Vindhyashila Cement's ROCE fell from 16% to 9% in three years while volume rose from 8 to 10 million tonnes and capacity from 10 to 15. Decompose the fall and say what caused it.

Of the seven-point fall, about 6.6 points come from the new plant's capital and depreciation and only 0.4 from operations. EBITDA per tonne falling from Rs 1,400 to Rs 1,100 costs 4.4 points, extra volume gives 4.0 back, the Rs 95 crore of extra depreciation costs 1.7, and Rs 3,000 crore of extra capital costs 4.9. Utilisation fell from 80% to 67%; at full output on today's margin ROCE would be about 15.5%.

EBIT over capital, and EBIT is volume times margin per tonne less depreciation. Change one input at a time, in a stated order, and recompute ROCE after each step; the difference at each step is that input's contribution. The order matters a little, so say it: margin per tonne first, then volume, then depreciation, then capital. A family comparing this year's household savings rate with last year's does the same: first the salary change, then the rent change, then the new car loan, so nobody argues about which one did the damage.

Start at EBIT of Rs 880 crore on Rs 5,500 crore, 16.0%. Drop the margin to Rs 1,100 a tonne on the old 8 million tonnes: EBITDA Rs 880 crore, EBIT Rs 640 crore, ROCE 11.6%, a 4.4 point loss. Add the 2 million extra tonnes: EBITDA Rs 1,100 crore, EBIT Rs 860 crore, 15.6%, 4.0 points back. Add Rs 95 crore of depreciation: EBIT Rs 765 crore, 13.9%, 1.7 points lost. Finally divide by Rs 8,500 crore instead of Rs 5,500 crore: 9.0%, a 4.9 point loss from the denominator alone.

Timing, on these numbers. Utilisation fell from 80% to 67% because capacity grew 50% while sales grew 25%; the capital is sitting in a plant that is a third empty. At full 15 million tonne output on today's Rs 1,100 margin, EBITDA would be Rs 1,650 crore, EBIT Rs 1,315 crore and ROCE about 15.5%, close to the old level. Capital per tonne of capacity barely changed, Rs 550 crore to Rs 567 crore per million tonnes, so the new plant is not expensive; it is early. Whether that was wise depends on how fast regional demand fills it, and whether rivals built at the same time, which would also explain the Rs 300 fall in margin per tonne.

Three things, in order of what they control. First, the route back to 16% is volume: every extra million tonnes at Rs 1,100 adds Rs 110 crore of EBIT, about 1.3 points of ROCE, so filling the plant matters more than any cost programme. Second, the margin fall of Rs 300 a tonne needs its own split between price and cost, because a price war that followed industry over-building is a different problem from a fuel cost rise. Third, the ROCEReturn on capital employed: operating profit divided by the capital invested in the business, equity plus debt, a measure of how hard the assets work. of a business mid-expansion always looks worst in the year the plant opens; the useful comparison is ROCE on the old capital base, which has fallen only to 13.9%. Say the limit: this treats capital employed as a single number, and a real analysis would separate the new plant's capital and earnings from the old plants'.