The Quiet Channel: India's Related-Party Loan Architecture

What the Framework Actually Reads

The R-axis is the universe's binding governance constraint in India. Mean 56.8, median 57, against B-axis mean 66.8 and T-axis mean 48.5. Among the 311 [R-weak] Chameleons identified in Note 1, the R-axis mean drops to 45 — these are the firms whose composite governance score is dragged specifically by the R-axis dimension.

The framework reads the R-axis through six sub-components in its specification: related-party transaction size, promoter remuneration, related-party loans and inter-corporate deposits, key managerial personnel turnover, dividend policy stability, and merger and acquisition governance. In the production snapshot, three of these sub-components carry universe-scale measurement; three currently sit on default values pending pipeline expansion.

The 56.8 R-axis mean is therefore not the framework's complete reading of Indian conflict-of-interest exposure. It is what the three currently-measurable sub-components produce. The headline RPT volume sub-component, which the regulation was designed around, is the one the framework currently cannot read at scale. The dimension through which most of the R-axis weakness signal flows in the current snapshot is the related-party-loan and inter-corporate-deposit sub-component.

The 695-Firm Cohort

The related-party loans and inter-corporate deposits sub-component reads with full variation across the Indian universe. The distribution is the following:

The high-risk band — 341 firms with material related-party-loan exposure exceeding ten percent of total assets, or carrying an active item-specific observation in the auditor's CARO 2020 report — sits at one in six firms. Combined with the Insufficient band, the population of firms carrying material related-loan exposure or partial-disclosure flags reaches 695 firms — thirty-four point five percent of the universe. This is the most populated R-axis weakness signal the framework reads. It is more populated than the pledging signal documented in Note 3, more populated than the promoter-remuneration signal, and substantially more populated than the framework's currently-defaulted RPT volume reading would otherwise produce.

The regulatory architecture this cohort sits in is, on paper, among the strictest in Asia. LODR Regulation 23 defines a related-party transaction as material if its value exceeds ten percent of the listed entity's annual consolidated turnover, or — for the top 1,000 firms by market capitalization — exceeds ₹1,000 crore individually. Material RPTs require audit committee approval, board approval, and Majority of Minority shareholder approval excluding promoter votes. Companies Act §188 imposes parallel requirements through the corporate-law channel^[3]; §177(4)(iv) mandates audit committee approval of all related-party transactions; §186 governs inter-corporate loans and investments specifically. The framework's challenge is not the rule — it is the measurability of compliance against the rule at universe scale. Annual-report related-party transaction tables file in unstructured PDF text rather than structured XBRL; the extraction pipeline that would parse RPT volumes against the materiality threshold across 2,012 firms is the same kind of pipeline whose absence Note 4 documented for the board-substance dimension. The 89.21 percent of the universe sitting on the R-01 default band is the visible consequence. What the framework does read at universe scale is the related-party-loan dimension — the sub-component captured in the auditor's report and financial statement notes, where the disclosure format is more standardized than the related-party transaction schedule itself. The 695-firm cohort is what that reading produces.

A pipeline note: the override pathway specifically targeting the related-loan mechanism — material loans combined with active CARO observation — currently fires zero times in the production snapshot. The mechanism specification anticipates this trigger; the CARO observation extraction pipeline is not yet at universe scale. The 341 firms in the high-risk band carry the loan exposure signal but not the combined CARO signal that would activate the override. This is one of the secondary override pathways referenced in Note 1 as pipeline-pending.

Sector and Segment

The high-risk related-loan cohort concentrates in capital-intensive sectors with elaborate subsidiary or special-purpose-vehicle structures.

Power, Realty, and Telecommunication lead the distribution. These are sectors where the standard architecture involves subsidiary special-purpose vehicles for project-level financing, group-affiliate funding for working-capital cycles, and parent-guaranteed bridging loans during construction or spectrum-payment cycles. The related-loan flow is structurally embedded in how these businesses are financed. Information Technology, Fast-Moving Consumer Goods, and standalone-listed Healthcare names sit at substantially lower related-loan concentration, consistent with their cleaner-cashflow profiles and lower reliance on group-affiliate funding architecture.

By index segment, the related-loan signal behaves differently from the pledging signal documented in Note 3. Pledging concentrated heavily outside the Nifty 500 — seventy-nine of the eighty-six pledging Kill Switch firings sat in the long tail. Related-loan exposure, by contrast, distributes more evenly: the high-risk band runs at 16.0% inside Nifty 500 and 17.2% outside; the Excellent band at 41.2% inside and 33.1% outside. The R-axis composite gap between segments is 5.6 points — small relative to the 16.4-point B-axis gap. Related-party-loan exposure is present at large-cap as well as small-cap. Reliance, Tata, Birla, and Mahindra group entities carry meaningful related-loan footprints by virtue of their group-structure architecture; the small-cap segment carries the loan exposure through different mechanisms (working-capital cycles, project SPV funding) but at comparable prevalence. The R-axis is, structurally, less segmented than the B-axis. The pathology runs through the listing tier rather than concentrating below it.

Four Firms, Four R-Axes

The four benchmark firms with full R-axis sub-component decomposition produce the most analytically interesting reading in the framework's R-axis output. Two contrasts within the cohort carry the empirical content.

Reliance is the framework's most consequential measurement gap. The firm carries the largest aggregate related-party transaction volume on the Indian exchange — flowing between the listed Reliance Industries, Reliance Retail Ventures, Jio Platforms, Reliance Strategic Investments, Network18, and the broader Jio cluster. The RPT Size sub-component, specifically designed to read this volume, sits at the universe default band because the annual-report-table extraction pipeline has not yet parsed the firm's FY2025 RPT schedule. The [R-weak] Chameleon reading at R = 51 therefore comes from sub-components other than the one that should be carrying the load — the related-loan penalty (Insufficient band) and the dividend variability flag combine with the default RPT score to produce the composite. The honest interpretation: Reliance's true R-axis position cannot be confirmed against a fully-parsed framework reading until the universe-scale pipeline catches up. The current R = 51 is real signal layered over a measurement absence.

The TCS-versus-Infosys contrast is the framework's most analytically interesting empirical finding. Both firms read at the Excellent band on promoter remuneration. Both share the dividend-policy flag at the Insufficient band. Both carry default readings on KMP turnover, M&A governance, and RPT size. The R-axis composite nonetheless lands fifteen points apart — TCS at 67, Infosys at 52. The single sub-component driving the gap is related-party loans and inter-corporate deposits. TCS reads at the Excellent band: zero or near-zero related-party-loan exposure as a fraction of total assets. Infosys reads at the High-risk band: the framework places its related-loan exposure at the highest-risk level, likely tracing to subsidiary-financing structures (EdgeVerve, Brilliant Basics, Lodestone) reflected in the FY2025 annual report financial notes.

The empirical reading is counterintuitive. TCS is wholly owned by Tata Sons and listed within India's largest business-house group structure; the conventional governance assumption would predict elevated related-party flow between TCS and the broader Tata group. Infosys is a standalone independent listed entity with founder-light governance; the conventional assumption would predict cleaner separation. The framework reads the inverse. The Tata-group firm's FY2025 financial notes show clean separation between the listed entity and the broader group on the loan dimension; the standalone IT firm's notes show subsidiary-financing exposure that crosses the framework's high-risk threshold. This is the empirical answer to a long-running question about how Tata-group-structure firms compare to standalone IT firms on related-party flow: in this snapshot, the Tata-group firm reads cleaner. TCS is the only Hidden Gem among the four anchor firms, and the related-loan sub-component is what produces the archetype distinction.

Zee Entertainment completes the contrast. The Kill Switch firing examined in Note 3 runs through the B-axis effective control collapse mechanism; the R-axis reads at fifty-seven, above the universe mean. Post-Subhash-Chandra-exit professional management has cleaned the compensation structure (R-02 Excellent), and the residual related-loan exposure (R-03 Insufficient) sits at the partial-disclosure band rather than the high-risk bottom. Zee is the inverse case to Reliance and Infosys: cleaner R-axis, distressed B-axis. The framework distinguishes the two pathologies through axis attribution.

Two Markets, Two R-Axes

Korea and India share an axis-level pattern: the R-axis is the binding governance constraint in both markets. The mechanism through which the R-axis weakness manifests, however, runs in opposite directions.

The Korean pathology is shareholder-return avoidance. Firms accumulate treasury stock without canceling it, then dispose of it selectively through transactions that are mechanically permitted under Capital Markets Act §165-3 but substantively dilutive to existing shareholders. The 1:1,066 disposition-to-cancellation ratio documented in the Korean Foundation Series Note 5 is the canonical figure^[4]. The Korean R-axis weakness is what firms refuse to do — return capital to minority shareholders in cancellation form rather than redirect it through the disposition channel.

The Indian pathology runs in the opposite direction because the Korean mechanism does not exist. Companies Act §68(7) mandates cancellation of all bought-back shares within seven days of buyback completion. The disposition channel that Korean treasury-stock pathology exploits is statutorily closed in India. Indian R-axis weakness flows instead through the related-party transaction architecture: 695 firms (34.5%) sit at the high-risk or partial-disclosure band on the related-party-loan and inter-corporate-deposit sub-component. Capital flows within group affiliates through structures that are LODR Regulation 23 disclosable, but whose materiality threshold leaves substantial sub-threshold transaction volume in the audit-committee-only approval channel. The Indian R-axis weakness is what firms quietly do — move capital through group structures via channels that are individually compliant but collectively material. Both R-axes are binding; the directions are inverse. Frameworks that conflate the two will misread both.

The Channel That Stays Quiet

Three readings follow. The R-axis 56.8 figure is what three of six sub-components produce when the other three sit on default values; the headline RPT volume reading specifically is the dimension the framework cannot currently measure at scale. The 695-firm finding is robust to those measurement gaps — the related-party-loan and inter-corporate-deposit sub-component reads with full universe-scale variation, and is the most populated R-axis weakness signal the framework currently produces. India's R-axis weakness operates through a mechanism (related-party loan flow within group structures) that has no Korean analog because the Korean regulatory architecture closes a different channel than the Indian architecture closes — the two markets' R-axis stories are comparable at the axis level, not at the pathology level.

The R-axis is the binding constraint on Indian governance. The sub-component carrying the binding signal is not the headline related-party transaction volume the regulation was designed around — that is the dimension the framework currently cannot measure at scale. The binding signal runs through the related-party loan and inter-corporate-deposit channel, where 695 firms sit at material exposure or partial disclosure, and where the Excellent reading distinguishes Hidden Gem from [R-weak] Chameleon at the archetype level. This is the channel that LODR Regulation 23's materiality threshold does not catch^[2] and the Companies Act buyback regime does not address — the channel through which capital quietly flows within Indian group structures, individually compliant transaction by individually compliant transaction. The framework currently reads it for one in three firms. The next pipeline expansion will read it for the rest.

The Apex G-Score framework currently covers 2,012 NSE non-financial mainboard listed companies as of the April 2026 production snapshot^[1]. Underlying data: FY2025 cross-section with quarterly shareholding patterns through Q3 FY2026 (December 2025). Three R-axis sub-components carry universe-scale measurement gaps in the current snapshot: RPT Size Anomaly sits on default for 89.21% of the universe pending an annual-report-table extraction pipeline; KMP Turnover and M&A/Investment Governance sit on default for 100% of the universe pending corresponding event-disclosure pipelines. The R-axis findings reported here reflect the working sub-components — promoter remuneration, related-party loans and inter-corporate deposits, and dividend policy consistency. The 695-firm high-risk cohort is robust at universe scale; the headline RPT volume distribution is not credibly measurable at universe scale and is not reported.