A price-down letter deserves a board paper

Every year, usually between January and March, an Indian auto-component maker receives letters from its vehicle-maker customers asking for price reductions on existing parts. The requests are framed as productivity sharing, and they typically ask for a few per cent across the board. At about the same time, the same customers invite quotations for new programmes that will run for six or seven years, often at prices that assume further annual reductions. Both processes are handled by sales teams and approved by management. In most companies I know, neither reaches the board except as an explanation for a margin.

I think that is a mistake, and I made it myself for years. A price-down is a transfer of margin from supplier to customer that must be recovered through productivity or not at all. A quotation for a seven-year programme is a capital commitment: it fills a line, often needs new tooling or machines, and locks the supplier into a price path before its costs are known. These are capital-allocation decisions. They belong in the same conversation as a new plant.

Thirty years of loyalty, priced in basis points

My father ran a small machine shop in Kolhapur. Most of its work came from one foundry, and when that foundry brought its machining in-house in 1989, his business lost two-thirds of its orders in six months. I tell the story because it is the simplest version of a risk every Indian component maker carries. Our industry grew by following a handful of vehicle makers, and many suppliers still earn more than a third of their revenue from one customer. That concentration was rational when the customer was growing quickly and the supplier was small. It becomes dangerous once the supplier has grown up, because the customer's purchasing team can now ask for almost anything, knowing that walking away would cost the supplier more than accepting.

At Kavirath, our largest customer accounted for 38 per cent of revenue in 2021–22. The relationship was thirty years old, built by our founder, and it had served both sides well. In late 2022 the customer asked for a 4 per cent reduction across existing parts and invited us to quote for gear sets for a new sport utility vehicle's six-speed manual gearbox, a programme worth about ₹1,100 crore over seven years. The price it wanted was about 9 per cent below our fully costed price.

Our sales team recommended accepting both. Their reasoning was the industry's reasoning: the volume would fill our Chakan II plant, productivity would close the gap, and refusing a customer of thirty years would damage everything else we sold to it. Our executive chairman, the founder's son, felt the weight of the relationship more than anyone. When we costed the programme properly, including the capital it needed and the reductions that would follow, it would have earned about 7 per cent on capital over its life, below our cost of capital. It would also have taken that customer above 40 per cent of our revenue.

Return on capital by customer, not by plant

The analysis that changed the decision was not complicated, but we had never done it. Indian component makers usually measure profitability by plant or by product line, because that is how their costs are recorded. Customers cut across both. A customer's parts are made in several plants, on lines shared with other customers, using capital approved at different times for different programmes. To see what a customer earns, you have to allocate capital employed, including inventory and receivables, and compare the return with what the same capital would earn elsewhere.

When we did that for our ten largest customers, the results were uncomfortable. Our largest customer earned a return on capital about four percentage points below the company average, mainly because of accumulated price-downs and the absence of raw-material indexation in its contracts. Two smaller customers, both tractor makers, earned well above average. Nobody had been hiding anything. We had simply never put the numbers in one place.

We declined to quote at the customer's price and offered a value-engineered design at 5 per cent below our own price instead. The customer awarded the programme to another supplier. Chakan II ran at 64 per cent utilisation for five quarters, and we carried about ₹26 crore of fixed cost that the volume would have absorbed. Relations with the customer's purchasing team were cool for a year, and I handled the communication badly: its senior leadership learnt of our decision from their buyers rather than from me. We did agree a 2.5 per cent reduction on existing parts in exchange for quarterly steel indexation, which the contracts had never included.

Why we set a ceiling on any one customer

After that decision the board agreed a ceiling: by March 2028, no customer should account for more than 30 per cent of Kavirath's revenue. It also asked to see return on capital by customer every quarter, beside the usual plant figures. The ceiling is not a rule against growing with a good customer. It is a statement that concentration above that level must be chosen deliberately, with its return shown, rather than accumulated quarter by quarter.

The results have been encouraging. By 2025–26 our largest customer's share had fallen to 31 per cent, and our top two customers' combined share from 58 to 47 per cent. Chakan II now makes tractor drivelines for export and electric reduction gears. EBITDA margin rose from 11.6 per cent in 2022–23 to 13.3 per cent in 2025–26, and indexation reduced the margin lost to raw-material lag from about 1.4 per cent of revenue to 0.4 per cent. The customer we declined remains our largest, and we have since won two smaller programmes from it at prices that earn their capital.

The opposing case deserves respect. A supplier that declines programmes may lose its place in the customer's next generation of vehicles, and some suppliers who accepted thin margins in the past were rewarded with long and profitable relationships. Capacity left idle has a real cost. And a board can turn a ceiling into a reason for timidity, refusing good business because of a percentage.

My answer is that the board should see the choice, not avoid it. For a component maker's directors, the practical questions are few. What share of revenue does each of our five largest customers represent, and what return on capital does each earn? How much margin did we give away in price-downs last year, and how much productivity recovered it? Which contracts lack raw-material indexation? A management team that can answer those questions is allocating capital. One that cannot is letting its customers do it.