A client asks your view on two stocks he holds: Rudravarna Steel at 7 times earnings earned at a peak margin, and Kalpavadi Software at 35 times earnings growing 18% a year. Which would you trim, and why?

Trim Rudravarna Steel first. Its 7x is measured on a peak margin; at its ten-year average margin earnings fall 59% to Rs 21 and the true multiple is 17.0x. If the cycle turns the stock can lose about 41%. Kalpavadi's 35x is demanding but rests on a stable margin; if growth slows the damage is nearer 20%. Keep Kalpavadi, sized so a derating is survivable.

Think of a mango seller who made a fortune in a season when the crop failed everywhere else. Valuing his stall on that one season's profit makes it look like a bargain; the next normal season tells you what it really earns. Steel is the same. For a cyclicalA business whose profits rise and fall with the economic or commodity cycle rather than growing steadily. business, a low multiple on peak earnings is usually the market pricing the fall that has not happened yet. Rebuild Rudravarna's earnings at its ten-year average margin of 15%: EBITDA of Rs 60 a share, less Rs 20 of depreciation and Rs 12 of interest, leaves Rs 28 before tax and Rs 21 after.

Notice the size of the drop. The margin falls by 40%, from 25% to 15%, but earnings fall 58.8%, from Rs 51 to Rs 21. Fixed costs, here depreciation and interest, do not fall with the margin, so earnings swing harder than revenue or margin. That operating leverageThe way fixed costs make profit move by a larger percentage than sales or margin. is why the headline 7x hides a normal-earnings multiple of 17.0x.

A high multiple is a bet on future earnings, so test the growth rather than the multiple. At 18% a year Kalpavadi's earnings reach Rs 65.7 in three years, and Rs 1,400 is 21.3x that. The question for a growth stock is how far it falls if growth disappoints, and the stable 24% margin means the earnings base is real, not borrowed from a cycle. The honest limit: if growth slows to 12% and the market re-rates it to 25x, next year's earnings of Rs 44.8 are worth Rs 1,120, a fall of 20%. That is a real loss, but not a collapse.

Put the two downside cases side by side, because the client's real question is which position can hurt him more. If the steel cycle turns and the market pays a normal 10x on normal earnings, Rudravarna falls to Rs 210, a loss of 41.2%, twice Kalpavadi's downside. On Rs 45 lakh that is about Rs 18.5 lakh. Kalpavadi's slower-growth case costs 20% of Rs 35 lakh, Rs 7.0 lakh. The upside is lopsided the other way: steel at best holds its price for another good year, while Kalpavadi delivering its growth at an unchanged multiple rises 18%.

Close the pitch the way a client can use it. The view is to trim Rudravarna first, perhaps by half, and keep Kalpavadi at a size where a 20% derating is a nuisance rather than a problem. Say the limit out loud: nobody knows when the steel cycle turns, and if margins stay high for two more years the trim will look early. That is a price the client pays for not owning a peak-earnings stock at full size.