01 — At a Glance
The Mid-Sized IT Player That’s Stealing the AI Narrative
- 52-Week High / Low₹3,037 / ₹2,025
- Q3 FY26 Revenue (USD)$451 Mn
- FY25 PAT (Full Year)₹1,702 Cr
- Q3 EPS₹23.21
- Annualised EPS (Q3×4)₹92.84
- Book Value₹505
- Price to Book4.39x
- Dividend Yield2.57%
- Debt / Equity0.21x
- LTM TCV (12 months)$2.1 Bn
The Opening Auditor’s Note: Mphasis closed Q3 FY26 with $451 million quarterly revenue (+1.5% QoQ, +7.4% YoY in constant currency). But the headline isn’t revenue — it’s the $428 million TCV wins in a single quarter with 69% AI-led pipeline. The company’s 52-week journey? Down 23.5% in three months (ouch), yet the machine keeps printing large deals. Blackstone reduced stake from 40.10% to 30.59% in November 2025. The stock is still searching for the bottom. But the pipeline is moving sideways into orbit.
02 — Introduction
The Transformation Partner That’s Playing the AI Zeitgeist to Perfection
Mphasis. Two syllables. One of those mid-sized IT services companies that never makes headlines until it suddenly does. The company was born from the EDS-HP buyout saga (2006 acquisition by EDS, later inherited by HP in 2008), and by 2016 — when most IT companies were obsessing over margin defense — Blackstone bought HP’s stake and positioned Mphasis as their “India-plus” transformation boutique.
Fast forward to 2025: TCS and Infosys are sweating over growth narratives in a mature BFSI market. Meanwhile, Mphasis is out here talking about “Savings-Led Transformation” and positioning GenAI as the structural reallocation of client tech spend away from effort-based billing and toward outcome-based deals. The pipeline has grown 2.5x since launching Mphasis.ai as a dedicated business unit.
Revenue base: ₹14,230 crore (FY25, up 7% YoY). PAT: ₹1,702 crore. ROCE: 22.7%. And here’s the kicker — the company doesn’t generate fortress-level returns like Castrol or TCS, but it’s executing better than its peers on the one metric that matters in 2025: AI deal win velocity.
The company serves 51 clients in the $5 million+ bracket, with 14 clients generating $20 million+ annually. Top 10 clients grew 11.8% YoY (LTM). But before you dismiss it as another IT services story, know this: 69% of the pipeline is AI-led. The last 12 months saw $2.1 billion in TCV wins (nearly doubled from prior-year levels). And while three-month returns are -23.5%, the business itself is on a trajectory that feels less “we’re growing at industry speeds” and more “we’re capturing disproportionate share of a structural shift in enterprise tech spending.”
Concall Note (Jan 2026): “We are unlikely to see discretionary spending come back in the same shape and form. Clients are recalibrating spend toward AI fabric build-out and governance.” — Management assertion that doubles down on “this is not a cyclical recovery story; it’s a demand reallocation.”
03 — Business Model: The Transformation Box-Ticker That Actually Transforms
Applications (75%). Outsourcing (25%). AI Sauce Everywhere.
The operating model is deceptively simple. Enterprise customers in BFSI (52% of Q3 revenue), insurance (15%), TMT (18%), and logistics (5%) hire Mphasis to modernize legacy estates, automate critical processes, and increasingly — build out AI governance and generative AI stacks. The company delivers through consulting, software development, infrastructure management, and data analytics across three primary verticals: Applications Services (75% of revenue), BPO (15%), and ITO (10%).
Geographically: North America is 83% of revenue (heavily BFSI skew), with emerging momentum in EMEA (9%) and Rest of World (7%, including India-centric GCC presence). The company sits at a sweet spot — the “not too big, not too small” scale where it can move faster than TCS/Infosys (average deal cycles of 3-5 months vs. 9-12 months for mega-cap IT) but with enough delivery infrastructure to handle $100 million contracts.
Differentiation claim: NeoIP, described as an “AI intelligence platform” integrating modernization, AI-ops, quality automation, and discovery across the IT estate. The platform includes proprietary agents (NeoZeta for relearning, NeoRaina for discovery, NeoCrux for testing) that allegedly compress modernization timelines from 5–7 years to 18–24 months and reduce relearning costs by 30–40%. Management says >50% of company revenue now comes from clients using NeoIP.
BFSI ClientsTop 15 US BanksCore Vertical
Direct Business98%Revenue Mix
AI-Led Pipeline69%Structural Shift
LTM TCV Wins$2.1 BnDoubled YoY
The NeoIP Bet: Management is betting its own platform becomes “sticky” as clients expand from entry-point use cases (say, one modernization deal) into enterprise-wide orchestration. The thesis: higher switching costs, natural wallet share expansion, margin leverage from platform reuse. Whether this pans out hinges on execution, client adoption breadth, and competitive response from Accenture/Cognizant/IBM. Still a bet, not a guarantee.
💬 Do you think Mphasis’ NeoIP platform can sustain competitive differentiation, or will hyperscalers (AWS/Azure) commoditize it? Thoughts?
04 — Financials Overview
Q3 FY26: The Numbers in INR Crore Terms
Result type: Quarterly Results (Dec 2025 Quarter) | Q3 EPS: ₹23.21 | Annualised EPS (Q3×4): ₹92.84 | Full-year FY25 EPS: ₹89.55
| Metric (₹ Cr) |
Q3 FY26
Dec 2025 |
Q3 FY25
Dec 2024 |
Q2 FY26
Sep 2025 |
YoY % |
QoQ % |
| Revenue | 4,003 | 3,561 | 3,902 | +12.4% | +2.6% |
| Operating Profit | 750 | 678 | 722 | +10.6% | +3.9% |
| OPM % | 18.7% | 19.0% | 18.5% | -30 bps | +20 bps |
| PAT | 442 | 428 | 469 | +3.3% | -5.8% |
| EPS (₹) | 23.21 | 22.56 | 24.65 | +2.9% | -5.8% |
The Earnings Math & Valuation: Full-year FY25 EPS was ₹89.55. Q3 FY26 delivered ₹23.21 (annualised: ₹92.84). That’s the metric that matters — the run rate is holding and growing despite a challenging macro for IT services globally. Current P/E (23.1x) sits at a slight premium to sector median (21.0x), justified by the AI pipeline momentum. Revenue growth of +12.4% YoY is well above the 6–7% guidance from larger peers. The only haircut: PAT growth is muted (+3.3% YoY) due to wage-cost inflation and platform investment drag.
05 — Valuation: Fair Value Range
What’s This AI-First Transformation Story Actually Worth?
Join 10,000+ investors who read this every week.
