Is it just math, or is there merit in studying the data? How much does quantitative research add to the issues at hand?

Old flats - once the storehouse of wealth for owners should their homes be chosen for en-bloc redevelopment - have become a source of anxiety for residents after the government two years ago cautioned against speculation of state buyouts of these assets.

The idea that the value of one's home can eventually run down to zero is a terrifying thought, one which increasingly more researchers are studying in an attempt to quantify the relationship between dwindling leases and Housing & Development Board (HDB) resale prices.

Crudely put, the studies so far have pointed to a unanimous finding - that the decline of resale prices is certain as the length of lease decreases, especially as limits on financing kick in.

If home owners here have found themselves still able to sell their flats at a profit, it is because the decline of the flats' theoretical value has been more than offset by rapid economic and property market growth since the 1960s.

Recent moves by the government to introduce the new Home Improvement Programme (HIP) II for units at the 60- to 70-year mark, and Voluntary Early Redevelopment Scheme (Vers) to allow flat owners to vote if they want the authorities to take back their apartments for redevelopment at the 70-year mark of their lease, appear to be some ways the state is trying to ease concerns.

And just last month, the government said that it is looking into relaxing Central Provident Fund (CPF) loan rules on the purchase of older resale flats.

Currently, a home owner can only use his CPF money for his flat purchase if his age plus the number of years left on the remaining lease of the property is at least 80 years, but even this remains subject to restrictions.

For flats whose remaining lease is less than 30 years, no CPF money can be used at all.

Is there value to research?

In recent months, studies looking into the effect of dwindling leases have been met with equal opposing forces of interest and dismissal from within the circle of government officials, academia, and property consultants.

Some consider it "mere math" that is not grounded in reality, especially when the price trend for HDB resale flats has been an upward one for most part of the property cycle.

Others recognise that amid the barrage of qualitative commentary, some truth can be gleaned from the objectivity of numbers.

It goes without saying, however, that every study comes with limitations. One real limitation is the scarcity of old flats in the market to make up a robust sample size to prove a trend.

But Michael Cho, founder of UrbanZoom, a homegrown artificial-intelligence-enabled research portal, believes that limitations do not diminish the value of such study.

A self-professed data geek, he says: "This topic is a highly quantifiable one and given the amount of data we have in Singapore, there's no reason why we shouldn't take advantage of this to anchor our discussions based on observable evidence."

He first thought of undertaking the study because he owns a HDB flat that is turning 40 soon.

"So this is something that I'm dying to find out myself too. This is after all the biggest financial asset of my family, so I definitely want to know if this is truly an asset that can hold its value over the years, and therefore whether it is worth passing on to the next generation."

Associate Professor Sing Tien Foo, Dean's Chair and director of the National University of Singapore (NUS) Institute of Real Estate and Urban Studies, says that empirical studies - as such studies usually are - are useful not just to show a trend but also to give important insight to explain why things happens a certain way.

"It's very different from 'crystal balling', where you try to see what will happen in the market in the next five to 10 years, because this is dependent on a lot of moving parts and conditions.

"This is not some theoretical argument of prices declining because of certain reasons, but we observe and quantify this impact, and the next challenge is to explain what causes these results.

For instance, we may understand that HDB flats survive depreciation effects better than private property beyond a certain age, but why?"

Asked to comment on the findings of a recent unpublished SRX Property study, HDB appears to still prefer caution in its approach to such studies, saying that "reports on housing data are market sensitive in nature" and that "such data should be published with due diligence, and any analysis done be fair, robust and properly contextualised".

SRX Property study

A recent study done by SRX Property, a subsidiary of Singapore Press Holdings which publishes The Business Times, found that from 2006 to 2018, the majority of the HDB flats sold made a profit.

This may not come as a surprise, given that the official HDB resale index has increased more than five times since 1990.

But SRX's study showed that even in the post-2013 bearish years, transacted flats still enjoyed capital gains, albeit lesser.

During that period, flat owners tended to hold their flats for a longer time, seemingly not in a hurry to sell unless a price match came along.

That said, over the entire 2006-2018 period, there was a distinct drop in capital gains for flats in the 46-55 year old category. They made average capital gains of 17.7 per cent, lower than the 30-40 per cent capital gains generated by the other age categories.

Their annualised appreciation rate (the yearly rate of return over the holding period) was also significantly lower at 1.7 per cent, compared to 5-7 per cent in the other age categories.

Such flats also seemed to have higher chances (four in 10) of making losses in absolute terms, versus a frequency of one in 10 for other age categories, although their quantum losses in absolute terms did not differ much from the rest of the age categories.

All in all, it seems that amid a rising market, very old flats don't appreciate nearly as much as their younger counterparts, and are in fact four times more likely to be sold at a loss compared with newer flats, while the value of flats that are 40 years old and newer don't seem to be affected much by age.

The limitations of the study include the sample size of 403 for transactions of older resale flats aged 46-55 years, which is significantly smaller than that for flats of other age groups, HDB points out.

"This is because the flats only came into this age group from this decade onwards. Hence, the analysis for flats in this age group would not be comparable with that for those in the other age groups."

Other industry watchers also noted the possibility of "selection bias", because the samples comprise only of units that were transacted, and home sellers tend to only want to sell at a profit.

The data thus does not cover "unrealised losses" of flats whose value continues to depreciate but which has not been realised because they have not been transacted.

NUS study

A study by NUS that published its findings in February this year also found that HDB flats survive depreciation better than private condominiums when they are 30 years and above.

It surmised that this was largely due to government efforts to upkeep old flats, which helps to preserve their value, whereas a lack of similar maintenance for ageing private properties leads them to depreciate more quickly in value.

Prof Sing, who was one of the authors of the study, says: "The problem seems more serious for leasehold private property owners who face both aging and lease-decaying effects, as ageing can hasten economic obsoleteness of older buildings."

Aside from this, the study also believes that subsidy grants of up to S$50,000 for first-time buyers of resale flats further mitigate the price depreciation for HDB flats as the properties age.

The study mapped age-related depreciation rates against transaction prices of resale properties using data obtained from the Urban Redevelopment Authority and HDB.

The team studied the depreciation rates of resale homes from 1997 to 2017 using historical resale transaction prices of HDB flats and private condominiums.

These transacted properties were divided into three housing types: HDB flats, 99-year leasehold non-landed homes and freehold non-landed homes.

The data was controlled for non-age-related factors such as housing size, housing type, distance to the nearest MRT station and to the Central Business District.

The study found that differences in depreciation among freehold and leasehold private homes and HDB flats appeared only after 10 years.

Prices for all three housing types showed decline as the property aged. After 10 years, private freehold homes depreciated at a slower rate relative to private leasehold homes and HDB flats.

In fact, private leasehold homes and HDB flats depreciated at similar rates for up to two decades.

At the 21-year mark, the depreciation rate for HDB flats was about 3 per cent while freehold private homes prices depreciated by more than 10 per cent.

Leasehold private home prices depreciated by more than 30 per cent when they reached the same age.

The findings also showed that the price declined at a much slower rate for HDB flats aged 30 years and above, compared to private homes.

But Prof Sing admits that there is inevitable selection bias in the study, which only included completed transactions.

Had some of the older units in more run-down buildings been sold, the depreciation effect measured could have been even greater.

The other constrain the identified was the researchers' inability to observe whether the owner of a unit had carried out renovation before the transaction, which would artificially reduce the depreciation effect and increase the value of the home.

UrbanZoom study

A study by UrbanZoom last year also found that HDB flats would depreciate, on average, 0.67 per cent per year across the first four decades as their leases run down, assuming a stagnant economy and property market with zero growth.

UrbanZoom combed through all completed HDB transactions from 1967 to 2013 to identify pairs of adjacent blocks that are within 100 metres of each other, and segmented them into buckets based on the difference in the lease commencement year between the corresponding pair.

What UrbanZoom found was that in the 10-25 year bracket, older HDB flats displayed a "non-trivial" median discount of 12 per cent compared to their newer neighbours.

The difference in value deepens with more years of difference in the length of lease, culminating in a median discount pushing close to 30 per cent for pairs of units that are at least 38 years apart in age.

Again, one natural limitation was that there were not many observable transactions beyond the 38-year range, given that HDB's public housing programme only started in the 1970s.

UrbanZoom also pointed toother factors that affected sales, such as changes in the borrowing limit and CPF usage limit for purchases of HDB flats when they cross the 40-year mark.

HDB also gave its feedback on the study, noting limitations such as a lack of control for location, storey height, floor area, extent of renovation, market conditions, and the economic growth of the nation.

What's clear from the studies so far, is that HDB property, like all leasehold property, will and does depreciate in value with time.

Painful as that fact is to stomach for home owners who have poured life savings into the asset and have to temper their expectations, the HDB market is one that has a solid base of demand that will continue to anchor prices.

Upcoming changes in financing terms could also act in favour of values.

This article was first published in The Business Times. Permission required for reproduction.