Pine Labs: operating units to equity value

Model specification,25 September2026. Internally money is INR million; aggregate monetary displays convert to ₹crore. Historical allocations are estimates, not observed unit facts.

E1. Reconcile operating units to the audited starting point

FY2026 audited DITP revenue is ₹1,836.82 crore: platform services ₹1,610.32 crore, traded devices ₹202.26 crore, and grants/recoveries ₹24.24 crore.

The allocation uses flow value ₹275,000 crore × approximately 33 bps = ₹907.50 crore, and rounded fintech share 3% × group revenue = ₹81.3177 crore. The remaining ₹621.5023 crore is an estimated terminal/online service pool. At ₹350 per month it implies 1.480 million equivalent paying devices, or 77.678% of the 1.905 million endpoint-average deployed devices. It does not establish an actual paying-device percentage. Applying ₹350 to every deployed device, then adding flow and fintech, exceeds audited platform revenue by ₹178.5977 crore. Online and OMC income remain inside these pools; no additional revenue is added for them.

IAP recognizes ₹352.54 crore of combined processing/distribution fees and ₹231.11 crore of principal card sales. For scenario analysis, allocate 30% of the non-principal portion of ₹64,000 crore IAP GTV to processing-equivalent value and 70% to agency distribution; allocate 40% of combined fees to processing and 60% to agency distribution. These are the model's historical attribution choices. GTV combines loads, certain redemptions and distributed card values; these are not verified distinct customer-throughput series. The implied unit fees are exposed for forecasting. No programme or active-card count is invented.

E2. Terminals: activity, retention, monetisation and sold devices

Closing active devices = opening devices × (1 − churn) + gross additions.

Gross additions are a selected fraction of the opening base. Net growth fades to terminal growth by FY2036, while churn stays constant. Average active devices are the midpoint of opening and closing counts. Equivalent paying units are average active units × paying share. Service revenue is paying units × monthly fee × 12. Monthly fee growth fades to zero.

Sold-device units equal gross additions × merchant-funded deployment share. The reference FY2026 gross additions are 250,000 net additions plus assumed 5% churn on 1.78 million opening devices. Applying 27.5% merchant funding gives estimated sold units; audited traded-goods revenue and rounded device COGS determine the reference selling price and purchase cost per modeled unit. These are calibrated estimates, not disclosed physical sales or prices; peripherals may be included. Device selling-price and procurement inflation are separate controls, both fading to zero. Owned/refurbished procurement has a different unit-cost assumption.

E3. Payment flow and fintech economics

Flow value grows at a selected activity rate, fading to terminal growth. Affordability is a subset of total flow. Revenue equals flow value × [affordability share × affordability fee + remaining share × other-flow fee]. The FY2026 affordability-value reference is ₹74,000 crore; the two-thirds revenue allocation used to derive its fee is illustrative, not a precise annual disclosure. The default forward mix of 22% is an assumption, against the historical volume proxy of 26.9% and management's discussion of faster low-fee UPI growth.

Flow direct cost is flow value × additional servicing cost in basis points. Net recognized fees already exclude relevant interchange/pass-through items; these are not deducted twice. Fintech revenue is transaction activity × estimated net fee per transaction. ₹0.7133 comes from the rounded broader fintech revenue allocation divided by 114 crore Setu transactions; it is not a quoted price.

DITP direct costs reconcile to ₹305.55 crore. Subtract rounded traded-device COGS of ₹176 crore, leaving ₹129.55 crore of network/service costs (the deck rounds this to ₹130 crore). Allocate 70% to active-device servicing and 30% to flow servicing. Forecast unit costs are independently adjustable and do not recalibrate when fees change. Other DITP revenue of ₹24.24 crore follows the active-device base; this is a carry-forward proxy, not a government-grant renewal forecast. Fintech's technology and staff costs remain in the shared capacity pools rather than an invented standalone cost allocation.

E4. IAP processing, distribution and float

The estimated processing, agency distribution and principal-card value baskets each have their own activity growth rate, fading to terminal growth. Processing and agency revenues are value × net fee. Principal cards recognize the sold value gross. Processing costs are value × service-cost bps; agency costs are value × incremental distribution-cost bps; principal-card purchases are a selected percentage of face value.

The historical cost allocation assumes processing costs of 0.1 bp and principal-card purchases at 95% of face value. The residual agency cost makes total IAP direct costs reconcile to ₹363.86 crore, with zero separately allocated historical float cost. These allocations are not disclosed. Their derived forecast rates are fixed inputs: changing a fee never resets costs to preserve an aggregate margin. Distribution costs here are additional expenses, not rebates already netted against recognized revenue.

Year-end customer funds equal total modeled IAP value × equivalent balance days / 365. This stock/flow measure mixes scopes and is not a disclosed redemption dwell time. Float income equals average opening/closing funds × eligible share × net yield. The reference eligible share uses the year-end floating-rate balance ₹5,322.88 crore / total earmarked ₹5,569.74 crore as a proxy. The reference yield reconciles ₹290.12 crore net income to estimated average eligible balances; it is inferred, not reported. Promotional payments are already netted; do not deduct ₹64.91 crore again. A separate float-servicing-cost control exposes the unsupported zero-cost allocation. Customer principal is excluded from shareholder cash and operating working capital.

E5. Staff, cloud, corporate costs and compensation

Expensed payroll equals normalized staff equivalents × expense per equivalent. The reference 5,007 is a closing employee count, not average consolidated FTE; dividing ₹916.54 crore expense by it provides a normalization, not verified employee economics. Staff growth and wage inflation are separate controls, fading respectively to terminal growth and zero.

Cloud costs begin at ₹207 crore. Each year's cost grows with the change in a weighted flow/fintech activity index, raised to a selected capacity elasticity, plus unit-cost inflation. The default activity weight is 50% each and elasticity 0.6. Corporate costs use a separate staff-capacity elasticity, initially 0.3. Both elasticities fade to 1 by FY2036, preventing perpetual productivity gains in the terminal value. These are cost-capacity proxies, not measured server usage. Remaining corporate costs are audited indirect costs less payroll, cloud and credit impairment. Credit impairment follows flow value from the ₹46.03 crore base; no additional settlement-loss charge is layered on top.

Future share-based compensation follows normalized staff equivalents and expense per equivalent, using FY2026 expense less the ₹41.53 crore modification charge. Existing outstanding options retain the exercise-proceeds dilution calculation. The allocation between existing awards and future replacement compensation is not disclosed; possible overlap remains a limitation.

E6. Investment and depreciation

Owned terminal replacements equal opening active devices × company-funded share × (1 − churn) / economic life. New owned terminals equal gross additions × company-funded share. Cash hardware investment is their sum × procurement cost. Replacements do not add active devices. Owned share plus merchant-funded share must not exceed 100%; the remainder represents partner-provided capacity. The 65% owned share, four-year life and ₹3,500 procurement cost are scenario choices; cohort ages and refurbished/international ownership mixes are unavailable.

Capitalized development uses normalized development equivalents and the same cost-per-equivalent proxy. The ₹104.03 crore base implies about 568 equivalents; these are not additional reported employees. Other IT investment begins at the ₹20.15 crore intangible-addition proxy. ROU renewal investment begins at ₹27.12 crore, the reported amortization proxy. Both scale with staff capacity and unit costs. These capital components differ from the timing of reported cash capex; the historical cash-flow line remains actual.

For FY2027 and FY2028, spending floors cover the March unspent IPO objects: devices ₹358.21 crore, IT/cloud ₹203.88 crore and technology ₹91.74 crore. Half is assigned to each year. A control splits IT/cloud between expense and capital. Ordinary spending counts first; only shortages are topped up. Extra hardware spending is unallocated deployment/capacity investment with no invented revenue benefit. The general-purpose/acquisition reserve is handled separately in the cash bridge. June utilization is subsequent evidence, not a partial rebase of March cash.

Opening depreciation of ₹270.13 crore runs off over an estimated remaining life. New investment cohorts use straight-line depreciation with a half-year convention: the selected terminal life for hardware and a separate common life for software, IT and ROU. That common life is an approximation. Forecast depreciation and capex are not revenue percentages.

E7. Working capital and early settlement

Early-settlement book = flow value × funded share × funding days / 365.

The default funded share reconciles the ₹918 crore opening book at an assumed 45 days. Neither this duration nor the resulting funded share is verified for the exact book. Existing fees remain in flow revenue; none are added again. The funding-cost hurdle is a diagnostic financing cost, excluded from FCFF. Funding-only break-even fee equals annual funding rate × days / 365; actual allocated fees, servicing and losses remain unknown.

Other working capital is service-funded operating assets plus goods inventories minus operating payables. The reference uses 75 days of service revenue, 30 days of device/principal-card purchases and a derived 10.615 days of eligible cash operating costs to reconcile management's ₹377 crore. The first component represents broader operating assets, not statutory receivables alone. These are economic proxies, not a reconstructed balance sheet. The alternate ₹366 crore and the ₹723 versus ₹383.08 crore trade-payables perimeter disagreement remain unresolved. Each forward duration is exposed.

E8. Cash flow, terminal economics and equity

Contribution is recognized revenue less direct unit costs. EBITDA subtracts shared cash costs and share-based compensation. EBIT subtracts depreciation. Tax is charged on positive EBIT; no benefit is assumed for losses. FCFF equals EBIT less tax plus depreciation, less cash investment and increases in other working capital and settlement funding.

The selected 13% all-equity hurdle discounts FCFF; it is not an independently estimated WACC. The terminal growth choice is 5%. FY2037 uses constant unit fees, costs and mix, activity/resource growth at terminal growth, no IPO top-ups, and sustainable replacement investment. Terminal depreciation is calculated from the steady growing investment cohorts and their lives, including the half-year convention. Terminal working-capital investment is growth × closing books. This avoids perpetuating temporary opening-asset depreciation runoff. Report both normalized and mechanical terminal cash flow. A negative terminal cash flow is a funding-deficit case, not a stable profitable franchise.

Equity is enterprise value plus cash and deposits, less customer float, borrowings, leases, the selected IPO reserve, contingencies and lien haircut. Roll 0.485 years from 31 March 2026 to the retained 24 September price date. Options use their exercise proceeds only when in the money. Negative equity, if produced, indicates a funding deficit rather than a tradable negative share price.

Sensitivities, evidence gaps and comparison

Scan every economic control over 21 points plus its selected value. Rank by sampled total-equity spread, separately label valuation versus operating assumptions, and report local elasticity. Range width affects rank; these are stress brackets, not probabilities. Test paying share × terminal fee, flow growth × affordability mix, hardware cost × life, and discount × terminal growth. Price, date and the diagnostic funding hurdle are excluded with reasons. Fixed historical attribution uncertainty remains model-form uncertainty and is not exhausted by the forward-control ranking.

The most useful missing evidence is actual average paying devices; owned/sold/refurbished cohorts and costs; additive subscription/online/flow revenue; processing versus distribution value, revenue and costs; billable fintech usage; average eligible float; and the settlement book's eligible activity, duration and fees.

The captured external research report of 30 July 2026 has a ₹210 target. Its undisclosed discounting and cash-bridge assumptions prevent exact reproduction. The external comparison does not calibrate or validate this unit model. No investment position or human adoption of assumptions is inferred.