The Vanishing Escape Valve: Reconciling Section 62’s Five-Year Assessment Window with Section 39(11)’s Three-Year Filing Bar – A Critical Examination of an Unresolved Fault-Line in the CGST Act
Summary: Article examines the interaction between Section 62 of the CGST Act, 2017 and Section 39(11), describing an unresolved issue where Section 62 permits best-judgment assessment within five years while Section 39(11), effective from 1 October 2023, bars filing returns after three years from the due date unless relaxed by Government notification. It explains that Section 62(2) provides for deemed withdrawal of a best-judgment assessment upon filing a valid return within the prescribed period, but this mechanism may become unavailable where the three-year filing bar has already expired. The article discusses textual and purposive interpretations of the provisions, notes that no reported High Court, GSTAT or Supreme Court decision has directly addressed this interplay, and refers to existing judicial guidance that failure to file within the Section 62(2) period leaves the statutory appeal under Section 107 available. It suggests practical approaches depending on when the assessment order is served, including filing the return where permissible, pursuing a Section 107 appeal, seeking a notification under the proviso to Section 39(11), or considering a writ petition, while describing the issue as an unresolved statutory gap requiring legislative or judicial clarification.
I. Introduction
Best-judgment assessment of non-filers under Section 62 of the CGST Act, 2017 has always carried a distinctive, almost benign, character within the assessment scheme of the Act. Unlike Sections 73 and 74 — which culminate in adversarial, merit-based demands — Section 62 is designed as a compliance-forcing device. It threatens the non-filer with a best-judgment order, but simultaneously hands him the key to undo it: file the return, and the order is deemed withdrawn. That design logic, intact since 1 July 2017, has quietly been placed in jeopardy by a provision inserted six years later — Section 39(11) — and the two now sit in the statute book in a relationship that the legislature does not appear to have thought through.
This article attempts to set out the issue, trace its statutory anatomy, and work through how it ought to be resolved — while candidly flagging that no reported ruling appears to have yet addressed it squarely.
II. The Issue Stated
Section 62(1) empowers the proper officer, notwithstanding Sections 73, 74 or 74A, to pass a best-judgment assessment order against a registered person who fails to file a return under Section 39 or 45 despite service of a Section 46 notice (FORM GSTR-3A). Critically, the officer has five years from the due date of the annual return under Section 44 to pass this order.
Section 62(2) softens this power with a self-correcting mechanism: if the assessee furnishes a “valid return” — defined in Section 2(117) as a Section 39(1) return on which self-assessed tax has been paid in full — within sixty days of service of the ASMT-13 order (originally thirty days, enlarged by the Finance Act, 2023), the order is deemed withdrawn. A further sixty-day window was added by the Finance Act, 2023 itself, on payment of an additional late fee of ₹100 per day of delay. Interest under Section 50(1) and late fee under Section 47 survive regardless, but the assessment itself disappears the moment the assessee actually complies.
Section 39(11), inserted by the Finance Act, 2023 with effect from 1 October 2023 and subsequently operationalised on the GST portal, introduces an entirely independent bar: a registered person “shall not be allowed” to furnish a return for a tax period after the expiry of three years from its due date, subject only to a Government notification, on Council recommendation, relaxing this for a specified class of persons.
The collision is immediate once the two timelines are placed side by side. Section 62(1) tells the department it may wait up to five years to assess a non-filer. Section 39(11) tells the very same non-filer that his ability to cure the default by filing a return dies at the three-year mark — regardless of when, within that five-year window, the department chooses to act.
Consider a straightforward illustration. A registered person fails to file GSTR-3B for July 2021 (due date 20 August 2021). The department issues a Section 46 notice and, thereafter, a Section 62(1) best-judgment order in, say, November 2025 — comfortably within the five-year limitation running from the due date of the FY 2021-22 annual return (31 December 2022), and therefore an entirely lawful order. But by November 2025, the three-year window under Section 39(11) for filing the July 2021 return (which expired on 20 August 2024) has already closed. The taxpayer is portal-locked. He can never furnish a “valid return” for that period, at any point, however promptly he tries. Section 62(2) — a right the statute holds out to him in the very same breath as the power to assess him — is not merely difficult to exercise. It is dead on arrival.
This is not a hypothetical corner case. Given that Section 62 assessments are frequently the product of departmental drives against dormant or defaulting GSTINs — often triggered years after the default, once data analytics or return-matching flags the non-filer — a meaningful proportion of Section 62 orders will fall in years four and five of the limitation window, precisely where the dichotomy bites hardest.
III. Is This a Genuine Conflict, or Merely the Failure of a Condition Precedent?
Before reaching for tools of harmonious construction, it is worth testing whether a “conflict” exists in the technical sense at all.
A. The textualist reading: no conflict, only an extinguished precondition
Section 62(2) does not confer an independent, free-standing right to file a return. It is a conditional consequence: “where the registered person furnishes a valid return… the said assessment order shall be deemed to have been withdrawn.” The provision is silent on whether the person is legally entitled to furnish that return — it assumes the capacity to file and merely prescribes what happens if that capacity is exercised within the stipulated window. Section 39(11), by contrast, goes to the anterior question: is filing possible at all? On a strict reading, Section 39(11) does not need to override Section 62(2); it simply forecloses the factual predicate on which Section 62(2) would otherwise operate.
This reading draws further textual support from what Section 62(1) chooses to say and what it omits. The opening non-obstante clause is carefully calibrated — it overrides Sections 73, 74 and 74A specifically. It says nothing about Section 39(11). Had Parliament intended the Section 62(2) filing window to survive the Section 39(11) bar, the natural drafting choice would have been to say so, either by amending Section 62(2) itself or by carving out an exception in the proviso to Section 39(11). Neither happened — and both provisions were, in fact, touched by legislative amendment in overlapping periods (Section 62(2)’s window was enlarged and Section 39(11) was inserted within the same Finance Act, 2023 cycle), which makes the silence more, not less, significant. Applying the ordinary canon leges posteriores priores contrarias abrogant — a later provision prevails over an earlier one to the extent of inconsistency — Section 39(11), being the later and more specific bar on filing eligibility, would ordinarily control.
B. The purposive reading: an absurd and self-defeating result
Against this, a purposive reading exposes a serious difficulty. Section 62 was structured from inception as a provisional substitute for self-assessment — a spur to compliance, not a final adversarial determination. That is precisely why it is carved out of the non-obstante sweep against Sections 73/74/74A: Parliament was signalling that best-judgment assessment under Section 62 is qualitatively different, meant to yield the moment genuine compliance occurs.
Reading Section 39(11) to silently disable that reversal mechanism for orders passed in years four and five produces a result that cannot rationally have been intended: two taxpayers, identically situated in every respect except the date on which the department got around to issuing the ASMT-13 order, are treated with radically different outcomes. The taxpayer assessed in month twenty-five retains a full opportunity to cure; the taxpayer assessed in month thirty-nine has none. The determining variable is not the taxpayer’s conduct — it is administrative timing entirely outside his control. A statute that makes the finality of a demand turn on how promptly (or tardily) an officer chose to issue a notice, rather than on anything the assessee did, sits uneasily with the presumption against arbitrary and unequal operation that Indian courts have consistently applied while construing fiscal statutes — most notably in the Supreme Court’s insistence, in K.P. Varghese v. Income Tax Officer, AIR 1981 SC 1922, that a construction producing manifestly unreasonable or absurd consequences must yield to one that advances the evident legislative purpose, even where the literal language could support the harsher reading.
The same instinct animates the well-established canon generalia specialibus non derogant — a general provision must yield to a special one governing the particular situation, to the extent of inconsistency. Section 39(11) is, in this framing, a general rule addressed to the ordinary universe of voluntary, self-initiated return-filing (its evident mischief being indefinite backlog filing that distorts ITC matching, revenue certainty, and the finality on which Sections 16(4) and downstream compliance obligations depend). Section 62(2), by contrast, is a narrow, specific mechanism addressed only to the discrete situation of a return filed in direct response to, and within a fixed statutory window triggered by, a best-judgment order. On this view, the specific provision should not be read as impliedly repealed by the general one.
C. Where the balance lies
Neither reading commands the field with certainty, and that, in itself, is the correct headline for a treatise: this is an unresolved statutory gap, not a settled question with an obvious answer. My own assessment, for what it is worth, is that the purposive reading deserves to prevail as a matter of principle, because the textualist reading converts what Parliament explicitly built as a provisional assessment mechanism into a silently final one for a defined subset of taxpayers, without any legislative articulation of that consequence. But a court applying strict rules of fiscal construction — which traditionally resist reading benefits into a taxing statute that the text does not confer — could just as defensibly hold the line on the literal wording of Section 39(11)’s “shall not be allowed.”
IV. Judicial Guidance on Section 62 — What Exists, and What Does Not
It bears emphasis that, as of this writing, no reported High Court, GSTAT, or Supreme Court decision has addressed this specific interplay between Section 62(1)’s five-year limitation and Section 39(11)’s three-year filing bar. Given that Section 39(11) is a comparatively recent insertion and GSTAT itself became operational only through 2025–26, this is unsurprising — the fact pattern needed for such a dispute to mature (an ASMT-13 order issued after the three-year filing window has already closed) has only recently begun to arise in volume.
What existing jurisprudence does establish is the consequence of a missed Section 62(2) window generally. In Hash Constructions v. Deputy Commissioner, 2021 (49) G.S.T.L. 236 (Ker.) [02-02-2021], it was held that where the assessee fails to file the return within the window prescribed under Section 62(2), the summary demand in FORM GST DRC-07 stands, and the only remedy available to displace it is a statutory appeal under Section 107 — the assessment does not become void merely because it was passed ex parte or without a personal hearing, since Section 62 best-judgment assessments are, by design, capable of being made without prior hearing.
This is significant for the present dichotomy because it confirms the fallback position for a taxpayer caught by the Section 39(11) bar: even though the return-filing route under Section 62(2) is closed to him, his substantive right to challenge the correctness of the best-judgment quantification through an ordinary Section 107 appeal (and thereafter Section 112, before GSTAT) remains wholly unaffected by Section 39(11), since that provision speaks only to the eligibility to file a return and has no bearing on appellate remedies.
V. Working Resolution for Practitioners
Pending legislative clarification or judicial pronouncement, the following framework is suggested:
1. Where the ASMT-13 order is served with the Section 39(11) window still open — broadly, before the three-year mark from the original due date, with sufficient residual time to utilise the sixty-plus-sixty-day cure period — Section 62(2) operates exactly as designed, and the assessee should file the valid return without delay to secure deemed withdrawal.
2. Where the order is served after the Section 39(11) cut-off has already elapsed, the assessee is portal-locked and cannot invoke Section 62(2) at all. The realistic remedies are:
- Statutory appeal under Section 107 on the merits of the best-judgment quantification — available regardless of the Section 39(11) bar, and the only tested remedy per Hash Constructions.
- A representation to the jurisdictional Commissioner or, through appropriate channels, to the GST Council, urging a notification under the proviso to Section 39(11) carving out an exception for returns sought to be filed specifically pursuant to a Section 62(1) order — this is, in principle, the cleanest and most administratively economical fix, since the proviso already contemplates class-specific relaxations.
- Constitutional challenge by way of writ petition, contending that the undifferentiated, notice-timing-dependent operation of the Section 39(11) bar on Section 62(2) situations offends Article 14 by making the survival of a specific statutory remedy turn on administrative delay rather than assessee conduct. This route remains untested and should be approached with the caution due to any first-impression constitutional argument.
3. Where the order falls in the ambiguous middle — served close to, but before, the three-year cut-off, leaving an attenuated but technically live filing window — the assessee should be advised to file immediately, since any delay materially increases the risk of being caught by the portal bar mid-way through the cure period.
VI. Conclusion
The Section 62 / Section 39(11) dichotomy is, at bottom, a case of two provisions — each individually coherent — being inserted and amended without cross-reference to the other, producing an outcome that appears to have escaped legislative attention: a statutory safety valve that functions perfectly for some non-filers and not at all for others, based on nothing more than when the department chooses to act. Until Parliament, the GST Council, or the courts step in to close this gap, practitioners must treat it as a live and material risk in any Section 62 matter approaching or beyond the three-year mark, and structure their advice — and, where the facts warrant, their representations — accordingly.
******
The author is a retired officer of the Indian Revenue Service (CBIC), and practises as an Authorised Representative before GSTAT, GST, Customs and Central Excise authorities. Views expressed here are his personal views only based on his reasearch and do not represent any legal opinion in any way.
Author Bio
