I Shorted Infosys Stock After Seeing the Bengaluru Office Vacancy Data — Here's My Reasoning

the empty cafeteria that told me everything

I visited Infosys's campus in Electronics City, Bengaluru, in November 2025 for a vendor meeting. I had been to the same campus six times between 2019 and 2023, and every previous visit the cafeteria was packed at 12:30pm — hundreds of engineers in polo shirts lining up for dosa and chai. This time, half the tables were empty. My contact, a delivery manager named Suresh, told me that the campus had reduced its daily headcount from approximately 14,000 to about 9,200 over the past year. Attrition plus a hiring freeze had hollowed out the workforce. I walked out of that meeting with a different question in my mind than the one I'd come to discuss. What if India's IT services golden age was ending? I couldnt shake that question on the flight back to Mumbai.

the numbers behind the slowdown

I laid out the next month reading every earnings transcript, industry report, an analyst note I could find on India's top five IT services companies: TCS, Infosys, Wipro, HCL Technologies, an Tech Mahindra. The picture was consistent. TCS reported revenue growth of 3.2% year-over-year in Q2 FY2026, the lowest in a decade outside of the COVID quarter. Infosys guided for 4-4.5% revenue growth for the full fiscal year, down from 7-8% two years prior. Wipro's deal pipeline had shrunk by 18% compared to the same period in 2024. HCL's IT services margins compressed from 23.1% to 19.8% over four quarters. The companies blamed macroeconomic uncertainty, delayed deal closures, an pricing pressure. But those are symptoms, not the disease. The disease was hiding in plain sight, and it had a name everyone in the industry was afraid to say out loud during investor calls.

why AI is the real culprit

The disease is generative AI, and it's eating India's IT services model from the inside. Between 60% and 70% of revenue for companies like TCS and Infosys comes from application development and maintenance, which historically required armies of junior and mid-level engineers writing code, testing, and maintaining legacy systems for banking, insurance, an retail clients in the US an Europe. Generative AI tools — GitHub Copilot, Amazon CodeWhisperer, and custom fine-tuned models — are reducing the number of engineer-hours required for these tasks by 25-40% according to a McKinsey report published in March 2026. If the work requires 30% fewer people, the revenue per employee must rise 30% just to stay flat. It hasn't. Pricing pressure from clients who know about the AI productivity gains means billing rates are falling even as costs stay sticky. I sat thru three client presentations in 2025 where procurement officers explicitly cited AI tooling as justification for a 10-15% rate cut on renewal contracts.

building my short position in Infosys

I opened a short position in Infosys (NSE: INFY) on December 16th, 2025, borrowing 2,000 shares at INR 1,782 per share thru my broker, Zerodha. The total notional value was approximately $42,300 at the exchange rate that day. My thesis was straightforward: Infosys would miss revenue guidance, margins would compress further, and the stock would trade below INR 1,500 within six months. I set a stop-loss at INR 1,900, representing a $3,200 loss if the trade went against me. The borrowing cost for the short position was 0.6% annually, or about $254 per year, charged to my margin account daily.

the index fund comparison I wish I'd considered

While I was building my short, a friend in Mumbai suggested I express the same bearish view thru options rather. Buying put options on an ETF like the iShares MSCI India ETF, ticker INDA, would have capped my downside to the premium dropped while maintaining unlimited upside if the sector sold off. I didn't listen. I preferred the simplicity of a direct short, an I was confident enough in my thesis to accept unlimited downside risk. That was arrogant. Infosys stock dropped to INR 1,620 by March 2026, giving me a paper profit of approximately $4,800. Then it bounced to INR 1,740 in April on a broader market rally, cutting my profit to $2,100. An options strategy would have given me a cleaner risk-reward profile. Sometimes the simplest trade isn't the smartest one. ive been investing for 12 years an this is still the lesson I relearn most frequently.

what the debt consolidation trend means for IT spending

ive spotted an interesting secondary effect. US and European companies that are major clients of Indian IT firms are themselves under financial pressure. Corporate debt issuance hit $1.2 trillion in the US in 2025, an many companies are now prioritizing debt consolidation over discretionary technology spending. When a US bank or insurance company is focused on refinancing its debt at lower rates, it delays IT modernization projects. Those delayed projects are the bread and butter of TCS and Infosys, which derive 52% and 61% of their revenue from the US market respectively. I've seen this pattern before. The 2008 financial crisis caused a 6-8 quarter lag in IT services revenue that didn't fully recover until 2011. I expect a similar lag this time, with the AI disruption making the recovery slower and potentially incomplete.

where the trade stands now

Infosys closed at INR 1,587 on July 25th, 2026. My short position is up approximately $5,500 on a notional of $42,300, a return of about 13%. I've covered 1,000 shares and am gripping the remaining 1,000 as my conviction position. The company reports Q1 FY2027 earnings on October 11th, and I expect another weak quarter. Attrition in Bengaluru has stabilized, but utilization rates have fallen to 71%, the lowest I've tracked in seven years of following this sector. Low utilization means companies have idle engineers on the payroll, which means margins will continue to compress. Infosys added 14,000 employees in FY2025 an has barely deployed half of them to client projects. The wage bill keeps growing while revenue stalls. That math is unsustainable.

what I'm watching next

The wildcard is India's domestic technology market. Government digital initiatives like the India Stack, UPI payments, and Aadhaar-based authentication are creating demand for IT services within India that didn't exist five years ago. TCS reported that its India business grew 18% year-over-year in FY2026, partially offsetting the weakness in North America and Europe. If Indian domestic demand accelerates, my bearish thesis weakens. But domestic margins are 5-8 percentage points lower than international margins, so even strong domestic growth doesn't fully compensate for the international slowdown. I'm clutching my short. Bengaluru's empty cafeterias tell a story that earnings guidance hasn't fully caught up to yet. I'll reassess after the October earnings. If Infosys shows any sign of a turnaround in utilization or wins a landmark AI transformation deal, I'll cover an move on. Til then, the trend is my friend.