Tech Mahindra: the specialist who bet on telecom
If you owned the whole company
Buy all of Tech Mahindra and the bill is about ₹1,37,698 crore. Last year's profit: ₹4,806 crore. So each ₹100 you invest earns about ₹3.5 a year — a 3.5% earnings yield, or a price-to-earnings ratio of 27.5.
Stop and stare at that number, because it is the strangest fact in this whole five-chapter family. TCS — the best business of the lot — sells for 14.2 times earnings. Wipro sells for 13.2. Tech Mahindra, the smallest and least profitable of the five, sells for 27.5 — nearly double. When a market charges the most per rupee of profit for the weakest earner, it is not pricing what the company is; it is pricing what buyers hope it will become. Last year's profit did jump 37%. The market has bought the recovery story in advance. Warren's rule for such situations is simple: hope is not a line item on the balance sheet.
What does this business actually do?
The same core trade as its four cousins — rent Indian engineers to global corporations, bill in dollars — but with a twist of concentration: Tech Mahindra grew up as the technology arm of the Mahindra group serving telecom companies. Phone and network operators around the world hire it to build and run their billing systems, network software, and customer platforms. Over the years it bolted on ordinary enterprise IT work (much of it via the absorbed Satyam business), but telecom remains its calling card. The Mahindra group holds 34.97% — a committed anchor owner, though a smaller stake than the Tata or Premji holdings at its cousins.
The science underneath
The base arithmetic is the industry's standard labour arbitrage — bill a programmer-hour abroad at a multiple of the Indian salary paying for it, and pocket the spread across a couple of lakh employees. But Tech Mahindra teaches a different first principle, and Charlie insists we spell it out: the mathematics of concentration.
A generalist like TCS sells hours to banks, retailers, drugmakers, and airlines; when one industry sneezes, the others carry the quarter. A specialist sells depth in one industry — and inherits that industry's fevers. Telecom is a brutal patron: operators carry heavy debts, their prices per gigabyte fall almost every year, and their spending surges in great waves (a new "G" every decade) then dries up between waves. A vendor whose clients live feast-to-famine will eat the same way.
The company's own ledger proves the theorem. TCS's operating margin sat in a serene 26–28% band for eleven straight years. Tech Mahindra's bounced from 19% to 14% to 18% to 9% (in March 2024) and back to 16% — of every ₹100 billed that terrible year, ₹9 remained, versus TCS's ₹27. Profit fell from ₹5,630 crore in 2022 to ₹2,397 crore in 2024 — cut by more than half — then climbed back to ₹4,806 crore. Same country, same engineers, same decade: the difference is who the customer is. Diversification of customers is not a slogan; it is a shock absorber, and Tech Mahindra rides without one.
And AI? The threat to the billable hour lands here as everywhere — code that writes itself means fewer programmer-hours sold. But there's a sharper edge for a telecom specialist: network operations is exactly the sort of rule-bound, repetitive work that automation swallows first, and telecom clients — forever squeezed themselves — will demand the AI discount faster and harder than a genteel private bank ever would. The most price-desperate customers force the repricing first, and Tech Mahindra has the most price-desperate customer list of the five.
The moat test
Give a rival ₹1,37,698 crore and ten years. Can they take this castle?
The defensible part: deep, decades-old entanglement in telecom operators' billing and network systems. Ripping out the vendor who runs your billing is a surgery no operator undertakes to save a few percent — the standard, genuine switching-cost moat of this industry, and it is why even in the awful year of 2024 the company stayed profitable. The Mahindra name adds trust.
The indefensible part: outside telecom, Tech Mahindra is a mid-sized generalist competing against TCS, Infosys, HCL, Accenture and a dozen others with no edge a client can name. And within telecom, the moat guards a neighbourhood where the houses keep losing value. Look at the tollgate's own earnings: a 17.5% return on equity last year and 13% averaged over three years — against Screener's own listed "con" of low ROE — and a 23.1% return on capital employed. Real, but modest. A ten-year rival probably couldn't evict Tech Mahindra from its telecom accounts; it could absolutely deny it growth everywhere else. Narrow, and narrower than HCL's or Wipro's when the wind blows from the telecom side.
The numbers Warren would check
| What we check | What it means | Tech Mahindra |
|---|---|---|
| Sales growth, 10 yr | Shop's growth rate | 8% a year |
| Profit growth, 10 yr | Owners' earnings growth | 5% a year |
| Return on equity | Profit per ₹100 owners kept in | 17.5% last yr; 13% over 3 yrs |
| Return on capital employed | Profit per ₹100 of money used | 23.1% |
| Operating margin | Kept of each ₹100 billed | 9%–19% swing; 16% now |
| Borrowings | Owed to lenders | ₹2,186 cr — almost debt-free |
| Cash from operations | Profit arriving as cash | ₹6,172 cr vs ₹4,806 cr profit |
| Dividend payout | Profit mailed to owners | 94% (150% in 2024!) |
| Promoter holding | Mahindra stake | 34.97% |
The good: honest cash (operating cash of ₹6,172 crore beats reported profit of ₹4,806 crore), an almost clean balance sheet, no dilution — equity capital has idled around ₹440 crore for a decade — and a 3.63% dividend yield. The trailing-twelve-month profit growth of 37% is the recovery the market is paying 27.5 times earnings for.
The bad: read the profit row like a heart monitor — ₹5,630 crore, ₹4,857, ₹2,397, ₹4,253, ₹4,806. Eleven years ago profit was ₹2,659 crore; a 5% compound crawl with violent swings en route. And note 2024's dividend payout of 150%: the company paid owners half again more than it earned that year. Generous, yes — also a business smoothing its shareholders' mood through a collapse in its earnings. Ten-year sales growth of 8% is the second-slowest of the five, ahead of only Wipro's 6%.
What could go wrong
Invert. What kills Tech Mahindra — or the price paid for it?
- The turnaround stalls. At 27.5 times earnings, the recovery isn't a hope, it's a prepayment. If margins settle at 16% instead of marching back toward 19%, today's buyer overpaid even if the business does fine.
- Telecom clients squeeze harder. Their economics worsen structurally; their vendors get the bill. This is the permanent weather of the neighbourhood.
- AI-driven repricing of network and code work — hits the routine end of Tech Mahindra's book first and hardest, as argued above.
- Another margin heart-attack. The 2024 slide to 9% happened once; nothing structural prevents a repeat, and the 13% three-year ROE shows how long the bruise lasts.
- Acquisition indigestion, again. The company was assembled through deals. A 94% payout leaves little retained cushion if managements buy trouble.
What management must do to keep the castle
Narrow moat and a rich price — the longest memo of the five:
- Finish the margin repair and prove it holds through one telecom downturn. The 9%-to-16% climb is progress; the test is the next famine, not this feast.
- Reduce dependence on telecom's capital-spending mood without pretending to be a generalist: pick two adjacent verticals (the manufacturing and mobility worlds next door at group companies are the obvious ones) and go deep, not wide.
- Never again pay out 150% of profit. Pay a sane dividend from actual earnings; a dividend is a report card, not a painkiller.
- Sell AI-run network operations as a fixed-price product before clients demand it as a discount. The specialist who automates his own specialty keeps the savings; the one who waits donates them.
- No acquisitions until organic margins have held above the high-teens for eight quarters. The house needs its own foundations checked before buying the neighbour's.
- Publish, and be measured on, revenue per employee. Selling smarter hours, not more hours, is the only exit from the arbitrage treadmill.
The verdict
Moat: narrow. Tech Mahindra owns a genuine switching-cost franchise inside the world's telecom operators — sticky enough that even its worst year (margins at 9%, profit halved to ₹2,397 crore) stayed profitable — plus a clean balance sheet and cash-backed earnings. But concentration in a chronically squeezed industry makes its earnings the lumpiest of the five, its decade of 5% profit growth is the family's second-weakest, and its 13% three-year return on equity is flagged even by the screener as low. Over five to ten years the moat should keep it alive; it has never yet made it prosper like its cousins. Dinner-table version: a decent specialist recovering from a bad fall, whose stock — at 27.5 times earnings against TCS's 14.2 — is priced as if the fall never happened. We'd admire the recovery from a seat further back.
Written in the style of Buffett & Munger for the Omaha Investments book project. Educational material, not investment advice. Numbers from Screener.in and live NSE data via Angel One as of the date above.