You pitched Kelvora Chemicals as a long. The portfolio manager grills you on Chinese prices, the new plant's utilisation and one customer at 40% of sales. Defend or concede each point with numbers, and say what would change your view.

Concede the China point in full, concede the plant to its contracted 60%, and defend the customer point for FY27 while admitting it belongs in the multiple. That takes FY27 EPS from Rs 60.00 to Rs 49.26, so the stock is on 24.4x rather than 20.0x. The long survives at a smaller size. The customer's renewal terms before FY30 are what would end it.

Think of a friend picking holes in a flat you want to rent: the lift is slow, the kitchen is small, the landlord owns the whole building. You do not argue each one with equal heat. You ask which ones cost you money, then give ground on those and hold on the rest. A grilled pitch works the same way: convert each objection into rupees of EPS first, because the size tells you which fight is worth having. Interviewers are not testing whether you can win every point. They are testing whether you know which points matter.

The three points hit EPS through different routes. A price cut drops straight to profit because no cost falls with it. Lost volume costs only the contribution marginRevenue less the costs that rise and fall with volume, such as raw materials and power. It is what each extra unit adds to profit., here 35%. A customer cut is also a volume loss, but a contract limits how far it can go in FY27.

Chinese prices: 30% of Rs 2,560 crore is Rs 768 crore of exposed revenue, and 15% off that is Rs 115.2 crore of lost profit before tax, Rs 8.64 a share after tax. Customs data shows the lower prices landing, so the point is big and true. Concede it and cut the estimate; arguing that customers are loyal to Kelvora's grades is the answer a stubborn analyst gives.

Plant utilisation: each 10 points of the new plant is Rs 80 crore of revenue, Rs 28 crore of contribution and Rs 2.10 of EPS. The bear's 50% costs Rs 4.20. But 60% is already under signed contracts, so the honest concession is 70% down to 60%, Rs 2.10. Defend with the contract, and give up only the part your evidence cannot cover.

The customer: if it cut to its 80% floor, Kelvora would lose 20% of Rs 1,024 crore, Rs 204.8 crore of revenue and Rs 5.38 of EPS. It has bought above contract every year, so for FY27 you defend. The risk is real but it lives at the FY30 renewal, which is a valuation question, not an FY27 earnings question.

At 24.4x the stock is no longer obviously cheap on FY27. The long now rests on FY28, when the plant should run near 85%: that adds about Rs 5.25 of EPS to the conceded base, about Rs 54.51, or 22.0x. So the thesis narrows from cheap on next year to cheap on the year after, if the plant fills. Say that plainly. A pitch that survives a grilling smaller but intact is more credible than one that claims nothing moved.

Name one development, not a list. For Kelvora it is the customer renewal: if the largest customer signals a second supplier or a lower floor before FY30, the 40% of sales that holds the business together becomes the risk, and no FY28 multiple survives that. A second leg down in Chinese prices would cut the estimate again but would not end the thesis. The renewal would.