Jeremy Sheppard explains some of the reliability issues you should consider when looking at any statistic while doing your property market research.
Examining the statistics is an excellent way to remove your partiality when making a decision as an investor. For example, you may be aware a suburb has a new shopping centre nearing completion. Another suburb may have a new estate opening up with many new houses for sale. These factors will affect the future demand and supply for property, but how and to what degree?
You can guess what influence infrastructure changes to a suburb will have on prices. You can assume you have all the facts. You can draw on the experience of others, too. But, in the end, the decision you make is always going to be a subjective one when it is based purely on factors like these.
Statistics, on the other hand, rule out subjectivity because they are calculated using a consistent formula. The computer performing the calculation didn’t grow up in one area, therefore biasing the algorithm slightly for that area. The computer isn’t short on caffeine at the time it makes its decision, or high on it either. The computer hasn’t had a lucky experience with one market and an unlucky experience with another, which plays on its mind. The computer is completely impartial.
Numbers
Statistics are also an excellent means of ‘quantifying’ the potential of a location. We know a vacancy rate of 1% is twice as good as a vacancy rate of 2%. These are numbers we can easily compare. If suburb ‘A’ has a yield of 4.85% and suburb ‘B’ has a yield of 4.75%, we can’t assume too much about their differences since they’re both very similar.
But we don’t know by how much a new bridge over the river in suburb ‘A’ will drive demand. And we don’t know by how much the lack of development activity suppressing supply in suburb ‘B’ will affect prices. Is suburb ‘A’ a better investment option than suburb ‘B’ and, if so, by how much?
These factors are not quantifiable and therefore do not lend themselves well to comparison.
The fundies
Things like a new bridge over a river or a lack of developer activity are what we at Redwerks call ‘fundies’. These are ‘fundamental drivers’,not statistics. Fundies are the basis of change in property demand and supply. These are the underlying factors affecting growth. They are very important to investors and shouldn’t be ignored.
However, fundies suffer from subjectivity and are unquantifiable. In an effort to refine our research procedure, we use statistics where possible to eliminate the subjectivity in decision-making and to allow easier comparisons of investment opportunities.
Are drivers affecting the market?
I’m leaning more towards the statistics to make final decisions and less towards the fundies. The major reason is because I want to know if a fundamental driver is actually having the influence on the property market that I thought it would.
I may think a major shopping centre will positively affect prices in suburbs around it. But what I may not know is perhaps residents in the immediate vicinity are actually worried about the peace of their neighbourhood being disturbed, more so than thinking about the convenience the centre will provide. What is obvious to me as an investor may actually be the opposite of what homebuyers are thinking.
By monitoring the supply and demand statistics I can know whether I am right about an assumption, or wrong. This is because the fundamental drivers are influencing the statistics. I can also know when the drivers of growth are starting to take effect by monitoring when they start to change the stats.
Some will say I’m getting into the market too late – after the data has been compiled and statistics calculated. You may have read the phrase, “Once you’ve heard about it, it’s too late”. Developers may often use a phrase like this to hurry you into making a quick decision because it suits them.
One thing I’ve learnt about real estate investment is it moves a lot slower than most people realise. My advice is there is no need to rush into anything with respect to property investment. Take your time and make sure this long-term investment, worth hundreds of thousands of dollars, is carefully thought out. Spend the time paying attention to detail, and you’ll be better off for it.
Following a trend is certainly possible in real estate investing and is a much safer option than buying too early and waiting for prices to start the climb you assumed would happen. And the truly good thing about the Demand-to-Supply Ratio (DSR) score is it is a ‘lead indicator’ rather than a ‘lag indicator’.
The DSR measures the imbalance between supply and demand. Prices usually go up when demand exceeds supply. That price increase eventually subdues demand, and balance is restored.
Because the DSR is a measure of the imbalance, it gives an indication as to what will happen in the future, not just the state of affairs right now.
Statistical reliability
If we’re going to allow statistics to influence our decision-making as investors, then we need to know if the statistics are reliable or not. In fact, the degree to which we allow the statistics to influence our decisions must come down to the reliability of the data.
Statistical reliability (SR) for any given datum comes down to a number of factors. Each statistic has its own unique considerations. Unfortunately, we can’t say, “This statistic is reliable and that one is not”. The available data changes from season to season, making a certain statistic more reliable in some cases than in others.
The upshot of all this is you need to examine each statistic for reliability each time you consider it in your decision-making.
SR considerations
When calculating the DSR score as published in this magazine, there are many considerations in determining SR for any single datum. There are even more considerations when combining data from multiple sources for greater accuracy. It gets quite technical and I’d like to keep some of it a secret, so I won’t go into too much detail. Besides which, you may not easily be able to replicate the same procedures for your own statistical considerations, so I’ll just summarise the key concepts you can use.
3D SR
SR can be summarised into three dimensions as shown in the graph overleaf (“A qualitative measure of a market’s capital growth potential”)
A qualitative measure of a market’s capital growth potential
Data measures capital growth potential and the ratio of supply and demand in three dimensions…
Breadth: Accessing a wide range of data
- How many stats indicating supply have been found?
- How many stats indicating demand have been found?
Depth: Accessing a large volume of data
- How many instances were considered, eg auction count?
- How wide was the area, eg LGA or suburb?
Height: Accessing metadata
- How up to date is the data?
- How reputable is the source?
- How susceptible to anomalies is the data?
Breadth
The breadth represents the range of statistics used. The wider the range of statistics, the more angles a market is viewed from and the less likely an important aspect is missed.
It is crucial that if you do use statistics as a means of determining where to invest, you don’t rely on only a few statistics. The more stats you have pointing in one direction, the more confident you can be the market of interest is heading in the right direction.
Getting a handful of stats is not too difficult. There are many publications that display the common stats. To get a more comprehensive range, however, you may need to visit a few websites and compile a list. It’s a little bit of effort that will be well worth it in the end.
Depth
The depth represents the size or volume of the market. If we were to calculate the auction clearance rate using results for only one auction, the result would have a low SR. In fact, the result in this case could only ever be zero or 100%. But if there were 10 auctions, the result would be more representative of the demand and supply of the market.
Knowing the size of a market of interest can quickly make you question the validity of the data. For example, looking at any housing data as opposed to unit data in the Sydney CBD would be quite unreliable simply due to poor trading volume. Units, on the other hand, would be very widely traded and the data would, therefore, be much more reliable. Check the size of the market used in the calculation of any statistic.
Depth consideration is a catch-22, in that you need a large market to get a good reliable statistic. But the larger the area, the harder it is to target a specific investment opportunity. This usually comes down to a choice about the geographical size of the area the data considers (more on this later).
Another depth consideration is the number of days over which trades took place. Data compiled over a quarter will have more trades than data over a single week. But again there is a balance needed; three-month-old data is more out of date than yesterday’s data.
As mentioned earlier, property markets move quite slowly. I have found monthly data usually contains enough volume of trades to be reasonably reliable for most markets. Quarterly information is also pretty good and isn’t at all out of date, except in the fastest of the fast-moving markets, which are very scarce. About seven weeks would probably be the ideal timeframe for the majority of markets, but nobody publishes data for that weird frequency.
Height
The height axis represents the nature of the statistics themselves and how they were calculated, and how the underlying data upon which they are based was compiled.
A median for property values in a suburb is more reliable if it is calculated based on recent sales than on sales six months ago. Timely data is more representative of the current state than old data is.
Another consideration that may be hard to gauge is the inherent reliability of the data provider. Some providers have a different idea about quality and what they publish.
For example, if there is a single sale of a property in a suburb for the month of July, the median for that suburb is the sale price for that single property. Technically, this is correct. But the median in this case is unlikely to be representative of that market as a whole.
Some providers will give you whatever they have and leave it up to you to gauge the reliability. Other providers prefer not to publish anything unless there is a significant degree of reliability. For example, they may not publish a median if that median was based on only one sale. You can argue pointlessly about which data provider is right, and what you’d prefer them to publish. But in the end an awareness of what the data is based on is all that’s really needed.
Geographic resolution
Let’s say a particular local government area (LGA) has shown exceptional population growth over the last three quarters. Given an LGA is a large area and you have three quarters of consistent data, you can assume the population data is reliable. But how do you capitalise on this knowledge? Where do you buy within that LGA?
An LGA can have multiple postcodes. The population growth figures you’ve obtained are unlikely to be evenly distributed across all of the postcodes within that LGA. And a postcode can have multiple suburbs. How do you know which suburb is getting the best of the population growth?
Most available data is at the suburb level. But there is some accumulative data you can obtain, down to individual streets. Street-level data is probably going to be unreliable because the size of the market is too small for an average or median to be representative of the street as a whole.
At the other end of the spectrum, many data providers give information at a national or state level. This is of little interest to me as an investor since I can’t buy an entire state, and each suburb within a state will have vastly different growth characteristics over the coming years. The question of where to pinpoint within a large area still needs to be answered. One suburb within a hot state may underperform another suburb within a cold state. The state itself is irrelevant.
To have any real significance to an investor, the data needs to be local. The microeconomy is more important than the macroeconomy. The macroeconomy will have some impact on your suburb, so it shouldn’t be ignored. But the further you zoom out, the less influence the data has on your intended investment property. The largest geographical area I would typically consider a statistic for would be the LGA, but I much prefer suburb-level data.
Demographic resolution
Property-type resolution needs to be considered as well. It is possible a certain suburb has very strong demand for houses, but the demand for units is not so flash. You may need to discern between statistics at property-type level.
You can go even further and distinguish between two-, three- and four-bedroom houses, for example. But the further you drill down into specifics, the less reliable the data becomes; this is because the finer the market resolution, the less volume there is to draw on for a reliable statistic.
I generally don’t bother with statistics at the number-of-bedrooms level. I only go down as far as the type of property and even then only into two broad categories: houses or units.
There are exceptions, however. For example, the data available for two-bedroom units versus one-bedroom units should be accurate enough in large unit markets such as the Sydney CBD or Melbourne CBD. But for the majority of suburbs Australia-wide, the data for number of bedrooms is too fine tuned to get reliable enough stats.
There are plenty of suburbs around Australia that have little or no unit market and mostly houses. So even going down to the level of type of property can yield unreliable results for many Australian markets.
The SR of the DSR
The statistical reliability of the DSR score published in the data section at the back of this magazine is a number out of 8. It is shown in the far right column of the tables. Make sure when browsing the DSR listing that you glance across at the SR column to keep things in perspective.
You won’t find a market listed in the data section that has an SR of below 4, which is the cut-off point. These market statistics are considered too unreliable to be of any use. The DSR score is calculated for about 30,000 markets in Australia. Obviously, we can’t publish all of these in this magazine. A larger list is available on the web/phone app called “Boomtown” (visit DSRscore.com.au for free access).
Conclusion
Personally, I won’t waste my time on a market that has poor SR. It may look promising, but risk is ignorance. Not knowing something about a market increases the risk of the investment.
Our treatment of reliability considers a lot more factors than those explained here. It can get quite complicated, but the most basic components have been addressed. Don’t place a lot of trust in any one statistic on its own without qualifying it thoroughly, even the DSR.
