Prevo

Singapore · Cooling measure

Singapore raises property-loan stress-test floors, cuts the HDB housing-loan LTV limit to 80%, and introduces a temporary 15-month HDB resale wait-out

MAS, MND and HDB jointly announced four instruments on 29 September 2022, effective 30 September: financial-institution medium-term rate floors raised 3.5% to 4.0% residential and 4.5% to 5.0% non-residential; a NEW HDB loan-eligibility floor at the higher of 3.0% or CPF OA plus 0.1pp; the HDB housing-loan LTV cut 85% to 80%; and a temporary 15-month wait-out for private-property owners and former owners buying non-subsidised HDB resale flats.

HIGH IMPORTANCEPRIMARY SOURCE CONFIRMED

Announced 29 September 2022 · Effective 30 September 2022

This article has been corrected

5 material corrections have been made. The full record is at the end of this article.

The 30-second brief

On 29 September 2022 Singapore announced four instruments effective the next day. Three tightened credit: financial institutions' stress-test floors rose to 4.0% residential and 5.0% non-residential, HDB created a new loan-eligibility floor, and the HDB loan LTV fell to 80%. The fourth was different in kind. Private-property owners and former owners now had to sell first, wait 15 months, and only then buy a non-subsidised HDB resale flat, reversing a sequence that had previously let them buy first and sell within six months. Official figures show that targeted class fell from roughly 34% to 12% of million-dollar HDB resale buyers. But HDB prices rose 6.2% over the following year and 9.7% in 2024, private volume fell four times harder than the HDB volume the package principally targeted, and displaced households were pushed into a rental market rising 17 to 23 per cent. The measure demonstrably changed who bought without cooling the market it aimed at.

Key numbers

What happened

MAS, MND and HDB jointly announced four instruments on 29 September 2022, effective 30 September: financial-institution medium-term rate floors raised 3.5% to 4.0% residential and 4.5% to 5.0% non-residential; a NEW HDB loan-eligibility floor at the higher of 3.0% or CPF OA plus 0.1pp; the HDB housing-loan LTV cut 85% to 80%; and a temporary 15-month wait-out for private-property owners and former owners buying non-subsidised HDB resale flats. TDSR stayed at 55% and MSR at 30%.

What changed

FI MEDIUM TERM RATE FLOOR

purchase or mwlresidential

3.5% → 4%

purchase or mwlnon residential

4.5% → 5%

HDB LOAN ELIGIBILITY FLOOR

fixed component

n/a → 3%

formula componentcpf oa rate plus 0.1pp

n/a → n/a

HDB LTV

max ltv hdb granted loans

85% → 80%

FI LTV

headline max no outstanding housing loan

75% → 75%

TDSR

threshold

55% → 55%

MSR

threshold

30% → 30%

WAIT OUT

PPO AND EX PPOnon subsidised hdb resale

0 months → 15 months

PPO AND EX PPOsubsidised housing

30 months → 30 months

The HDB loan-eligibility floor is a NEW instrument and a formula, not a raised number: the higher of 3.0% or CPF OA rate plus 0.1pp. It is an assessment floor, not the rate charged; the HDB concessionary rate remained 2.6% for Q4 2022. The FI residential floor was raised and the HDB eligibility floor was created: two floors, different lenders, one raised and one created, one a fixed number and one a formula, neither changing the rate charged. Rows with from equal to to record parameters the package explicitly did NOT change, stored so that a reader cannot infer a change from silence. The wait-out row uses months, not per cent; unit_override marks it. The previous route was purchase first then dispose within six months, so from equal to 0 records the absence of a prior wait, not a prior wait of zero length in an otherwise identical sequence.

Why it matters

Mortgage-based measures cannot directly constrain a purchase completed without mortgage financing. The 15-month eligibility bar could reach private-property right-sizers even where they held sufficient cash or CPF resources and were not constrained by LTV, TDSR or MSR. This is the archive's clearest case of an instrument that reaches buyers credit and tax instruments cannot, and its clearest case of a package whose effect on composition is demonstrable while its effect on the market is not.

We've seen this before

5 July 2018 · 6 July 2018

Singapore raises ABSD by 5 to 10 points and tightens LTV limitsSTRONG RELEVANCE

Why relevant

Precedent for multi-instrument packages that combine leverage tightening with a buyer-class-targeted instrument, and for reading a package's effect against the market it actually targeted rather than the market that moved most.

Where the comparison breaks

  • July 2018 targeted the private market; September 2022 directed three of four instruments at HDB
  • July 2018's targeted instrument was a tax, which reaches only financed and unfinanced buyers through price; September 2022's was an eligibility bar, which reaches cash buyers a tax cannot
  • July 2018's absorption finding is a supply story; September 2022's cross-market finding is a rate-cycle and supply-shortage story
  • July 2018 had a same-evening pull-forward; September 2022 had one day of notice and no comparable rush is recorded

What happened after

July 2018 established that a volume ratio can move for supply reasons rather than demand reasons. September 2022 needs the same discipline in the opposite direction: private volume fell four times harder than the HDB volume the package principally targeted, and the explanation is the rate cycle rather than the package.

26 April 2023 · 27 April 2023

Singapore raises ABSD, doubling foreign-buyer rate from 30% to 60%MODERATE

Why relevant

Forward direction. The April 2023 ABSD package is the next round in the sequence and the reason cross-market analysis of September 2022 is not valid beyond 26 April 2023: from 27 April 2023 the private market was under a materially changed regime while the HDB market was not.

Where the comparison breaks

  • September 2022 was overwhelmingly HDB-directed; April 2023 was a private-market buyer-class tax escalation
  • September 2022 combined credit tightening with an eligibility bar; April 2023 was primarily a single-instrument tax change
  • April 2023 truncates September 2022's cross-market window, which is a methodological relationship rather than an analytical similarity

What happened after

April 2023 marks the boundary beyond which September 2022's HDB-versus-private comparison measures two policy changes rather than one. Its own outcome profile, composition change without aggregate price decline, is the closer analogue for September 2022's buyer-composition finding than either event's price outcome.

Not a precedent for: Framework introductions; Prohibitive foreign-buyer taxation; Private-market-only packages; Measures with a long announcement-to-effective gap; Any analysis treating flat size as a proxy for buyer type

What happened next

Q4 2022 - Q3 2023 against Q4 2021 - Q3 2022

HDB resale market, first four calendar quarters containing the post-event period

CALENDAR · Q4 2022 - Q3 2023 against Q4 2021 - Q3 2022HDB resale, national

Transaction count 27,885 to 25,755, a fall of 7.6%, on symmetric FLOW windows. Transaction-level median resale price 515,000 to 545,000, a rise of 5.8%, uncontrolled for flat mix. HDB Resale Price Index 168.1 to 178.5, a rise of 6.19%, mix-controlled, point to point Q3 2022 to Q3 2023. Every flat type rose more than the aggregate median: 2-room 10.3%, 3-room 7.8%, 4-room 7.6%, 5-room 6.1%, executive 7.8%, against a headline 5.8%.

Interpretation

The by-type pattern is the signature of an adverse mix shift: larger flats fell hardest by volume, dragging the uncontrolled median below true like-for-like growth. The mix-controlled RPI at 6.19% against the median's 5.83% confirms the direction, but the gap is modest at 0.36 percentage points, smaller than the by-type spread implies. State the RPI as the price outcome; the median understates. DO NOT equate the RPI with the transaction-level median: the RPI is a hedonic index controlling for flat mix and the median is not. Rule established during Data Load 1.

Causality: MEDIUM-LOWDecember 2021 measures still working through the marketglobal rate-hiking cycle raising mortgage rates independently of the floor changefour instruments introduced simultaneouslyCOVID-related BTO construction delays pushing demand into the resale marketconstruction-delay recovery and BTO supply ramp-upflat-type and town composition moving the transaction-level median independently of like-for-like prices
Why this grade

The file grades this outcome's two halves separately and the schema holds one grade per row. Volume decline: LOW-MEDIUM, stored here as MEDIUM_LOW. Four instruments commenced simultaneously; the rate cycle and BTO supply were also moving; the seasonal anomaly at Q4 2022 is consistent with the package but does not isolate it. Prices: LOW. Prices rose 6.19% over the window and continued rising for three years, reaching 9.70% in 2024, so no price-suppression effect is demonstrable.

Million-dollar HDB resale transactions

CALENDAR · Q4 2022 - Q3 2023 against Q4 2021 - Q3 2022HDB resale at or above S$1 million, national

Count 362 to 428, a rise of 18.2%. Share of all resale transactions 1.30% to 1.66%. Median within the subset 1,088,000 to 1,088,000, unchanged. Monthly counts across the anchor ran 45 in September 2022, then 38, 26 and 28 through Q4 2022.

Interpretation

High-value activity continued growing after the package. The unchanged subset median is the diagnostic: the high end did not pull away, more sales simply crossed a fixed line.

Causality: NOT A POLICY EFFECTthe S$1 million threshold is nominal and not inflation-adjustedcompositional change as newer, larger or better-located flats enter the resale stockthe overall median rose 5.8% over the same window
Why this grade

The count rose 18.2%. With the overall median up 5.8% against a fixed nominal threshold, and with the subset median unchanged, the rise is substantially mechanical. A rise in the count is not by itself evidence of anything.

HDB resale volume by flat type

CALENDAR · Q4 2022 - Q3 2023 against Q4 2021 - Q3 2022HDB resale by flat type, national

2-room up 48.7%, 3-room down 1.7%, 4-room down 3.9%, 5-room down 17.3%, executive down 24.8%, multi-generation down 57.1%, on symmetric FLOW windows. Larger flats bore almost the entire volume decline.

Interpretation

Suggestive of right-sizer displacement and diagnostic of nothing. The senior exemption required age 55 or above AND right-sizing to a 4-room or smaller flat, and non-senior private-property owners remained subject to the wait-out across all non-subsidised resale flat types. Flat size does not identify buyer status.

Causality: LOWflat size does not identify buyer typethe nominal S$1 million threshold interacts differently with each flat typeflat-type composition of the resale stock changed over the windowBTO supply and MOP flow altered the mix of flats reaching resale
Why this grade

THIS IS NOT A CAUSAL TEST. Flat size does not identify buyer status. A compositional pattern within these figures was examined and rejected because it failed the consistency test across adjacent flat types; see the rejected_inferences row on this event.

Private residential rental volume (the volume half of the canonical file's O7)

CALENDAR · Q4 2022 - Q3 2023 against Q4 2021 - Q3 2022Private residential rental contracts, island-wide

Why this grade

URA exposes rental contract records through a per-month interactive search interface returning contract-level rows, not as a published historical volume series. No free historical series exists. This is a verified absence, not a failed search.

1 January to 29 September 2022

Buyer composition among million-dollar HDB resale buyers

CALENDAR · 1 January to 29 September 2022Buyers of HDB resale flats sold at S$1 million or more, national

Private-property owners and former owners represented approximately 34% of buyers of HDB resale flats sold for S$1 million or more between 1 January and 29 September 2022.

Interpretation

This is the event's strongest mechanism evidence and it is not derivable from the loaded transaction dataset, which records no buyer attributes. It supports a buyer-composition effect consistent with the intended mechanism. It does NOT prove that affected buyers were paying entirely in cash, that affected demand disappeared rather than being delayed or diverted, that the measure caused overall HDB prices to fall, that it reduced the total number of million-dollar transactions, or what would have happened without it.

Causality: MEDIUMthe two reference windows are different lengths and neither is a calendar yeardifferent positions in the rate cyclemortgage rates rose sharply between the two observationsprivate-property values, and therefore right-sizers' available proceeds, changed over the interval
Why this grade

The wait-out targeted exactly this buyer class and no other instrument in the package reaches cash buyers. Capped at MEDIUM: this figure is a pre-measure baseline rather than an effect, the counterfactual is unobserved, and the comparison that gives it meaning is outcome 12, which rests on weaker sourcing and a non-comparable window.

calendar 2023, 2024, 2025 and 2026 to date

HDB Resale Price Index, annual trajectory after implementation

CALENDAR · calendar 2023, 2024, 2025 and 2026 to dateHDB resale, national

2023 up 4.94% (171.9 to 180.4). 2024 up 9.70% (180.4 to 197.9). 2025 up 2.88% (197.9 to 203.6). 2026 to date down 0.39% (203.6 to 202.8), a half year, Q4 2025 to Q2 2026.

Interpretation

HDB resale prices continued rising for three full years after implementation, with 2024 the strongest year of the period at 9.70%, stronger than 2022 itself. The first negative reading is a half-year 2026 figure, well past the point where attribution to this package is void. Any narrative in which the September 2022 package cooled HDB prices is contradicted by the 2024 figure alone. Do not manufacture a price-decline story.

Causality: LOWinterest rates through the periodBTO supply ramp-up and delayed-project completionsMOP flow expanding resale supplythe April 2023 ABSD packageother HDB eligibility, grant and supply policiesremoval of the wait-out effective 28 July 2026
Why this grade

Prices rose through the entire attributable window and accelerated two years after implementation. No price-suppression effect is demonstrable. The 2025 and 2026 figures lie outside any window in which this package can be attributed; they are recorded for trajectory, not for attribution.

Attribution to this package is void from 27 April 2023 for cross-market analysis and mechanically altered from approximately 30 December 2023 when the earliest affected cohort completed the wait-out. The 2024 to 2026 figures are recorded as trajectory only. See the outcome_windows ladder on this event.

Q4 2022 - Q3 2023 against Q4 2021 - Q3 2022; prices point to point Q3 2022 to Q3 2023

Private residential market, prices and volume

CALENDAR · Q4 2022 - Q3 2023 against Q4 2021 - Q3 2022; prices point to point Q3 2022 to Q3 2023Private residential, island-wide and by market segment

Prices, point to point Q3 2022 to Q3 2023: overall PPI 187.8 to 196.0, up 4.37%; CCR non-landed 144.4 to 142.5, down 1.32%; RCR non-landed 198.7 to 213.0, up 7.20%; OCR non-landed 224.0 to 237.4, up 5.98%. Volume, island-wide, symmetric FLOW windows: resale 17,723 to 12,011, down 32.2%; all types of sale 29,162 to 19,245, down 34.0%; new sale of uncompleted units 10,537 to 5,986, down 43.2%; sub-sale 649 to 974, up 50.1%.

Interpretation

New sales fell hardest at 43.2%, the segment most exposed to the raised financial-institution stress-test floor. Consistent with the credit instruments biting, though the rate cycle confounds it identically and cannot be separated. Sub-sales were the only growing private category, up 50.1% from a base of 649. A sub-sale is a buyer offloading an uncompleted unit before completion, so growth here while new sales collapsed is the classic signature of buyers who committed under one financing regime and could not complete under another. CCR was the only declining price segment; see the CCR pattern note, which treats this as structural rather than as an effect of this package.

Causality: LOWglobal rate-hiking cyclethe April 2023 ABSD package altered the private market from 27 April 2023December 2021 measures still working through the marketlaunch timing and project mixstructural CCR underperformance from causes independent of this package
Why this grade

The file grades these separately and the schema holds one grade per row. Private market effects, volume down 32.2% and new sales down 43.2%: LOW for this package, because the rate cycle is the dominant candidate and new-sale exposure to the raised stress-test floor cannot be separated from market rates rising through the same period. Sub-sale growth: LOW-MEDIUM, directionally consistent with buyers unable to complete under tightened financing, which includes the stress-test floor change.

Cross-market comparison, HDB against private

CALENDAR · Q4 2022 - Q3 2023 against Q4 2021 - Q3 2022; prices point to point Q3 2022 to Q3 2023HDB resale and private residential, national

HDB resale price up 6.19% and volume down 7.6%. Private resale price up 4.37% and volume down 32.2%. Private volume fell roughly four times as hard as HDB volume while HDB prices grew faster, despite three of the package's four instruments targeting HDB.

Interpretation

The package's HDB-targeted instruments coincided with the mildest volume decline in the market they targeted. That is not what the instrument design predicts. The better explanation is that the rate cycle hit the private market hardest, because private buyers are far more mortgage-dependent than HDB buyers, many of whom are cash-and-CPF upgraders, while the HDB market faced structural supply shortage from BTO construction delays that supported both prices and transaction resilience. Two different forces, one window.

Causality: INCONCLUSIVEglobal rate-hiking cycle, which bears far harder on the more mortgage-dependent private marketBTO construction delays creating structural HDB supply shortagefour instruments introduced simultaneouslythe two markets have different buyer financing mixes, so identical policy would not produce identical effects
Why this grade

The comparison is the strongest single argument in the event and it resolves nothing about this package. Two independent forces, the rate cycle and BTO supply shortage, are each sufficient to produce the observed asymmetry, and the package's own instruments cannot be cleanly isolated from either.

point to point Q3 2022 to Q3 2023

Private residential rental index

CALENDAR · point to point Q3 2022 to Q3 2023Private residential rental, island-wide and by market segment

Non-landed whole island 139.5 to 163.3, up 17.06%. Non-landed CCR 133.4 to 152.7, up 14.47%. Non-landed RCR 145.4 to 172.2, up 18.43%. Non-landed OCR 144.3 to 172.6, up 19.61%. Landed whole island 121.1 to 164.2, up 35.59%.

Interpretation

The landed figure is an outlier against a 14 to 20 per cent range across every other series and is a thin, heterogeneous segment. Recorded, not claimed, on the same basis as the CCR noise rule. Nothing in this event turns on it. The All-Residential rental index could not be used: URA discontinued it after Q2 2021, and the compute layer refused the anchor as outside coverage rather than returning a truncated window. Non-landed whole island is the aggregate figure for this event.

Causality: NOT A POLICY EFFECTCOVID construction delays and delayed completionsreturning expatriate, student and foreign worker demand after border reopeningthe same shock appears in the HDB rental market, which the package's instruments do not reach
Why this grade

A market-wide rental shock with identified independent causes: construction delays, delayed completions, and returning expatriate, student and foreign worker demand after border reopening. Rents rose two-and-a-half to four times faster than prices in both tenures simultaneously. No plausible reading attributes this to the September 2022 package.

Q3 2022 to Q3 2023

HDB rental market, median rent

CALENDAR · Q3 2022 to Q3 2023HDB rental, 78 town and flat-type cells with data in both quarters

Aggregate up 23.5%. 3-room up 22.7%. 4-room up 23.6%. 5-room up 24.6%. Coverage is 3-room through 5-room only: 1-room, 2-room and executive return no towns with data in both quarters.

Interpretation

HDB suppresses cells with too few contracts to publish, with 2,217 suppressed 1-room cells, 2,107 2-room and 1,412 executive. The absence at both ends of the size range is itself information: those cells are too thin to publish, not missing by accident. Tight clustering across flat types, minimum 1.8% and maximum 40.0% with a narrow central mass, indicates a market-wide movement rather than a segment effect.

Causality: NOT A POLICY EFFECTCOVID construction delays and delayed completionsreturning expatriate, student and foreign worker demand after border reopeningHDB suppresses thin cells, so coverage is not a random sample of the rental stock
Why this grade

The same market-wide rental shock recorded in outcome 7, appearing in a tenure the package's instruments do not reach. This outcome matters to the event because the wait-out required disposal before purchase, so affected households had to rent through the 15-month period into this market. That interaction is an observation about the instrument's design meeting contemporaneous conditions, not a measured outcome of the instrument.

Q4 2022

Seasonal anomaly at the anchor quarter

CALENDAR · Q4 2022HDB resale, national

HDB resale Q4 volume as a percentage of September volume times three: 2019 108.6%, 2020 97.2%, 2021 99.2%, 2022 81.7%, 2023 106.7%, 2024 93.5%. Q4 2022 is the outlier against a five-year seasonal baseline, roughly 15 points below the weakest comparison year.

Interpretation

The honest statement is that Q4 2022 is anomalous against a five-year seasonal baseline, not that the package caused it.

Causality: INCONCLUSIVEHDB transaction dates are approval or registration dates and lag the sale by weeksthe rate cycle was biting hardest in the same quartera single year against a five-year baseline is a small sample
Why this grade

HDB transaction dates are approval or registration dates lagging the sale by weeks, and September 2022 is almost entirely pre-announcement since the measure landed on the 29th, so the anchor is smeared across the quarter boundary. A 15-point deviation in one year is consistent with the package and equally consistent with the rate cycle, which was biting hardest in the same quarter.

January to November 2024

Buyer composition, the January to November 2024 observation (split from outcome 3)

CALENDAR · January to November 2024Buyers of HDB resale flats sold at S$1 million or more, national

Private-property owners and former owners represented approximately 12% of buyers of HDB resale flats sold for S$1 million or more between January and November 2024.

Interpretation

SPLIT FROM OUTCOME 3 ON THE FOUNDER PASS, 13 August 2026, and the reason is the finding. Outcome 3 cited the February 2024 MND parliamentary answer for BOTH figures. That document CANNOT support this one: the measurement window ran to November 2024 and the document was published in February 2024. A source cannot report a measurement that had not finished when it went to press. The two figures are separately sourced, of different strength, and are now separate records so that the weaker one cannot borrow the stronger one's provenance.

Causality: MEDIUMthe two reference windows are different lengths and neither is a calendar yeardifferent positions in the rate cyclemortgage rates rose sharply between the two observationsprivate-property values, and therefore right-sizers' available proceeds, changed over the intervalby January 2024 the earliest affected cohort had already completed the 15-month wait-out, so this window is mechanically altered
Why this grade

Directionally consistent with the wait-out having changed buyer composition, and held at MEDIUM for the same reasons as outcome 3 plus one more: the sourcing here is secondary. Multiple credible reports of a statement by the Minister for National Development, not a primary document the archive has read.

The case for

The buyer-composition effect is real and officially evidenced: private-property owners and former owners fell from approximately 34% of million-dollar HDB resale buyers in the period to 29 September 2022 to approximately 12% in January to November 2024. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The two halves are SEPARATELY SOURCED and of different strength: the 34% rests on the February 2024 MND parliamentary answer, the 12% on multiple credible reports of a ministerial statement, because the parliamentary answer predates the end of the 2024 measurement window and cannot contain it. The wait-out targeted exactly that class and no other instrument in the package reaches cash buyers. Sub-sales grew 50.1% from a base of 649 while new sales collapsed 43.2%, the signature of buyers who committed under one financing regime and could not complete under another.

The case against

HDB resale prices rose 6.19% over the outcome window and kept rising for three years, accelerating to 9.70% in 2024. Million-dollar transactions rose 18.2%, substantially mechanically given a fixed nominal threshold and an unchanged subset median. Private volume fell four times harder than HDB volume despite three of four instruments targeting HDB, which is not what the instrument design predicts. The two official buyer-composition windows are not like-for-like, and the rate cycle confounds every credit finding identically.

What this view assumes

  • the RPI and the transaction-level median are different measures and must never be equated: the RPI controls for flat mix and the median does not
  • the S$1 million threshold is nominal and not inflation-adjusted, so a rising count is not by itself evidence
  • flat size does not identify buyer status
  • quarterly aggregations are labelled as calendar quarters containing the post-event period, never as elapsed windows

What we don't know

  • whether affected households actually rented through the wait-out, which requires buyer-level data the public datasets do not contain
  • how much of the 34% to 12% fall is the wait-out and how much is the credit instruments, rates, or changed private-property proceeds
  • whether displaced demand disappeared, was delayed, or was diverted
  • the counterfactual path of HDB prices and volumes without the package
  • the 12% figure rests on reporting of a ministerial statement rather than a primary document the archive has read; the February 2024 parliamentary answer cannot support it because it was published before that measurement window closed

Prevo view

Interpretation

The September 2022 package combined credit tightening with a targeted eligibility restriction. The higher assessment floors and lower HDB LTV reduced borrowing capacity for financed purchases. The temporary 15-month wait-out addressed a different problem: private-property owners and former owners who might possess sufficient cash or CPF resources to remain active despite mortgage restrictions. Later official evidence indicates the share of million-dollar HDB buyers who were private-property owners or former owners fell substantially after implementation, while million-dollar HDB transactions continued to increase. The defensible conclusion is that the wait-out changed the COMPOSITION and TIMING of demand from private-property right-sizers. It is not defensible, without additional evidence, to conclude that it stopped the growth of million-dollar HDB transactions or caused HDB resale prices to decline. Beyond that, the package's design imposed a cost the rule's face does not convey: because disposal had to precede purchase, affected households had to rent through the 15-month period, and they did so into a rental market rising 17 to 23 per cent in both tenures. That interaction converts a structural observation about the instrument into one with a measured magnitude, while remaining an observation about design meeting contemporaneous conditions rather than a measured outcome of the instrument itself.

Confidence: MEDIUM

What would change this view: If official buyer-profile data showed no decline in private-property owner and former owner participation after implementation, the buyer-composition mechanism weakens materially. If participation declined but had already been falling before 30 September 2022, attribution to the wait-out weakens. If the reduction occurred only after mortgage rates rose materially further, a credit or rate-cycle explanation is supported over an eligibility explanation. If participation fell while total million-dollar transactions kept rising, the correct conclusion is that composition changed while high-value activity was sustained by other buyers. If 5-room and executive transactions weakened relative to smaller flats, that is consistent with right-sizer displacement but does not prove it, because the dataset does not identify buyer type. If private or HDB rental activity increased after implementation, that is consistent with displaced households renting during the wait-out but does not prove causation. If buyer participation increased immediately after the July 2026 removal, the temporal-eligibility mechanism is strengthened, subject to the changed 2026 environment. A clean causal conclusion would require buyer-level data identifying previous private-property ownership, disposal date, HDB purchase date, financing method, flat type, purchase price, and whether an exemption or waiver applied; the public dataset contains none of these fields.

Evidence check

17

Material claims

17

Primary confirmed

0

Corroborated

13

Qualified

0

Government estimates

View evidence report
#ClaimTypeClassVerificationEvidence location
1The medium-term interest-rate floor used by financial institutions to assess residential property purchase loans and mortgage-equity withdrawal loans was raised from 3.5% to 4.0% per annum.NumberObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022; MAS Notice 645
2The medium-term interest-rate floor for non-residential property purchase loans and mortgage-equity withdrawal loans granted by financial institutions was raised from 4.5% to 5.0% per annum.NumberObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022; MAS Notice 645
3The revised financial-institution floors applied to relevant property purchase loans and mortgage-equity withdrawal loans, and were used when computing Total Debt Servicing Ratio and, where applicable, Mortgage Servicing Ratio.RegulationObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022; MAS Notice 645
4HDB introduced an interest-rate floor for determining the eligible HDB housing-loan amount, set at the higher of 3.0% per annum or 0.1 percentage point above the prevailing CPF Ordinary Account interest rate.RegulationObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022; HDB reproduction
5The actual HDB concessionary interest rate remained 2.6% per annum at implementation, continuing to be set at 0.1 percentage point above the prevailing CPF Ordinary Account rate and subject to quarterly review.NumberPolicy characterisationVERIFIED PRIMARYHDB concessionary rate schedule, Q4 2022
6The maximum loan-to-value limit for housing loans granted by HDB was reduced from 85% to 80%.NumberObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022
7The financial-institution housing-loan loan-to-value framework was unchanged by this package, including the headline 75% maximum for qualifying borrowers without an outstanding housing loan.NumberObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022
8The Total Debt Servicing Ratio threshold remained 55%.NumberObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022
9The Mortgage Servicing Ratio threshold remained 30%.NumberObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022
10A temporary 15-month wait-out period was imposed on private residential property owners and former owners purchasing non-subsidised HDB resale flats, counted from disposal of the private residential property.PolicyObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022; HDB eligibility conditions
11Singapore Citizens aged 55 and above, including their spouses, moving from private property to a 4-room or smaller resale flat were exempt from the 15-month wait-out.PolicyObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022; HDB eligibility conditions
12The wait-out period for private property owners who are first-timers and wish to apply for the CPF Housing Grant and Enhanced CPF Housing Grant for their resale flat purchase remained unchanged at 30 months.PolicyObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022
13HDB retained case-by-case discretion for private-property owners and former owners facing extenuating circumstances, including financial difficulty, regardless of the applicant's age.PolicyObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022
14The package took effect on 30 September 2022.DateObserved factVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022
15The Government stated the package was intended to promote sustainable property-market conditions by ensuring prudent borrowing and moderating demand.Policy characterisationPolicy characterisationVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022, title and opening
16The Government expressly described the 15-month wait-out as temporary, intended to moderate HDB resale demand while broader housing supply and market conditions stabilised.Policy characterisationPolicy characterisationVERIFIED PRIMARYJoint MAS/MND/HDB media release, 29 September 2022
17The 15-month wait-out was removed with effect from 28 July 2026.DateObserved factVERIFIED PRIMARYHDB announcement removing the temporary wait-out

Claim → Evidence → Source

Sources

Primary sources

Corrections

  • 13 August 2026

    Original: The existing 30-month wait-out applicable to private-property owners and former owners seeking subsidised housing remained unchanged.

    Corrected: The wait-out period for private property owners who are first-timers and wish to apply for the CPF Housing Grant and Enhanced CPF Housing Grant for their resale flat purchase remained unchanged at 30 months.

    THE STORED SCOPE WAS BROADER THAN THE RELEASE STATES. The release confines the unchanged 30-month period to PPOs who are FIRST-TIMERS applying for the CPF Housing Grant and Enhanced CPF Housing Grant on a resale purchase. The wider framing may still be correct under HDB's operative eligibility rules, but this release does not establish it and the archive follows the document it cites. evt_sg_2026_hdb_waitout_removal_jul claims 6 and 7 keep the broader framing on different documents; the divergence is recorded on both sides rather than harmonised, because harmonising would mean choosing a scope neither document alone supports. Found on transcribing the implementation annex.

  • 13 August 2026

    Original: Implementation conditions differed by transaction route: purchase of a new flat; purchase through open booking; purchase of a resale flat; taking over ownership of an existing flat. [four routes, annex not transcribed]

    Corrected: Three routes, transcribed verbatim, with the BTO and Sale of Balance Flats row carrying two conjunctive conditions. Plus the footnote-2 commercial-rate population and the senior 2-room Flexi and Community Care Apartment provision, neither previously in the archive.

    THE EVENT EXISTED PARTLY TO WARN AGAINST COLLAPSING THIS TABLE, AND THE ARCHIVE HAD COLLAPSED IT. Transcribing proved the warning right in an unpredicted way: the route COUNT was wrong, because the release groups BTO with Sale of Balance Flats and groups open booking with resale. The archive's four-route reading is a reasonable reading of the policy and is not how the release groups them. Two further populations were missing entirely. A summary can be wrong about a table's CONTENT and its SHAPE at once.

  • 13 August 2026

    Original: Persons aged 55 and above, and their spouses, moving from private property to a 4-room or smaller resale flat were exempt from the 15-month wait-out.

    Corrected: Singapore Citizens aged 55 and above, including their spouses, moving from private property to a 4-room or smaller resale flat were exempt from the 15-month wait-out.

    Later official HDB material identifies the qualifying senior as a SINGAPORE CITIZEN. "Persons" is broader than the eligibility rule and would tell a non-citizen aged 55 that an exemption reached them. Three conditions are required, not two: citizenship, age and right-sizing. Found in the founder primary-source pass.

  • 13 August 2026

    Original: Outcome 3 cited the February 2024 MND parliamentary answer for BOTH the 34% figure (1 January to 29 September 2022) and the 12% figure (January to November 2024).

    Corrected: Outcome 3 keeps the parliamentary answer for the 34%. The 12% is split into outcome 12, sourced to multiple credible reports of a ministerial statement.

    A DOCUMENT CANNOT SUPPORT A FIGURE MEASURED AFTER IT WAS PUBLISHED. The 2024 measurement window ran to November; the document was published in February. This is a new failure class for the archive, and unlike the April 2023 provenance defect it IS structurally checkable, because publication dates and measurement windows are both stored. Assertion A23 now checks it archive-wide. The 34% was never in doubt; what was wrong was that the weaker figure was borrowing the stronger figure's provenance. Found in the founder primary-source pass.

  • 13 August 2026

    Original: Some market chronologies describe this package as Singapore's 13th round of cooling measures.

    Corrected: DROPPED. No replacement claim.

    Dropped as context rather than evidence. Round counts depend on which announcements, refinements and reversals are treated as separate rounds, the underlying measures and dates are the archive's record, and no analysis in this event rests on the ordinal. This claim was the mildest of the three, already sitting at PARTIALLY_VERIFIED and therefore not blocking publication, which is precisely why dropping it was a choice rather than a necessity: the archive should not carry an ordinal it has told itself twice not to rely on. One of three identical drops across evt_sg_2013_tdsr_jun, evt_sg_2022_cooling_sep and evt_sg_2023_absd_apr.

Prevo provides research and informational analysis only. It is not a broker, investment adviser or fiduciary, and nothing on this site constitutes investment, legal, tax or financial advice. Verify independently.