In a group discussion you are handed a contract drug manufacturer with an order book of three times revenue, 38% of revenue from one client and a 27% EBITDA margin. Argue the long case, then the strongest bear case, and land on a view.

Long, but sized as a position with one large, named risk: the order book that makes the bull case is the same client concentration that makes the bear case. Three years of signed work and a 27% margin in a business clients cannot switch quickly are real strengths. But 55% of the backlog rides on one molecule, and losing half of Client A would cut EBITDA by about 39%. The view flips on that molecule's patent life and pipeline.

Lead with the one fact a bull would put first. An order book of three times revenue means Rs 7,200 crore of work is already signed, so the next three years are about execution, not about winning customers. Contract manufacturing is sticky: a drug company that moves production must requalify the new plant with regulators, which takes a year or more and risks supply. That switching cost is why a 27% EBITDA margin can last. Add that a growing roster of more than twenty smaller clients shows the plant's quality record travels.

A food stall outside one office building does brilliant business until the office moves. The bear case is that the visibility and the concentration are one fact: Client A supplies 38% of revenue and 55% of the order book, all from one molecule. An order book is a promise to buy, not cash. If that drug loses share to a rival therapy or meets a safety problem, Client A cuts orders and the backlog shrinks faster than it was built.

Quantify it, because a group discussion rewards the person who puts a number on the risk. If Client A halves its orders, revenue falls by Rs 456 crore to Rs 1,944 crore, down 19%. With 45% of costs variable, each rupee of lost revenue takes 55 paise of EBITDA with it. EBITDA falls from Rs 648 crore to about Rs 397 crore, down 39%, and the margin from 27% to 20.4%. That is operating leverageThe way fixed costs make profit swing by a larger percentage than revenue, in both directions. working against you.

Pick a side, then hand the group the test. Long, as a position sized for a single-client risk, because the switching cost and the smaller-client roster are durable while the concentration is a known, checkable exposure. Three facts would decide it: how many years of patent life Client A's molecule has left, whether Client A has a second product moving to Vyomika, and whether the order book excluding Client A is growing. If the molecule faces generic competition within two years, the view turns to avoid.

In the room, the scoring is on the structure as much as the view. Argue the long case as if you believed it, then argue the bear case better than anyone else at the table, then show which fact reconciles them. That is what a portfolio manager listening wants from an analyst.