The case that decided when an unpaid bill can — and cannot — put a company into insolvency. Sit in the Gallery for a TV dance show, an angry email, and a ₹20-lakh standoff. Step up to the Bar for the three questions and the plausible-contention test. Take the Bench for the over-readings already loose in the market.
Most readers start in the Gallery — and climb before they realise it.
The show. Star TV’s Nach Baliye — viewers vote for their favourite dancers by phone. Mobilox runs the tele-voting, and subcontracts the hard part to Kirusa, who builds the software and the toll-free lines (purchase orders, late 2013).
The bills. Kirusa does the work and raises monthly invoices through 2014 — about ₹20 lakh goes unpaid. It chases by email. The parties also sign a non-disclosure agreement: keep our client work confidential.
The freeze. 30 January 2015: Mobilox writes — you displayed our confidential client campaign on your webpage; that breaches the NDA; every payment is withheld till this is resolved. Kirusa denies it, demands payment, threatens legal action… and then goes silent for over a year.
The new weapon. Late 2016: India’s brand-new Insolvency Code arrives — and the old “can’t pay your debts” ground for winding up a company disappears. Kirusa sends a formal demand notice under the new Code for ₹20,08,202.55. Mobilox replies within days: serious disputes exist; this notice is a pressure tactic.
The seesaw. The NCLT throws Kirusa’s petition out — there’s a notice of dispute. The appeals tribunal reverses: the “dispute” is “vague, got up and motivated”. The case reaches the Supreme Court — the Code is barely a year old, and nobody yet knows what “dispute” means.
The verdict. The Court writes the rulebook: the tribunal is not to try the case. If the debtor points to a real, pre-existing dispute — not bluster, not moonshine — the petition must be rejected, however the dispute might eventually end. Mobilox’s NDA defence qualifies. The petition dies (21 September 2017).
Epilogue — the axe learns to wait. The three questions from this judgment are now recited in virtually every supplier-versus-company insolvency case in India. Insolvency is a rescue machine, not a money-recovery hammer: where a genuine quarrel exists, it must be tried in an ordinary court first — “before the axe falls.”
Because insolvency is not a debt-collection tool. The Code’s door opens only for undisputed operational debts — the moment a real quarrel exists over whether the money is payable, the fight belongs in an ordinary court or arbitration, not in insolvency.
And because the quarrel came first. Mobilox’s NDA complaint was on paper from January 2015 — almost two years before the demand notice. A genuine, pre-existing, documented dispute is a complete shield, whoever might eventually win it.
If you got both — you already understand what the appeals tribunal itself got wrong in this very case.
A judgment is authority only for what it decides. Fix the questions before you read a single answer.
The trigger: when must the NCLT reject an operational creditor’s s. 9 petition — and is the “and” in s. 8(2)(a) conjunctive, so that only a pending suit or arbitration counts?
The standard: how real must the “dispute” be — must it be bona fide, likely to succeed, or merely plausible — and must it exist before the demand notice?
The width: is “dispute” confined to the three sub-clauses of s. 5(6) — or does any real quarrel about payment count (here, a breach of an NDA)?
Two layers, so you always know whose voice you are reading: Our reading is this page’s interpretation — we state it and we stand behind it. The Court is the judgment verbatim, para-numbered from the certified copy.
Our readingA s. 9 petition stands or falls on three questions: an operational debt above the threshold; documents showing it due and unpaid; and no pre-existing dispute (or suit or arbitration). Fail any one, and the petition must be rejected (Para 25).
Our readingA dispute need not already be in court. Reading s. 8(2)(a) conjunctively would shield only debtors who had already sued or arbitrated — an absurdity for quarrels born days before the notice. So the “and” is disjunctive: the existence of a dispute alone is enough (Para 29).
Our readingThe dispute must be pre-existing — alive before the demand notice or invoice arrived. A defence conceived inside the 10-day reply window, with no earlier trace, is not an “existing” dispute (Para 24).
Our readingThe test is plausibility, not probability. The tribunal checks that the defence raises a plausible contention needing further investigation and isn’t bare assertion — then stops. No merits, no mini-trial, and no smuggling “bona fide” back into a section Parliament deliberately stripped of it (Paras 35, 40).
Our readingSection 5(6) is an inclusive definition — any real dispute about payment counts, even one (like an NDA breach) outside its three sub-clauses. Mobilox’s two-year-old, documented NDA quarrel qualified; the petition should never have been admitted (Paras 43, 45, 47).
1 · Purpose first. Operational debts are usually small, and the Code must not hand suppliers a hammer: the scheme exists so that operational creditors cannot “put the corporate debtor into the insolvency resolution process prematurely or initiate the process for extraneous considerations” (Para 29). Insolvency is for resolution, not recovery.
2 · Text bent to serve purpose. The “and” of s. 8(2)(a) becomes “or” because the conjunctive reading would protect only those already in litigation (Paras 29–32); and “bona fide”, deliberately deleted by Parliament from the definition of dispute, cannot be judicially smuggled back in (Para 35).
3 · A standard borrowed, then calibrated. From Australia’s “genuine dispute” cases and English winding-up practice (Paras 37–39), the Court distils one line of Indian law: plausible contention, no merits, no mini-trial (Para 40) — then applies it: two years of documented NDA correspondence is a real dispute, whatever its eventual fate (Paras 43–46).
“All these circumstances go to show that it is right to have the matter tried out in the present case before the axe falls.” — Para 46: the image every insolvency lawyer now carries — the axe waits while real disputes get tried.
We read the certified copy end to end so you don’t repeat what the summaries got wrong. Both findings are checkable by anyone with the PDF.
Wrong. The tribunal must “separate the grain from the chaff and… reject a spurious defence which is mere bluster” — and the dispute must be pre-existing. Mobilox won because its NDA complaint had a two-year paper trail predating the demand notice; a defence invented in the reply, with no earlier trace, is chaff.
The move: date-stamp the dispute. Ask one question of every defence: where is it in writing before the notice arrived? No pre-notice trace, no shield.
✓ verified against certified copy · Para 24 · 40Wrong. “The Court does not need to be satisfied that the defence is likely to succeed” and does not “examine the merits of the dispute”. The NCLAT was reversed in this very case for branding the defence “vague, got-up and motivated” — a merits judgment it had no business making.
The move: when a tribunal (or opponent) starts assessing who would win the underlying quarrel, hand up Para 40 — existence is the question, never the answer to the quarrel itself.
✓ verified against certified copy · Para 40 · 45A judgment is a tool with two edges. From the Bench you must see both — you will not always be on the same side of it.
Is the debt an operational debt above the threshold, shown by documents to be due and unpaid, with a complete application?Para 25, conditions (i)–(ii) — the threshold before the dispute is ever reached
Did the debtor’s dispute arise before the demand notice or invoice was received?Para 24 — “must be pre-existing”
Does the defence raise a plausible contention requiring further investigation?Para 40 — plausibility, not probability of success
Is it supported by some record — more than bare assertion or a patently feeble legal argument?Para 40 — “an assertion of fact unsupported by evidence” is chaff
Answer the four questions to see what the adjudicating authority must do.