The 562-Day Hole: The Standard Governing General-Insurance Transfers Sunsetted by Mistake, and the Federal Court Kept Applying It

On 20 July 2017, in the Federal Court in Sydney, Gleeson J made orders confirming seven schemes that moved the insurance books of CGU, Swann, WFI, IAG Re, Mutual Community, CGU-VACC and HBF Insurance into a single licensed insurer, Insurance Australia Limited. Seven IAG-group insurers became one: on the evidence recorded in the judgment, the transferring insurers had approximately 3,244,900 active policies on issue as at 30 June 2016. The reasons, published a month later, work carefully through the statutory preconditions and find each one met. At paragraph 45 they identify the instrument those preconditions depend on:

The relevant prudential standard, for the purposes of s 17C(2)(a) and (b) of the Act, is Prudential Standard GPS 410, Transfer and Amalgamation of Insurance Business for General Insurers (“GPS 410”).

The sentence is in the present tense. On 20 July 2017 there was no Prudential Standard GPS 410. It had been automatically repealed 292 days earlier — a count computed here from the register’s own dates — on 1 October 2016, by the sunsetting machinery of the Legislation Act 2003, and it would not exist again until APRA remade it on 16 April 2018. The judgment contains no occurrence of the words “sunset” or “repeal”. Neither of the matter’s two published judgments mentions the instrument’s status; the parties’ submissions and the transcript are not public, and this analysis makes no claim about what was said in them.

This piece traces how that happened, from the register error APRA says caused the lapse, through the eight confirmations the Court decided while the standard was missing, to the statutory review — tabled in the same months — that acknowledged instruments could sunset without ever reaching a list, without a case to point to.

How a standard vanishes#

The sunsetting framework is a deliberate piece of regulatory hygiene: under section 50 of the Legislation Act, a legislative instrument is automatically repealed roughly ten years after registration unless it is remade or exempted. The safety net is section 52: the Attorney-General tables lists of instruments due to sunset, agencies check the lists, and anything still needed gets remade in time. The 2017 statutory review of the framework describes that tabling arrangement as “the primary mechanism by which responsible ministers, the Parliament and agencies are reminded to take action on legislative instruments of continuing relevance” — while noting it “operates as an administrative mechanism only and has no legal effect on the sunsetting of legislative instruments”.

GPS 410 fell through the net from upstream. The standard was made on 7 February 2002 as part of a single determination bundling seven prudential standards — GPS 110, 120, 210, 220, 230, 410 and the transitional GPS 900 — registered as F2006B01543. Through 2006, APRA revoked and replaced most of that bundle: the register’s own annotations record GPS 210, 220 and 230 revoked with effect from 14 February 2006, GPS 120 revoked, and GPS 110 revoked by Insurance (prudential standard) determination No. 7 of 2006 with effect from 1 January 2007. On APRA’s account, it is that last revocation the register got wrong. The explanatory statement for the 2018 remake records:

The original GPS 410 was mistakenly labelled on the Federal Register of Legislation as having been repealed on 1 January 2007 by Insurance (prudential standard) determination No. 7 of 2006 – Prudential Standard GPS 110 – Capital Adequacy. GPS 410 was consequently not included in the lists of sunsetting legislative instruments prepared by the Attorney-General under section 52 of the Legislation Act 2003 … It was not therefore identified for remaking and was repealed on 1 October 2016 by operation of subsection 50(1) of the Legislation Act. It was not APRA’s intention that the original GPS 410 be allowed to sunset.

On APRA’s account, an instrument the register recorded as already dead could not appear on a list of instruments about to die: the arrangement the 2017 review says exists “to ensure that lists are accurate and that nothing sunsets ‘without knowing’” was undone by the register record it draws on. The explanatory statement adds that the Office of Parliamentary Counsel “have acknowledged the error but advised there is no way in which the operation of GPS 410 can be continued other than by remaking it”. The register’s status record for F2006B01543 today carries two entries: in force from 7 February 2002, repealed 1 October 2016, reason: sunset. The 2007 mislabel itself is attested by APRA’s account; no contemporaneous capture of the register’s display was locatable, and this piece attributes it to the explanatory statement throughout.

As late as 30 June 2017 — nine months into the lapse — APRA’s own prudential standards page still listed GPS 410 among the current general insurance standards, with no note of its repeal.

What the statute needed the standard for#

The gap would not matter if GPS 410 were peripheral. It is not. Division 3A of Part III of the Insurance Act 1973 requires a transfer or amalgamation of general-insurance business to go through a scheme confirmed by the Federal Court, and it delegates the procedural content of that process to the prudential standards. The compilation in force in July 2017 reads, at s 17C(2):

An application for confirmation of a scheme may not be made unless: (a) a copy of the scheme and any actuarial report on which the scheme is based have been given to APRA in accordance with the prudential standards; and (b) notice of intention to make the application has been published by the applicant in accordance with the prudential standards; and (c) an approved summary of the scheme has been given to every affected policyholder.

Section 17E(2) adds that an application for confirmation “must be made in accordance with the prudential standards”. APRA’s explanatory statement is explicit about which instrument gave those words content: “GPS 410 was the prudential standard that set out those procedural requirements” — when scheme documents go to APRA, how the notice of intention is published and what it must say, the minimum content of the policyholder summary, when the application can be made, and what is lodged after completion. Remove GPS 410 and the phrase “in accordance with the prudential standards” points at nothing.

Eight confirmations, fourteen schemes, one missing rulebook#

The IAG matter was not an outlier. A full-text sweep of the Federal Court’s judgments database found eight matters in the 562-day window in which the Court published reasons confirming Insurance Act transfers — fourteen schemes in all, from the Chubb–ACE merger integration in October 2016 to the Progressive–Hollard transfer ordered in November 2017. The full text of all eight judgments was captured for this analysis and is checked by the committed verification script: every one cites GPS 410 as an operative standard, and none contains the word “sunset” or “repeal”. The ACE reasons use the same present-tense formula as the IAG reasons (“The relevant prudential standard … is Prudential Standard GPS 410”); the Great Lakes reasons record APRA giving approvals “under paragraph 9 in Prudential Standard GPS 410” after dispensation orders made on 14 December 2016, mid-lapse. Among the judgments the sweep returned, the first reasons to note the standard’s status came four days after the remake: in Sunderland Marine, decided 20 April 2018, Lee J records the notice requirements of “APRA Prudential Standard GPS 410 (which took effect from 16 April 2018)” — accurate, and silent on what preceded it.

Inside the IAG reasons, the lapsed instrument does real work. The Court finds the schemes and actuarial report were given to APRA “in accordance with GPS 410 para 5” (on 26 September 2016 — days before the sunset, when the standard was still law). It finds the notice of intention was “published in accordance with paras 9 and 10 of GPS 410 on 11 March 2017, 18 March 2017, 5 April 2017 and 8 April 2017” — four dates on which those paragraphs were not in force. It records that “as required by para 11 of GPS 410” the notice preceded public inspection on 8 May 2017, and a heading cites “para 16 of the Prudential Standard GPS 410” for the inspection requirement. Regulator and court speak the same way: APRA appeared at the hearing, and its counsel’s submission, recorded at paragraph 114, told the Court APRA “is satisfied there has been material compliance with the procedural requirements under the Act and the standards and with the court’s dispensation orders”. The submission speaks of “the standards” generally and does not name GPS 410 — APRA’s other general-insurance prudential standards were in force throughout — but the Court’s own compliance findings at paragraphs 54 to 60 name GPS 410 by paragraph number. At paragraph 116 the judgment places “substantial weight” on APRA’s non-objection to the schemes; it does not say what weight it gave the compliance submission. And in the AXIS–Swiss Re matter, the reasons state that “GPS 410 at [8] requires an insurer to secure APRA’s approval of the summary of the scheme before publishing a notice of intention”, and record that “APRA approved the scheme summary and notice of intention on 7 December 2016” — sixty-seven days into the lapse, a step whose stated requirement came from the lapsed instrument. (The Act itself, at s 17C(1), makes APRA the approver of scheme summaries; what GPS 410 supplied was the requirement to secure approval before publication.)

Section 17E(2) — the provision requiring the application itself to be made in accordance with the prudential standards — is not addressed anywhere in the IAG reasons; the section appears only through s 17E(3), APRA’s right to be heard. Whether s 17C(2)‘s preconditions were unsatisfiable without a standard, or vacuously satisfied because there was nothing to comply with, is a question neither set of published reasons reaches. Whether it was raised in submissions or argument is not something the published reasons record.

What policyholders actually got#

The counterweight, and it is substantial: nothing in the two published judgments, APRA’s explanatory statement or the other records searched for this analysis indicates any policyholder received less because the standard was missing. The notification that actually reached people flowed from the Court’s own orders. On 17 October 2016 — sixteen days into the lapse — Gleeson J dispensed with the requirement that an approved summary go to every affected policyholder and instead ordered an extensive notification program: emails, letters, newspaper notices in eleven mastheads on three occasions, a Gazette notice, webpages, social-media posts, a call centre.

The delivery numbers are large: 739,394 notification emails, 100,729 posted notices to one brand’s policyholders, scheme information inserted into 1,852,051 pieces of routine correspondence, 10,743 notifications to brokers. The response numbers are close to zero: 7,276 visits to the dedicated webpage between 1 November 2016 and 14 July 2017, six requests for the scheme summary (four dispatched; two requesters had not supplied an address by the hearing), no attendees at any of the eight public inspection sites, one foreshadowed objection never pursued. The Court reviewed claimed deviations from its orders and found substantial compliance, fortified by APRA’s own review of the program.

So the practical protection machinery ran — under court orders and APRA supervision, at scale, to a measured response close to zero. What did not run was the thing the statute actually names: a prudential standard in force.

Did it matter, in law?#

The case law the IAG judgment itself canvasses points both ways, and the piece states it plainly. In Re Armstrong Jones (1997), Emmett J took the view that the equivalent notice precondition in the Life Insurance Act was not a condition precedent to the Court’s jurisdiction; Katz J followed that approach in Royal & Sun Alliance (2000) even where a notice regulation had not been complied with; and in Re Insurance Australia (2004), Lindgren J — with some doubt — applied the same reasoning to s 17C(2)(a) itself, confirming a scheme in a case where the actuarial reports had not been given to APRA in accordance with the prudential standards. On that line of authority, a defect in these preconditions does not destroy the Court’s power to confirm. The transfers stand; nothing in this piece suggests otherwise.

But the authorities describe a court knowingly weighing a compliance defect. That is not what the record here shows. The reasons in eight matters make affirmative findings of compliance with a repealed instrument, in the present tense, without the repeal being mentioned — which is a statement about the text of the judgments, not about what anyone knew. APRA’s explanatory statement gives two reasons for skipping external consultation on the remake: GPS 410 was “being remade unchanged”, and this was “also in part due to the urgency of reintroducing a procedural framework for the transfers of insurance business”. The statement does not say what APRA considered the legal effect of the gap to be. The remake operates prospectively only — “This instrument commences upon registration on the Federal Register of Legislation” — and neither the instrument nor its explanatory statement contains any provision validating, or any statement addressing, what was done in the gap.

The review that described the outcome, without the case#

The timing of the oversight system’s own self-examination is the sharpest part of the record. In September 2017 — month twelve of the lapse — a statutory review of the sunsetting framework, required by section 60 of the Legislation Act, was completed; it was tabled in the House on 23 October 2017 and the Senate on 13 November 2017, while GPS 410 was still missing. Its 78 pages contain zero occurrences of “APRA”, “GPS 410” or “prudential” — searched in full for this piece, across text, footnotes and appendices. APRA is not among its 25 listed submitters.

The report nonetheless acknowledges, in a note to a statistics table in Appendix E, that instruments can sunset without ever appearing on a tabled list: its counts “include instruments that were not listed for sunsetting because of changes in their status advised after list preparation”. That is the outcome GPS 410 met, though not by the mechanism the note describes — the note speaks of a status change advised after a list was prepared, where on APRA’s account GPS 410’s wrong status had been sitting on the register for years before any list for the October 2016 sunset date would have been drawn up. The report’s chapter on the register itself reports that “submissions generally provided positive feedback on the operation and utility of the FRL”. And its Recommendation 22 asks the Office of Parliamentary Counsel to “consider mechanisms for generating sunsetting alerts through the FRL in addition to the tabling of sunsetting lists”. An FRL-generated alert would draw on the same register status; the report does not discuss what happens when that status is wrong.

Whether the review’s authors knew of the GPS 410 lapse is not something the public record answers, and this piece makes no claim about it either way.

What the pre-registration said, and what held#

Six hypotheses were registered before the analysis script was written; the script re-verifies each against the committed documents and refuses to emit chart data if any fails. All six held: the 562-day arithmetic; the eight in-window confirmations citing GPS 410; the IAG judgment’s zero mentions of “sunset”, “repeal”, “17E(2)” and “17C(5)”; the point-in-time statutory text; the sunsetting review’s zero mentions of APRA, GPS 410 and “prudential” alongside its Appendix E note; and the verbatim presence in the judgment of every notification and uptake figure used here.

One strengthening of the evidence base happened after registration and is recorded plainly: as registered, the claim about the seven non-IAG judgments rested on a full-text sweep of the judgments database. During pre-publication verification, the desk captured the full text of all seven (plus the IAG dispensation judgment and the post-remake Sunderland Marine judgment) and extended the script to search each directly — every one cites GPS 410, and “sunset” and “repeal” appear in none of them. The claim published here is the strengthened, directly-checked version.

What could explain this instead#

The lapse may have had no operative legal effect. On the Re Armstrong line of authority, compliance with the s 17C(2) preconditions is not a condition precedent to the Court’s jurisdiction — a court that had been told the standard was gone might well have confirmed the same schemes anyway, treating the lapse as a formal defect. If so, the hole changed no outcome, and the finding is about the record’s silence rather than any decision that would have gone differently.

The substance of the protection was delivered regardless. The mechanisms the standard existed to guarantee — documents to APRA, published notice, public inspection, and a court-designed substitute for the policyholder summary — happened, under the Court’s orders and APRA’s active supervision, on a scale the judgment documents in detail. (The s 17C(2)(c) summary-to-every-policyholder requirement itself was dispensed with, so on that limb the notification program replaced the mechanism rather than delivering it.) On this reading the standard was redundant in the one period it was missing, because the court orders and the regulator reproduced its content.

The key steps predated the sunset. For the IAG schemes, the documents that s 17C(2)(a) requires were given to APRA on 26 September 2016, and APRA approved the notice and summary on 28 September 2016 — while GPS 410 was still in force. Only the later steps (publication, inspection, the application and confirmation itself) fell inside the gap, so the strongest compliance findings partly describe acts done under a live instrument.

The silence may be an artifact of what judgments record. Reasons address what is contested. No party had an interest in raising the instrument’s status — the applicant wanted confirmation, APRA supported the schemes, and no policyholder appeared. An uncontested, apparently satisfied precondition may attract a single sentence and no scrutiny; the absence of any mention of the sunset is evidence about the reasons’ text, not proof of what anyone knew or would have decided.

The framework never relied on the register alone. The 2017 review records that sunsetting lists “are prepared in consultation with portfolios”, and reminds agencies that tabling “does not replace the responsibility of agencies to monitor the sunsetting dates of their own legislative instruments”. On that description the lapse is not purely a register-metadata failure: the portfolio was consulted on list preparation, and the instrument’s own agency carried an independent duty to track it. APRA’s explanatory statement identifies the register label as the cause and does not address either point.

What this is worth#

The finding is institutional, and it is narrow. On APRA’s account, a prudential standard that statutory preconditions depend on ceased to exist for 562 days after a register label recorded it as already repealed, and the section 52 lists — which could not list an instrument the register recorded as already dead — did not catch it. The Federal Court and the regulator continued to apply the vanished instrument by name, in eight published sets of reasons covering fourteen schemes and, in one matter alone, seven insurers with roughly 3.2 million active policies on issue at the last count the judgment records. The framework’s own contemporaneous review acknowledged that instruments can sunset without reaching a list, and named no case. No policyholder harm is evidenced, no transfer is impugned, and GPS 410 is in force today — remade in 2018, and replaced again by a further determination that commenced on 1 January 2026. Why none of those checks surfaced the error is not something the published record answers. What the episode shows is quieter: the machinery that keeps delegated legislation alive runs on register metadata, and for as long as one label was wrong, the standard it described was invisible to the process built to save it.

Comment was not sought from APRA, the Attorney-General’s Department, the Office of Parliamentary Counsel, the Federal Court or IAG; the desk publishes on the public record and states that plainly. Each of those institutions may hold records bearing on what was known and when — APRA on when it discovered the lapse; OPC on the register correction; the parties on what was before the Court beyond the published reasons. If any record provided by those institutions materially changes the picture reported here, a dated correction will be published via the corrections page.