Read Time: 6 min

Transfer Pricing

April 2026

Transfer Pricing - Fintech & Digital Payments Pricing Algorithms

Fintech & Digital Payments: Pricing Algorithms, Compliance Functions & Market Side Value Creation

Introduction

India’s fintech landscape has scaled at break‑neck speed with UPI now processing ~20–21 billion transactions a month and daily volumes in the ~640–700 million range. This ubiquity, driven by QR code merchant acceptance and low-cost digital rails, has transformed how value is created and monetised in payments. For Transfer Pricing (TP), this means profits are no longer explained by tangible assets alone but by risk engines, KYC / AML controls, data pipelines, switching platforms and user‑side network effects.

At the same time, OECD TP guidance (2022) consolidates BEPS Actions 8–10 (intangibles, profit split, HTVI) and financial transaction rules, sharpening how DEMPE and risk control determine returns for unique intangibles and integrated platforms. For India‑centric fintech groups, RBI’s KYC Master Direction (updated repeatedly through 2024–25) and Payment Aggregator rules define critical regulated functions whose performance and control have TP consequences. 

Why Fintech & Digital Payments Complicate Transfer Pricing

Multiple value drivers coexist in tightly integrated stacks:

  • Core platforms (switching, settlement orchestration, routing)
  • Algorithms (fraud detection, credit scoring, chargeback management, dynamic routing)
  • Compliance engines (KYC / AML, sanctions screening, V‑CIP) under RBI directions
  • Data assets (merchant telemetry, device / behavioural signals) that feed ML models
  • User / merchant network effects that amplify monetisation (P2M) as UPI scales

These features create unique and valuable intangibles with cross‑border DEMPE contributions and few clean external comparables pushing taxpayers and authorities toward profit‑split or contribution‑based approaches rather than one‑sided cost‑plus alone. OECD guidance recognises such settings highly integrated operations, unique intangibles, multi‑party contributions as classic use cases for transactional profit split and HTVI safeguards.

Key Fintech Intangibles Relevant for TP

  1. Fraud & Risk Engines (Real‑Time) – ML models that ingest transaction, device and behavioural data to assess risk and route authorisations.
  2. Credit Underwriting & Scoring – used in BNPL / merchant cash advance / embedded lending, often trained on proprietary data.
  3. KYC / AML Orchestration & Sanctions Screening – technology and process IP aligned to RBI Master Directions and FATF‑aligned updates (including V‑CIP).
  4. Acquirer / Aggregator Orchestration Logic – smart routing and failover; uptime architecture.
  5. Cyber‑Security & Tokenisation – controls that preserve trust and acceptance.

Under OECD Chapter VI, these are unique intangibles; the associated returns should follow DEMPE, who develops, enhances, maintains, protects and exploits and who controls risks and has the financial capacity to bear them.

On the compliance side, the RBI KYC Master Direction (2016, updated through 2024–25) codifies CDD, ongoing due diligence, sanctions screening, V‑CIP and central KYC usage functions that, if performed in India with decision‑making authority, indicate substantive contributions that may warrant non‑routine returns relative to mere routine processing.

TP Implications Across Common Fintech Functions

Payment Processing, Switching & Aggregation

  • Characterisation: Payment aggregation / processing + platform enablement; typically, a mix of routine facilitation and non‑routine platform intangibles.
  • Regulatory anchor: RBI Guidelines on Payment Aggregators / Payment Gateways require authorisation, net‑worth, merchant onboarding, settlement timelines, grievance redressal and technology baselines evidence of substantive regulated activity (and associated control) in India.
  • Pricing:
    • Cost‑plus for routine ops (merchant onboarding, settlement ops, customer support) where functions / risks are limited.
    • Residual profit split where proprietary routing / fraud engines and multi‑jurisdiction DEMPE jointly drive margins.

Data Use & Monetisation

  • Use‑cases: risk analytics for merchants, performance dashboards, cohort insights.
  • Constraints: DPDPA 2023 + DPDP Rules 2025 define consent, legitimate use, significant data fiduciary obligations, localisation / transfer conditions impacting control over data and hence who economically exploits it.
  • Pricing: where third‑party CUPs are scarce, consider allocation keys (e.g., data volume, active merchants, model feature contribution) or profit‑split. Use HTVI guardrails if projections are highly uncertain.

Digital Lending & BNPL

  • Intangibles: underwriting models, collections strategies, alternative data models.
  • Regulatory overlay: KYC / AML, underwriting governance; RBI guidance on financial transactions/credit risk has analogues in OECD Chapter X for intra‑group financial transactions, emphasising control over risk and pricing guarantees/credit support.
  • Pricing: split between routine servicing centres (cost‑plus) and residual returns for IP owners and risk controllers (profit‑split). Where an India unit controls data pipelines and decides on model updates, its share increases.

Cross‑Border Gateways & Remittances

  • Issues: allocation between sending / receiving entities; compliance costs in each jurisdiction; with UPI’s international tie‑ups expanding, sourcing rules and user‑side contributions matter.
  • Pricing: platform fee split based on value drivers (liquidity provision, compliance, corridor risk, FX spread decisioning).

OECD & Indian Regulatory Perspectives (Quick Map)

  • OECD 2022 TP Guidelines: consolidation of intangibles, profit split, HTVI and financial transactions; focus on accurate delineation and DEMPE.
  • HTVI: allows ex‑post outcomes to inform ex‑ante pricing where uncertainty is high, relevant for algorithm licences / model transfers.
  • India’s User / Market Side: While global Pillar One (Amount A) remains in progress, India’s historic policy stance on market/user jurisdiction is visible in SEP and (earlier) Equalisation Levy; these influence expectations around market‑based returns in TP positions pending Pillar One’s finalisation.
  • RBI: KYC Master Direction (periodically updated), Payment Aggregator guidelines codify substantive regulated functions that anchor DEMPE and risk control locally.
  • Data Protection: DPDPA 2023 + Rules 2025 shape legal/economic control over data used in ML training and monetisation.

User / Market Side Contributions in India

India has argued that local users and merchants create location‑specific value (network effects, behavioural data, advertiser / merchant attraction). Historically, India used Equalisation Levy (6% on digital ads; 2% on e‑commerce supply) and SEP thresholds (₹20 million revenue or 300,000 users) to defend market jurisdiction rights in the context that influences TP audits on market‑side returns, even as DST‑like measures are expected to be rationalised under Pillar One once implemented. (Note: CBDT indicates scope changes and sunset interactions over time.)

For fintech models monetising local P2M scale (e.g., UPI), this translates into uplift expectations for Indian market entities where they drive user / merchant acquisition, perform KYC / AML or iterate fraud/credit models using Indian datasets.

Practical Challenges for Fintech MNEs

  • Comparables Scarcity: No two risk engines or payment stacks are alike; CUPs for algorithm licences are rare; databases for service fees need careful functional filters.
  • Regulatory Fragmentation: KYC / V‑CIP changes and PA authorisation norms shift cost/risk profiles frequently across entities.
  • Data Governance: DPDPA requirements affect data flows, storage and control central to DEMPE and profit attribution.
  • Pillar One Uncertainty: Market‑side allocations may evolve; TP policies should remain adaptable.
  • Overlap with SEP/DST History: Evaluating interactions to avoid double taxation and position defensibly.

Recommended TP Strategies for the Digital Payments Stack

1) Strengthen DEMPE Analysis (Algorithms & Data)

Document who builds and updates models, who sets acceptance/risk thresholds, where training / inference occurs and who approves releases. Map data lineage and access rights under DPDPA; align these with economic ownership of the algorithms. Use change logs and model cards as TP evidence.

2) Robust Inter‑company Agreements (ICAs)

Explicitly address data usage rights, model training rights, algorithm ownership, updates, service‑level and compliance obligations (e.g., KYC/V‑CIP) and shared platform cost allocations compliant with RBI norms. For risk/control functions, embed decision rights and indemnities.

3) Method Selection: Don’t Default to Cost‑Plus

  • Routine Operations (call centres, merchant onboarding, reconciliations, dispute back‑office): Cost‑plus.
  • Integrated Platforms with Multi‑Jurisdiction DEMPE (fraud engines, routing, underwriting): Residual Profit Split, with allocation keys tied to objective usage metrics (e.g., transactions scored, model features originated, A/B tested improvements). This mirrors OECD profit split guidance in complex, integrated settings.
  • Licensing Algorithms / Analytics: Use valuation techniques + HTVI guardrails when forecasts are uncertain; CUP feasible only for standardised modules.

4) Benchmarking Playbook (Payments & Services)

Where market references are needed (e.g., processing/aggregation fees, risk services, KYC utilities), build narrow, function‑matched searches; consider specialty datasets for service fees and triangulate with TNMM as a cross‑check. Maintain India‑specific filters where location savings or regulatory burdens are material.

5) Location Savings & Market Premium

India‑based capability hubs may deliver location savings; document if and how these are already competed away (e.g., salary arbitrage captured by market rates) before layering any uplift. Conversely, for market premium (UPI‑driven P2M scale), articulate a facts‑based rationale for market entities’ shares when they perform on‑the‑ground DEMPE (acquisition, compliance, risk tuning).

6) Align with Pillar One Trajectory

Track Pillar One’s Amount and MLC status and its interplay with India’s legacy DST/EL measures. Where in scope, expect destination‑based reallocation to markets and plan for corresponding TP policy adjustments to avoid mismatches.

7) Documentation: What Indian TP Files Should Show

  • Functional analysis that clearly separates regulated compliance work (RBI‑mandated) from routine ops and non‑routine IP‑driven functions.
  • Model governance artefacts (design docs, risk appetite statements, MRM sign‑offs).
  • Data governance artefacts (consents, purpose limitation, cross‑border transfer assessments under DPDPA).
  • Benchmarking workpapers + profit‑split worksheets with transparent allocation keys.

Conclusion

Fintech and digital payments businesses in India sit at the intersection of regulated compliance, unique platform intangibles and market‑side network effects. The 2022 OECD TP Guidelines and India’s RBI / DPDPA regimes together imply that returns should follow real substance—who designs and controls risk engines, who governs data and compliance, and who builds and exploits the user/merchant base.

A hybrid TP architecture usually works best:

  • Cost‑plus for clearly routine, rule‑bound operations;
  • Residual profit split for jointly developed algorithms / platforms;
  • Careful market‑side attribution where Indian entities demonstrably drive adoption and compliance; and
  • Adaptive policies to anticipate Pillar One.

With disciplined DEMPE documentation, purpose‑built ICAs and evidence‑rich benchmarking, fintech MNEs can minimise disputes and align profit attribution to how value is truly created in India’s real‑time payment economy.