An introduction
The problem, named
Every mid-market company above a certain scale carries the same invisible cost. The price your procurement team negotiated, the price on the revised purchase order, and the price actually paid — these three numbers are almost always different. Nobody notices until month-end reconciliation. By then the working capital is gone, the dispute window is open, and a finance manager has four days of forensic work ahead.
This is not a discipline problem. It is an architecture problem. There is no layer in your stack that enforces which version of truth was authorized at the moment payment moved.
What India's listed companies disclose
Schedule III of the Companies Act 2013, as amended in March 2021, requires every Indian company to disclose trade-receivables and trade-payables aging in five buckets, each split into four sub-categories including disputed and doubtful. The bottom-right cell of that disclosure — disputed AND aged more than three years AND doubtful — is the validity-gap residue. The evidence is twenty-five listed Indian companies deep, organised across the four sector buckets that map onto where this product deploys first.
The four cases below are the deepest-disclosed within the substrate. Each is audited, SEBI-filed, and traceable to a primary regulatory document.
These are listed, audited, SEBI-disclosed companies. Large companies absorb disputes because they have the balance sheet to. A mid-market company at ₹50 to ₹500 crore in revenue does not have that absorption capacity. A ₹50 lakh dispute aged for two years is a meaningful working-capital cost. A ₹15 lakh ITC permanently lapsed is permanent cash gone. The validity gap kills smaller firms faster, because there is no margin to absorb it.
Why the disputed-payables line is empty
A reasonable challenge to the dispute-residue thesis: if disputes are structural, the Schedule III disputed-payables line should make them visible. The line is empty for most mature companies. Two institutional patterns explain why — and both signal capacity that mid-market companies do not have.
HCC Contract Solutions Limited has operated for seventeen years as a dedicated claims-management subsidiary. On 31 March 2026, HCC executed a deed transferring an arbitration-awards portfolio aggregating ₹1,979.09 crore from the parent to the subsidiary. Disputed receivables exit the parent's trade-receivables disclosure entirely. The line item reads zero. The exposure is real and is sitting one level removed.
A different institutional response: ₹400 crore QIP in April 2024, ₹89 crore realized through Vivad se Vishwas FY24, ₹162.87 crore Mauritius-project completion proceeds May 2024. Direct balance-sheet carriage of residual exposure: ₹2,681 crore contingent liabilities, approximately ₹1,200 crore arbitration awards in stages. Active realization, not reclassification — and equally inaccessible to a mid-market company.
The recognition rule
A state S transitions only when every invariant of the governing agreement holds.
The same rule is evaluated at three moments in an agreement's life — before payment, where the expected state is frozen so a payment that does not match is held before it settles; at the bank event, where the movement is recognised or held; and at closure, where retention and guarantees are released. Validity is enforced across the whole lifecycle, not checked once.
If any invariant of the agreement version fails — amount mismatch, counterparty mismatch, timing window violation, missing approval — the transaction does not finalise. Invalid states are structurally impossible, not procedurally repairable.
The same rule governs payables and receivables. Inbound bank credits resolve against open invoices on the receivables side; outbound payments resolve against approved invoices on the payables side. One architecture, both directions.
Every recognition produces a cryptographic decision hash that links to the agreement version, the invariant set, the bank event payload, and the approval trail. The full decision is reproducible from frozen inputs years later. Auditors verify; they no longer infer.
The state model
The state vocabulary is deliberately constrained. A transaction is in exactly one of these. There is no "Payment Pending," no "Awaiting Approval," no soft transitional limbo. The state is the truth.
How a transaction moves
Sizing the gap, before the conversation
The cost of the validity gap is the sum of five independent sizing components, gated by a sixth: the contract-governed share of spend. We use floor coefficients — the 25th-percentile point across each evidence distribution — so the sizing under-states rather than over-states. For an organisation with X crore annual revenue and Y crore annual procurement spend, the formula reads as below.
| Anchor | Floor coefficient | |
|---|---|---|
| 01 | Procurement leakage Five-bucket model — material price variance, quantity or measurement variance, scope-change leakage, duplicate or inflated invoicing, reconciliation or timing variance. Interpolated between Xelix 0.35% best-in-class and WorldCC ≈8.6% full-lifecycle erosion. | Y × 2.55% 0.0255 × annual procurement spend |
| 02 | Dispute carrying cost Schedule III aging floor, derived from the twenty-five-company evidence substrate. The bottom-right cell — disputed, aged three years or more, doubtful — is the validity-gap residue this captures. | Y × 1.0% 0.010 × annual procurement spend |
| 03 | Reconciliation labour Three AP FTEs per ₹50 crore of revenue at ₹40,000 fully-loaded per FTE per month — conservative against industry-median four-per-50 staffing benchmarks. | X × ₹28,800 (3 FTEs ÷ 50 Cr) × 12 × ₹40,000 |
| 04 | GST input tax credit lapse GSTR-2B mechanics. Permanent ITC lapses from vendor-side mismatches uncovered after the recovery window closes. Implied 18% GST rate against procurement, 0.5% lapse-floor at the 25th percentile of GSTN reconciliation studies. | (Y × 18%) × 0.5% implied ITC × lapse floor |
| 05 | TDS reconciliation loss Form 26AS mismatches. Implied 2.5% TDS rate against revenue, 0.5% loss-floor at the 25th percentile. Compounding exposure under Section 40(a)(ia) disallowance. | (X × 2.5%) × 0.5% implied TDS × loss floor |
| 06 | Transaction mix ratio Contract-governed share of total spend. Eligibility gate, not a sizing component — the framework works where the contract substrate exists. Below 40% contract-governed share the substrate is too thin and the methodology does not apply; between 40% and 50% it applies with a contract-formalisation path first. | ≥ 40% eligibility threshold |
Each anchor is sourced from primary evidence. Schedule III aging precedents from twenty-five listed Indian companies. GSTN reconciliation studies from CBIC notifications, IRIS GST, ClearTax. TDS reconciliation studies from BDO India and Terra Insight. Procurement leakage interpolated between Xelix and WorldCC industry data. The methodology is open; the sizing is conservative; the sources are named.
A worked example
A mid-market company at ₹50 crore annual revenue and ₹32.5 crore annual procurement spend — a typical 65% procurement-to-revenue ratio.
| Anchor | Calculation | Annual cost | |
|---|---|---|---|
| 01 | Procurement leakage | 32.5 × 0.0255 = 0.83 Cr | ₹ 82.88 Lakh |
| 02 | Dispute carrying cost | 32.5 × 0.010 = 0.33 Cr | ₹ 32.50 Lakh |
| 03 | GST ITC lapse | (32.5 × 0.18) × 0.005 = 0.0293 Cr | ₹ 2.92 Lakh |
| 04 | TDS reconciliation loss | (50 × 0.025) × 0.005 = 0.00625 Cr | ₹ 0.62 Lakh |
| Total annual validity gap at the floor | ₹ 1.19 Crore | ||
| Reconciliation labour — operating leverage, shown separately 50 × 28,800 = ₹14.40 Lakh · scales with revenue, never summed into the gap |
₹ 14.40 Lakh | ||
Two-point-three eight percent of revenue, at the floor — the recognition-integrity cost alone, with reconciliation labour carried separately as operating leverage. The actual cost is almost certainly higher — Indian project-overrun data consistently runs two to three times global averages, and the 25th-percentile coefficient is calibrated below the median for defensibility. The same formula scales linearly. A ₹500 crore organisation, at the same procurement intensity, runs an annual validity gap of approximately ₹11.9 crore at the floor.
Size the gap in your own books →A note for the EPC reader
On 12 January 2026, the Ministry of Road Transport and Highways issued Circular H-25011/02/2025-P&P, amending dispute-resolution provisions across BOT, HAM, and EPC contracts. Disputes valued at ₹10 crore or above are now excluded from arbitration. They proceed through a tiered administrative settlement process, with civil court litigation as the fallback. NHAI's pending dispute exposure at the time the circular issued was already material.
And the instruments that ring an EPC contract — performance guarantees, mobilisation advances, retention, earnest-money deposits — are governed too. Each is a recognised instrument with a face value and a validity period, released only when the agreement's closure conditions hold. No guarantee keeps accruing bank commission past the date it should have been returned; no retention sits unreleased after the defect-liability period closes. The commercial edges of the contract, not just its invoices, become accounting truth.
Disputes that previously cleared via arbitration in two to four years will now sit longer in Schedule III's "more than three years" aging bucket as they migrate to civil-court litigation. The arbitration-as-dispute-clearing mechanism that some EPC players had institutionalised becomes less effective. Axiom Layer does not help you resolve disputes faster. It prevents the conditions that produce them.
What we are. What we are not.
The operating-leverage argument
Zoho makes your books fast. We make them right.
Your accounting tool automates the routine. Three-way matching, recurring-payment tagging, vendor master rules. Modern accounting platforms run at seventy to eighty percent automation today. That is not the problem we solve.
Your tool processes data after the data lands. It does not validate whether the data has the right to exist. It books the bank event and moves on. The validity question lives upstream of the ledger.
Where finance-team cost concentrates is the edges — new vendors, variation orders, retention release, defect rectification, period-end recovery cycles. Every one of those is a validity question, not a data question.
As deployment compounds, the contract-governed share of spend grows. Tail-spend vendors that should be formalised get surfaced. The architecture you are buying is, over time, the surface area on which it operates.
The moat is not the software. It is committing to a world where invalid states cannot exist.
In summary