. The company's enterprise and data centre segment, which it expects to scale to 30% of revenues, is directly tied to hyperscaler spending patterns. Any slowdown in global AI capex would disproportionately impact its optical fibre and networking equipment revenues. Transcripts +1
Netweb Technologies appears relatively insulated with only 5% of revenue from exports. Its AI server demand is primarily driven by domestic government initiatives, particularly the India AI Mission which has expanded from 10,000 to 25,000 GPUs with potential to scale to 100,000. However, the company remains dependent on global technology partners like NVIDIA, AMD, and Intel for chipsets, creating indirect exposure to global spending cycles. Transcripts +1
Perhaps the most concerning aspect is the lack of measurable returns on AI infrastructure investments.
This creates significant demand uncertainty for infrastructure suppliers.
Netweb Technologies and E2E Networks both show accelerating revenue growth but deteriorating profitability. Netweb's AI segment margins remain undisclosed, while E2E's net profit growth is negative across all FY26 quarters despite revenue growth accelerating from 23% to 72%. This suggests companies are investing heavily in capacity without corresponding profit improvement.
Sterlite Technologies' capital efficiency has deteriorated significantly. Asset turnover stands at just 0.27x compared to industry averages of 0.5-0.8x, while free cash flow generation remains volatile. The company's high-performance computing solutions are not delivering returns that justify its extreme valuation premium.
Anant Raj's data centre business shows more promise but lacks transparency. The company reports ₹176 crore in data centre revenue from 28 MW capacity, targeting ₹1,200 crore at 63 MW capacity. However, segment-wise return metrics are not disclosed, making it difficult to assess whether data centre investments are generating the 18-22% ROE needed to justify current valuations. Transcripts +1
Sterlite's 512x P/E represents a 59x premium to what its 8.65% ROE would justify, while HFCL's 83x P/E trades at 19x above its 4.31% ROE. These valuations are completely disconnected from operational reality.
The second-order risks are severe. If execution delays occur in AI infrastructure projects, Netweb Technologies and E2E Networks could face 50-70% stock price corrections. The high valuation base creates unrealistic growth expectations that compound execution challenges into existential threats.
India's AI-linked stocks trade at premiums 3-34x above historical averages. The Indian tech sector trades at 20.5x earnings, while Sterlite trades at 512x and Netweb at 118x. Even compared to global tech benchmarks like the S&P 500's 33x P/E, these valuations appear extreme.
This provides some fundamental support, but current valuations appear to discount perfection with no margin for error.
Anant Raj's valuation appears most sustainable among the group. Its 35x P/E represents a 2-3x premium to historical averages, which could be justified if it captures 2-3% market share of the projected $45.7 billion market by 2033. The company's improving ROE and strong domestic focus provide resilience against global tech corrections.
Netweb Technologies has the strongest balance sheet with debt-to-equity of just 0.02 and a current ratio of 2.61, giving it high resilience to weather global tech corrections. However, its 118x P/E still represents 4-6x above historical averages, requiring sustained 35-40% growth for years to justify.
Sterlite Technologies and HFCL Limited represent 70-90% near-term valuation excess with only 10-30% long-term AI opportunity. Their extreme valuations, weak balance sheets, and high global exposure make them vulnerable to severe corrections if AI sentiment reverses.
The analysis suggests investors should exercise extreme caution. Sterlite Technologies and HFCL Limited should be avoided or significantly reduced due to extreme valuation risk. Netweb Technologies could be held or reduced given its strong balance sheet but high valuation. Anant Raj represents the most sustainable opportunity with reasonable valuation and improving fundamentals. E2E Networks should be avoided due to poor fundamentals and high execution risk.
The key is distinguishing between genuine long-term AI opportunities and near-term sentiment-driven excesses. India's data centre industry growth is real, but current valuations discount perfect execution for extended periods. Investors should focus on companies with strong balance sheets, improving capital efficiency, and reasonable valuations rather than chasing AI sentiment alone.
