PhysicsWallah Ltd FY26: Growth at ₹3,900 Crore, Profits Still Learning
General information and entertainment, not investment advice. The author is not a SEBI-registered adviser or research analyst. No recommendation, no promised returns. Markets carry risk including loss of capital. Figures may not be current. Consult a registered adviser before acting.
1. At a Glance
PhysicsWallah hit ₹3,900 crore in FY26 revenue—a 35% leap on ₹2,887 crore the year before. The engine is running. The problem: it’s still burning cash on the bottom line. PAT landed at -₹22.5 crore, which is better than FY25’s -₹216 crore loss, but profitability remains a promise rather than a fact.
The market rewards this story anyway—the stock trades at ₹102.73 per share, valuing the company at ₹29,378 crore. That’s a P/E that can’t be calculated (negative earnings), but it’s a multiple game nonetheless.
Offline centers are the simmering story. Management claims they’ll turn profitable in FY27. Q4 showed operating profit fell to ₹29 crore from ₹183 crore in Q3—a reversal that muddies the “all is well” narrative.
The balance sheet holds ₹1,126 crore in cash. Debt stands at ₹1,052 crore. Not a crisis, but not fortress-like either. The company bet heavily on its IPO story. Now it executes, or it becomes a cautionary tale about scale without sustainable returns.
2. Introduction
PhysicsWallah didn’t exist as a company until 2020. Before that, it was a YouTube channel called “Physics Wallah”—Alakh Pandey teaching physics for free starting in 2014. The popularity was real. By 2020, he and Prateek Maheshwari decided to build a business around it.
Fast forward to November 2025. PhysicsWallah listed on the BSE and NSE, raising ₹3,480 crore. The IPO priced at ₹272 per share. The stock now sits 62% lower, at ₹102.73, having peaked at ₹162 in the early listing euphoria.
The company operates across three formats: purely online (YouTube, app, website), offline centers (Vidyapeeth, Pathshala brands, and others), and hybrid models. It serves 13 education categories—JEE, NEET, UPSC, foundation, state boards, vernacular batches, and newer upskilling domains like data science and software development.
Paid user base crossed 4.46 million by FY26. Engagement metrics moved north: average app downloads at 60.18 million cumulative, average engagement time per student at 107 minutes in FY25 (latest disclosed). The community is sticky. The business model, however, is still finding its feet.
3. Business Model: WTF Do They Even Do?
PhysicsWallah’s core offer is affordable test preparation. A one-year JEE course costs ₹4,500. Peer players charge ₹75,000–80,000. NEET courses run ₹4,800 here against ₹63,000–67,000 elsewhere. UPSC at ₹18,000 versus ₹1,10,000 at the top five competitors.
This is not a pricing strategy born of altruism. It’s distribution leverage. The YouTube moat—13.7 million subscribers by July 2025—funnels free viewers into a paid funnel. A million subscribers cost a fraction of a TV ad buy. The conversion math works differently when your top-of-funnel is free and massive.
The revenue mix is split nearly 50-50 between online and offline, with offline inching ahead in FY26. Online grew 39% YoY; offline grew 31% YoY. Management guides for 55% online penetration in 3 years. That’s a slow rebalance, not a pivot.
Offline is the capital story. PhysicsWallah operates 198 offline centers across Vidyapeeth (the legacy brand, 79 centers) and Pathshala (47 centers) and smaller brands. Each center is a physical lease, staff, infrastructure play. Opening a center requires capex. Filling it requires acquisition and retention spend. Margins are negative until the center matures—management says 12 to 24 months.
The new-age categories—state boards, Curious Junior, foundation courses, “12 after 12 for UPSC”—are throwing homeruns on small bases. State board revenue jumped 9x in FY26. Curious Junior (small cohorts, two-way online) went 4x. These are where growth is explosive but currently small revenue bricks.
Pricing within segments has shifted from volume to layering. Base batch prices held steady. But premium tiers—Infinity, Infinity Pro, Power Batch—grew faster. ACPU (average course price per user) rose 11–12%, management claims, almost entirely from premium mix, not base pricing hikes.
AI is the B-side. Management launched Ask AI (3 million queries solved in Q4), AI Guru (100 million questions answered), and AI Grader (2 million answer sheets evaluated). These are product features aimed at stickiness and differentiation. The model is Aribhatta SLM (Small Language Model)—4 billion parameters, open-sourced, cheaper to run than frontier models. By FY27, management plans an AI Tutor that remembers past errors and coaches in a Socratic style. Monetization of these features is a roadmap item, not FY26 revenue.
4. Financials Overview
Figures are consolidated, in ₹ crore.
Metric
FY24
FY25
FY26
YoY Change (FY25→FY26)
Revenue
1,941
2,887
3,900
+35.1%
EBITDA
19
31
1,459
+4,680%
PAT
-1,040
-216
-22
+89.6%
EPS (₹)
-173.4
-0.99
-0.08
N/A
The headline: EBITDA went vertical. At ₹1,459 crore, FY26 EBITDA is 47% of revenue. Management attributes this to operating leverage—direct cost savings of 2.4%, employee cost discipline (-1% YoY), and marketing efficiency gains.
The caveat: PAT is still negative. Depreciation ate ₹437 crore. Interest cost ₹102 crore. Taxes are complex—FY26 showed an effective tax rate of 344%, a sign of prior-year loss carry-forwards and timing mismatches. The company paid ₹34 crore in tax on a ₹10 crore pre-tax profit; that’s a filing artifact, not a cash drain.
Q4 FY26 Specifics (Jan–Mar 2026, ₹ Cr)
Revenue: ₹919 crore (+51% YoY)
Operating Profit: ₹29 crore (vs. ₹184 crore in Q3 FY26)
PAT: -₹75 crore (vs. +₹101 crore in Q3)
OPM: 3% (vs. 22% in Q3)
The Q4 collapse is the puzzle. Seasonality plays a role—Q4 (Jan–Mar) is post-admission season, fewer enrollments, lower visibility. But the magnitude is jarring. Revenue was soft (-₹163 crore QoQ), and costs didn’t decline proportionally. Management cited “one-time expenses” (labor code compliance, IPO costs) for the profit swing.
Management Commentary (May 2026 Concall)
Management guided FY27 revenue growth at “more than 30%” and EBITDA improvement of “upwards of 100%” on an annual basis. Offline profitability is promised for FY27 (not just quarterly, but full-year). The CFO stated the company would have been PAT-positive in FY26 without one-time items, though the precise number was not quantified.
5. Valuation Discussion: Fair Value Range (Educational Only)
What follows is a walkthrough of how three valuation methods work, using this company’s numbers as the example — not a target, not a forecast, not advice.
Method 1 (P/E Multiple): Annualised EPS for FY26 is -₹0.08. Since this is negative, a P/E is not applicable in the traditional sense. For a peer band, comparable edtech firms trade at 18–48x forward earnings (MPS at 18.5x, Jaro Institute at 24x, Sodhani Academy at 28.6x). If we assume PhysicsWallah achieves ₹10 crore full-year PAT in FY27 (per management’s directional guidance), that implies ₹0.03 EPS on 286 million shares. Applying the peer median of 24x produces a price of ₹0.75. This reveals the arithmetic: earnings matter enormously; the market is currently pricing in high future profitability, not current performance.
Method 2 (EV/EBITDA): FY26 EBITDA was ₹1,459 crore. Peer EV/EBITDA multiples range from 12–48x (sector median ~22x). Applying 22x produces ₹32,098 crore enterprise value. Net cash is ₹74 crore (cash ₹1,126 crore minus debt ₹1,052 crore). Equity value is ~₹32,172 crore, implying ₹112 per share. This is close to the current market price.
