How would you value Doorsanchar Networks, a telco at 7x EBITDA with capex equal to half of EBITDA, against Pankhudi Software, growing 35% at 8x revenue? Which methods fit each, and why is EV/EBITDA misleading for the telco?

Value the telco on cash after capex and the software company on revenue or gross profit, and never compare their EBITDA multiples. Doorsanchar's 7.0x becomes 14x on EBITDA less capex, and its equity yields about 4.9% of free cash flow after interest and tax. Pankhudi's 8.0x revenue is 10.7x gross profit and about 80x EBITDA, paying for 35% growth. A DCF works for both but needs different horizons: steady cash for the telco, a path to a mature margin for the software company.

Because the network does not maintain itself. Doorsanchar must spend Rs 2,600 crore a year, half its EBITDA, just to keep customers on a working network, so EBITDA overstates by two what the business actually throws off. An autorickshaw driver earning Rs 2,000 a day is not earning Rs 2,000 if the vehicle needs Rs 1,000 a day of fuel and repairs to keep running; the fare is EBITDA, the fuel is capex. On the honest number, EBITDA less maintenance capexThe capital spending needed to keep revenue and capacity where they are, as opposed to spending that adds growth., the telco's EV of Rs 36,400 crore is 14x Rs 2,600 crore. That is not cheap.

A cash yield on a levered asset. Net debt of Rs 15,600 crore is 3.0x EBITDA and 43% of the enterprise value, so equity is Rs 20,800 crore. Interest at 8% is Rs 1,248 crore; with depreciation equal to capex, taxable profit is Rs 1,352 crore and tax Rs 338 crore. Free cash flow to equity is Rs 2,600 crore less interest and tax, about Rs 1,014 crore, a yield of 4.9% on the equity value. That yield, its growth and its stability are the valuation: a DCF of steady cash flows, a dividend yield against peers, and EV to EBITDA less capex as the cross-check. Spectrum renewals, which arrive in lumps, are the number to ask about next.

Because its EBITDA margin is a choice, not a result. Pankhudi earns a 75% gross margin and then spends most of it on sales and engineering to grow 35%; its 10% EBITDA margin could be 30% tomorrow if it stopped growing. So investors value what is durable, revenue and gross profit, and judge the price by growth: 8.0x trailing revenue is 5.9x next year's revenue, and growth plus margin, the Rule of 40A rough screen for software companies: revenue growth rate plus profit margin, in percentage points, should sum to 40 or more., is 45. A DCF still applies, but its heart is the year the margin matures: say what mature margin you assume, because that is the whole valuation.

The answer becomes absurd, which is the test. Pankhudi at the telco's 7.0x EBITDA would be worth Rs 1,050 crore, 0.7x revenue, a price at which a rival would buy it for its customer list alone. Doorsanchar at 8.0x revenue would be worth Rs 104,000 crore, nearly three times its current value for a business that cannot grow faster than the population. Say the limit too: multiples are shorthand for a DCF, and the two companies need different DCFs. The telco's risk is in the terminal value and the next spectrum auction; the software company's is in whether 35% growth persists and what margin it lands on. Comparable transactions help the telco, where towers and operators change hands at known multiples, more than the software company, where every deal is a growth story.