Why you should be analyzing condo price data by floor area

Last week, we looked at the MRT stations with the most number of condos and we found some interesting spots in Aljunied, Farrer Park and Potong Pasir.

This week, we are going narrow our search. We are going to make use of the square footage filter on the MRT screener to limit unit size.

Problems with looking at just PSF

One of the biggest issues with comparing PSF is that smaller units tend to have much high PSF and larger units (think 2000 - 3000 sqft penthouses) tend to have much lower PSF. This often distorts the actual numbers.

Looking back at the list we made last week, you’ll notice a large variance in the min and max psf. Without any indication of size, it can be difficult to draw decisive conclusions from it.

A large variance between the minimum and maximum numbers indicates a wide range of values in the dataset, showing significant variability and spread among the data points.

Adding size as a filter

In our MRT screener, you will find the Advanced Filters on the right (on mobile, at the top). By setting a range here, you are filtering the transactions that are aggregated to the range you’ve selected. Transactions of units that are too big or too small for you will be excluded.

We are going to set the size range to 900 - 1200 sqft, the average size of a 4-room HDB flat in Singapore. In this size range, you’d typically find apartments with 2-3 bedrooms depending on the developments.

Once you’ve applied the filters, you will be presented with an unordered list of condos. We’re going to sort them by average PSF in ascending order. In doing so, you might end up with something like this (below). This is normal when we are setting a small timeframe (12 months in this case).

To make this table more useful, go ahead and check the box to the right that reads “Exclude stations with no transactions” to eliminate these results.

By now, you should have arrived at a list similar to this:

As you might have noticed, the variance in PSF have significantly tightened. On top of that, with the floor area we have chosen, we can easily multiply the PSF by 1000 to get a rough gauge of the quantum. For example, the 14 transactions made in Yew Tee are like to be around $1.03M, from $870K to $1.13M.

Why you should try it

Using this technique significantly reduces noise in the dataset and helps us determine which neighborhoods we can afford based on the size of the apartment we need.

Floor area or living space is often the thing we can’t really compromise on. We can live a little further from the MRT, on a lower floor, have 1 less bathroom, but if you are married with 3 kids, a 500 sqft one-bedder is definitely not ideal.

If living space is a non-negotiable for you, we suggest factoring it in as early as possible in your property search.

When does living space not matter?

As you are reading this, you’re probably thinking, “How outrageous, when would it not matter?!” Well, never really, but there are some cases where its simply not the number one factor. For example, when the property buyer has no intentions of living in it but is in fact renting it out.

However, it’s never wise to completely ignore living space. Case in point, a lot of shoebox units (300 - 350sqft) perform poorly, both in rent and capital gain, precisely because of the lack of living space.