Singapore · Macroprudential
MAS introduces Total Debt Servicing Ratio framework for property loans
MAS introduced the TDSR framework for property loans on 28 June 2013, effective 29 June 2013: a 60% TDSR threshold on total debt servicing against gross monthly income, exceedable only exceptionally and subject to credit-committee approval, a 3.5% specified medium-term residential rate used where higher than the prevailing rate, a haircut of at least 30% on all variable income including bonuses and on rental income, tightened borrower/mortgagor rules, and simultaneous LTV refinements.
Announced 28 June 2013 · Effective 29 June 2013
Sign in to save this event to your research. It is free.
Original rule
as at 29 June 2013
Current position
as at 13 August 2026
TDSR threshold
60% of gross monthly income
Residential stress rate, financial institutions
3.5% specified medium-term rate
Non-residential specified rate
4.5%
Non-residential specified rate
No amendment recorded in this archive
Variable and rental income haircut
At least 30%
Variable and rental income haircut
No amendment recorded in this archive
This article has been corrected
2 material corrections have been made. The full record is at the end of this article.
Key numbers
In brief
Singapore's Total Debt Servicing Ratio framework took effect on 29 June 2013, requiring lenders to test every property borrower against total existing debt, a stressed mortgage rate and discounted variable income. Private price growth stalled within one quarter and turned negative in the next. The standalone causal effect of TDSR is not identified, because seven earlier cooling rounds and an expanding supply pipeline ran alongside it. The framework is still in force, tightened once in December 2021 and stress-tested against a higher floor since September 2022. What it changed was not the cost of buying but the arithmetic of qualifying, which is why it reached buyers that transaction taxes had left alone.
Why it mattersInterpretation
TDSR is the clearest Singapore example of the difference between segment-specific demand control and system-wide credit-capacity control. It changed the question from "How do we make investment property more expensive?" to "How do we stop households from taking on debt that may become unaffordable when rates normalise?" This distinction drives all later comparisons.
What changed
Parameters at introduction
Before: Debt-servicing assessment was not standardised across financial institutions. Lenders applied their own treatment of existing debt, variable income and rate assumptions.
| Applies to | At introduction |
|---|---|
tdsr threshold | 60% |
medium term rateresidential | 3.5% |
medium term ratenon residential | 4.5% |
variable income haircut | 30% |
The 60% figure is a THRESHOLD, not an absolute maximum: loans exceeding it were regarded as imprudent and could be granted exceptionally, subject to the institution approved procedures and credit-committee approval. The 30% variable-income haircut is a regulatory MINIMUM, not a fixed rate, and applies to all variable income including bonuses and to rental income. Framework introduction: no prior standardised values, so from is null. Medium-term rates are used where higher than the prevailing market rate (intentional stress testing, not a rate forecast). Borrower/mortgagor rules and the simultaneous LTV refinements are non-numeric components; see event summary.
Full event recordDates, regulator, scope, every stored claim value, the position before and the current status
Event facts
- Announced
- 28 June 2013
- Effective
- 29 June 2013
- Announcement to effective
- 1 day
- Regulator
- Monetary Authority of Singapore
- Instruments and scope
- All property loans granted by regulated financial institutions to individuals, including residential and non-residential property. Residential-market analysis covers private residential property and HDB purchases financed by financial institutions.
- Total monthly debt against gross income
- 60%[MAS press release 28 June 2013; MAS Notice 645]
- Residential stress rate
- 3.5%Non-residential property loans used a separate specified rate of 4.5%.[MAS Notice 645]
- Claim 4
- 30%Not the 30% Mortgage Servicing Ratio for HDB loans, a separate measure.[MAS Notice 645]
- Before this framework
- Debt-servicing assessment was not standardised across financial institutions. Lenders applied their own treatment of existing debt, variable income and rate assumptions.
- Positioning at introduction
- Permanent structural prudential standard, not a cyclical cooling measure. Stated in the 28 June 2013 release at paragraph 11 and reaffirmed by MAS in 2017.
- Current status
- Amended. TDSR threshold changed 16 December 2021; Residential stress rate, financial institutions changed 30 September 2022.
Market context
We've seen this before
26 April 2023 · 27 April 2023
Why relevant
Reverse direction of the anchor comparison: when a credit-capacity question is analysed, ABSD 2023 shows what a segment-targeted transaction tax does instead (changes who buys and where demand lands, not aggregate credit capacity).
Where the comparison breaks
- Mechanism: credit-capacity constraint vs transaction tax
- Coverage: broad vs highly differentiated by buyer class
- Immediate channel: maximum borrowing capacity vs acquisition cost
- Observed response: broad price weakness vs foreign-participation collapse without broad price correction
- Policy character: structural prudential framework vs demand-management cooling measure
What happened after
ABSD 2023 demonstrates the alternative mechanism's outcome profile: composition change without aggregate price decline.
15 December 2021 · 16 December 2021
Why relevant
Forward direction of the structural-ancestor link: December 2021 is the major later parameter change within the TDSR framework (60% to 55%, 16 December 2021). There was no general TDSR reduction in 2018.
Where the comparison breaks
- Parameter change within an existing framework vs framework introduction
- Multi-instrument package vs single structural framework
- Continued annual appreciation vs sustained multi-year price decline
What happened after
Shows how the framework's parameters were later used cyclically while the framework itself remained structural.
Not a precedent for: ABSD increases; foreign-buyer taxes; SSD changes; isolated LTV adjustments; land-supply changes
What happens next
Q3 2013
Immediate transaction-activity decline
MARKET OUTCOMEOBSERVED OUTCOMECausality: HIGHShow detail
Immediate transaction-activity decline
MARKET OUTCOMEOBSERVED OUTCOMEDeveloper transaction activity weakened materially relative to the high sales rates seen before TDSR. URA PPI rose +0.4% QoQ versus approximately +1.0% in Q2 2013: price growth stalled rapidly; prices did not decline in Q3.
Interpretation
The immediate slowing in credit-sensitive transaction activity is consistent with the effect expected from TDSR, but Q3 price data alone does not establish a price decline.
Why this grade
TDSR directly affected borrowing capacity; it applied broadly across property borrowers using regulated financial institutions; implementation was immediate; transaction activity weakened sharply after introduction; contemporary lenders, buyers, developers and market analysts identified financing capacity as a major constraint. Qualification: activity was also affected by the cumulative impact of earlier cooling measures and changing market expectations.
Q4 2013 - Q1 2014
Initial private price inflection
MARKET OUTCOMEOBSERVED OUTCOMECausality: MEDIUM-HIGHShow detail
Initial private price inflection
MARKET OUTCOMEOBSERVED OUTCOMEURA PPI approximately -0.9% QoQ in Q4 2013, the beginning of the sustained decline in the overall index, followed by approximately -1.3% QoQ in Q1 2014. Full sequence around the event: Q2 2013 +1.0%, Q3 +0.4%, Q4 -0.9%, Q1 2014 -1.3%.
Interpretation
The timing and financing mechanism make TDSR a highly plausible major contributor to the inflection.
Why this grade
The sequence is striking (+1.0 -> effective 29 June -> +0.4 -> -0.9 -> -1.3) and the financing mechanism fits, but a single policy cannot be isolated perfectly from the cumulative effect of earlier cooling measures and rapidly changing expectations.
calendar 2014
Calendar 2014 price and sales declines
MARKET OUTCOMEOBSERVED OUTCOMECausality: MEDIUMShow detail
Calendar 2014 price and sales declines
MARKET OUTCOMEOBSERVED OUTCOMEURA PPI approximately -4.0% for the year. Developer private residential sales excluding ECs: 2012 22,197 units; 2013 14,948; 2014 7,316. 2014 developer sales were approximately 51% below 2013 and approximately 67% below 2012.
Interpretation
TDSR was an important component of the regime that initiated and maintained tighter borrowing conditions through 2014.
Why this grade
The 2014 declines cannot reasonably be attributed to TDSR alone; concurrent factors include the housing completion pipeline, cumulative earlier restrictions and changing demand.
Q4 2013 - Q2 2017
Fifteen-quarter sustained price decline
MARKET OUTCOMEOBSERVED OUTCOMECausality: MEDIUMShow detail
Fifteen-quarter sustained price decline
MARKET OUTCOMEOBSERVED OUTCOMEPrivate residential prices declined for 15 consecutive quarters, Q4 2013 through Q2 2017, returning to quarterly growth in Q3 2017. Cumulative decline approximately 11%-12%, depending on exact comparison points.
Interpretation
TDSR materially contributed to the market turning, while supply and broader market conditions helped determine the depth and duration of the downturn.
Why this grade
TDSR was an important component of the regime that initiated and maintained tighter borrowing conditions, but the duration of the downturn cannot reasonably be attributed to TDSR alone given the concurrent supply and demand factors.
H2 2013 onward
HDB resale downturn, multiple policy exposures
MARKET OUTCOMEOBSERVED OUTCOMECausality: LOWShow detail
HDB resale downturn, multiple policy exposures
MARKET OUTCOMEOBSERVED OUTCOMEHDB resale prices entered a prolonged downturn beginning around this period.
Interpretation
Classified as observed outcome, multiple policy exposures, not direct TDSR outcome.
Why this grade
LOW in isolation: material concurrent policy changes included the additional HDB measures introduced on 27 August 2013 (mortgage servicing capacity, permissible loan tenure, PR eligibility for resale HDB purchases).
The 27 August 2013 HDB measures land inside any reasonable outcome window for this market.
Prevo analysis
Prevo view
InterpretationTDSR was a major structural contributor to the inflection that ended the post-GFC private residential upswing, operating alongside the cumulative effect of seven earlier cooling rounds and an expanding housing supply pipeline.
Confidence: MEDIUM-HIGH
The case for and the case against2
The case for
The inflection sequence is striking and mechanism-consistent: +1.0% (Q2), effective 29 June, +0.4% (Q3), -0.9% (Q4), -1.3% (Q1 2014). Transaction activity weakened immediately, and contemporary lenders, buyers and developers identified financing capacity as the constraint. Causality for the immediate transaction slowdown is HIGH.
The case against
The full 2014-2017 downturn cannot be attributed to TDSR alone: a substantial completion pipeline, rising vacancy, softer rentals, cumulative earlier restrictions and later rate normalisation all contributed. HDB outcomes are LOW-confidence in isolation because of the 27 August 2013 HDB-specific measures.
What this view assumes2
- The 3.5% floor functioned as intentional stress testing rather than a rate forecast
- The commonly used seven-round chronology is a convention; underlying dates are the record
What we don't know2
- TDSR's exact contribution versus earlier cooling rounds cannot be isolated
- The claim that TDSR caused developers to shrink unit sizes is a plausible second-order interpretation, unverified without project-level evidence
Evidence behind this event
5 claims, 5 verified
- Causally established outcomes
- 0
- Interpretive sections, not claim-verifiableWhy it matters, Prevo View, The case for, The case against
- 4
Every claim, by type
Rates, figures and counts3
- Claim 1
MAS established a TDSR threshold of 60%. Property loans exceeding the threshold were regarded as imprudent and could be granted only exceptionally, subject to the financial institution's approved procedures and credit-committee approval.
VERIFIED PRIMARY[MAS press release 28 June 2013; MAS Notice 645]Exceptional cases were permitted subject to institutional approval procedures and credit-committee approval.
- Claim 3
Specified medium-term residential mortgage rate set at 3.5%, with financial institutions required to use the higher of the specified medium-term rate or prevailing market rate when computing debt servicing.
VERIFIED PRIMARY[MAS Notice 645]Non-residential specified medium-term rate: 4.5%. The 3.5% is the rate used where higher than the prevailing rate, not a rate forecast.
- Claim 4
Financial institutions were required to apply a haircut of at least 30% to all variable income, including bonuses, and to rental income.
VERIFIED PRIMARY[MAS Notice 645]Regulatory MINIMUM haircut, not a fixed rate; applies to variable income and rental income.
Dates1
- Claim 2
Framework effective from 29 June 2013.
VERIFIED PRIMARY[MAS press release 28 June 2013]
Characterisations and comparisons1
- Claim 5
MAS regarded TDSR as a structural prudential measure rather than a temporary cyclical property cooling tool.
VERIFIED PRIMARY[MAS press release 28 June 2013, paragraph 11]
How this is scored
Counts are by provenance, meaning who established the claim, not by how confident we are. A policy fact is one the regulator's own document states. A market observation comes from a named data series. A derived calculation is one we computed, with the working recorded on the claim.
Interpretations are counted, never netted out. This page will not display zero unsupported claims while interpretive sections sit outside the claim ledger, because that number would be true only by excluding the material most likely to be wrong.
A claim of one type is only treated as verified by a source of the matching type. A market observation is not verified by a regulator press release.
Claims are grouped by the type recorded on each one. Grouping hides nothing: every claim is in exactly one group, in full.
Sources
2 documents
Primary sources2
- MAS Introduces Debt Servicing Framework for Property Loans
Monetary Authority of Singapore · Published 28 June 2013
Original source link unavailable
- MAS Notice 645: Computation of Total Debt Servicing Ratio for Property Loans
Monetary Authority of Singapore · MAS Notice 645 · Published 28 June 2013
Original source link unavailable
Event checked against its primary sources on 13 August 2026. Each claim keeps its own verification status.
Revision history
v1.1
14 August 2026
- Rebuilt against Template v2: added the status strip, the structured event facts block and the parameters-at-introduction block.
- Key-number captions rewritten to fit their cards. Previously the claim text was clipped mid-sentence at render time.
- Added the status line on every historical parameter and the disambiguation separating the 30% income haircut from the 30% HDB Mortgage Servicing Ratio.
- Surfaced the 4.5% non-residential specified rate, which previously appeared only inside a claim qualification.
- Evidence check recounted by claim provenance rather than by verification tier.
- No claim, number, date or causality grade changed in this revision.
Corrections and clarifications
13 August 2026
Correction
Original: TDSR followed seven earlier rounds of residential property cooling measures implemented from September 2009 through January 2013 under the commonly used market chronology.
Corrected: DROPPED. No replacement claim.
Dropped as context rather than evidence. Round-counting conventions vary between commentators, the underlying policy dates are the archive's record, and no analysis in this event rests on the ordinal. The claim had already been downgraded to UNVERIFIED on the founder primary-source pass earlier the same day, because it asserted verified against the primary source with no primary document linked. Dropping it rather than re-sourcing it during the Tier 1 backfill removes a permanent verification liability for a statement nothing depends on. One of three identical drops across the June 2013 event, the September 2022 event and the April 2023 event.
13 August 2026
Clarification
Original: TDSR threshold set at a maximum of 60% of gross monthly income.
Clarified: MAS established a TDSR threshold of 60%. Property loans exceeding the threshold were regarded as imprudent and could be granted only exceptionally, subject to the financial institution's approved procedures and credit-committee approval.
The 28 June 2013 release describes a threshold above which a loan is regarded as imprudent, not an absolute cap. Exceptions were permitted subject to the institution's approved procedures and credit-committee approval. Describing it as a maximum overstates the constraint and misrepresents how the framework operated at the margin. Found in the founder primary-source pass.
13 August 2026
Clarification
Original: Made particularly explicit by MAS in subsequent official explanations and speeches; do not present as a verbatim quotation from the 28 June 2013 press release unless the exact wording is separately sourced.
Clarified: MAS press release 28 June 2013, paragraph 11
The structural-prudential characterisation appears directly in paragraph 11 of the 28 June 2013 release. The stored qualification hedged our own sourcing, attributing to later speeches something the primary document says itself. Not material to what a reader concludes, but a provenance record that understates its source is still wrong. Found in the founder primary-source pass.
13 August 2026
Correction
Original: Variable income generally subject to a 30% haircut.
Corrected: Financial institutions were required to apply a haircut of at least 30% to all variable income, including bonuses, and to rental income.
Two errors in one line. 30% is a regulatory MINIMUM, not the rate applied, so an institution could apply more; and the coverage was understated, because the requirement reaches all variable income including bonuses and reaches rental income. The stored number did not change; what it means did. Found in the founder primary-source pass.
13 August 2026
Verification change
Was: verified against the primary source
Now: UNVERIFIED
The seven-round chronology is substantively supported under the commonly used sequence (Sep 2009, Feb 2010, Aug 2010, Jan 2011, Dec 2011, Oct 2012, Jan 2013), but no primary supporting document is linked and none was checked. Verified against the primary source asserted a document that does not exist in the record. Re-link when those events are written in Tier 1 backfill. Found in the founder primary-source pass.
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.