Redington DCF — equation sheet

Public edition of the model's equation sheet, 26 September 2026; the equations are unchanged from the 10 September 2026 build. Each equation names the data rows it consumes (A = the Research firm B initiation tables — a sell-side initiation dated 10 September 2026, named only by this label here — I = Research firm A's note of 9 September 2026, T = company earnings-call transcripts, B = reconciliation and boundary identities from the model build, S = structured data points captured from those sources). Units in the model data: INR millions (₹ crore figures in prose are presented at a scale of ten to the crore); rates as fractions; days as days. Labels E1 onward are cited by the model code. Research-firm names are withheld on this site; the labels are consistent across the article, the model data and the code.

0. What the source data supports, in one paragraph

Two statement years, FY25 and FY26, at group level (A1–A3), one FY24 headline row (I1–I3), and five quarters of headline lines (T1–T3). Growth is disclosed by business line: segment revenues for FY25 and FY26 with Research firm B's path to FY40 (A5), quarterly segment growth and shares from management (T7, S revenue-growth-yoy, revenue-share), SSG's banked and renewal mechanics (T8, A18). Economics are disclosed by geography, not by business line: gross margin, EBITDA margin and working-capital days for SISA and RoW (A6), with the FY26 segment-by-geography mix as the bridge (A8, A9). The geography margins do not reproduce the group (B6), so they enter as differentials while the level is the statement. Ten years of margin history and eleven of days history (A13, A15) anchor the extrapolations. Arena has its own revenue, PAT, debt and finance-cost history (A21, T9). Research firm B's DCF is fully stated (A25) and its bridge convention is identified (B9), so a calibration scenario can be run from Research firm B's own inputs. Nothing in the source data gives unit volumes, prices, headcount or the breakdown of other expenses.

1. Axis, units, conventions

• Fiscal years April to March. FY25 and FY26 are actual columns computed from inputs; FY27 to FY29 are the explicit forecast; FY30 to FY40 fade; the terminal value follows the horizon H (FY40 by default, matching A25; FY36 as an option).
• Consolidated throughout. Arena (49.4% owned, A21) is inside every consolidated line; its PAT, debt and finance cost are carried as a block so that the minority share and the exit case can be computed (E17). PAT is consolidated; the attributable figure and EPS follow the minority (B2).
• Revenue is revenue from operations; total income is revenue plus other income (B1). Working-capital days are on their own bases: receivables on revenue, inventories and payables on COGS (B4). Debt is borrowings excluding leases unless the bridge option says otherwise (B5). Finance cost is the P&L line; factoring sits in opex (B10).
• Two cuts run side by side: growth by segment (E1–E2), economics by geography through the FY26 mix (E3–E5). The bridge weights are inputs from A8, never solved.

2. Identities that close at the base years

Each is computed in the model for FY25 and FY26 and printed as a tie-out line (E30). From the source data: EBITDA = gross profit − employee − other expenses (B3, ±1 mn); EBIT = EBITDA − depreciation; PBT = EBIT + other income − finance costs − exceptional (FY25 exceptional is a gain, sign corrected, B2); PAT = PBT − tax; attributable PAT = PAT − minority share of Arena's result, which reproduces Research firm B's EPS and Research firm A's adjusted profit (B2); NWC = receivables + inventories − payables at 35.8 days of revenue FY26 (B4); gross RoCE 16.3%, net RoCE 19.7%, post-tax RoCE 12.3% (B7); equity roll-forward leaves 4,742 mn of OCI in FY26 (B12); Research firm B's geography margins miss the group by 12 to 17 bps (B6).

3. Equations

Revenue, by segment (the growth cut)

E1 R_{s,t} = R_{s,t-1} (1 + g_{s,t}) s ∈ {MSG, TSG, SSG, ESG, OTH}; FY25 and FY26 from A5; Others is the residual of A5 (10bn FY26) FY27–FY29: g_{s,t} from the assumption arrays; FY30–FY40: g_{s,t} = g_{s,FY29} + (g_LR − g_{s,FY29}) · (t − FY29) / (FY40 − FY29) E1a ESG: g_ESG,t = (1 + u_t)(1 + p_t) − 1 u = unit growth, p = price change (S product-price-change, T12, A20) E1b SSG: R_SSG,t = max( b · R_SSG,t-1 , R_SSG,t-1 (1 + g_SSG,t) ) b = banked share 63.5% (T8) E1c TSG: g_TSG,t from the array; LD_t = large-deal share of group revenue (A12: 2.1% FY26) feeds E4 E1d OTH: g_OTH,t = growth of the other four segments E2 R_t = Σ_s R_{s,t}

Geography (the economics cut) and the bridge

E3 R^SISA_t = Σ_s w_{s,t} R_{s,t}; R^RoW_t = R_t − R^SISA_t w_{s,FY26} from A8: ESG 0.54, TSG 0.65, MSG 0.55, SSG 0.44, OTH 1 (assumption); FY26 check ≈ 0.55 of revenue (A10) w_MSG,t: 0.55 → 0.72 by FY29 (A9; +6 points a year from FY23–FY26), then held; the others held at FY26 E3a s^SISA_t = R^SISA_t / R_t; s^SSG_t = R_SSG,t / R_t

Gross margin and gross profit

E4 m_t = m_FY26 − Δ_GM (s^SISA_t − s^SISA_FY26) + κ_SSG (s^SSG_t − s^SSG_FY26) − δ_LD (LD_t − LD_FY26) m_FY26 = 4.915% (A1, held); Δ_GM = m^RoW − m^SISA = 1.3 points (A6, FY26; 1.2 in FY25); κ_SSG = SSG margin premium over the group, 0.85 points (A18, T8); δ_LD = large-deal dilution (T4 ordering, A12), assumption The geography margins enter as the differential Δ_GM only, because their levels do not reproduce the group (B6) E5 GP_t = m_t R_t; COGS_t = R_t − GP_t diagnostic split: m^SISA_t = m_t − (1 − s^SISA_t) Δ_GM; m^RoW_t = m_t + s^SISA_t Δ_GM

Operating expenses and EBITDA

E6 OPEX_t = F_t + v R_t + X_t; F_t = F_FY26 (1 + π)^(t − FY26) v = (OPEX_FY26 − OPEX_FY25) / (R_FY26 − R_FY25) = 2.20%; F_FY26 = OPEX_FY26 − v R_FY26 = 10,043 (B11: the two statement years fix both) π = inflation on the fixed base (assumption); X_t = expensed programme spend (T8 capability spend), assumption, default 0 E7 EBITDA_t = GP_t − OPEX_t checks: FY27 against management's 2.2–2.4% ex-Arena (T3) and Research firm B's 2.03% (A6); the quarterly opex elasticity 0.3–0.55 (T3) against v/(OPEX/R) = 0.72

Below EBITDA

E8 D_t = d R_t d = D_FY26 / R_FY26 = 0.173% (A1, held; FY25 0.219%) E9 EBIT_t = EBITDA_t − D_t E10 OI_t = o R_t o = OI_FY26 / R_FY26 = 0.155% (A1, held; Research firm B takes it to 0.05% by FY28E) E11 INT_t = r_d · ½ (Dc_{t-1} + Dc_t) + INT^A_t Dc = core debt = borrowings − Arena debt; r_d = (INT_FY26 − INT^A_FY26) / average core debt FY25–FY26, the FY26 implied core rate (A1, A21, T6), overridable; INT^A from E17; the solver method (average or opening) is an assumption E12 EXC_t = 0, except the Arena exit write-off (E17) E13 PBT_t = EBIT_t + OI_t − INT_t − EXC_t E14 TAX_t = τ PBT_t τ = TAX_FY26 / PBT_FY26 = 24.3% (A1, held); FY25 on the corrected PBT is 22.0% E15 PAT_t = PBT_t − TAX_t; MI_t = (1 − 0.494) PAT^A_t; PAT^attr_t = PAT_t − MI_t (B2) E16 EPS_t = PAT^attr_t / N N = 782 (A25)

Arena block

E17 R^A_t: FY26 68bn (A21); FY27 = R^A_FY26 (1 + g^A_FY27) with g^A_FY27 = −50% (S revenue "next year" US$500mn, T9), then grows with the group PAT^A_t: FY26 −4,100 (A21); continuation → 0 by FY28 (A21, T9); remediation → 0 by FY30; exit at t_x: 0 after t_x Debt^A_t: US$130mn FY26 (S debt, Q2 FY26) → US$90mn FY27 (A21) held; remediation → 0 by FY30; exit → 0 INT^A_t = r_A Debt^A_t r_A = INT^A_FY26 / average Debt^A (T6: about ₹180 crore of interest after factoring; A21) exit: R^A leaves R^RoW, MI_t = 0 after t_x, proceeds = 0.494 · V^A − W^A (V^A = US$45mn market value, A21; W^A = write-off, assumption) into cash at t_x

Working capital, cash and financing

E18 REC_t = R_t · DSO_t / 365; INV_t = COGS_t · DIO_t / 365; PAY_t = COGS_t · DPO_t / 365 NWC_t = (REC_t + INV_t − PAY_t) · (1 + λ (s^SISA_t − s^SISA_FY26)) DSO, DIO, DPO held at FY26 (B4: 66.1, 26.3, 58.1; FY25 63.7, 24.4, 51.9; A15 for the eleven-year range) λ = −10 / 34.5 per unit of SISA share (A6: SISA 30 days against RoW 40 on Research firm B's basis), the mix effect on days E19 ΔNWC_t = NWC_t − NWC_{t-1} E20 CFO_t = EBITDA_t − TAX_t − ΔNWC_t pre-interest E21 CAPEX_t = c R_t c = CAPEX_FY26 / R_FY26 = 0.153% (A3, held; FY25 0.164%) E22 DIV_t = payout · PAT^attr_{t-1} payout 30% (T10, held; 40% stated at 1QFY25 as the alternative) E23 Cash_t = Cash_{t-1} + CFO_t − INT_t (1 − τ) − CAPEX_t − DIV_t + ΔDebt_t + proceeds_t ΔDebt_t = max( φ R_t − (Cash_{t-1} + CFO_t − INT_t (1 − τ) − CAPEX_t − DIV_t + proceeds_t), −Debt_{t-1} ) φ = Cash_FY26 / R_FY26 = 0.94% (A2, held): borrow to the cash floor, repay surplus E24 FCFF_t = EBITDA_t − τ EBIT_t − ΔNWC_t − CAPEX_t other income excluded as interest income (A3)

Valuation, reverse questions, returns

E25 EV_0 = Σ_{t = FY27..H} FCFF_t / (1 + k)^{n_t} + TV_H / (1 + k)^{n_H}; n_t = t − FY26, less ½ under mid-year timing TV_H = FCFF_H (1 + g) / (k − g), or under the value-driver option NOPAT_H (1 + g)(1 − g / RoCE_T) / (k − g) with RoCE_T an assumption k = 13% (A25, Research firm B's cost of equity applied to firm cash flows, as its own DCF does); g = 6% (A25); H = FY40 (A25) E26 Equity_0 = EV_0 − NetDebt_0 − MI_0; value per share = Equity_0 / N NetDebt_0 options (B5, B9): borrowings − cash 15,160; with leases 17,202; Research firm B's printed 45,982 MI_0 options: book 2,633 (A2); market 0.506 · US$45mn · fx (A21) E27 Reverse: g such that value = price (A25 CMP 378; T price rows); a growth scaler on FY27–FY29; a working-capital days level E28 Returns (B7): CE_t = E_t + Debt_t; gross RoCE_t = EBIT_t / avg CE; net RoCE_t = (PBT_t + INT_t − EXC_t) / (avg CE − avg Cash); post-tax RoCE_t = EBIT_t (1 − τ) / avg CE; RoWC_t = EBITDA_t / NWC_t; RoE_t = PAT^attr_t / avg (E_t − MI^book_t) E_t = E_{t-1} + PAT_t − DIV_t (OCI ignored; B12)

Calibration scenario and tie-outs

E29 scenario = 'firm_b' replaces the desk drivers with Research firm B's stated ones (A5, A6, A1, A3, A25): segment growth FY27–FY29 from page 33 and the FY30–FY40 fade by segment from its revenue points (year growth to FY32, then the FY32–35 and FY35–40 CAGRs; group FY40 7,617bn); margin mode 'ebitda_margin_path' with 2.03 / 2.15 / 2.18 then 2.2 / 2.3 / 2.3 / 2.4 (FY35) / 2.5% (FY40) interpolated; D&A 0.16%; other income 0.06%; tax 25%; NWC days 35 / 36 / 37 as closing NWC on revenue (its FY27E balance sheet gives 34.8), unscaled; capex 0.10%; payout 40% of same-year PAT; k 13%, g 6%, H FY40; net debt = Research firm B's printed 45,982; MI book The test: value under this scenario against ₹470 (A25). The residual is reported, not tuned away. E30 Tie-outs, FY25 and FY26 (printed as diagnostics): EBITDA identity residual; PBT residual (FY25 with the sign corrected); PAT − minority − EPS × N; NWC from components against the statement; equity roll-forward residual (OCI); cash-flow working-capital change against ΔNWC (FY26: −18,084 against −14,757).

Regime switches to expose as scenarios

1. Endpoint normalisation: p_t negative after the FY27 ASP year, u_t flat (A20, T12).
2. Large-deal share LD_t and its dilution δ_LD (A12, T4).
3. SSG growth against the banked floor (T8).
4. Arena branch: continuation, remediation, exit (A21, T9).
5. Working-capital days 28–40 and the cash floor (T4, A15).
6. The geography mix path w_MSG,t (A9).

4. Extrapolations

Values carried forward from the company's own history by a stated rule. Each is an assumptions.json entry with basis extrapolation.

5. Assumptions

Values no source supports, each with its chooser and range in assumptions.json (basis assumption), or a claimant's statement adopted as the default (basis guidance or estimate).

6. Gaps, and which Source fills each