What a component maker's chief executive has to know well

An electric car needs no multi-speed gearbox, so the transition removes the product that earns most of our revenue. It also creates demand for reduction gears, rotor shafts and motor housings, which must be quieter and more precise than today's gears because no engine noise masks them. I allocate capital between those two facts. Our lines for petrol and diesel vehicles are run for cash, with spending limited to safety, quality and programmes customers have already nominated. Electric-vehicle investment is staged: each stage needs confirmed customer nominations covering at least 60 per cent of its capacity before the board releases the money. Between April 2021 and March 2026 we spent ₹540 crore on electric-vehicle lines and ₹1,110 crore on everything else. I present both to the board as one portfolio, with the expected decline of each older programme shown beside the ramp-up of its replacement, so that directors see the transition as one plan rather than two competing budgets.

Vehicle makers expect annual price reductions from suppliers, and they award new programmes through quotations that commit a supplier for six or seven years. Both habits are reasonable from the buyer's side. From ours, each quotation is a capital commitment and each price-down is a transfer of margin that has to be earned back through productivity. Since 2022 we have measured return on capital employed by customer, not only by plant, and the board sees it every quarter. We added quarterly steel and aluminium indexation to every new contract, which cut the margin lost to raw-material lag from about 1.4 per cent of revenue to 0.4 per cent. The board has also set a ceiling: by March 2028 no single customer should exceed 30 per cent of our revenue. Our largest was at 38 per cent in 2021–22 and stood at 31 per cent in 2025–26. Concentration is not a sin in this industry, but it should be chosen, priced and reported.

Kavirath employs 7,800 permanent people and about 4,900 contract workers across five plants, and four of those plants have recognised unions with long-term wage settlements. I treat industrial relations as an operating discipline rather than a human-resources function. I meet each union's leadership at least twice a year without the plant head present, and every settlement since 2022 has linked part of the wage increase to measures workers can see and influence: first-pass yield, attendance and safety. After the dispute at Chakan I in 2023, I began moving core production roles from contract to permanent employment through a skills test that contract workers already doing those jobs could take. The change adds about ₹34 crore a year to wage costs at full run-rate. In return we have lost no production days since the 2024 settlement, and trained operators now stay. I do not believe a plant can sustain precision work when half the people at its machines are hired through somebody else.

I learnt gear manufacturing from the cutter outwards: hobbing and shaving at a tractor plant, then grinding after heat treatment and end-of-line noise testing at a German supplier. That background matters more in electric vehicles than it did before, because a reduction gear turning at more than 15,000 revolutions a minute is audible inside a quiet cabin, and the tolerances that decide its noise are measured in microns. As chief operating officer I brought our plants onto one operating system, with common measures for equipment effectiveness, scrap and customer complaints. Customer-reported defects fell from 410 parts per million in 2017–18 to 95 in 2020–21, and to 38 in 2025–26. All five plants are certified to IATF 16949, the automotive quality-management standard, but I judge them by the audits our customers' engineers carry out after a field complaint. I still ask to see the grinding-wheel dressing records before the plant's financial summary.

The Ghaisas family founded Kavirath, owns 51.4 per cent of it and is represented on the board by the executive chairman and one non-executive director. I am accountable to the board as a whole, not to the family, and the distinction is useful only if both sides practise it. The chairman and I agree each board agenda together and exchange papers before they are circulated, so that our disagreements reach directors as options rather than surprises. Related-party matters, chiefly castings bought from Ghaisas Castings Private Limited, a promoter-group company, go to the Audit Committee every year with an independent price benchmark. I have recommended against a promoter proposal twice in five years and supported the family's position many more times. I am also working with the Nomination and Remuneration Committee on a succession plan that measures the family's next generation and professional managers by the same written criteria. A promoter-led company is strongest when the family's patience and the manager's discipline both show in the minutes.

Exporting precision driveline parts from India is less about price than about qualification. A European tractor or off-highway maker typically takes eighteen to twenty-four months from first sample to series supply, and it expects stock held near its plant on long payment terms. At Kesmore Forgings I learnt to plan for that wait, and at Kavirath I set up Kavirath Europe B.V., a small engineering, sales and warehouse subsidiary near Tilburg in the Netherlands, so that customers deal with engineers in their own time zone. Exports grew from ₹610 crore in 2021–22 to ₹1,370 crore in 2025–26, sold to seventeen customers in nine countries, and the largest export customer's share fell from 60 to 34 per cent. The cost was working capital: export inventory rose from 48 to 71 days, and the subsidiary took four years, not three, to break even. I now give the board a range of qualification times for every new market, never a single date.