Singapore’s HDB Resale Price Index reached 203.4 in Q1 2026 — its first quarterly dip since Q2 2019. The index is correct. But it measures the flat’s gross market value, not the seller’s net cash. A compounding CPF principal-plus-interest refund sits between the headline and the hand.
RPI 203.4, −0.1% q-o-q, +1.2% y-o-y (RAOSCAFF arithmetic: 203.4 ÷ 201.0 − 1 ≈ +1.2%), across 6,285 resale transactions (HDB final, Q1 2026). The index tracks the flat’s gross transaction price. Before cash reaches the seller, the outstanding loan is repaid and the CPF principal plus accrued interest at 2.5% p.a. is refunded to CPF. Index-gain and cash-gain are different populations of outcome.
HDB published its Q1 2026 Resale Price Index as 203.4 — down 0.1% quarter-on-quarter, described as the first decline since Q2 2019 — while still +1.2% year-on-year (RAOSCAFF arithmetic: Q1 2026 RPI 203.4 vs Q1 2025 RPI 201.0 = +1.194% ≈ +1.2%; derived from the published HDB RPI table). In Q1 2026 there were 6,285 resale transactions (HDB final), −4.6% year-on-year and +19.6% quarter-on-quarter. HDB published a flash estimate on 31 March 2026 and the final data on 23 April 2026.
The index is a gross asset-value measure — it tracks resale transaction prices and reports a moment in the flat’s market value. It says nothing about what any individual seller walked away with in cash. Where CPF savings funded part or all of the flat’s purchase price — and on sale, that money must come back to CPF first.
This brief decomposes that gap. It critiques the measure, not HDB, CPF, any estate, project, agent, or person. Every figure traces to a primary HDB or CPF Board source.
The index sits above the waterline. Below it: the outstanding loan, then the CPF P+I refund, growing with every year held. What’s left — if anything — is the seller’s cash.
Axis 1 — gross (the index). RPI 203.4 records the transaction prices of flats resold in Q1 2026. It is a market-value measure at a moment: the price a buyer paid. From that price the seller must first service the outstanding housing loan. Only after the loan does the CPF refund enter — and only after the CPF refund does the seller see any cash.
Axis 2 — the deduction that grows (CPF P+I at 2.5% p.a.). On sale, CPF’s rule is explicit: “The amount that you have to refund is the CPF principal amount you withdrew and the accrued interest (P+I).” That interest accrues at the CPF Ordinary Account rate — 2.5% per annum, the statutory floor, confirmed unchanged for 1 Apr–30 Jun 2026. It compounds on the principal withdrawn from the date of withdrawal until sale. A flat held longer has a larger CPF wedge between its gross price and the seller’s net cash — with no change to the index.
Axis 3 — the zero-cash floor (negative sale, written off). If the proceeds after the loan are insufficient to cover the full CPF P+I refund and the flat is sold at market value, CPF’s rule applies: “You do not need to top up the CPF housing refund shortfall in cash, as long as the property is sold at market value.” [S4] The shortfall is written off — the seller nets zero cash. The refund amount returns to the seller’s own CPF account; it is not a loss of wealth. But it is not cash in hand. The price index measures the flat’s gross market value and is silent on this deduction; the seller’s net cash is a separate figure.
| Published headline (HDB) | What it does not measure | The verified mechanism |
|---|---|---|
| RPI 203.4, −0.1% q-o-q, +1.2% y-o-y† — a gross index [S1][S2] † +1.2% y-o-y = RAOSCAFF arithmetic (203.4 ÷ 201.0 −1) |
A price is not cash | Net cash = sale price − outstanding loan − CPF P+I refund [S3][S4] |
| One “price” number | A statutory refund obligation (P+I) sits between the price and the seller’s cash; it grows with holding period | CPF OA accrued interest 2.5% p.a. (statutory floor, 1 Apr–30 Jun 2026), compounding withdrawal to sale [S3] |
| A gross price level for the period | Index-gain ≠ cash-gain; the index measures X (flat’s market value); the seller’s net cash also depends on Y (loan balance + CPF P+I refund) | At market value, CPF shortfall written off → zero-cash “negative sale” despite a positive index [S4] |
If you read one thing: a resale price index reports the flat’s gross value — not the seller’s net cash after the CPF P+I refund.
Singapore’s HDB Resale Price Index for Q1 2026 is 203.4 — correct, published, and useful. It records the gross market value of flats transacted. It does not record the seller’s net cash: a resale seller must first repay the outstanding housing loan, then refund the CPF principal plus accrued interest (P+I) to the CPF Ordinary Account. That refund’s interest leg compounds at 2.5% per annum — the statutory floor, confirmed 1 Apr–30 Jun 2026 — from the date CPF savings were withdrawn until the flat is sold. The longer the hold, the larger the CPF wedge between the index number and the seller’s hand. Where proceeds after the loan are insufficient to cover the full P+I and the flat is sold at market value, CPF writes off the shortfall — the seller receives zero cash, a “negative sale” (CPF Board rule, [S4]). The aggregate RPI is silent on any individual seller’s net cash position; the index and the seller’s cash are two different defined quantities measuring one housing-market idea. The refund returns to the seller’s own CPF account; it is not a loss of wealth — but it is not cash. Mirror Brief SG-01 makes one claim: a price index is not a seller’s cash, and which of the two you cite is a choice, not a fact about the world. No causal claim about why the index dipped is made; no policy or scheme is rated good or bad; no recommendation is offered; no person, estate, project, or firm is named as performing well or poorly. All figures trace to primary HDB and CPF Board sources; derived figures are labelled RAOSCAFF arithmetic.
Mirror format — RAOSCAFF anchors on the publisher’s own primary data and decomposes the figure they print. SG-01 is a gross-to-net decomposition: the HDB Resale Price Index (Q1 2026, flash 31 Mar 2026 / final 23 Apr 2026) sets the gross axis; CPF Board primary pages for the OA interest rate (1 Apr–30 Jun 2026) and the CPF refund-on-sale rule set the deduction axis. Every figure used in the report traces to one of six primary-verified sources listed in §H of FACTS.md. The “live-render caveat” is logged: HDB pulse pages render the data body via JavaScript; the RPI figures are corroborated from independent reports citing HDB and match HDB’s own page titles and published dates — treated as primary-confirmed. CPF figures were fetched directly from cpf.gov.sg. No dollar-value worked example is asserted as a primary figure; any such figure would require seller-specific assumptions not published by HDB or CPF.
Politically neutral throughout: no causal or policy claim; no minister, government, or party named; the “~6% of flats crossed S$1m” framing (traceable only to a Parliamentary statement) is excluded. Defamation-disciplined: the object of critique is the metric’s gross-vs-net framing, not HDB’s or CPF’s competence and not any market participant. Predict-not-recommend: no buy/sell/hold view; no “should you sell”; no security or fund cited. Reference-period precise: RPI = Q1 2026; CPF OA / accrued-interest rate = 2.5% p.a. for 1 Apr–30 Jun 2026. The “% profitable” and “8 in 10 zero-cash” framings are excluded as they are not primary HDB statistics. FACTS.md is the source-of-truth file; every number in this report is anchored there.
Primary sources S1–S5 are HDB and CPF Board official publications, verified Phase-0 GREEN. S6 is independent corroboration used only to confirm JS-rendered HDB data body figures. Full citations and the RAOSCAFF arithmetic register are in the companion FACTS.md, §H.