S&P 500 vs Private Property: Which Performed Better Over 10 Years?

Most people who bought a $1.5M condo in 2013 and sold it a decade later will tell you they "roughly broke even." They paid $1.5M. They sold for maybe $1.35M. Small loss, they say — held through the cooling measures, survived the cycle, not bad.

That calculation is wrong. Not slightly off. Structurally wrong.

The mistake is comparing the full property value to the full property value. But you did not put $1.5M into that condo. You put in $375,000 — your downpayment. The bank put in the other $1.125M. The correct comparison is $375K in property versus $375K in the S&P 500 on the same day, held for the same number of years.

When you fix the methodology, the 2013 cohort result is not "roughly broke even." It is a loss of more than your entire original downpayment.

The table that changes everything

Here is the corrected comparison across four entry years. Property ROE is the net gain or loss on the $375K equity stake after carrying costs. S&P 500 return is what the same $375K would have grown to, dividends reinvested.

A note on the data: URA PPI values from 2010, 2013, and 2018 are from the publicly available URA historical series (base Q1 2009 = 100). From 2021 onwards, index values are confirmed by Coalwood's own tool reads — Q1 2025 at 205.4, Q1 2026 at 210.8. S&P 500 total return figures are drawn from the public historical record (Macrotrends equivalent). Carrying costs are scenario estimates, not audited buyer calculations.

The 2013 row is the one to read twice. The URA price index peaked near 207 in early 2013 and sat around 186 a decade later — a confirmed nominal loss at the index level of roughly 10%. Add $350K in carrying costs on top of a capital loss, and a buyer who put in $375K effectively lost $502K. The S&P 500, over the same 10 years, turned the same $375K into roughly $1.06M. That is not a small gap. That is a $1.2M difference in outcomes.

Where the $350,000 comes from

The carrying cost figure is not a rounding error or a pessimistic assumption. It is the mechanical result of owning a $1.5M condo for 10 years.

On a $1.5M condo with a 25-year mortgage, you will pay roughly $350,000 in interest, taxes, and fees over 10 years — nearly as much as the original downpayment.

This is what makes the 2013 cohort's result so severe. A nominal capital loss of $152K is tolerable. A nominal capital loss of $152K plus $350K in carrying costs is not. The leverage that makes property exciting on the way up — you control a $1.5M asset with $375K — is the same mechanism that makes carrying costs lethal on the way down. Every dollar of interest and tax is paid out of your equity, not the bank's.

These figures exclude ABSD. The representative case is a Singaporean citizen purchasing their first private property. If ABSD applies — for a second property, a PR buyer, or a foreign purchaser — add a further $75K to $225K on a $1.5M purchase. That shifts every row in the table further against property.

What real OCR/RCR buyers actually got

The index tells one story. Individual projects tell a more granular one. Here are confirmed resale outcomes from actual OCR and RCR transactions, sorted by entry cohort.

The pattern is not subtle. Projects bought in the 2013–2017 window — Greenwich, Hillier, Hillview Peak, Waterbay — delivered 1.2% to 2.4% per year in nominal capital gains. The S&P 500 returned roughly 12–15% per year in total return over the same windows. Even at the project level, before the carrying cost adjustment, the gap is enormous.

Projects bought in 2019–2020 — Treasure at Tampines, Parc Esta, Clement Canopy — delivered 4.5% to 5.4% per year. Closer, but still behind the S&P 500's roughly 15–17% annual total return over the same 5-year window.

The leverage adjustment is what makes property competitive on the upside. A 5.3%/yr return on a $1.5M asset, earned off a $375K equity base, theoretically translates to roughly 21%/yr ROE before carrying costs. That is how property can beat the S&P 500 in good years — and why the 2020 cohort comes close. But in the 2013 cohort, there was no capital gain to leverage. There was a capital loss. And $350K in carrying costs was extracted from a pile that was already shrinking.

The rent question: does it change the math?

It can. Significantly.

Rental data from Treasure at Tampines — 751 confirmed transactions — shows a typical 2-bedroom unit renting for roughly $3,300 per month. Over 10 years, that is $396K gross. Almost exactly equal to the total carrying cost estimate.

An OCR investor who bought in 2013, rented continuously at $3,300/month, and sold in 2023 would see their carrying costs substantially offset by rental income. Net of tax, vacancy, and agent fees — roughly 20–30% less than gross — effective rental income over 10 years comes to around $280–$320K. That shifts the 2013 cohort's equity result from −134% to something closer to −30% to −40% ROE. Still worse than the S&P 500's +183%. But not catastrophic.

The fork matters: the article's central comparison applies cleanly to owner-occupiers. For full-period landlords with good occupancy, the comparison narrows — though it does not reverse for the 2013 cohort.

For owner-occupiers, the counter-argument is real but different in kind. The S&P 500 does not give you a place to live. The $3,300/month in rent you did not pay is a genuine economic benefit — it just cannot be simultaneously invested in an index fund. Whether that matters is a lifestyle question as much as a financial one.

The S&P 500 comparison is also USD-denominated. The SGD has generally strengthened against the USD over the past 15 years, which modestly reduces the SGD-converted S&P 500 return — partially compressing the gap, though not reversing the conclusion for the 2010 or 2013 cohorts.

The 2020 cohort: genuinely too close to call

The 2020 row is the honest exception. A buyer who entered at a URA PPI of around 153 and is exiting now at around 205 has seen roughly 34% nominal capital appreciation on the index. On well-chosen projects, considerably more — Parc Esta at 33%, Treasure at Tampines at 27%, both confirmed across hundreds of actual transactions.

Over five years, carrying costs are lower — roughly $175K rather than $350K. That leaves a net property gain of around $335K on the $375K equity stake, or about +89% ROE.

The S&P 500 roughly doubled over 2020–2025. On $375K, that is +$375K, or +100%.

The gap is $40K on a five-year comparison. For a landlord who rented the property out with good occupancy, rental income likely closes that gap and possibly flips the result. For an owner-occupier who saved $3,300/month in rent they were not paying, the lifestyle value makes the comparison even more ambiguous.

This is the one window where the case for property holds up under the equity-adjusted test — not because property outperformed the S&P 500 on pure numbers, but because the margin is thin enough that the non-financial benefits make a credible difference.

The honest answer

The entry year is almost everything.

Buyers who entered in 2013 — near the top of the pre-ABSD cycle — faced the worst possible combination: a nominal capital loss at the index level, and a decade of carrying costs that consumed more than their original downpayment. Even accounting for rental income at optimistic occupancy rates, it is very difficult to construct a scenario where a 2013 OCR/RCR condo buyer matched the S&P 500 on an equity-adjusted basis over 10 years. The confirmed resale data from projects like The Greenwich (2.0%/yr over 11 years, 76.5% win rate) and Kingsford Hillview Peak (1.2%/yr over 8.4 years) is consistent with that conclusion.

Buyers who entered in 2020 had the opposite combination: genuine capital appreciation running at 5%/yr or better on well-chosen projects, five-year carrying costs that had not yet accumulated to crushing levels, and a strong rental market to offset costs. That cohort's equity-adjusted outcome is genuinely close to the S&P 500 — particularly for landlords, and particularly on projects that outperformed the index.

The URA PPI sitting at 210.8 in early 2026 — confirming a 5-year gain of +32.7% from mid-2021 — means that buyers entering now are likely closer to the 2018 profile than the 2020 one. Not a bad entry cohort, but one where the S&P 500's historical total return means the equity comparison will be competitive rather than clearly in property's favour.

The question has never been which asset is better in the abstract. It has always been whether the timing and financing structure allowed leverage to work for you — or against you. For 2013 buyers, it worked against. For 2020 buyers, it mostly worked for. The difference between those two outcomes is not luck. It is the year the index was at 207 versus the year it was at 153.

If you bought between those two years and have been measuring success by comparing your sale price to your purchase price, you are almost certainly looking at the wrong number.