Headline ROIC (40%) is the average return on all capital ever invested, most of it old and cheap. Investors eat the future, and the future is built by new capital. The sharper question, from M2.03: what is each incremental rupee earning?
Incremental ROIC (ROIIC) = ΔNOPAT ÷ Δinvested capital (over a multi-year window)
Reinvestment rate (RR) = Δinvested capital ÷ cumulative NOPAT over the window
Growth ≈ RR × ROIIC (the engine equation; the M3.04 forecasting law)Compute it over windows, never a single year (single-year ROIIC is noise: a capex-heavy year reads as "value destruction" while the plant is still being built).
Window 1: the golden years, FY16→FY24 (normalized NOPAT):
ΔNOPAT = 4,974 − 1,922 = 3,052
ΔIC = 12,500 − 4,800 = 7,700
ROIIC (8-yr) = 3,052 ÷ 7,700 = 39.6%Cumulative NOPAT over the window ≈ ₹25,600 cr (sum of the annual NOPATs); so RR = 7,700 ÷ 25,600 ≈ 30%. Growth check: RR × ROIIC = 0.30 × 39.6% = 11.9% vs actual NOPAT CAGR = (4,974/1,922)^(1/8) − 1 = 12.6%. The identity closes (the ~0.7-point gap is timing and margin drift; the equation is a consistency check, not a law of single years). Read: Asian Paints reinvested only ~30% of its profit, because incremental returns were ~40%, it could grow ~12% while paying out the other 70%. That is the platonic ideal of a compounder: high incremental ROIC with enough runway to deploy a chunk, and the discipline to return the rest rather than force it into low-return growth.
Advertisement
Window 2: the recent years, FY20→FY25 (the story that matters now):
ΔNOPAT = 3,777 − 2,768 = 1,009
ΔIC = 13,200 − 7,500 = 5,700
ROIIC (5-yr) = 1,009 ÷ 5,700 = 17.7%Here is the whole analysis in one contrast. Headline ROIC is still a gorgeous ~29–40%. But incremental ROIC has halved, from ~40% to ~18%. The marginal rupee, spent on new capacity as decorative growth slowed, on the sub-scale home-décor build, and defended into a margin war, is earning less than half what the legacy business earns. Eighteen percent still clears the 12% hurdle, so the company is still creating value at the margin; but the direction is the point. A business whose incremental returns are compressing toward the hurdle, right as a ₹10,000-crore competitor arrives, is a business whose future looks less like its past than the headline ROIC suggests.
This single number, not the P/E, not the brand, not the ROE, is the quantitative heart of the Asian Paints debate, and you would never see it if you stopped at headline ROIC.
The normalization that makes this honest. Because we held tax at 25.2% across the decade, none of this ROIIC is the FY20 tax cut in disguise. Had we used actual taxes, FY20 NOPAT would have jumped ~15% on the rate change alone and Window 1's ROIIC would read artificially high. Always normalize the tax rate before you compute incremental returns across the 2019–20 break. This is the most common way a well-meaning Indian ten-year build overstates a compounder.
The refinements (from M2.03, applied): (1) capital spent this year earns next year, so for fast growers a lagged ROIIC (ΔNOPAT led one year vs ΔIC) is fairer; on this data it lifts the recent number a couple of points but does not change the story. (2) State the answer as a range: "recent incremental returns in the high-teens, down from ~40% in the prior decade", which is precision enough for the only question that matters. In your workbook, this two-window table is a mandatory row for every company you ever study; it is the compounder test, and it is where a rebuild by hand beats a screener every time (the screener will show you the flattering headline ROIC and never compute the ΔNOPAT/ΔIC that tells the truth).