A resale 2 bedroom condo returned more on the money put in than a new launch in every year from 2017 to 2021

A two-bedroom buyer with about $1.1 million can buy a new launch, paying in instalments across three years of construction, or a resale, paying in full and collecting rent from the first month. On price the two look alike. As investments they are not alike, because they commit different amounts of money at different times, and money committed earlier has to earn for longer.

Measured as a return on the capital actually committed, month by month, financing at 2.5%, the resale won in every one of the five purchase years. A two-bedroom bought as a resale in 2020 returned 15.1% a year; one bought at a launch that year returned 9.4%. Across 2017 to 2021 the resale returned between 11.0% and 15.1% a year and the launch between 5.2% and 9.4%.

Most of the gap is what the money bought. In 2020 the median launch two-bedroom cost $1,200,000 for 689 square feet and the median resale cost $1,090,000 for 904 square feet. The resale buyer took a third more floor area for less money, rented it from the first month, and sold it four and three-quarter years later on a median gain of 29.7% against the launch buyer's 21.2%. What follows is where that holds, where it narrows, and the one cut where the launch wins.

Key takeaways

• The resale returned more in every purchase year: 11.0% to 15.1% a year against the launch's 5.2% to 9.4%, financing at 2.5%. What each returned
• The launch buyer pays late but earns later still, nothing at all until the building is finished. The 2021 launch cohort is the extreme: 5.2% a year against the resale's 14.2%. Why the launch starts behind
• In the CCR the launch cohorts returned about nothing or less at 2.5%, from 0.6% down to a loss of 9.7% a year, while CCR resales returned 7.1% to 9.1%. By region
• A higher mortgage rate hurts the resale more, and in one cut it reverses the answer: RCR purchases made in 2018 returned 7.5% a year at a launch against the resale's 7.3%, once financing reaches 3.5%. What a higher rate does

What each returned

Every two-bedroom bought between 2017 and 2021 and resold from 2022 on, matched on the unit itself, split by whether the purchase was a new launch or a resale. 2,325 launch purchases and 1,363 resale purchases.

Bought	Bought as	Pairs	Median price	Median size	Median gain	Median hold
2017	New launch	307	$1,028,000	710 sqft	17.4%	5.2 years
2017	Resale	266	$1,100,000	936 sqft	22.4%	5.1 years
2018	New launch	397	$1,083,000	667 sqft	19.8%	5.0 years
2018	Resale	360	$1,080,000	904 sqft	21.1%	4.9 years
2019	New launch	632	$1,122,500	689 sqft	20.9%	4.8 years
2019	Resale	224	$1,130,000	915 sqft	27.0%	5.1 years
2020	New launch	694	$1,200,000	689 sqft	21.2%	4.7 years
2020	Resale	270	$1,090,000	904 sqft	29.7%	4.8 years
2021	New launch	295	$1,296,453	710 sqft	13.6%	4.4 years
2021	Resale	243	$1,165,000	893 sqft	26.8%	4.4 years

The launch buyer paid about the same money for roughly a quarter less floor area in every year, and by 2021 was paying about $131,000 more for it. That is the launch premium, and it has to be earned back before anything else happens.

The return below is the internal rate of return on the monthly cash flows, annualised: every payment out, the mortgage drawn and serviced, rent in after costs, and the sale at the end. It is not the price gain, and it is shown at four financing rates because no single rate is right for a hold that ran from 2017 into 2026.

Bought	Bought as	1.5%	2.5%	3.5%	4.5%
2017	New launch	8.1%	7.0%	5.9%	4.7%
2017	Resale	13.2%	11.3%	9.4%	7.4%
2018	New launch	9.4%	8.3%	7.3%	6.2%
2018	Resale	12.9%	11.0%	9.0%	7.0%
2019	New launch	10.3%	9.3%	8.3%	7.3%
2019	Resale	14.6%	12.8%	11.0%	9.1%
2020	New launch	10.3%	9.4%	8.4%	7.4%
2020	Resale	16.9%	15.1%	13.2%	11.3%
2021	New launch	6.2%	5.2%	4.2%	3.1%
2021	Resale	16.1%	14.2%	12.3%	10.4%

• Bought at a new launch
• Bought as a resale

The resale is ahead in all twenty cells. The narrowest gap is 2018, 11.0% against 8.3% at 2.5%, and the widest is 2021, 14.2% against 5.2%. Nothing about the ranking depends on the financing rate here; only the size of the gap does.

Why the launch starts behind

Three things separate the two, and only one of them is price.

The launch buyer commits money late, which helps. Five per cent at booking, fifteen more when the option is exercised, then the balance across eight construction stages. On a $1.2 million purchase the buyer has put in $240,000 by the third month and the mortgage covers the rest as the stages fall due, interest-only until the building is done. The resale buyer has committed the whole downpayment and the stamp duty on day one.

The launch buyer earns nothing for three years, which hurts more. The resale collects rent from the first month; the launch collects its first rent in month 38. Over a hold of about four and three-quarter years that is roughly three years of rent the launch never sees, on a unit the buyer is already paying interest on.

And the launch buyer bought less. A median 689 square feet against 904 in 2020, for more money. Rent follows floor area, so the smaller unit earns less every month it is let, and the higher price per square foot is the thing the resale market has to re-rate before the launch buyer is even level.

The 2021 cohort shows all three at once. Those buyers paid the most per square foot of any year, took delivery latest, and sold soonest, at a median hold of 4.4 years. Their rent had barely started when they sold. 5.2% a year against a resale buyer's 14.2% is not a statement about the projects; it is a statement about how little of that hold was earning.

By region

• Bought at a new launch
• Bought as a resale

The pattern holds in all three regions and it is not equally strong in each.

CCR, bought	Bought as	1.5%	2.5%	3.5%	4.5%
2017	New launch	-0.6%	-2.0%	-3.5%	-5.0%
2017	Resale	10.5%	8.7%	6.7%	4.7%
2018	New launch	-8.1%	-9.7%	-11.3%	-13.0%
2018	Resale	9.0%	7.1%	5.1%	3.0%
2019	New launch	1.1%	-0.2%	-1.6%	-3.0%
2019	Resale	10.4%	8.6%	6.6%	4.7%
2020	New launch	1.3%	0.1%	-1.2%	-2.5%
2020	Resale	10.9%	9.1%	7.1%	5.2%
2021	New launch	1.7%	0.6%	-0.5%	-1.6%
2021	Resale	9.8%	7.8%	5.7%	3.7%

The CCR is the one region where the launch buyer did not merely trail, but lost. At 2.5% every CCR launch cohort returned between 0.6% and a negative 9.7% a year, while every CCR resale cohort returned between 7.1% and 9.1%. These are thin cells, 11 to 41 pairs each against 35 to 70 on the resale side, and the 2018 launch figure rests on 11 purchases at a median $2,386,660, the most expensive cell in the study. Read it as a warning about a small number of very large cheques, not as a regional rate.

RCR, bought	Bought as	1.5%	2.5%	3.5%	4.5%
2017	New launch	8.1%	7.0%	5.9%	4.7%
2017	Resale	12.5%	10.6%	8.7%	6.7%
2018	New launch	9.6%	8.6%	7.5%	6.5%
2018	Resale	11.2%	9.3%	7.3%	5.2%
2019	New launch	10.9%	9.9%	8.9%	7.8%
2019	Resale	13.9%	12.1%	10.2%	8.3%
2020	New launch	11.7%	10.8%	9.8%	8.9%
2020	Resale	16.2%	14.4%	12.5%	10.6%
2021	New launch	5.9%	4.9%	3.8%	2.7%
2021	Resale	14.8%	12.9%	11.0%	9.0%

OCR, bought	Bought as	1.5%	2.5%	3.5%	4.5%
2017	New launch	9.4%	8.4%	7.3%	6.2%
2017	Resale	14.3%	12.5%	10.6%	8.6%
2018	New launch	9.6%	8.6%	7.5%	6.5%
2018	Resale	14.6%	12.7%	10.8%	8.8%
2019	New launch	10.2%	9.3%	8.3%	7.3%
2019	Resale	17.1%	15.2%	13.4%	11.5%
2020	New launch	9.8%	8.8%	7.8%	6.8%
2020	Resale	19.5%	17.7%	15.9%	14.1%
2021	New launch	8.1%	7.1%	6.2%	5.2%
2021	Resale	18.0%	16.2%	14.3%	12.3%

The best return in the study is an OCR resale bought in 2020: 17.7% a year at 2.5%, on a median $900,000 purchase of 883 square feet. The best launch return in the study, an RCR purchase in 2020 at 10.8%, is below every OCR resale cohort. Where a launch came closest was the RCR, where the launch premium over a resale was smallest.

What a higher rate does

Every cell falls as financing gets dearer, and the resale falls faster. The pooled 2017 resale drops from 13.2% to 7.4% across the four rates, 5.8 points; the launch drops from 8.1% to 4.7%, 3.4 points. The launch has borrowed less for less of the hold, so it is less exposed to the rate.

That is enough to reverse the answer once. RCR two-bedrooms bought in 2018 returned 9.3% as a resale and 8.6% at a launch when financing at 2.5%. At 3.5% the launch is ahead, 7.5% against 7.3%, and at 4.5% it is well ahead, 6.5% against 5.2%. It is one cut out of twenty and it needs a rate above 3%, but it is the shape of the thing: the launch's case is strongest when money is expensive and the resale's is strongest when money is cheap.

Nowhere else does the ranking change. In the other nineteen cuts the resale is ahead at all four rates.

What this means for choosing

On this data, for a first property, held about five years, bought between 2017 and 2021 and already sold, the resale returned more. The finding is stronger than the price gain alone suggests, because the price gain understates it: the launch also gave up about three years of rent.

Two things stop that from being a rule. The first is that a launch buyer is buying a different thing, a new building with a fresh lease and no immediate repairs, and this study prices none of that. The second is timing: these are units bought in 2017 to 2021 and sold from 2022, a window in which the resale market re-rated sharply. A launch bought today completes into a market nobody has measured yet.

The practical reading is narrower and more useful. The launch premium per square foot is the whole question. Where a launch asks 40% more per square foot than nearby resales, three years of lost rent on top of that is very hard to make back in five years. Where the premium is small, as it was in the RCR, the two get close, and a dear mortgage can tip it.

What this does not tell you

• Second properties. Additional buyer's stamp duty is not charged in any figure here. Realila's policy records hold no rate history, so a 2017 to 2021 purchase cannot be stamped at the rate that applied on the day, and a second-property table would be a guess. Every return here is a first-property purchase.
• Twenty-one launch pairs. Twenty-one launch pairs were bought after their building was completed and are modelled as if rent began at month 38; they are too few to move any cell.
• The rate you paid. Four rates are shown because the real one moved through the period and differed by borrower. None of the four is a forecast.
• The unit you would buy. Every cell is a median of a cohort, not a unit. A specific launch against a specific resale can sit anywhere inside these ranges.
• Rent that stood still. The 2025 rent per square foot is applied flat across the whole hold. Rents fell and then rose across this window, so cells that earned most of their rent early are flattered and those that earned it late are not.
• Units that have not resold. Only completed pairs are counted. Launch buyers still holding, including everyone whose building finished recently, are not in the data.

Method and data notes

The pairs are the construction used in the two-bedroom resale note: every unit bought and later resold, matched on the unit itself, private condominiums and apartments, with the same suspect and mislabelled classes excluded. Here the window is narrowed to purchases made in 2017 to 2021 with holds of four to six years, so that launch and resale cohorts are compared over the same span. Bedroom count and region are taken from the purchase caveat. Whether a purchase was a new launch or a resale is the sale type recorded on that caveat, the same split the two-bedroom against three-bedroom note uses when it compares price gains.

Return on capital committed is the internal rate of return on the modelled monthly cash flows, annualised. It is not the price gain. A resale buyer pays the downpayment, the buyer's stamp duty and legal fees at the start, draws a mortgage at 75% of the price over 360 months, and receives rent from the first month. Launch payments follow the standard progressive schedule, with the booking fee and downpayment from Realila's policy records and the eight construction-stage shares stated as an assumption; rent is credited from month 38, three years after purchase, which is the cohort's median time to completion. The launch mortgage is drawn as the stages fall due, is interest-only until completion, and amortises over the remaining term after that.

Buyer's stamp duty is the registered schedule. Rent is the 2025 two-bedroom rent per square foot from the rental yield note, pooled or regional to match the cell, applied flat across the hold and scaled by each cell's own median floor area. Vacancy is 3% of the year and the letting agent takes half a month a year. Maintenance is $280 a month at 474 square feet, pro-rated on area, an assumption carried from the shoebox work and registered nowhere. Property tax is the registered non-owner-occupied schedule applied to an annual value taken as the modelled annual rent, which is a proxy: a real assessment for any one unit will differ. Selling costs are 2% commission plus GST and $3,000 in legal fees at each end.