Jeremy Sheppard on Why the Same Property Data Tells Two Stories

What is the secret to capital growth and how should it be measured? Over what sort of time period? And if we agree that capital growth should be our primary property investment goal, what data can we rely upon to make our purchasing decisions? space with lots of differing opinions.

And while we may think of data as fact and a safe basis for making decisions, it's how we slice and dice it that matters. In fact, the same data set can tell two or maybe even more conflicting stories. So how can we know what to rely on?

Welcome to the elephant in the room. This is the podcast where we love to talk about the big things in property that never usually get talked about.

I'm Veronica Morgan, real estate agent, buyer's agent and buyer's agent mentor, co-host of Foxtel's Location Location Location Australia, author of Auction Ready and co-host of Your First Home Buyer Guide.

And I'm Chris Bates, mortgage broker, recently ranked number five in Australia out of over 18,000 brokers in the annual MPA Top 100 Mortgage Broker Awards. Before we get started, I need to let you know that nothing we say here can be taken as personal advice.

we always recommend you engage the services of an appropriate and experienced professional.

Today we're talking with another property data specialist, Jeremy Sheppard. Jeremy's passion is using property data to find great capital growth locations. And in 2010, he created the Demand to Supply Ratio, or DSR, which is a predictor of capital growth.

And winding back from there, he's very open about buying 15 investment properties in the noughties before subsequently offloading a number of them. To quote Jeremy, I think I've made all the mistakes I can for now. From here, it's all about helping other investors avoid the mistakes I made.

Thank you so much for joining us today, Jeremy. It's a great place to start learning from our own mistakes. It's why we have a property demo every episode. Yeah, well, I could have an entire episode of Property Dumbo's all-personal experience.

So, yeah, I've got a baker's dozen of huge mistakes that I've made in the industry, yeah.

And it's not just yourself, Jeremy. Thanks for coming on. I think the property market, we know that 78% of property investors only buy one. A lot of them don't hold them that long.

And you only have to look at CoreLogic's pain and gain report to see that a lot of people in property don't do well. But I think it's always doing what I've done since 2012. It's really chatting to every people and seeing what decisions they've made over time and then

back testing that and saying okay well that did that really work or didn't work or you know and hearing individual property stories so i mean before we got to go a bit more about your to plant demand supply ratio and all lots of things we're going to talk about today um means there's sort of some key mistakes that you think that are good takeaways for listeners to to avoid

Oh, wow. Yeah, where do I start? Down your alley, Chris. Financing. So I was very aggressive with my finance. I was after the best possible loan I could get. I had some peculiar circumstances with trusts and so on. And...

If only one lender will touch you, you're in a risky position because they've got you over a barrel. So yeah, that was one of the big mistakes and that had me selling properties. So there's one.

Should I go on?

Well, actually, yeah, do go on. But even just leading into that, I mean, obviously people at the moment might be feeling the pinch of increasing interest rates and so that's sort of a lending style issue, right, on the same sort of theme.

So you were sort of forced to sell down because it was those many decisions? No one else would touch me. No? No, okay. No one else would touch me, yeah. But like you said, it led you to sell down some. How did you then decide which ones to sell?

Well, interestingly, at that point in time, I had the demand to supply ratio, so that became invaluable, except for the properties in New Zealand. And I sold those off because, well, I just didn't have the intel.

And I realised after creating that algorithm just how valuable it was. So It highlighted to me that I've got a risk with these properties in New Zealand. I don't actually know what's going on there. And so it just made more sense to get rid of the risky.

So the lack of knowledge was the risk in a way. That's right. Yeah, risk is where you don't know. It's ignorance, yeah. Yeah.

I mean, the finance risk is definitely how you structure any type of investment, whether it's buying your first home or upgrading or your first investment property or when you are leveraging up, how you structure that and maintain the risks with things like interest only.

And you're right, you've got to be dangerous of going to second tier and third tier lenders that may increase your borrowing capacity a little bit. Yeah. You know, do you become a mortgage prisoner and, you know, do their rates shoot up? Are they got better, you know, worse funding lines, et cetera.

So I think there's, you know, you're right. I could go on and on about that.

On the property per se though, so I'd like, is there some sort of key takeaways where you said, like, I'll never buy another off the plant apartment or I'll never, you know, is there certain things that you've sort of learned there?

Well, I never did buy anything off the plan or in any Greenfield estate. But some of the mistakes I have made is, you know, it's funny. When I had 16 investment properties, I still considered myself a bit of a noob because I purchased a report.

Here are some hot locations around Australia, a few reports, in fact. And I just thought, yeah, they're well written. There seems to be a decent amount of research going in here. But the author really had no idea what they were doing.

And I bought in Gladstone in 2008 on the back of one of these hotspot recommendations. I sold that one late 2022. Was it early this year? 14 years later for 20% less than what I paid for it.

So I estimated the opportunity cost there to be around about a million dollars. Whilst the rest of the nation was just going great over that period of time, more than doubled in value, this place actually went backwards.

And there were four other locations that I picked out from those reports that I was planning to buy in.

I only bought in one and it was a similar story and the other two were a similar story too. So that was a real wake-up call for me about, you know, what is an expert? There are lots of self-proclaimed experts.

There are experts at marketing themselves as experts and then they are nothing like experts and there's a lot of them in the industry. So I'm

uh very skeptical nowadays and i think i probably tuned my attention to to research just knowing how easy it is to to go the the totally wrong direction it's a really interesting point right so this is one of the things we've seen time and time again right the the next best hot spot or you should buy in this suburb for x y z reasons um

and it's like this hidden gem and then it's on a report etc um people aren't actually looking for that advice actually all you're doing is selling you know bananas to monkeys to be honest because the monkeys are looking for that they they come to you asking hey where should i invest where should i buy um and then they build a report for that um and they can sell thousands of them because people are looking for that information but you know time and time again i think the results are similar to what you're talking about it's not driven by anything besides

you know, not backed by any research, I guess. What do you think was driving that? Well, hang on.

It is backed by research, but it's just how they've applied that research and, you know, because How to Live With Statistics, that book that Kent recommended to me that I've written in 1954, it's gold.

If anyone wants to understand how statistics can be used to mislead us or to sell a dream, I guess. Um, go and read that book. It's a very easy read. Um, so yeah, it's research, but it's just the, the conclusions that have been arrived at aren't necessarily helpful. So that's right.

Were you, obviously you were into property in the sense that you bought so many of them and you were obviously actively buying these reports and looking, were you always a data guy? Like, were you in like a data specialist in a different field at the time?

Like what actually led you to get into the property data space? I studied electrical engineering at uni and before I even finished uni, I was getting jobs in IT. This is before the dot-com crash, so rates were better.

I got into share investing, realised I was terrible at that and switched over to property because I thought it just made more sense. I'm a human being. I know where I want to live.

uh so i guess the um yeah the data analysis yeah i don't have any history with it i don't have any formal qualifications and i'm not a data scientist some people mistake me as that i would call myself as a property investor first things and then secondly as maybe a property investment data analyst

But it was a penny dropped in late 2009 when back in those days there used to be magazines, things called magazines, and they had data in the back of them. And I used to just look at the data and I thought, how could I use it?

any of this to help make a more informed decision. And then I realised this is all about supply and demand. Every commodity, product, service, the price changes based on the demand relative to supply. And so I started looking at metrics from that perspective. Is this a metric of demand or of supply?

How can I pull those all together and then acquire data for every suburb around Australia so that I can use the same algorithm to make an objective decision on on where to buy. And that led me to looking for more and more metrics.

Now, when I first spoke to you, it would have been back in 2018, I think, and I was trying to work out long-term capital growth. How do we predict long-term capital growth?

And if I hate to, I mean, I'm relying on paraphrasing here because I can't remember the exact, but I fairly certainly, you said to me back then, look, you know, the best I'll put money on is two years, right? That's as far out as I can prepare to put my money on.

So is that what DSR, which is demand supply ratio, is that what it does? Basically, you've looked at all the different inputs and said, right, these are what contribute to this ratio.

This ratio is what's the greatest predictor in terms of how the relativity between supply and demand works. Is it still two years? Is it still to say that, you know, this is what I see in the near future? It's actually three years.

So most markets seem to play out in about three years, but it can go longer. It could be like half a dozen or it could be over in 18 months. So when there is one of these imbalances in supply and demand, it doesn't take long for the market to respond to that.

And so forecasting using this approach is very difficult over a longer period of time. And although there are some metrics that lead to higher capital growth over the long term, there are a couple of issues with pursuing long term growth.

I mean, first of all, you need data from, let's say your long term period is 20 years. So you need at least 30 years of historical data in order to find some metrics. And things like auction clearance rates, vacancy rates, things like that, they're just not that old.

So the amount of data that you have, historical data from 20, 30 years ago, is a little bit threadbare. And then there's the issue that over time, things change.

So we don't have any data now about where the key locations are for transport via teleport, because that thing hasn't been invented yet. But you can bet that if there's something... Maybe it's... magnetic levitation tubes or something, vacuum levitation. Who knows what the future holds?

But what we do know is that a significant change to technology can result in a significant change to property markets and what features people value. There are some that haven't changed, and I would never bet against the long-term trends.

But, yeah, it is difficult but not impossible to pick outperformers over the long term.

So when you talk about you need 30 years' worth of history to be able to sort of look out in the, you know, ahead 20 years, isn't it... And then so much changes, of course, that makes it very, very difficult. And isn't it then...

to understand what the underlying fundamentals are so you can sort of make a better gamble on an area that shows consistency of what we do know contributes to capital growth as opposed to what might happen in the future.

Because I know there's a lot of speculative investing in, you know, like greenfield sites and airports and, you know, fast train lines and all that sort of stuff. But there is risk inherent in that.

Whereas if you're actually looking at locations that already have a certain amount of amenity or connectivity or lifestyle or whatever, and you think, okay, well, that's likely to continue over time, those other areas need to have that major infrastructure to be put in place or to be created so that those areas share the same sort of amenity.

Wouldn't you sort of to say, okay, well, there's ways to predict some areas are going to be a safer bet for for the future or are you sort of more focused on? That's all we can do. All we can do is look at historical data.

So if there is any evidence in, say, for example, proximity to CBD. So this is a hotly debated topic and one that I've looked at recently

dare I say it, exhaustively and there's nothing in it that I can see in the data. I mean, I'm always looking to try and improve the algorithm. Distance from CBD is very easy to calculate but it's just not a predictor of future capital growth. So...

Yeah, all we're doing is looking for these historical metrics that we can rely on without any speculation about, you know, magnetic levitation vacuum tubes or driverless cars or anything like that.

So with the, I mean, if you're trying to predict something over three years and you're basing that decision of whether to buy or not over what's going to work over 10, 20, 30 years, or is it purely just, hey, this area looks really good over three years because, hey, very low listings, growing population, all these metrics that you want to sort of use, which we can talk about.

But is the decision then to sell, like is the... is that get in while you can, get in before the market knows it's a good place to buy and basically get out before the market, you know, decides the run's over. Well, actually, the market has probably already started.

So in order for demand to greatly exceed supply, it's already gone through marginally exceeding supply. So there may have already been some capital growth. But the good thing about investing in real estate is it moves like an oil tanker, not a runabout. So it takes a long time to turn around.

So if you do have your finger on the pulse, you can jump in without missing out on too much growth. And typically what I've seen is growth, you have a growth surge of, say, three years. Now, that's, you know, everything varies dramatically, but that's sort of what I'd say is typical.

And then you see, say, the next three years would be pretty much flat or very mediocre. So it's very lumpy, the growth. There's nothing smooth about it. So the idea is recognise a market where demand is smashing supply, enter that market, and

And then if you're keeping your finger on the pulse of that market, it will reach a point where it's pretty clear that it's peaked.

And there's good evidence to suggest that going forward, that's going to be a fairly flat market and you could actually move your money elsewhere, depending on if there's an elsewhere that's so much better. And what do you think the minimum return that you need is

over that short timeframe to make this a good bet, right? Because transaction costs, let's just call it circa 10%. And you're going to be paying capital gains tax, so you're going to lose 25% of the gain. So, you know, what return? Are you saying 30% is the minimum?

Well, you nailed it with the CGT. So if someone has seen very little CGT, I put the figure at 12%. You said 10% for transaction costs. I put it around about 12%. So that's without even considering CGT. So if you bought a dud five years ago, it's been flat.

One fortunate circumstance, if you can call it that, is that you don't have to pay CGT when you sell. So then you're... then you are looking at those transaction costs of about 12%.

So let's say you forecast over the next three years the replacement property market is going to grow by 40% and the current property market where you are invested is going to be flat. Then that's a no-brainer. That's really easy.

But once you introduce CGT, that changes a lot of things. There's a property, the first property that I bought, I've got so much capital gains tax liability there that every time I do this analysis, the data tells me you've got to hold.

But there are circumstances, you'd think it's quite infrequent, but there are plenty of circumstances where there is an opportunity for investors to sell and it's clear. Well, I mean, there's no point buying it.

If you're going to try to buy something over a short time and you're not focusing on the long term, my worry is that you create a problem for yourself, right? You put a rod for your own back, right? So you buy something and 500, it goes up 40%.

So it goes up from 500 to 700. You get it right, right? You've timed it perfectly. And you're probably right. Potentially it was 450 and you had to buy it at 500 because the market already moved. It was already moving. You were chasing a moving ball a little bit.

So it's got significant growth this year. But once you take out the transaction cost 60K, you made 140K, then you're going to pay 25% of that in CGT, probably, or 20, 25, depending, because you hold it more than 12 months. So now you've got 100 grand in your pocket.

And that's if it goes really well on this scenario. But you can't hold it because you know that it's going to be lumpy. So... You're better off just getting, paying that, you know, 35K of CGT to then potentially go and recycle it.

Is that sort of the way that this really works is you're constantly trying to find the next best spot? And how does it, if you, you know, because if, for example, you do sell, are you then at the mercy of the market of whatever it is in 2026 trying to find another asset?

And, you know, the whole, you've got to go through that whole process again. It's a lot of legwork trying to find good assets, right? You can't just all of a sudden pluck one. That's right. Yeah, well, obviously it all comes down to the numbers. And you mentioned a term there, recycling costs.

So that's the term that we put for CGT, agents commission, legal fees, and then re-entering the market, paying stamp duty, more legal fees. So all of those costs is the trouble that you're going to. And you can factor in what your time is worth.

Like you said, you know, there's a lot of headache involved in selling and then finding another market and buying again. What we're doing is comparing recycling costs with opportunity costs. So we forecast the growth for the market that you're currently in, forecast the growth for the alternative, you know, the best.

And if there's a significant difference, like opportunity costs heavily outweighs recycling costs, then it makes sense to sell, to move on. If, on the other hand, they're even just close, then, yeah, the best advice is just sit tight.

Because, yeah, I could see that the DSR would work as a really great tool in deciding when to sell or even as you, you know, in terms of your priorities.

If you were going to sell some properties, if you had more than one that you wanted to sell, it could be a very useful part of that decision-making process to work out how you set your priorities there.

In terms of buying with an intention to recycle, yeah,

then that just amps up the risk hugely right so you've got to have be a particular type of investor with a particular appetite for risk to do that because what if you're wrong on your on your choice of your next one you've just undone absolutely everything and the other thing too that in your your dsr is that for a suburb or for sa3

That's for suburb. Yeah, we can aggregate it to SA3, but it is at the suburb level. But what you're saying there about the risk, that's why you would want to factor in a significant margin for error.

You would want the difference in the two forecasts, your existing property and the replacement to be quite considerable. And what would you, like if you're going to put a number on that, what would you say like in terms of a percentage?

Oh, if the DSR was 70 versus 35, I mean, that to me is a no-brainer, yeah. So, and on that though, also, how do you account for within any suburb, you will have properties that outperform and properties that underperform?

So, because this really then, for this to be successful, if you're going to use the DSR to pinpoint a location to invest, to then monitor DSRs to work out, well, what's next after it sort of hits its peak and plateaus?

If you don't master the art of choosing an individual asset, you could undo it all by picking an underperformer or you could supercharge it if you knew how to actually select an overperforming asset.

Where does that come – because that's something that I haven't come across data that actually will pinpoint individual property. Is that something that you're working on? Yes, and –

I originally, you've probably heard the phrase, I've heard this so many times, the suburb does 80% of the heavy lifting. And I assumed something similar. I thought that's probably a good guess. But I knew, I know that everyone who says it is just guessing. So nobody has done any research on it.

So Luke Metcalf, founder of MicroBurbs and I, have started work on this just recently.

And I must admit that I was a bit surprised at initial results, but there's something to be said for, well, yeah, I don't want to spoil the announcement that we might have later in the year. But the point is what you're saying, asset selection is definitely important.

And, yeah, the suburb is the first step, and you could undo what you've done. Whenever I have clients ask me, you know, where's the suburb that I should be buying in, Jeremy? Here's our brief. And so I'll do the research for them and give them a list of suburbs.

Straight away, after I've given them the suburb, they start asking, well, what should I buy here? Which property? And the only advice that I can give at the moment is... buy something that's typical because that is what the DSR is based on. We're aggregating this sort of data.

So if you buy something that's atypical, you could, like you said, Veronica, supercharge it or you could undo it. So play safe and just buy something that's typical is the best advice I can give them. But so, Veronica, what is it that you're looking for then in a property?

Like as opposed to the suburb, you've got the suburb figured out, what are the sort of features that you'd be looking for? Well, there are two features, I guess, fall into two categories. First of all, you've got to look at universal features.

For example, you don't want to be on a main road, right? You don't want to have the wrong aspect, particularly if you're in Brisbane, for instance, you don't want a garden that faces west. If you're in Melbourne, you don't want a garden that faces south. So you want good natural light.

You want to be close to transport without being too close to it, you know, because you don't want to be impacted by the negative aspects of, say, a train station, but you certainly want to be within an easy walk to that. So there's things that are sort of...

universal under along those lines and then there's things that are specific to every location like for example my office I often talk about Balmain if you bought a weatherboard cottage in Balmain then you'd be doing you know that's great people love weatherboard cottages in Balmain they're pretty and and you know and they're really in keeping with the area but

And there's a number of different types of typical properties, though, there. You know, if a weatherboard cottage is one, a Victorian terrace is another. You know, there's a number of different types of properties that you might call typical.

But if you went and bought a weatherboard cottage, say, in another suburb, maybe you go to Gladesville, and a weatherboard cottage there is not as typical, not as well regarded because it's a bit atypical, you know. So...

That local understanding of knowing what actually works in one suburb will not work universally in other suburbs. That's quite key to it because people, and I know we see this certainly in our areas that we operate in,

as a buyer's agency and we see buyers that come from out of the area and what they gravitate towards and they often will see an opportunity because obviously no other buyer's interested in it because the locals don't like the atypical on the lower end of the scale and they'll think they've got a great bargain.

But when they come to sell it, they'll have to find another buyer like them and there's less of them around. So that typical thought pattern is what underpins what we look for in a property. Yeah, well, I mean, all the DSR is looking at at this point is, yeah, suburb level.

So let's get in the right ballpark first, that sort of thing. Yeah, you have to. That's the question and answers, yeah.

Jeremy, if you were looking at a suburb, just what's the, like, go through your metrics, I guess, and I've only got a few of them, so it could take a while, but what are some of the main ones you think that, you know, like when I'm looking at a suburb, I'm just straight on a satellite, right?

I'm just like, where is this property located? Like, I want to know exactly what's around it. You know, and, and so what's your go-to, like, where do you initially, someone pulls a random suburb, I just say, right, Brunswick in Victoria, right? Everyone knows Brunswick or what most people do.

What are the things you want to know about Brunswick to know whether it's got a good DSR score, right? Like what are the, what's the first one you go to? Well, I really just look at the DSR. I don't really look much at the base metrics.

So the algorithm is built in such a way that it puts all this together. And quite often I'll have people contact me and say, Jeremy, how come this suburb has a 0% auction clearance rate and yet it's got a very high DSR? And that's because there are multiple metrics.

I've never found a market that has every metric is above like the benchmark, the norm. There's always good and bad. And half of that isn't because there's this tug of war between supply and demand. It can often be simply because the data is stuffed up. There's some anomaly in the data.

And that's why you need to look at so many, to look at a property market from so many different angles. Each metric on its own is unreliable, but each metric tells a little story about the possible nature of supply and demand.

But to try and answer that question without totally dodging it, market cycle timing is one of the better metrics.

I would probably have a look at that sometimes. But I want to be careful that I don't allow my own bias to interfere with what the DSR is saying. Because I've done that in the past. The media have contacted me and said, Jeremy, give us your top 10 for 2020.

And I've looked up the DSR and I've filtered out a few because I didn't like them for whatever reason. And if I just left them in there, the result would have been better. So there's so much bias in what we do, it's really hard to be completely objective as human beings.

But we are getting to a point in this AI age where sooner or later we are going to have to trust the machine.

So you're lifting, if you lift the lid on the DSR score, what are some of the elements that are in there then? I mean, you've got, I mean, market cycles, they speed up, they slow down. I mean, they're pretty erratic these days.

And also different suburbs are on different supposed property clocks and all that sort of palaver. But also, I mean, I think, tell me if I'm wrong. I think like, for example, I seem to remember that one of the metrics that goes into your algorithm is some search data, property search data.

Is that correct? Yeah, yeah, online search interest. Yeah, so it's the number of searchers per property that are for sale. So it's some kind of vague reflection of demand relative to supply.

And this is searches in Google though? Like this isn't searches on domain or real estate? Real estate, realestate.com.au. Oh, you can get access to that data, can you? Yeah, right, on a suburb level. Yes.

Some of the other metrics, ripple effect potential, neighbour price balancing is another one where we're comparing the price of this market to its neighbours, unit to house value comparisons.

There's all the usuals, vacancy rates, auction clearance rates, selling speeds, you know, days on market, those sorts of things. There's a large number of metrics in DSR version 2. So back in 2010, I launched the first version of DSR. It got an update, version 2, in 2015.

And Luke and I will release version 3 in a few months. Cool.

And what about demographics? Yeah. Is that part of it? Yeah, there are some demographics that we found useful. A lot of them are useless, but yeah, some are.

I'd be keen to understand what you've come up as being useless because, you know, there's a lot of myths around property data and there's a lot of misinformation, false police biases, all that sort of stuff that you've alluded to and you've mentioned there.

What sort of things do you think that people would expect to be in a measure like this, in a metric like this, but perhaps would be red herrings? Yeah. Population growth. Population growth is one that everyone just thinks, oh, of course, the more people there are, the more demand you've got.

The problem is at the suburb level where you've got to buy practically, it's more a reflection of supply than it is of demand because you can't have... a high forecast population growth without there being properties there for people to move into.

So the properties come first, the supply comes first, and then if demand is sufficient, the people move into those properties. So that's one metric that's not there. Another one is incomes and income growth.

So looking at the census data on family household income, got into some... pretty intense arguments with people in the past, the old school, here's what the data is saying, this is what we have to accept. It's just an unreliable metric.

And keep in mind that I'm looking for anything I can to make this algorithm better. I mean, I use it for my own portfolio. So If incomes, that particular data set, if that led to being a predictor of capital growth, it would be in the DSR straight away. But unfortunately, it doesn't.

I'm on a personal mission to help more people make better property decisions. And you can find out all about what I'm working on at veronicamorgan.com.au. And there you'll find resources for first home buyers, details about my Buyer's Agent Mentoring Program,

access to suburb help for investors, or if you're looking to buy your dream home or an investment property in Sydney's inner west, eastern suburbs, or lower North Shore, you can connect with my team at Good Deeds Property Buyers.

If you're thinking about buying your first home, upgrading to a new one, or purchasing an investment property anywhere in Australia, we would love to carefully guide you through this journey and importantly, get the finance right. Please reach out via our website, wealthful.com.au.

Don't forget that you can download our free full or forecaster report. Which experts can you trust to get it right? Theelephantintheroom.com.au.

Yeah, let's just go through them. So one of the beliefs is that, you know, people need population growth causes prices, right? And you say that that also will be more supply, right? People will build more places, right? Because you can't have population growth without more supply. Like otherwise...

Um, but we, potentially you can, you can under build, right? Like you can, so, you know, you were saying what was driving Gladstone, like what was driving it back when you bought, um, and then the story of what happened after, right? So there was oversupply was built for the population growth, but

You know, if you're talking sort of a regional town that isn't... Just before you go to regional, say for argument's sake you're looking at population growth and you assume a certain area's going to go off and it doesn't build enough, but that's a particularly desirable suburb, then the population growth is forced to compete more or those people that want to move there are forced to compete harder to be in that suburb and then that forces a ripple effect.

Do you know what I mean? So I guess...

Even though you're right, they can't buy a property that doesn't exist, but you could see how it could impact areas that are in demand if the population, if the new entrants or they might displace previous residents or the previous locals or perhaps it creates a ripple effect into other suburbs.

So it could still have an impact. I'll give an example that, you know, the Chinese buyer and there's a big mythical story

a lot of misinformation around Chinese buyers, but there's been certain suburbs in Sydney and certain suburbs in Melbourne where there has been quite a lot of Chinese money come in and really changed the whole makeup of a suburb over some time.

So them buying into an area that they weren't previously locals in have displaced people that potentially would have wanted to buy there and potentially got priced out by them. You know what I mean? So even though more houses weren't created,

the population growth still did change the makeup of the area and has resulted in some price growth in a lot of those areas. Well, I mean, if you've got a population of 3,000 in a suburb and then 300 move out but 300 move in, then you've got the same population.

Yeah. Yeah, but you got more competition. At a city level. So, you know, um, Sydney's population goes from five to six million people. Yeah, right. Okay. Newtown's population might be the same, right.

But Sydney's population has meant that people buying into Newtown now are different to what we're buying into Newtown five, 10 years ago. So. Um, because they're, you know, the, maybe they were just new kid buying to, you know, born today in 20 years time.

That's a population growth that is going to compete for that same house in new town. And so it doesn't have to be, you wouldn't probably go to the suburb. That was really hard, right?

Same as income growth, because like, that's what you can talk about the next point, but population growth of the city, um, you are creating more demand and you are changing the buyer pools in each market. So do you agree with that at a city level it matters?

At a macro level, at a macro level particularly international migration.

because the properties have not been built for them and they they all reduce vacancy rates to to zero uh until more properties are built but when you see these um these they call them growth corridors the property developers and they've got a forecast for population growth those sorts of things and of course an investor has to be practical about it let's say yeah sydney is getting all these uh migrants from melbourne is

which suburb you're going to buy in. So at some point you've got to make a micro decision and that's where population growth is not helpful. I think you're actually bang on. So when someone says, oh, you've got to buy in this area because population's going up, right? Absolutely.

That's, you've hit the nail on the head.

However... We're not fans of growth corridors. No, absolutely not.

But if forever the...

you know australia's population people living longer right that's going to lead to more population growth because we're going to have a higher population because we're living longer right um demographics we have more kids um or less kids you know etc so i think that so the income growth thing is

Just because we had an episode with PK, which I'm sure you watched, right? And we got into a bit of debate around this income growth. And a lot of people listen to your chart. Apparently, there's this chart that I haven't seen, which I'd love to see.

But is that based on census data, like the income growth every five years in a suburb?

isn't a good predictor of growth because uh so income growth doesn't matter is that sort of where that belief comes from because if you look at that census incomes don't go up that much um and so then growth shouldn't be that much whereas you you know some people believe that you need income growth for prices to go up is that what you're sort of thinking

Well, all we can do is look at a data set and see whether it is a predictor of future capital growth. So if there's some problem with that data set, then, yeah, we can't use it. And that's the case. I looked at higher income areas versus lower income areas.

I looked at income versus estate for a suburb, suburb income versus estate. And I looked at change in income over time. And after trying a few things, I couldn't find anything that worked. There are better metrics out there to spend time on.

But if a novice investor is rummaging around looking at census data and trying to calculate their own income change, it's just not going to help them. I think you're right there too, because absolutely the census data is pretty pointless, right?

Because in that suburb over that five years, things only change not that much because only 15, 20% of properties probably transact. You've got some at the low end, which is maybe the apartment, some of the houses and incomes in the area don't really change that much because only a small portion are changing.

then potentially they hit their peak earning capacity, right? And the next lot of people moving in is earning higher incomes than they were five years ago, but that doesn't change the metric at all much as well because 85% of the people are still on the same incomes as they were five years ago.

So I think you're right. When you're taking a census data to predict income growth,

the it doesn't match up right you'll see that you know incomes don't change in the suburb that much over time however that doesn't mean prices aren't changing and the people who are buying in that suburb today aren't different to what they were five years ago um and that's real income growth in my mind is the people who are buying in the market today what are their incomes versus who were buying in the suburb 10 15 years ago is do you align with that thought process or does it not make sense for you

That's one of the problems I suspect with why it doesn't work is because you don't have to live in the suburb to buy in the suburb. So how do you know the incomes of the people that are about to make a purchase? But it doesn't matter what their incomes are.

If they turned up to buy, they've spoken to their mortgage broker, they can afford to buy there. And when you think, how does, you try and think of the story, how does income growth channel through into, flow into capital growth.

Like, you know, someone comes home from work, busts through the door, announces to their partner, hey, sweetheart, I got a pay rise. And their partner replies, I did too, same day. Let's ring up CoreLogic and see what the new value of our house is.

No, no, they would race out and they'd buy in a better suburb. Exactly. Yeah, exactly. That's another reason why it probably doesn't work. Yeah, because basically higher incomes mean that people can pay higher prices.

It brings in a different type of buyer potentially or it means that that buyer moves from one suburb they're looking at to a different suburb. And so it's volatile. It's static, although there might be only a small percentage of properties that transact. Over time, the make-up of that area –

you know, will make a difference to prices. I mean, also, you know, you think higher incomes mean that people can actually renovate their houses and improve their houses, and that's actually going to lead to price growth, but then that's not ever taken into account. You know, do you actually do that?

Do you look at price growth and say, well, there's a percentage that's possibly through improvements, right? Yeah, so that would be capital injection, but it can still come out in the figures.

So let's say, for example, in January you have, well, let's say you have a suburb where there's only two possible values for a property. It's either $500,000 unrenovated or $600,000 and it's a renovated property.

Now, if three of those properties sell in January, let's say two of them are unrenovated and one of them was renovated. So you've got Two sales at $500,000, one at $600,000, so the median is $500,000.

Now, in December, you still have another three sales, but two of them were renovated, $600,000, and one of them was unrenovated, $500,000. So the median has grown by 20%. I know. Prices have gone up. But of course, it's just capital injection.

Now, that's very difficult to detect those sorts of fake capital growth. But in order for there to be this indicator of incomes growth flowing through into the property value, a lot of ducks need to line up. So first of all,

Out of the residents in a particular suburb, 33% of them are going to be tenants. So they're not going to spend their increased income on the property. So you can forget about them. Out of the remaining 70%, when you get your pay rise, what are you going to spend it on?

Yes, you can renovate your property, but you can also just pay down your mortgage. You could buy some shares. You could buy a car, a house. You could go travel, go on holidays. You could buy furniture. There's so many things you can choose to spend your money on.

Only a small portion are going to spend their money on renovating the property. And then even once they have got their property just how they like it, Are they going to sell it or are they going to live in it? Because if they don't sell it, nobody knows about it.

So it can't influence capital growth. So there's a lot of ducks that need to line up in order for that income to flow through into the value of the property.

But what would be a really big mistake is if you looked at your research and said, oh my God, there's no correlation between income growth and capital growth. Therefore, it doesn't matter what the incomes are in the area in which I'm buying. Because those two things aren't necessarily the same, are they?

Well, gee, I don't know how incomes can be useful at this point in time. No, but what I'm saying is that the inverse isn't true. It isn't that they're useful. We agree. They're not useful based on what we've just all been talking about.

But equally, it's not like you say, oh, but I'm going to totally disregard income. I'm going to go and look in low socioeconomic areas because income doesn't count. I think income will play a part, but a different part to what potentially a lot of investment specialists might be inferring, perhaps.

Yeah, well, it's interesting you bring up the lower socioeconomic demographics because I haven't found anything yet to suggest that they underperform.

Okay, now I did a little bit of research before our meeting today because I thought I'm going to pluck out the suburbs that all three of us live in. And so that's random enough, isn't it? So now I've got it right though, Jeremy, you're in Mount Druitt, is that correct? Yes.

Yes, I'm in Newtown. I also added in Balmain because I talk about it all the time, it's where my office is, and I added in Bayview, which is where you are, isn't it? If I've got that right, Chris?

I'm in Bokala, but there ain't another Bayview. Oh, no, I got it wrong. Let's say I live in Bayview. I'm happy. Bayview's a nice suburb.

I thought for some reason you went to Bayview. So I looked back over the last 30 years, and so I just thought this would be good to have a discussion around. I don't know the answers to this. All I can just say is, so I should have checked the CEFA scores.

So that's the socio- Economic Index for Australia. Yeah. You probably could check that really super quickly, Jeremy, because you've got a handle on that. Yeah, I've shut down the database, so maybe not so quick.

So my guess, and I'm going to totally guess this, but I'm guessing in Mount Druitt it might be sitting around 6 or 7 perhaps, and Bayview and Newtown, Newtown maybe about 9 maybe or 8 or 9, Bayview, Balmain probably around 10 is my guess, right? So that's just guesstimates.

If we look at the annual growth rate, so the annual, so an average over 30 years, Bayview is 8.1%, Newtown 7.9%, Mount Druitt 7.5%, Balmain 8.2%.

So if you extrapolate that, though, I mean, this is the thing that if you – so they're all, you know, within a band there, which is tracks according to the aggregate for the whole of the country, right?

But if you looked at opportunity cost because – and I'm only going on this because you said that incomes don't matter –

that if you said, right, well, the income, the average income for Mount Druitt, which I should have checked, but I know it would be lower than Balmain, for example, you know, if you look at Balmain being the best performer, Mount Druitt being the lowest performer, there's an opportunity cost

in Mount Jewett of that difference between 7.5% annualised growth and 8.2% of $186,000. But in Balmain, if I flipped it the other way and said instead of growing at 8.2%, it only grew at 7.5%, that would be a difference over the 30 years of $500,000. So it does make potentially a difference.

So this is, look, I've just plucked these randomly and this is just what the data tells me. All right. I'll tell you what you're doing wrong with your research. Okay. And this is a mistake that a lot of analysts make.

And I've seen this done by the RBA, Real Estate Institute of Australia, SGS Economics and Planning, a whole bunch of them. What they do is they only look at one timeframe.

Now, probably about, I don't know, maybe eight or nine years ago, maybe it wasn't that far, I was looking for two property markets where one was diverging from the other. So you plot their capital growth as a cumulative growth chart, so not their actual dollar values.

And I was trying to find one that diverges just continually over a long period of time from the other. And so I might try Mossman versus Mount Druitt, Sydney versus Launceston, any combination. I kept going and going and going, and I couldn't find one. Well, I could, but it took forever.

And one thing that I observed over this decade of looking at these historical growth charts is there is a common or pattern in the data.

One market will diverge, outperforming the other, and the other will catch up. It'll be like a pinch point, sometimes crossing over and beating the original one, only to be caught up again in the future. So they either crisscross all the time or they have these pinch points where they come together.

the winner of a capital growth race is determined by when you set the start and finish dates, not where you buy. So you are looking at one 30-year timeframe. In the world of data science, what they would do, they call this cross-validation.

So if you had a 30-year horizon, a 30-year data set, you would run a 20-year window across that multiple times. to see whether anything changes. Have you got the same pattern, regardless of where you start and finish that 20-year point in that 30 years? Do you have the same pattern?

So those figures you're reading out now, five years ago, they could have told completely the opposite story, and five years from now, they could tell completely the opposite story. And that's what I have found, that there is a tendency over the long term for all property markets to grow at the same rate.

So if I analyse this better, I would be saying, right, let's look at it after five years, so 1993 to 1998, then look at it after 10 years, say 1993 to 2003, then look at it at 15, and look at it at five-year increments, and then we could see whether it followed the pattern that you're talking about.

Well, you can keep it to just 20 years. If you've got 30 years of data, you could probably even use 25 years. But you just want to roll that across each month, do it repeatedly again and again.

It's a sampling. It's called sampling. So you take a subset of the full population of data and then you see, can I split this up in some way where it tells me a different story? And you will find that.

So if I was going to look at comparing two areas and one of them was Gladstone, would that tell us the same story?

Well, there are always exceptions to the general pattern that follows through, always. I mean, you know, the general rule about playing the lottery is you lose. But there are exceptions to that.

But we can't ask people who have won, what is it you did differently on the day when you filled out that coupon? That's just completely random. So Gladstone has had a specific circumstance which has interfered with its last, you know, 10, 15 years of growth. But long term, it will come good.

They have two specific circumstances, really. The first one is that they're into the mining, you know, boom. The first one was that investors and investor advisors were basically pushing it and so heaps of people bought in there. Do you know what I mean?

So in a way, it's what got people into the market in the first place and then basically what undid them. And I would think that potentially...

I'm wondering whether as a metric in your DSR, maybe a slightly different metric would be increased investor activity in an area and, you know, at the same time decreased stock, you know, rapidly decreasing stock. Is that a predictor of what's going to be, you know. Over supply. Over supply.

Well, yeah, or not even an oversupply. It's an over demand.

Yeah, it's interesting you bring this up because I've heard a lot of people in the past talk about these self-fulfilling prophecies about, you know, if you pump up a property market, a particular advisor pumps up that property market and that pushes it up.

and it just keeps going up because there's more demand. The DSR, this market cycle timing metric, is sort of self-regulating in this respect. So if there's been too much recent growth, the score becomes lower.

So if people were to invest in this market, push up the demand, because I announce XYZ Suburb is going to be the next growth suburb and I really pump it up, then the DSR will come lower and lower the more and more growth you get in that area.

And so it would be less and less on people's radars. So it does sort of self-regulate. There is no possibility for a pump for three years. You can't keep promoting a market as being an outperformer for three years or two years even.

Now, I don't think any particular investment advisor has got the volume of individual buyers probably to tip a market like that. But these groupthink in certain forums, for argument's sake, and you do, you get in certain investor forums. I don't spend an awful lot of time in there.

But whenever you do go in there, you do certainly see there's absolute trends in terms of what people think about is, you know, the place to invest next. I mean, currently, for instance, a lot of investor money has moved from Adelaide to Perth. And, you know, so therefore it's not just one advisor.

There's a whole bunch of people. There's people that have an advisor. There's people that don't have an advisor. They're all sort of following that trend. So how does it impact your score? To have more people buy into an area and less available stock, why wouldn't that send your DSR score up?

Well, it would. Yeah, if there is legitimate demand there, more buyers. It might not be sustainable, but it's still legitimate. These people have made a decision that that's the place that they believe is going to go, you know, is the best place to invest in.

Your early warning system wouldn't work because it would still be reflecting the fact you've got competition and more people flooding in and less stock, right? Have I got that right? Yeah. Yeah, yeah. If people are buying there and supply is weak, then prices are going to go up.

And so the DSR is measuring that.

But the question is... Is it being artificially inflated by some forum? Well, you know, I'll take artificial equity any day. If I can sell and make a profit. Yeah, if you're selling, that's great.

But if you're buying in there because you're thinking, I'm on the beginning of a growth cycle and it's been pushed by, it's been manipulated effectively. Right. There are other metrics apart from simply how many people are buying right now. You know, that's like transactional metrics.

There's other things like ripple effect and market cycle timing. And if those metrics don't line up because this has just been pumped up for whatever reason on a forum, then it won't appear with a high DSR.

So Jeremy, with the short term, so let's say someone doesn't want to try to, you know, basically take a punt on short term.

Yeah, with the DSR and we're backed with evidence and they don't want to say, I want to buy in this location that I think is going to do well over three years.

if someone's thinking sort of 30 years right because a lot of people like you say they don't have the time or the energy to actively trade property and they don't want to take the risk of it right because we talked about when it goes well yeah you make a little bit of money but if it goes bad doesn't grow opportunity costs wasted borrowing capacity you know loss etc the when you make a bad property decision you may not buy another one etc right so there's risk involved with just making bad decisions

Do you believe that then they should just almost like just pick any suburb really? Because the belief is, you know, it doesn't really matter where you buy over the longer term because they're all going to go back to the mean.

Like, does it matter whether you, like if you've got, let's say you're buying the inner north of Sydney, does it matter if you buy in the established suburb where they're not building anymore or the suburb where they're building lots of apartments, right? And the demographics changing or the suburb where,

It's got a very strong school zone or it's a bit of a black zone for public transport or it doesn't matter because they're going to grow the same rate over 30 years. Is there no point in overlaying any thought to that because they're going to go back to the mean?

Well, yes and no. So first of all, there is a general trend towards the reversion to the mean, but there are some indicators of outperformance over the long term. It's just very difficult, that's all.

So if you were to enter a market with a short-term outlook and things didn't turn out so well. You can just sit back and hold and you might not want the aggravation, the effort of trading.

You can just sit there and you will get some decent growth eventually. And One thing that surprised me about the DSR, it's been around for a while now, well over a decade. I was looking at its performance over 10 years on the assumption that's somewhat long term.

It hasn't been around for 20 years, so I can't do that analysis.

But I was surprised to see that although all this capital growth mostly happened in the first three years, still after 10 years, it had been outperforming the national growth rate, not because it's continuing to grow faster than the national growth rate, but because of the first three years of exceptional growth, it was still ahead of the national growth rate.

So even over 10 years, it's not that bad. And... To answer your question, a very large number of suburbs, you can go pretty much eeny, meeny, miny, moe.

There are some exceptions to that, like Gladstone, like mining towns, some fringe suburbs, some greenfield estates, and in particular, unit markets, because you've got no control over future supply. So if you buy in a built-up area, you buy a house in a built-up area, you buy an established house, like an old house,

in a built-up area, it's almost impossible to fail over 30 years. Okay.

And then does it matter, though, because you said, like, running those time series, does it matter when you buy in these suburbs in terms of... And, like, you know, for example, in 2021, right, or 2017, or, you know, at the height of booms, you can buy poor streets and weird aspects and weird blocks and...

for pretty high prices right um so does it matter when you're buying like you know for example you might buy in the poorer streets and the weird aspects at better prices in 2018 or 2022 right um so does it matter when you enter the market at all or do you think that ultimately you shouldn't worry about anything you should just just pick anything it doesn't really matter

Well, it's hard to say definitively it doesn't matter because you used the term weird. I haven't analysed weird properties. I don't know how to classify them. I don't know how to numericalise that.

So it's possible there could be some feature that a property has which over the long term will underperform that hasn't been analysed. But in order to know which features... to pursue, it has to be analysed. We can't just guess. It's 21st century now. We don't do that anymore.

And so you think the local knowledge of

agents in an area that can see when properties go for premiums in booms they go for they hold value in downturns they always have a lack of stock they always have ridiculous demand out of area buyers want them do you think that those factors that they see on in through working in that maybe that's hard for them to analyze because you're analyzing individual properties inside that's hard but

Do you think that that's all a bit of a waste of time? Or do you think there's not enough data, there's not enough evidence for you to put this into DSR? So for your point of view, yeah, it's hard. Yeah, there might not be enough data now.

I mean, eventually this is an inevitability that eventually these machines will know more than we can possibly know. And whatever we can personally observe in our careers, first of all, we're subject to our own bias.

You know, you attend an auction at the start of the morning and the fourth one happens By the end of the day, you're caffeine depleted. You have a completely different appraisal of what happened because of subjectivity, because of bias.

And the beautiful thing about the algorithms is that they approach data in the same way, whether it's in Coogee or Koolangatta. It doesn't matter. They're completely objective. The second thing, of course, is that there's literally thousands of suburbs, and we cannot personally get around to that many suburbs.

And, of course, there's the history. I mean...

versus a career that's maybe only 15 years. And humans, we can only weigh up in our own minds maybe half a dozen variables, whereas algorithms can cope with hundreds.

So even if the data isn't available now, and I would say that it definitely is to make a difference to property investors, it's available now, it is an inevitability. This is going to happen. We are at some point going to have to trust the machines.

Have you ever done the sort of segmenting a market, right, and saying, okay, well, the top 20% of properties within a suburb, their growth rate versus the bottom 20%? Has that proven anything at all? Yeah, there's still a fair bit of work to do there.

One thing I was looking at just recently, just late last week, was the maximum sale price of any property in a suburb versus the minimum sale price of any property in a suburb, looking at that over 30 years. How has that changed?

Because if there is this thing like an investment-grade property that outperforms over the long term and there's the complete polar opposite in the same suburb, you know, a trash property that continually underperforms, then we should be able to see the percentage difference between min and max for a suburb widen over 30 years.

But it had actually narrowed.

uh marginally narrowed and i've done this sort of research before back in 2018 where i looked at the standard deviation of properties over the over the long term within the same suburb and yeah things had definitely narrowed they hadn't widened which tends to suggest uh that that over the long term it is yeah it is a pattern in the data that more often than not uh this you know properties will grow at the same rate that's not to say that you can't find exceptions

This is fascinating, Gerry. I think that this is fodder for another conversation. And I think because I've done research on individual properties, I've tracked many properties in different suburbs over multiple sales, and I can see the gap widening on individual properties.

But to say that the aggregated data doesn't show that is quite fascinating. I'd love to explore more why that might be the case. But, you know, this has been a very interesting chat, Jeremy. I hate to pull it up short. We're under a time constraint today.

But let's hope to have this chat in, you know, more fullness in the not too distant future. Sure. And thanks very much for inviting me on your show, guys. Thank you, Jeremy.

If you have a question that you'd like us to answer in an upcoming Q&A episode, you can send us a voicemail or written question via the website, theelephantintheroom.com.au or you can email us directly at questions at theelephantintheroom.com.au.

If you like what you're hearing, please share this episode with others you feel would benefit. And while you're at it, why not leave us an iTunes review? Five stars would be great.

I know that sounds a bit cringy but we have it on good authority that every review helps make it easier for other people to find out about us and hear what our amazing guests have to say.