Sales to capital ratio

This is the lever amateurs skip. Growth isn't free. Next year's extra revenue has to be manufactured by this year's new capacity, stores, servers, inventory, and receivables. A forecast that raises revenue without charging for the assets that produce it is internally false. There are two ways to charge for it properly.

Route A: how do you forecast reinvestment with explicit lines?

Route A uses explicit lines. You forecast capex, D&A, and ΔWC separately, each from its own logic:

• Capex goes in as a % of revenue, anchored to history and capacity math. Split it mentally into maintenance, which replaces what wears out and is often proxied by D&A, and growth, which is new capacity. Cross-check growth capex against announced projects and unit costs. Cement capacity costs ~₹6,000 to 9,000 cr per 10 mtpa depending on brownfield/greenfield, and a big-box store $30 to 60m. Verify per sector.
• D&A goes in as a % of revenue or, better, as a rate on the asset base. It has to track capex with a lag. If capex runs far above D&A for years, the D&A line will catch up and squeeze future margins. If capex stays below D&A for years, the asset base is shrinking. Eventually revenue will shrink too, because you can't forecast growth off a liquidating asset base. That's the capex-versus-depreciation red flag you learned earlier, run in reverse.
• ΔWC goes in as a % of incremental revenue, anchored to the company's history of non-cash WC to sales. Cement runs ~5%, FMCG is often negative, capital-goods/EPC is brutal. Phase 2's cash-conversion cycle told you which is which.

Route A is the right tool when each line can be forecast on its own, as with mature companies running visible capex programs. It's also what a full 3-statement model does natively. Build 2 uses it.

Route B: what is the sales-to-capital shortcut?

Route B is Damodaran's sales-to-capital shortcut. It collapses all three lines into one question: how much revenue does ₹1 of net new capital produce?

Sales-to-capital (S2C) = Δ Revenue ÷ Reinvestment          (the marginal ratio)
⇒  Reinvestment_t = Δ Revenue_t ÷ S2C
⇒  FCFF_t = NOPAT_t − Δ Revenue_t / S2C

Now one assumption prices the growth. Anchor S2C three ways:

• Use the company's own marginal history. Sum 3 to 5 years of ΔRevenue and divide by the same years' summed reinvestment.
• Check the sector table. Damodaran publishes sales-to-capital by industry on the same data page. The illustrative shape is asset-light IT services/software ~2.5 to 4+, retail ~2 to 2.5, FMCG ~2.5 to 4, cement/steel ~0.8 to 1.3, utilities/telecom ~0.3 to 0.7. Verify it live.
• Use the identity below, which ties S2C to ROIC and makes the whole forecast auditable.

Route B is the right tool for young companies, whose capex, D&A and WC lines are noise. It suits quick models too, such as the Day-2 model in the 3-day deep dive. And it works as a cross-check on Route A. The two routes must reconcile: capex − D&A + ΔWC ≈ ΔRev/S2C. In other words, growth capex plus WC build should come to about ΔRev/S2C when maintenance capex ≈ D&A. Build 1 reconciles them on real-scale numbers.

How do you check sales to capital against ROIC?

One identity welds margins to reinvestment. You know it from Phase 2: ROIC = NOPAT/Invested capital = (NOPAT/Sales) × (Sales/Invested capital), which is after-tax margin × capital turnover. The same decomposition holds at the margin:

Incremental ROIC ≈ after-tax operating margin × sales-to-capital

A company with a 16.5% EBIT margin, 25% tax, and S2C of 1.2 is implicitly claiming incremental ROIC ≈ 0.75 × 16.5% × 1.2 ≈ 14.9%. Change any one of margin, S2C, or claimed ROIC and the other two must move. Hold on to this identity. In the growth-quality tie, the core idea of the week, it becomes an audit weapon.

Keep invested-capital books in every model. Carry a line, IC_t = IC_{t−1} + Reinvestment_t. It costs one spreadsheet row. In return you get the implied-ROIC audit, NOPAT_t ÷ IC_{t−1}, plus the reinvestment-rate series. Later in the phase you also get the economic-profit view of the same forecast. A DCF without an invested-capital line can't be audited, and that's usually why it doesn't have one.