I'm Jeremy Sheppard. And I'm Damian. We use data to expose deceitful property experts and their marketing BS. This is the Expert Busting Series. This is episode 14 of the Expert Busting Series. Today we're looking at infrastructure projects and do they make a difference to capital growth?
Jeremy. Yeah, research of infrastructure projects has got to be one of the biggest sinkholes that a property investor can pursue. And it's largely unnecessary as you're about to discover through this analysis of historical data. Bit of a long one today, but uh we had to show a lot of evidence because people just won't believe us. And we How long have you spent on this Jeremy? Just to let everyone know.
>> months uh compiling all of this data together and coming to this conclusion that infrastructure research is overrated. But let's uh let's look into it. >> have a look at the ABS engineering construction activity. Yeah, so the ABS has this data set. Uh here's a list of some of the things included. You can see roads bridges railways pipelines harbors, etc.
This data set dates back to the late '80s '80s which means we have a a long history we can examine covering eras of both good time uh and bad, booms and busts. The problem however is that the data is only at the state level. So even if it does correlate closely to capital growth, we can't pinpoint precisely the areas that are going to deliver the best capital growth, but uh it will help to discover whether or not infrastructure does relate to capital growth. And when we talk about infrastructure, you might hear it at the barbecue or there's a new um airport coming through or there's a new train station. To keep it really simple, right? There's a new shopping center.
So that's what we talk about infrastructure at a a high level. Yeah, and this uh I mean that list from the prior prior slide was a was a good indicator of the sort of things this data set is showing. So it's specifically engineering construction activity. And then the other thing is it could also relate to a bit of noise, right? Like if there's a new shopping center in so you've got an established suburb there and then there's a shopping center. That could negatively impact potentially that property if it's across the road because maybe the noise levels, but obviously you got that convenience.
So it's Yeah, it's a Yeah, some infrastructure projects can negatively affect capital growth, but this is an analysis trying to find infrastructure projects that uh positively affect it. All right, so this chart shows the relationship between engineering construction activity and the capital growth that followed over the following 3 years. Now every one of those tiny white circles is a state or territory's engineering construction activity for a quarter sometime in the last 25 years. And the turquoise horizontal x-axis along the bottom uh shows the amount of money spent on engineering construction activity for that quarter per person. So we calculate per person because some states are larger than others. They have uh more people in them.
So we want to know per capita. Uh and also that dollar amount has been adjusted for inflation. That's to account for increasing construction costs over the period. >> Jeremy, like if we look at the $1,000 at the bottom of the the table, $1,000 and you've got like a little white dot there. What does that represent? So let's say there's call it I don't know a 100 people within that Is it suburb we're talking about or It's a state.
>> State. So a state at some point in the past there was some engineering construction activity. It's usually measured in the the billions of dollars. You divide by the population of that state. So New South Wales is an example. >> Yes.
Okay. Yeah, all right. Um so I was going to point out um the lowest circle. Can you find the lowest circle? I would say would it be this one? Maybe over here?
The lowest is over in the bottom right. Okay. Well, just to your left a little bit there. So just above It's about 4,400 bucks. >> From a growth perspective, not a dollar perspective. >> Oh, sorry.
Yes, yes. For the lowest one there from the purple vertical uh y-axis is showing the amount of capital growth which took place over the 3 years after there was a record of engineering construction activity that the ABS um took note of. So there's a big one over here. I'll let you go. Yeah, so basically towards the right of the chart means high engineering construction activity and towards the left means low. And a circle towards the top of the chart means high capital growth, towards the bottom means low capital growth.
So if there was a relationship between high engineering construction activity and high capital growth we would expect a spattering of circles starting in the bottom left and going up towards the top right, but instead the trend line, that's the white dotted line, is almost dead flat. And this means the general rule is that engineering construction activity has no relationship to 3-year capital growth. Now I looked at a few other capital growth time frames and they showed a similar sort of um horizontal white dotted trend line. Uh so the next thing to try was a change in engineering construction activity and that's on this next slide. Excuse me. So this chart shows the change in engineering construction activity over a 3-year period and compares it to the capital growth that followed that over the next 3 years.
So the purple vertical y-axis on the left is still the same uh but the horizontal x-axis along the bottom, it's no longer dollars but change in dollars in percentage terms from 3 years earlier. Let's look at an example. Uh see the white circle on its own on the far right? Uh it has an increase in engineering construction activity over 3-year period of about 100%. So over that 3-year period there was a doubling. Now that could have been measured in New South Wales from 2003 to 2006 or it could have been uh for Tasmania from 2010 to 2013.
I don't know which state it was. That's uh not im- uh important, but if you run your eye over from that circle to the purple left vertical y-axis, you'll see the growth over the next 3 years was about 16% per annum. So in that one case the massive increase in engineering construction activity did result in a significant increase in house prices over the 3 years that followed. But that engineering construction activity Jeremy, that's um is that just one project or that's multiple projects within that year? Are we looking Multiple projects. >> So that's potentially could be like a a new train station at Rouse Hill.
It could be um potentially the airport work or something like that. >> widening, yeah. Resurfacing, yeah. All sorts of stuff. stuff. So it's the the spend inflation-adjusted spend per capita.
Per capita for that year. Uh quarter. Quarter. Quarterly for a state. All right, so if we expected a strong correlation between change in engineering construction activity and capital growth we should see a spattering of these observations, these tiny white circles, starting in the bottom left and stretching up to the top right, but instead we see again an almost flat trend line. And that means there is no relationship between 3-year change in engineering construction activity and the following 3-year house price growth.
Now I tried a few other combinations and saw something interesting. It's on the next slide. So this is the same kind of chart, but I've doubled both the time frames. So instead of a 3-year change in engineering construction activity, I measured 6-year change. And similarly instead of 3-year house price growth, I've measured 6-year house price price growth. Uh and now we can start to see some sort of correlation, not a flat trend line but it's in the wrong direction.
So this trend line suggests that to get better capital growth, you need a decrease in engineering construction activity, not an increase. And there were other combinations with even uh stronger correlations than this, but all of them were inverse correlations where the trend line went down to the right, not up to the right as you would have expected. And it gets more interesting on the next slide. And you can represent any story or we could show any story that we like, right? Like you could have just picked five data points to show that uh infrastructure does make a difference to areas with capital growth, but you've obviously picked a lot of data points to see what is the probability, right? Or the averages.
>> That's right. Yeah, so I didn't cherry-pick um is the expression. Yeah. Yeah, we want and we want the averages, right? It's the best way to look at the data. Yeah, so we won't we're trying to gauge what's the general trend here.
What's the pattern? So 6-year engineering construction activity. >> Yeah, so in this chart I swapped the sequence of events. Notice the title change at the top? So instead of change in engineering construction activity followed by change in house price growth, I measured change in house price growth followed by change in engineering construction activity. So th- this is not just simply flipping the axes around.
This is changing the sequence of events. And now we can see a correlation, but it's the correlation between house price growth and then engineering construction activity. So what this data set is telling us is that capital growth leads to an increase in engineering construction activity, not the other way around. Mm. And you might be wondering how how is that possible? I've got some ideas on the next slide.
So, one logical explanation is that state and local councils derive their income from land values. So, state government obtains revenue from stamp duty whenever a property is sold. Uh and the amount of stamp duty is based on the property's value. So, if property values go up, stamp duty revenue goes up, too. And then the state government government can then allocate that increase in revenue to new infrastructure projects. And it's similar for local councils, as well.
The council derives revenue from council rates, which are based on land values. So, if the land value goes up, then the council gets more revenue, which they can then allocate to infrastructure projects. Uh note that federal government revenue mostly comes from income tax, so it's not as heavily influenced by property prices. There is capital gains tax, of course. Uh that is a federal tax, but it's not a significant contributor to the Fed's purse as income tax is. Uh and also, uh although the federal government funds some ma- major infrastructure projects, they usually do so in cooperation with the state government.
So, it could be like a uh we'll we'll contribute $2 for every $1 you contribute to the project. So, it's still these infrastructure projects, even when the Feds get involved, it's still linked closely to how much the states can contribute. And the state government, their revenue is heavily influenced by uh land values. So, again, it all ties back to uh these land values. Uh anyway, to explore this potential explanation, and of course, I don't know exactly. There's no audit trail.
I can't say that I know the cause of engineering construction activity increases. All we can do is say there's a correlation, not a cause. Uh so, to explore it further, I removed federal and state government engineering construction activity from the data set. And I focused so- solely on the local council data, which is almost purely dependent on land values. Uh both state and federal governments have other sources of revenue that contribute to their budget. But the council is almost entirely dependent on rates, which are almost entirely dependent on land values.
Uh so, after removing federal and state government contributions, uh the correlation improved even further. Uh which tends to confirm the suggestion uh that it is increased revenue from capital growth that funds infrastructure projects. In other words, infrastructure projects do not push up prices as much as prices push ahead projects. Bit weird, but that's what the data says. Uh all right. One of the shortcomings of that engineering construction activity data set from the ABS is that it does not include things like universities and hospitals, commercial and office buildings, shopping centers, business parks, industrial areas.
Engineering construction activity is more like civil work works like roads, bridges, pipelines, things like that. Not buildings. But the ABS has another data set that covers those types of things with and it's called building activity, and it's specifically non-residential, so as not to confuse with housing. Uh the categories included uh you can see the list there. Uh um although many of those sort of items you're looking at, they might not You might not ordinarily consider them to be infrastructure. Uh some of these sorts of projects could have a similar effect to infrastructure, especially things like universities and hospitals.
Uh they appear in the list on the right as education buildings and health buildings. All right. So, I analyzed this data set, too. The next uh chart it shows the relationship between non-residential building activity and 3-year capital growth in house prices. Building activity is shown in dollars per person per quarter across the turquoise horizontal x-axis. That's at the bottom of the chart.
Uh again, this figure has been adjusted to take into consideration inflation. Uh capital growth appears in purple up the vertical y-axis on the left, like with previous charts. And the trend line is the dotted white line. And because it is almost dead flat, it says there's no correlation between these two variables. Now, I did try a few growth periods other than 3 years, but they showed the same relationship. So, once again, I tried looking for a change in non-residential building activity.
But as you will see on this next chart, >> So, the more activity and the more spend there is, as we can see here, the capital growth is maybe declining slightly. So, That's such a flat line. >> There's There's nothing in it. Nothing in it. >> Yeah. All right.
Uh the correlation begins to reverse again for this data set, like with the previous. Uh with engineering construction activity. So, this is a correlation between 6-year change in non-residential building activity and 6-year house price growth. >> Isn't that interesting? So, more being spent in the particular pockets, there's like less capital growth. >> Less capital growth, yeah.
Again, the correlation is around the wrong way. Uh well, it could come down to uh again, well, the theories that I showed in that previous chart of of rising land values triggers um more of this sort of infrastructure-related projects. So, yeah. Let's um price growth follows lower non-residential building activity. And on the next chart, this is where I swap the sequence of events around again for this data set. There's a slight correlation between 2-year hou- house price growth followed by 2-year non-residential building activity change.
So, once again, we cannot use this data set as a predictor of future house price growth. It's the other way around. So, at this stage, you're probably thinking uh infrastructure is completely useless. Uh but there is another data set, third one, again from the ABS. Uh this one's capital expenditure or capex, for short. Uh now, note this is not government-funded.
This is not public sector spending. It's spending by the private sector. So, the revenue to fund these infrastructure projects should be, theoretically, less dependent on rising land values. And you can see here a list of sort of things this data set measures. The major player here is mining. Uh in fact, in terms of billions of dollars, private capex spent on mining dwarfs pretty much every other category.
And now, the ABS obtains this info by surveying big companies uh in each of these uh industries. There are some industries not included in the survey, like um agriculture, forestry, and fisheries. All right. Next slide. This chart shows the correlation between 3-year change in private capital expenditure and 2-year capital growth that followed. Again, the data is adjusted for inflation and is per capita to consider smaller states.
And finally, we get a correlation that matches our expectations. However, you can see it's uh pretty weak correlation. The trend line is only climbing slightly, and if you look at those tiny white circles, uh they look pretty random. There's no obvious pattern to the naked eye. But it's something, at least. So, at this stage, we can't completely write off infrastructure as being uh totally useless, but you're getting a picture from historical data that it's not all it's cracked up to be.
Uh and there are enough alarm bells from what we've seen so far to suggest that uh infrastructure is definitely overrated. But the analysis so far is what we would do if we were going to incorporate infrastructure data into an algorithm, like the DSR. What we did next is we analyzed individual projects, uh looking at them at on a case-by-case basis. So, Damo, uh did you want to take us through some examples? Sure. So, I think the way that I looked at this was We'll go through the example shortly, um but historical charts.
So, looking at um the commencement and when the project finished, but looking at a period before it was sort of completed to after completion to see if there's any variance in in the data. So, the first one is the Brisbane Airport runway. It was about a $1.1 billion uh um project. So, it's a pretty much a second major runway at the Brisbane Airport, and it commenced around that 2012 to 2020. So, a quite a long what, 8-year 8-year period. It created around 2,700 plus jobs.
Um And like I said, completed around 2020. And the way I looked at this was you pick out certain suburbs so impacted versus controlled. So, markets like Hendra, Nundah, Banyo, Bundall, Northgate. So, looking at the period before July 2020, the 4 years prior that. So, over here, I'm looking at the median values, and the change overall was around 14% versus the controlled markets markets which are a bit further out around sort of 14%. So, I think if you think about it, this was sort of the impacted markets were closer to the airport.
So, I think it could potentially would have a negative effect potentially due to the noise. Would you prefer to be closer in or further out because these controlled markets are further out so you're not most probably getting that noise. And I also had the Brisbane local government area factored in. So, pre-completion, there was pretty much no impact on those controlled and impacted markets. And then afterwards, the 4 years after July 2020 after the project was complete, let's have a look at the total growth. Pretty much the same thing with those controlled and impacted markets.
You know, 54% and 55%. So, pretty much no difference from that that point of view which typically what you see in a lot of these SA3s or these big cities is the percentage growth is very similar, right? Like if you're buying Hendra for example versus further out, you sort of do eventually get that ripple effect. Anyway, so just looking at this, infrastructure has not made really any difference to capital growth. And if there was something minor, how can you put that towards it being for the project? >> All right, could have been something else.
>> Yeah, and we can do this in our suburb data platform and I'll show you after I go through these slides how you can do it yourself. It's really simple. If you want to look at your own scenario, the Albion Park Rail Bypass. So, this is down in the Illawarra of New South Wales about an hour and hour and 20 from the Sydney CBD South. Now, there was a new bypass that was put in, spent about 630 million dollars and it was just to help the traffic flow from the Princess Highway through to the Illawarra. So, yeah, 630 and about created about 550 jobs cuz a lot of people think, "Oh, there's a new bypass coming in." I heard it a lot, "All the prices are going to go up." And did they?
Well, let's have a look. So, I guess before 2021 October 2021, the 4 years prior, so it opened up on October 2021. So, we want to look at the 4 years prior prior that and commencement started around 2019. So, there would have been a couple years where there might have been an announcement or maybe the government approved it, but it didn't actually start getting constructed until what, January 2019. But so, 4 years I think it's a nice little average to look at. But areas I picked out, Albion Park Rail, Albion Park, Oak Flats, Yallah Dapto, all the markets that would be impacted by this new bypass, overall there was capital growth around 21% or 5% per annum for that 4 years.
So, 5% capital growth every year compared to the controlled markets I picked which sort of had a bit more capital growth of around 25% compared to the 21%. So, areas like in Woonona which is sort of in that Wollongong local government area or Warrawong, Warilla, Mount Warrigal, Barrack Heights, they wouldn't have been impacted by that new bypass. So, pretty much again, the markets sort of grew very at a similar rate. Now, let's look at after construction, the 4 years after, what was the capital growth? Overall for those impacted markets for that bypass project, around 20% total capital growth compared to the controlled markets where the project didn't impact but sort of still in the similar type of area, had around 25%. So, no difference at all and the controlled markets actually had better capital growth compared to the impacted markets.
But the story here is that if you said, "Oh, I'm buying in Albion Park because of the the bypass is going to do so much better." You would have been worse off buying in for example Warrawong or Warilla cuz you're still down that South Coast corridor. And so, we move to the next project. I had a look Bruce Highway. So, 9.5 billion dollars. So, it's commenced back in July 2013 and completed in 2020. So, again, let's have a look at the impacted markets.
I won't go through all of them. You can have a look at the presentation, but Queensland. Queensland, yeah. 5 years leading up to that December 20 20, I had about 5% capital growth those impacted markets. Controlled markets had about 7% and then right after that period, same sort of story. Impacted markets had 72% capital growth versus 70%.
So, pretty much negligible, right? No difference whatsoever. And I know you're going to go over to the new Bendigo hospital, but before you do, Jeremy, I just want to show everyone I want to show everyone the how you can actually do this. So, if you go to our platform, you can head over to historical charts. I've got my default saved criteria here already. So, you can set your own default criteria.
And then all you do is you can just pick any market that you like. So, you can add the plot. You can delete, but this is just for example all the significant urban areas around Australia. And then you can go back as far as you like and look at how the markets have actually performed. You can get the data. So, over here I'm going from January 2021 just as an example.
I'll do it in a summary and then I can see the start value and the end value for the 5 years leading up to 2026. And in this example, I can see that obviously Brisbane and Perth were the higher and Adelaide was sort of the high fliers through that period. So, if you want to do your own research, just go to go to historical charts. I think that was about it. Was there anything else? And yeah, historical charts.
Anything to add, Jeremy, on that? Now, let me talk about the new Bendigo hospital. Yeah, so this was redeveloped in two stages. Stage one completed in 2016 and stage two in 2018. And the total project value was 630 million. And the chart you see here shows the growth in property values in Bendigo compared to Melbourne.
Now, if the new hospital stimulated the Bendigo economy, it had no impact on property prices. Values did start climbing in 2020, but this coincided with COVID, not the completion of the stage two of the hospital and stage one somehow caused no growth. So, there's another major piece of infrastructure which had no impact on property prices. Well, that's a one you'd hear a lot of there's a new hospital going in. Mhm. Yeah, another one, Canberra International new international airport.
The cost was $2 billion. Once completed, which was in 2016, it created about 300 additional jobs, not many. And as you can see from the chart, there was no sudden rise in property values. The it's a bit hard to see there. The Canberra Queanbeyan curve is the blue one. So, a bit of a busy chart, but I wanted to show all the comparisons of around the rest of the the state.
Well, not just ACT, New South Wales. So, Sydney's in there, Wollongong, Nowra, Bomaderry, etc. And it shows no discernible benefit was gained by Canberra for this project. The next one is the WestConnex. And this was a large road project in Sydney. Total cost was expected to be 1.6 billion.
That was expected back in 2017. And at that time, it was thought to bring a big boost to the Sydney even New South Wales economy. There were estimates of something like 10,000 jobs being created indirectly just during construction. And it was expected that more than $20 billion would be injected into the New South Wales economy from this one project. But the chart shows just how lacklustre growth was around Sydney in the years that followed. Sydney and Western SA4s, that's statistical area level four, they're unable to get up to even 5% per annum.
So, yeah, another big project, nothing there. Now, in the process, can you just take us to the next slide, Demo? In the process of examining these projects, these are just examples. We analyzed 120 projects in total. This took us months to do. This is a snapshot of the table of contents from that analysis document.
And as you can see, most of them are a fail. So, some got marked as a maybe. Very few are a clear success. The table of contents actually goes on to the next slide Demo. Here are some more. So, very early on we realized when we're doing this analysis that smaller projects are unlikely to shift the capital growth needle.
A small project must be in either a very remote area to have any impact and it requires a noticeable number of workers relative to the population of that remote area that are forced to move there and place a strain on real estate. And then the project needs to be quite a lengthy one to have time to impact that real estate and push prices up. So, it can't be just a road upgrade because after the project completes, which might only take 6 months, uh, that demand for rental accommodation is going to drop as well. Uh, so towards the end of the research, we started rejecting many projects because they simply were not significant sounding to have any impact. So not every project was a fail, but roughly, um, one in 10, very roughly, one in 10 had any discernible impact. So definitely infrastructure projects are overrated according to this research.
And sorry, just to jump in there. It's we're not saying it's a bad thing either. I just wouldn't be following property investment advice or allocation advice just because oh, there's a new infrastructure project going in. I don't think it's a bad thing having a a new hospital close by or better transport like So, yeah, I think that's what we're trying to say is that from all this, the it barely, like from a probability percentage point of view, it makes no difference to capital Yeah. Yeah, we're not saying we're not busting infrastructure research, we're just saying it's overrated. I mean, prior to this, uh, presentation, if you were to pick on a scale of 1 to 10, how influential do you believe infrastructure, uh, is for pushing prices, by the end of it, I'm sure you'll have a lower number.
All right, on the next slide is the third phase of our research. So we are not experts at researching infrastructure. Perhaps we are picking the wrong type of projects and perhaps we're ruling out projects that have more importance than we realize. So to cover that potential blind spot, I thought I'd assess the performance of someone who has published a lot of infrastructure-based hotspot reports. Uh, this person seems to be revered in the industry for their research. Uh, they've been publishing these sort of reports for a long time and there is a large sample of their past hotspot recommendations that we can analyze.
So that's what I did. I looked at their past recommendations to see how they performed. Here's a snippet in the next slide from the first report that I analyzed. Now this report was published in late 2007 and it's the first one that I have a copy of. Run your eye down the list of 12 hotspots picked out in the report and you'll see there's a focus on infrastructure. You got statewide mining activity, five major mines, $2 billion pulp mill, $10 billion in major projects.
Solar power station, water and gas pipelines, gas plant, etc. Now, I'm not going to show this level of detail for all hotspots that they recommended because there are over 500 of them. Uh, what I did was analyze the performance of the area mentioned in the report in the years that followed the report's publication. And let's start off by analyzing the performance of Adelaide, which is the first market recommended in the first report. All right, the turquoise line shows the per annum growth rate of Adelaide starting from when the report was published, which was October 2007. And the purple line is the national growth rate.
So Adelaide was ahead of the nation for capital growth and that lasted for quite some time. Uh, but as as is always the case, Adelaide had some rather mediocre growth around the middle of the chart. Uh, recently it's been outperforming again. So if you had randomly bought anywhere in Adelaide, which was the first recommendation, uh, you'd have been better off following this recommendation. So this is a win. I'm counting this one as a win.
The turquoise line is mostly above the purple line. Now, rather than eyeball a chart like this, uh, which can get subjective in some cases, I'm going to calculate the degree degree of outperformance more precisely using a number called the alpha. The alpha is a term they use in the share market. It's the percentage of growth above some baseline or benchmark. Now as an example, let's say you picked, uh, a hotspot and it had 10% growth, uh, while the broader market, that is the benchmark, had 6% growth. Then in this case, the alpha would be 4%.
4% above, uh, the benchmark, 4% more growth. So the alpha gives us an idea of how much better a hotspot recommendation was compared to random suburb selection. Random suburb selection requires no expertise, right? You just go eeny meeny miny moe, so that's the baseline. So when you hire the services of an expert, you're expecting something above that baseline. Now if the alpha is zero, it's the same as random suburb selection.
If the alpha is negative, it means the market had less growth than the benchmark. Uh, this is worse than random suburb selection. So you want the alpha to be as high as possible and the higher the alpha, the more successful, uh, the recommendations were. Now with that in mind, on the next chart, So for example, like yeah, as of 2026, you would say over the last 10 years on average the alpha, like let's would be around or the national average growth rate might be what, 7% per annum, call it. So if you had an alpha of 3% every year for like for the last 10 years, it'd be 10% capital growth. Per annum.
>> Per annum. That's right. Yes. >> So alpha's always the above whatever the average is to get to the >> is the amount above the benchmark. Yeah. All right, so in this next chart, I'm highlighting the difference in growth between uh, Adelaide and the benchmark.
So this is trying to highlight the alpha over time in a chart. So that turquoise shaded area is the area when Adelaide was outperforming the benchmark, the national growth rate. And that vertical gap is the measurement of the alpha at that point in time. Now this chart only shows when the alpha was positive. There was a period of time when Adelaide's performance was lower than the nation and so the turquoise area disappears behind the purple shade. That's from about 2016 up to and including 2022.
Now we're going to use the alpha to assess the performance of the infrastructure expert more comprehensively, but before I go on, I just want to clarify what I'm using as the benchmark. Now I mentioned a term called the benchmark. It's supposed to be the broader market. In the share market, that benchmark might be the ASX 200 or it might be the All Ordinaries Index. In property, the benchmark might be the national growth rate or the growth of properties just in the state capitals. In my calculations, I used the growth in median house prices for all suburbs within Australia's significant urban areas.
So there's roughly 100 significant urban areas around, uh, Australia. So they're things like all the state capitals are a significant urban area, but they also include regional markets like, uh, Geelong, Bendigo Ballarat Wollongong Newcastle, Maitland, Central Coast, Toowoomba um Gold Coast, Sunshine Coast, Rockhampton, Bunbury, um, yeah, so there's loads of these. A significant urban area is defined by the Australian Bureau of Statistics. It's sort of a cluster of suburbs that have a total population of 10,000 or more people. So I've used that instead of all Australian suburbs and that's because 2/3 of Australian suburbs are extremely remote. There's very little real estate there.
There's very few transactions and so you can get some very unreliable capital growth calculations. So that's why I'm restricting it just to the significant urban areas. All right, let's have a look at the second market that the expert, uh, recommended. Uh, this was Cloncurry. It's a remote area in Queensland and apart from a couple of months early on and a few more recently, for most of the time, this market has underperformed the national growth rate. For most of the period since the recommendation was published, Cloncurry has been overshadowed by the benchmark.
The turquoise shaded area even goes into the negative growth zone for about 15 years. So this, the second, uh, recommendation was a clear fail and a big one, too. So looking at the first chart in isolation, you might have thought, um, well, uh, the this expert knows what they're doing, but after looking at the second one, which was terrible, uh, it sort of balances out. So what we need to do now is assess all their recommendations and combine them. You can't get all right, so this is understandable to have one fail, though it is pretty monstrous fail. So I analyzed over 500 more recommended hotspots by this expert.
Uh, I can't show you a chart of each one, so I'm going to speed up the assessment now by showing a chart that summarizes this expert's past performance, but before that, I just want to clarify how I've made these calculations. So firstly, I measured the capital growth for each property market that the expert recommended. Then I compared that growth to the benchmark, that's the national growth rate, and that gives me the alpha. But growth can be measured over a range of time frames, 1 year, 2 years, 3 years, etc. So I calculated the alpha for a wide range of different growth periods and put it all together in this one single chart. All right, a little bit of a peculiar chart, so let me just explain this.
So, this chart shows how the experts 500 plus recommendations performed as measured by alpha and for a range of different growth periods. Let me start with the axes. Uh so, the period over which growth was measured it is shown up the left vertical Y axis. Starts in the bottom left at 1 year and goes up to 18 years. So, recommendations that this expert made in 2007 have up to 18 years of growth. Oh, by the way, we're recording this in early 2026.
Uh we can measure growth the hotspot recommendations from 2007 over 1 year, the year after the recommendation, over 2 years, 3 years, all the way up to 18 years. But, recommendations made in 2022 may only have 1 year growth, 2 year growth, 3 year, maybe 4 year growth depending on when during 2022 the recommendation was made. But, not any higher since 2022 was only 3 or 4 years ago. All right. Now, the alpha is shown across the horizontal X axis. Sorry, Jess.
Just quickly, the experts alpha like that's up in the recommendations made in 2020, correct? From 2007 up to 2020. >> they made a recommendation in 2020 and it's 2026 now, we'd only have that 5 6 years obviously perfect. That makes sense. >> have also have 1 year, 2 year, 3 year growth. >> Correct.
>> But, only up to yes, 6 years. That's right. Uh all right. So, looking at the 1 year growth at the bottom, you can see the alpha is about 0.3%. Uh this means out of all the recommendations, they typically had 0.3% more growth than the broader market over the first year. Now, that year could have been 2007 uh to 2008, or it could have been from 2010 to 2011, or from yeah, 2020 to 2021.
So, it's any 1 year period from the month in which a recommendation was made. And as an investor, what we want is those sort of 3 4 5 years. You want this the first That's when you want optimize your capital growth, right? Like early on is where we can sort of predict with data a lot more heavily. But, after that throw a dart. It's guessing really.
Anyone that says that they know, um obviously you could factor things in like if you've got a a 5 plus year time horizon of 10, you go to the bigger cities just for maybe potentially a bit of stability. But, the growth is what we want up front, right? So, it's interesting to me to see that this isn't a lot higher in the sort of first 5 years. Mhm. That's what you'd expect, right? >> Yeah.
Yeah. Uh I'll show you more where where we can achieve that, but um if you just go back, sorry. I haven't finished with that one. Mhm. Uh yeah. So, now because there were over 500 recommendations in total, that 1 year alpha uh it's the median alpha for over 500 recommendations.
So, it's a very reliable measure. But, the alphas appearing towards the top of the chart, they are based on much fewer samples. So, only a dozen recommendations were made in 2007. So, the alpha for an 18 year growth period is unreliable. Mhm. And the best growth period for the 500 plus recommendations made by this expert was uh 7 years.
That's the widest bar stretching to the right. And the median alpha there was about 1.3%. So, that means markets recommended had 1.3% more growth per annum over those seven annums following the recommendations. Now, converting a per annum alpha to total growth over that 7 year period, let's say the broader market had 60% growth over those 7 years, then the infrastructure experts recommendations had 75% total growth over that same period. So, it's 15% more growth in total than the broader market, which is what you get if you compound an extra 1.3% per annum. Uh and that's looking at all 7 year periods, regardless of whether they were picked in 2007 or 2010 or uh 2000 um 15.
Okay. So, 7 year growth was their best. Their worst period was 13 years, where the alpha was negative by about 1%. So, that means if you held a property in any one of the recommended areas for 13 years, you typically >> underperformed. >> would have had about 1% less growth than most Australian property owners over that 13 year period. >> Isn't that wild?
So, you're you're going off this report and it it's very interesting that why would this particular expert produce a report like this if they haven't done their own analysis? It's just like, "Here, buy my report. Cash grab." Potentially. >> Yeah. Well, back in the day, certainly back in 2007, it was all just opinion related um guesswork rather than data. The data didn't start until well, 2010.
>> to sell these reports, don't they? Yeah. Yeah. Yeah, they sold one to me. Uh and you can see some uh cyclical nature in the performance. Uh the alphas snake upwards having eras of above average growth and eras of below average growth.
Uh and that reflects the way in which many markets grow. They have surges of good growth and then uh followed by flat periods of underperformance. Okay. So far, this suggests that infrastructure research by this expert is a bit hit and miss. >> Mhm. Uh certainly over particular time frames.
Uh gee, you may need to exit the market typically after 7 years to be better off, but when the benefit's only about 15%, that's not going to be enough to cover costs like CGT, agents commission, cost of buying elsewhere like stamp duty. So, this data shows that historically you would have probably been no better off following the infrastructure compared to random suburb selection. Now, uh I'm not saying infrastructure research is pointless. I'm just saying it's overrated, and here's a good example of why. This chart is the same alpha calculations, but instead of using the experts recommendations, I've used the recommendations of the basic DSR. Uh it's the top 20 house markets of each month.
Uh so, just to clarify, this is version one. Did the basic DSR came out in 2010, January 2010. It was superseded by version two in 2015, which was also superseded by version three in 2025, and I'll show you charts um about versions two and three in a minute. But, just looking at the basic DSR for now, um I'm showing the basic DSR because it has absolutely no consideration whatsoever of infrastructure. Uh straight away, you can see that the DSR is a more reliable way to pick hotspots. Uh and you can see that typically gets off to a flying start.
The bottom bar shows that on average in the first year of ownership, there is about 5% more growth than the benchmark. Over 2 years, a little less than 4% per annum extra growth. Uh most of the outperformance happens in the first 3 years. Although, 10 years isn't too bad either. The alpha over 10 years is about 1.5%. Is that an average over that 10 years, Jess?
Or is that at that 10 year mark how it performed at that 10 year mark? >> Yeah. Uh yeah. So, it's the average. So, it's you're using compounding, so these are all per annum. So, it's 1.5% per annum more more capital growth.
But, definitely a trend line, right? Like it's that three like I mentioned before, that three to four years. This is what people want. They want the capital growth ASAP. Yeah. That's it.
Yes. And and when you are measuring supply and demand, there's what we're doing here is measuring markets where there's an imbalance between supply and demand, but that then balances out. So, over the long term, you'd expect uh the you know, 15 years, 16, 17 years into the future for it to match the national growth rate. It's just in those earlier years where there is tremendous imbalance of supply and demand that you're going to get that fast faster rate of growth. >> aren't cherry-picked either, right? Like there there's a lot of um top 20s in here, correct?
Yeah, every month. Um so, from 2010, what were the top 20 from Sorry, January 2010, February 2010, March 2010. Just keep going. So, there's yeah, many. And the thing is there might be one or two in there where you might have bought that might not have performed as good, but overall, the probability is that you would have done really well if you picked the right suburb. >> That's right.
There were failures amongst this, but this is the aggregate, the the average. Uh so, yeah, although 10 years I was I was going to point out 10 years um of 1.5% uh if the national growth rate over the over those 10 years was 6% per annum, that equates to 80% total growth due to the compounding. >> Mhm. But, an alpha of 1.5% would result in 106% growth in total. So, an extra 26% growth, which for a uh you know, $500,000 property is well over $100,000 difference. So, yeah, pretty cool.
>> expect it's average out, but the one thing is like you said, over that 15 year period, you're outperforming the average, which which is what we want, right? Yeah. Now, this is yes, and this is also the sort of growth that's typical for So, this is assuming you buy any house in any suburb of any one of those top 20 at any point in time over the last um well, since January 2010. Mhm. So, um yeah, it is comprehensive. It's a large data set.
So, it's quite a reliable count of sample size. Anyway, so you can see from the chart that underperformance was not typical for any growth period. That doesn't mean that none of the suburbs picked out by the DSR one underperformed. It just means that on average they outperformed. And yeah, just like to point out the DSR does not consider infrastructure at all. It was a very simple algorithm.
Now, it was superseded by DSR version two, also known as DSR plus in 2015. And this is the alphas for the DSR two. The worst growth period was 13 year growth with an alpha of a bit more than 1.3%. Again, you can see the DSR two does not typically have any negative alphas. And note the trend towards lower alphas the longer the time frame, which is common for all real estate. The longer you hold a property for, the more likely you're going to get average growth in line with the the benchmark.
And for more on why that happens, check out episode number 11 in this series. It's called apples and oranges. Also, episode 12, which is about why short-term wins better than long-term. All right, note that the DSR two does not consider infrastructure either. And yet it's it's still outperforming quite well. And I mentioned the DSR plus came out in 2015.
However, its scores um are backdated to 2010. So, a small portion of the history that you're looking at right now has been backdated. It's not actual. Whereas all of the DSR one were actuals. Now, DSR two was superseded by DSR three in mid-2025. So, almost all of the DSR three is back-tested.
Just something to keep in mind. Anyway, these are the alphas for the DSR three. And the worst growth period is 15 year growth right at the top. The alpha is about 1.3% for 15 years. These alphas are what you could expect if you randomly chose a house from a suburb that was also randomly chosen from one of the top 20 suburbs as picked out by the DSR for any month in the past from 2010 to the end of 2024. Again, the DSR three does not currently consider infrastructure as one of its variables.
Because as you're beginning to get the idea, infrastructure is overrated. Now, in the future perhaps it might be integrated, but at the moment there are far more promising improvements to the DSR than a focus on infrastructure. And how many metrics go in the DSR three, Jeremy? It's hard to say. I mean, there's one metric in there which is itself an algorithm of many metrics. And some of those metrics are combination of some raw data.
So, yeah, I don't know how to keep count. But it's a lot more than obviously DSR two. So, there's more data sets. So, the more DSR variations we're going to have or upgrades, they're going to have just more and more data sets to make it better. Yeah, that's not the only way you can make it better, but yeah, more data certainly helps. All right.
Now, we all learn from our mistakes. It's unfair for me to show the latest DSR algorithm which has all of its improvements backdated without also showing more recent research from this expert. On top of that, I did notice some more recent reports from the infrastructure expert in 2021 and 2022 that were specifically titled infrastructure led. And here's a sample from one of those reports. Each report contained 10 hotspots. None of them are suburbs.
They are at least as large as local government areas. For example, in this list you can see the Brimbank LGA recommended halfway down list. Brimbank is a local government area in Melbourne. And in some cases an entire significant urban area was recommended. For example, Townsville and Tamworth at the bottom. Now, I found four reports like this, each with 10 recommended hotspots.
Keep in mind that 40 recommendations is a very small sample. And they only span a very short era. This will highlight another problem, which I'll come to in a minute. For now, here are the alphas combined set of recent infrastructure led hotspots. So, the first thing you'll notice, there are less bars, and that's because we can only calculate alphas for growth periods ranging from one to four years. And that's because of how recent these recommendations were made.
And the second thing you'll notice is that the performance has improved dramatically. The alphas are all over 4% now, whereas previously the best was 1.3%. That's from the full set dating back to 2007. And the third thing that changed is the x-axis. It now stretches across to a maximum of 15%. On previous charts I had it limited to only 10%.
I had to stretch it out because of the next chart. This is the alphas for the same period for the DSR three's top 10 local government areas. So, I've changed it from suburbs to local government areas to make it closer like-for-like comparison with the recent reports of the infrastructure expert. Now, even though the recent sample for the infrastructure expert was a a noticeable improvement on his career alpha, the DSR three alpha for the same period was still about three times larger. And this applies more weight to the argument that infrastructure focused research is not time well spent. By the way, just to put into perspective those DSR three alphas, the worst growth period for the DSR three for this era was four year growth.
That's the bar at the top. The four year alpha is around about 11%. The national growth rate over the same period was 6%. So, the typical local government area picked out by the DSR three top 10 had about 17% growth per annum. That's 11% alpha on top of a 6% benchmark. Now, 6% growth per annum equates to total growth over four years of 26%.
But 17% growth per annum over four years is 87% growth. So, 61% more growth. That's more than triple the growth of the rest of Australia for that period. And this is by randomly choosing any house in any randomly chosen suburb in an LGA that was also randomly chosen from the top 10 by the DSR. Now, that sounds really impressive, but just keep in mind that this is a small sample. And it's a for a very specific era.
And this was an era in which pandemically low interest rates fueled rapid price growth. A lot of markets boomed. It was quite easy to get high alphas during this era. And in boom conditions alphas are much higher than in than in tough times. So, don't expect this kind of performance into the future. This is like a one-off.
The previous chart we showed where there were many alphas for the DSR three, that will give you a much more typical idea of what you could expect in the future. Covering good times and bad. I think that's the best way to do your analysis, right? When you're starting out, whatever your price point is, I'd say SA3s, but you can start with local government areas based on my price point. You can go into suburb search and say, "Look, I want to spend up to 700,000." And then you hit you want to find the top 10 local government areas, and that can easily be done for any month. But obviously, if you're investing now, you look at the current month's data.
And that's a really good starting area. And then from there, then you can start diving into those individual suburbs you want to invest in. Because if you just go to top 10 suburbs, they can be all over the place. They can be all over the place. And another reason why we aggregate to these larger areas like LGAs and SA3s is because you get so much more reliable data. So, it'll put you in the right ballpark.
And then like Damo said, you can then drill down to the suburb level. Reliability, I like it. Yeah, all right. So, in conclusion, analysis of historical infrastructure spending from the ABS showed a weak correlation to capital growth. In fact, it's more likely that capital growth pushes infrastructure projects rather than the other way around. Secondly, we looked at promising infrastructure projects on a case-by-case basis trying to find ones likely to push prices along.
We found that most of them didn't shift the needle at all. And we did not include in that analysis where capital growth actually went backwards. Like a negative like being under the flight path. And then thirdly, we examined the past performance of an infrastructure expert and found the performance to be rather lackluster. So, the conclusion is infrastructure is overrated. I'm curious to know now, on a scale of one to 10, how influential would you rate infrastructure projects as pushing capital growth compared to uh the the scale of 1 to 10 you would have had prior to this episode.
Nice research, Jeremy. So, in our next episode of the expert busting series, episode 15, vacancy, what reflects a balanced market? Thanks for watching.