# Dcf India > Build a discounted cash flow valuation calibrated to Indian markets — G-Sec risk-free rate, India equity risk premium, Indian corporate tax regimes, rupee inflation-consistent terminal growth, and the reality checks that stop a DCF becoming a spreadsheet with a target price at the end. Use when valuing an Indian listed company, building or reviewing a DCF or FCFF/FCFE model, deriving WACC or cost of equity for an Indian company, or when the user asks what a company is worth or what a fair value is. - Skill: `sharma23yash-oss/dcf-india` (Agent Skill) - Install (CLI): `npx skillmds@latest add sharma23yash-oss/dcf-india` - Raw SKILL.md: https://api.skillmd.com/api/skills/sharma23yash-oss/dcf-india/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Marketing & Growth - License: MIT - Author: sharma23yash-oss (https://skillmd.com/u/sharma23yash-oss) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/sharma23yash-oss/dcf-india --- # DCF for Indian Companies A DCF is an argument about the future written in arithmetic. This skill sets the India-specific inputs and imposes the discipline that makes the argument honest. ## Choose the right cash flow **FCFF** (firm) discounted at WACC → enterprise value → subtract net debt → equity value. Default choice for a non-financial company. ``` EBIT × (1 − tax rate) + Depreciation & amortisation − Capex − Change in non-cash working capital = FCFF ``` **FCFE** (equity) discounted at cost of equity → equity value directly. Use for banks and NBFCs, where debt is raw material rather than financing — but for lenders prefer an excess-return or residual-income model; see `bank-nbfc-analysis`. Never mix: FCFF discounted at cost of equity, or FCFE at WACC, is the single most common modelling error. ## Cost of equity — India inputs ``` Ke = Rf + β × ERP ``` **Risk-free rate (Rf):** the 10-year Government of India benchmark G-Sec yield. Use the current yield, not a historical average, and state the date. As a sanity band it has traded broadly in the mid-6% to low-7% range through 2026. Sources: RBI, CCIL, or the FBIL benchmark. Do not use the US Treasury yield unless you are building the whole model in USD, in which case you must also convert the cash flows and use a USD-consistent inflation differential in terminal growth. **Equity risk premium (ERP) for India:** approximately **7.0%**, with a defensible range of 6.5–7.5%. This is a *mature-market premium plus a country risk premium*, and independent Indian valuation practice converges near 7% (Incwert's multi-method assessment, January 2026, recommends 7.0%; Damodaran's country premium tables give a comparable figure derived from India's sovereign rating and relative equity volatility). State which ERP you used and why. An analyst who cannot defend their ERP cannot defend their target price. **Beta:** avoid a raw regression beta from a single index over a single window — Indian small and mid-cap betas are unstable and liquidity-distorted. Prefer: 1. A **bottom-up beta**: unlever peer betas, average, relever at the subject company's target capital structure. ``` βunlevered = βlevered / (1 + (1 − t) × D/E) βrelevered = βunlevered × (1 + (1 − t) × D/E target) ``` 2. Or an adjusted beta: `0.67 × raw + 0.33 × 1.0`. For most Indian large caps a bottom-up beta lands between 0.7 and 1.3. A beta outside 0.5–1.8 needs justification. **Small-size and illiquidity premium:** for companies below roughly ₹5,000 crore market cap with thin trading, an additional 1–3% is common Indian practice. Disclose it as a separate line; never bury it in the ERP. ## Cost of debt and WACC ``` Kd = current marginal borrowing cost (not the historical average implied rate) WACC = We × Ke + Wd × Kd × (1 − t) ``` Get Kd from: the company's disclosed borrowing cost in the notes, its latest credit rating and the corresponding corporate bond spread over G-Sec, or a recent NCD issuance. Weights should be **market value** of equity and book (or market, if traded) value of debt, at a target structure — not the current snapshot if the company is mid-deleveraging. ## Tax rate Use the **effective statutory rate the company actually pays**, and reconcile it to the disclosed tax note. | Regime | Base | Effective incl. surcharge & cess | |---|---|---| | Concessional domestic (22%) | 22% | **≈25.17%** | | New manufacturing regime (15%), where the company qualifies | 15% | **≈17.16–17.47%** | | Default corporate rate | 30% | **≈34.94%** | | MAT (where applicable) | 15% of book profit | plus surcharge & cess | Most large Indian companies have migrated to the 22% concessional regime, which is irrevocable once elected. If a company's effective tax rate is far below statutory, find out why — SEZ benefit, brought-forward losses, tax holiday — and model the **step-up when the benefit expires**. A DCF that assumes a 15% tax rate in perpetuity because of a tax holiday expiring in three years is wrong by a large margin. The Income-tax Act 2025 applies from 1 April 2026 and renumbers sections without, in the government's own description, changing the substantive burden. Verify rates for the year you are modelling. ## Forecast horizon - **Explicit forecast: 5–10 years.** Use 10 only when you can defend the year-8 revenue number. - Forecast **revenue growth, EBITDA margin, capex/sales, and working capital days** — not FCFF directly. If you forecast the answer you have not built a model. - **Fade to steady state.** A company earning 25% ROCE cannot do so forever; competition compresses it. Model the fade explicitly rather than assuming it away. ## Terminal value — where the India errors happen ``` TV = FCFF(n+1) / (WACC − g) ``` **Terminal growth (g) must be nominal-rupee consistent.** India's nominal GDP growth has run in the high single digits (roughly 4–6% real plus 4–5% inflation). Therefore: ``` g ≤ long-run nominal GDP growth ≈ 5–6.5% ``` Two absolute rules: - **g < WACC.** Always. If not, the model is telling you the company outgrows the economy forever. - **A rupee DCF cannot use a USD terminal growth rate.** A 2–3% g in a rupee model implies the company shrinks in real terms relative to the Indian economy. This is the most frequent error in Indian DCFs built from Western templates. Sanity: terminal value should typically be **60–80% of enterprise value** on a 10-year forecast. Above 85%, the model is a terminal-value assumption wearing a forecast as a costume — say so. Cross-check the terminal value with an **exit multiple**: implied terminal EV/EBITDA should be plausible against where the sector actually trades. If your perpetuity growth implies a 35× terminal EBITDA multiple for a cement company, the growth rate is wrong. ## Bridge to equity value ``` Enterprise value − Gross debt (including lease liabilities under Ind AS 116) + Cash and cash equivalents (only genuinely surplus cash) + Value of non-consolidated investments / associates (at market or fair value) − Minority interest (at fair value, not book) − Contingent liabilities judged probable and quantifiable − Unfunded pension / gratuity obligations = Equity value ÷ Diluted share count (include ESOPs in the money, warrants, convertibles) = Value per share ``` India-specific bridge items to check every time: - **Ind AS 116 lease liabilities** are debt. Retail, aviation, hotels and QSR companies carry large ones. If you count the lease liability as debt, EBITDA must be post-Ind AS 116 — do not mix pre- and post-116 conventions. - **Cross-holdings in group companies** are common in Indian conglomerates. Value separately, then apply a holding-company discount (see `relative-valuation-india`). - **Not all cash is surplus.** Deduct the cash the business needs to operate, and any cash that is restricted or lying in subsidiaries where repatriation carries a tax cost. ## Mandatory sensitivity Never present a single number. Produce a matrix: ``` WACC 10% 11% 12% 13% g 4.5% 5.0% 5.5% 6.0% ``` And a scenario table (bear / base / bull) driven by **revenue growth and terminal margin**, not by the discount rate. The discount rate sensitivity shows model behaviour; the operating scenarios show the actual argument. ## Reality checks before you publish 1. **Does the implied multiple make sense?** Back out implied forward P/E, EV/EBITDA and P/B at your fair value. Compare to the company's own 5-year and 10-year trading range and to peers. 2. **Reverse DCF.** Solve for the growth and margin the *current* price implies. Then ask which is more plausible — the market's assumptions or yours. This is often more informative than the forward DCF. 3. **Does cumulative capex support the revenue?** Check implied asset turns in the terminal year against history. 4. **Does reinvestment support growth?** `g = reinvestment rate × ROIC`. If your terminal g is 6% and terminal ROIC is 12%, you need a 50% reinvestment rate. Is that in the model? 5. **Sum the forecast FCFF.** If the company has never generated positive FCF and your model has it doing so from year 2, explain the mechanism. ## Output format ```markdown # — DCF, ## Inputs | Input | Value | Source | | Rf (10Y G-Sec) | x.xx% | | | ERP (India) | 7.0% | | | Beta (bottom-up) | x.xx | | | Ke | xx.x% | | | Kd (pre-tax) | x.x% | | | Tax rate | 25.17% | | | WACC | xx.x% | | | Terminal g | x.x% | | ## Forecast | ₹ cr | FY26E | FY27E | ... | Terminal | | Revenue | | | | | | EBITDA | | | | | | EBIT | | | | | | NOPAT | | | | | | + D&A / − Capex / − ΔWC | | | | | | FCFF | | | | | ## Valuation - PV of explicit FCFF: ₹X cr - PV of terminal value: ₹Y cr (Z% of EV) - EV: ₹ cr → Equity: ₹ cr → **₹ per share** - vs CMP ₹ : +/− % ## Sensitivity ## Reverse DCF The current price implies . Assessment: ... ## Key judgements 1. ... ``` ## Hard rules - **Every input carries a source and a date.** An unsourced ERP or Rf makes the whole model unauditable. - **State g < WACC explicitly** and show terminal value as a % of EV. - **Never present a point estimate as a price target without a range.** The output of a DCF is a distribution. - Do not tune assumptions until the answer matches the market price, then present it as independent work. If you had to, say so. - This is a valuation exercise, not investment advice. Say so in the output, and note that the user should form their own view or consult a SEBI-registered adviser.