Selling A Hougang 4-Room At S$622,000 To Buy In District 19 — The Sale Nets S$369,682, The Upfront Bill Is S$399,300

This is an illustrative example on the quarter's median prices and a stated household, not a real family and not advice on whether to sell, buy, keep or rent out. It lays out the figures the position engine computes for each side of the question; the decision is the reader's.

Most upgrade questions get argued as a yes or no. The figures do not work that way. There are three separate sides to lay out, and each has its own set of numbers. One is what the flat returns after the outstanding loan and selling costs. One is what the condo asks for upfront and every month. The third is what the flat would earn if it stayed and went on the rental market. We ran all three through our position engine at this quarter's medians. Here is what stood out to us: the sale nets a large sum, and it still lands below the cash and CPF needed before the loan is drawn.

According to data.gov.sg's HDB resale records, the median 4-room price in Hougang was S$385,000 in 2Q2020 across 67 sales. In 2Q2026 it was S$622,000 across 143 sales. For the same quarter, URA Data Service caveats put the median price of a 60 to 99 sqm private unit in District 19 at S$1,430,000, from 140 sales, with a median area of 888 sqft. The illustrative couple bought in 2020 with a 75% HDB loan. Every figure below is indicative and drawn from recorded transactions and approvals. None of it is a valuation of any particular home.

The flat's median has moved 61.6% since 2Q2020, and the loan balance takes a large share of that

Start with the side that is already settled. The median has moved from S$385,000 to S$622,000, a change of S$237,000, or 61.6% on the engine's measure. The 75% loan taken in 2020 was S$288,750. After six years, the balance stands at about S$235,488. Selling costs, covering agent and legal fees, come to about S$16,830. Seller's stamp duty is S$0 at this holding period. What is left is about S$369,682 in net proceeds.

Two things the table cannot settle. The medians sit on different transaction counts, 67 sales then and 143 now, so the two quarters are not measured on the same base. And a median says nothing about storey, condition, renovation or remaining lease on any one flat. The engine is reading the middle of the recorded market in Hougang, not a specific unit.

The table below sets out the flat side in full.

Buying before the flat is sold puts a S$286,000 stamp duty line into the same table

Now the purchase. At the District 19 median of S$1,430,000, buyer's stamp duty is S$41,800. If the condo is bought after the flat is sold, additional buyer's stamp duty is S$0. If it is bought before the sale, the engine applies the 20% rate under the rules as recorded, which is S$286,000. Remissions are not modelled here. That single line sits in the table twice for a reason: the order of the two transactions changes it, and nothing else about the purchase changes with it.

The loan is capped at S$1,072,500, which is 75% of the price. The binding constraint in this example is the loan-to-value limit, not income. That leaves S$71,500 in minimum cash and the rest in CPF or cash. Added up by the engine, S$399,300 is needed before the loan is drawn. Against S$369,682 from the sale, the engine records a shortfall of S$29,618 to find elsewhere. The table below shows both stamp duty scenarios side by side.

The monthly instalment is the line that moves the most

The upfront numbers are one-off. The monthly ones repeat. On the flat, the instalment runs at about S$1,310. On the condo, at 3% over 27 years, it is about S$4,834. The engine puts the difference at S$3,524 a month. That is the figure most of the decision tends to turn on, and it is the one that keeps turning up for the full tenure.

The engine also runs both sides at a test rate of 4%. This is a sensitivity check written into the rules, not a view on where rates go. At that test rate, the flat instalment reads S$1,476 and the condo instalment reads S$5,418. Both numbers assume the maximum loan is taken in each case. A smaller loan, or a shorter tenure, would produce different figures that the table does not show.

Here are both, at the loan rate and at the test rate.

Renting the flat out is the third side, and it carries its own set of figures

The third position keeps the flat. Using data.gov.sg rental approvals, the engine estimates a monthly rent of S$3,250. That estimate comes from 21 recorded approvals, which is a thin base, and approvals do not record unit condition or furnishing. On the current median price, that works out to a gross yield of 6.3% on the engine's measure.

From there the costs come off. Property tax at the non-owner-occupied rate is S$1,638 a year. Maintenance is S$900 a year. After vacancy, tax and maintenance, net rent is S$33,225 a year. Interest on the remaining loan is S$6,123 a year. The net carry the engine records is S$27,102 a year, or 4.4%.

Note what this side does not include. It assumes the flat is retained, so the proceeds in the first table are not realised and the purchase in the second table is a different question entirely. The table below shows the rental side as the engine computes it.

Holding the condo for five years costs S$254,927, and S$150,873 of that is interest

The last figure is the one that gets left out of most upgrade sums. Over a five-year hold, the engine totals S$254,927 in costs on the condo. Of that, S$150,873 is mortgage interest from a standard amortisation schedule at the stated rate. The rest covers selling costs at exit, property tax and maintenance on the engine's configured norms. To cover that total, the price would need to rise 18.2%. That is a break-even figure, not a view on prices.

The rules behind all of this are versioned. Stamp duty, LTV, TDSR and property tax rates were checked against IRAS and MAS publications on 18 September 2026. The newest version applied here took effect on 4 July 2025. Rules change with each round of measures. Run the same household through the engine under an older rule set and some of these lines read differently.

What the figures leave open is the gap between a median and an actual unit. The flat side rests on 143 recorded sales in Hougang in 2Q2026. The condo side rests on 140 caveats in District 19, at a median of 888 sqft. A specific flat on a high floor, or a specific unit with a larger layout, sits somewhere other than the middle. The rental estimate rests on 21 approvals, which is the thinnest count in the pack. In the next quarter's data, the lines worth watching are the two medians and their counts, the rental approval count for Hougang 4-rooms, and whether any rule version changes the stamp duty or LTV rows.

One number rarely settles a question like this. The S$369,682 from the sale reads well on its own, until it sits next to the S$399,300 the purchase asks for first. The S$3,524 monthly difference reads differently again next to a net carry of S$27,102 a year. We lay out all three sides because each one answers a different question, and the household asking is the one that knows which question it is actually asking.

Method: the flat is a 4-room in the town, bought at the median of Q2 2020 with a 75% HDB loan at 2.6% over 25 years, of which 6 years have been paid. Its value today is Q2 2026's median. Selling costs, seller's stamp duty, equity, the rent estimate and the carry are the position engine's under the rules recorded as in force. The condo is a 60 to 99 sqm resale in the nearest district with enough sales within 2.5 km of the town. It is taken at Q2 2026's median. The buyer is a citizen aged 38 with a household income of S$15,000 a month and no other debt. sale at today's indicative mid value after 5 years. Seller pays 2% commission and S$3,000 legal fees, plus 9% GST. Mortgage interest from a standard amortisation schedule at the given rate. Loan of S$1,072,500 (75% of the price), the maximum available. Property tax on an assumed annual value of S$35,700 (85% of market rent). Maintenance at the configured norm. Rates are versioned data in config/rules.yaml, checked against IRAS and MAS publications on 2026-09-18. The newest version applied here took effect 2025-07-04. Rules change with each cooling-measure round. Private figures are from URA Data Service caveats, HDB figures from data.gov.sg resale records, both loaded weekly by this site's ingest. Medians are of the non-excluded transactions in the period named. Sales flagged by the hygiene rules (related-party transfers, bulk deals, outliers) are left out. Every figure is indicative and drawn from recorded transactions, not a formal valuation.