Attraction - Price Gradients: multiple non-linear relationships?

This post is the product of further thought since identifying and describing the 'beach-price gradient' here

http://www.somersoft.com/forums/showthread.php?t=19895

Much advice on investment property selection focuses on its location and ther area's future prospects for capital growth and attractiveness to tenants.

Subject to council rules, the investor can control the property (including renovating, extending, or knocking down) but has limited control over the location. Location can set a limit to the capital growth possible and the potential to add value without overcapitalising.

To avoid overpaying and to ensure maximum growth potential, careful location assessment must precede location selection.

So what are the ingredients of assessing a location?

Almost any book on property investment will have a list something like this:

1. Proximity to CBD
2. Proximity to (good) schools and other educational facilities
3. Proximity to transport
4. Proximity to shops etc
5. Proximity to beach or river
6. Proximity to recreational facilities (which may include 5 above)
7. Proxmity to hospitals/health facilities

All this is a bit vague, as 'proximity' could mean anything between 500 metres and 5 kilometres. Also in some cases, just because 'close' is good, 'even closer' is not necessarily better, and could be worse.

I suspect that in many cases the relationship is non-linear. In some cases 50 metres is worth three times as much as 500 metres away, but in other cases it could be worth the same or less. Also there may not always be much difference between 500 metres and 1000 metres, but a considerable difference between 1km and 2km.

Each point will be discussed in turn:

1. CBD

Generally property prices fall with increasing distance from the CBD. The rate of fall off can vary; it depends on how fashionable inner-city or suburban living is at the time. Other factors also play their part; a single house marooned in a firmly non-residental warehousing precinct with few local facilities may not be particularly desirable (unless the land was of value for some other purpose).

2. Schools and education

Schools are everywhere. What are regarded as 'good schools' (either private or state) are more scarce and may lift property prices in surrounding neighbourhoods (refer to 'The Millionaire Mind' for more discussion on this).

Government regulations (such as school catchment zoning for state schools) can also conspire to increase land prices in favoured zones if the school is highly regarded (the use of govt legislation to transfer wealth to selected landowners is another issue that is out of scope here).

A house across the road from the school gates is undesirable (think of things in letter boxes!). Around the corner and 5 minutes walk away is very good. 10 minutes walk away is also very good and in practice no worse than 3 or 5 minutes away. Subject to other transport, 40 minutes walk away is bad and the 'good school' effect is unlikely to be very strong at this distance.

3. Transport

Like schools, quality is important, as all suburbs have roads and most have some form of public transport.

Car transport is less fussy; will a suburb really be valued much less if it is 3km from the freeway exit than 1km? I would think not. Other attributes would be more important.

As with schools, extreme proximity to a noisy road or rail line is a minus, especially if it's just over the back fence, and the graffiti taggers are just a short fence-hop from the house.

Public transport is more selective. For public transport to be good it must be (i) frequent, (ii) operate througout the day, and (iii) fast. For inner suburbs, as distances are so short you can compromise on speed, but the first two are still vital.

In most suburbs of most cities this means trains. Trams are probably acceptable in inner-Melbourne and parts of Adelaide. Buses may be OK in inner parts of Sydney or Adelaide where services are good. But they have a low prestige in Melbourne and are generally ignored (except on a handful of routes).

'Good public transport', like 'good schools' are not in every suburb so have a scarcity value. In the case of Melbourne about 20% of residents are within walking distance of trains and maybe 20% near trams, so this criteria excludes well over 50% of houses and some entire suburbs.

500 metres is definitely better than 1000 metres, and far better than 3km. 250 metres could be better still! People will walk further for better service, so 800 metres to a train is likely to be superior than 400 metres to a bus.

4. Shops

These are pretty much everywhere, so I can't see that they are as scarce as 'good schools' or 'good public transport'. Walking to a reasonable selection is good if possible. But a 2-4km drive to a good supermarket is not unacceptable (especially if there is a closer convenience store) if other proximity factors like schools or beach apply.

5. Beach or river

Unlike schools 50 metres distance is much much more valuable than 300 metres which is more valuable than 3km. If views are involved, there aren't bogs or mosquitoes, and it's near other facilities then the scarcity value of the waterfront is extreme and prices will be high. As an example, for coastal regional cities, a 3km distance to the beach might have no value premium, whereas it would in a capital city as there are many suburbs 5, 10 or 20km inland.

6. Recreational facilities

7. Hospitals/health facilities

Both have a greater scarcity value than shops. Comments on proximity to shops similar, but walking or wheelchair access may be highly prized for seniors facilities.

Consideration of these points show that a linear relationship between distance and value does not apply in all cases. There is such a thing as 'too close' to some facilities, whereas for others like beaches, there is no such thing. The relative significance of each of these points also varies.

In case you're wondering where I'm leading, I intend to integrate the above into a possible formula to compare suburbs on a metro-wide basis in a future posting.

Rgds, Peter
 
Fwiw and, as with all my posts I haven't given this much thought, I think there is at least one other factor that needs to be considered.

Lets be honest here, beach property has a certain sex appeal. But even this goes beyond sex appeal. Even if you swim as well as a brick and you're the size of a house, (as a general rule) people do like to live or be seen either at the beach or very close to it.

And living on the right beach has a certain FIG JAM about it, without the JAM.

But not all beaches are created equal.

Beaches, like celebrities, command a premium depending on whether they are A-listers, B-listers and the like.

A possible sample of A list beaches:

- Bondi
- Manly (Sydney)
- Merewether (Newcastle)
- Wategos (Byron Bay)
- Noosa


I also suspect that the presence of an A-List beach in an area will drag up the prices of surrounding beaches more than any financial model that does not seek to take account of the FIG aspects of beach life.

Just some thoughts.

Mark :)

ps. In the rare event that there are people who don't know what FIG JAM means, it stands for "F*ck I'm Good, Just Ask Me"
 
Pitt St said:
And living on the right beach has a certain FIG JAM about it, without the JAM. But not all beaches are created equal. Beaches, like celebrities, command a premium depending on whether they are A-listers, B-listers and the like.

Mark, you are quite right.

One could also add that the prestige of a beach has nothing to do with its inherent swimming, surfing, dogwalking, fishing, shellfinding or sandcastle-building qualities.

This point is confirmed by the number of beachside mansions that have their own swimming pools.

Hence analysing beach prestige on the fundamentals ('fundamental beach analysis') is probably doomed to fail. Instead prestige is a mixture of history, folklore and prejudice. This is repeated by tourist brochures, postcards and locals who want to make a buck. It also means that analysing beach prestige has more to do with historical and cultural factors than anything to do with the sand, sun and water.

Events like the Melbourne cup are far more than horses. Indeed, apart from the underpaid feeders and horsehandlers for whom they are a livelihood, the horses themselves are third to gambling 'n' garb.

In a similar way, Bondi-lore is far more than the beach; it encompasses a particular portrayal of men and lifestyle lived. Of course a richness in history is no guarantee of suburban prestige; the inner-western suburbs of Melbourne are steeped in early industrial history, but as property prices in suburbs such as Sunshine show, this has not given them a price premium, except where a 'port' is involved, as with Port Melbourne, Williamstown and Fremantle.

Your point has also unintentially undermined the basis of my next proposed posting, by exposing a particular assumption that I was to make. However it adds a valuable counterpoint and strengthens some comments made in the 'regression to the mean' thread.

Rgds, Peter
 
Having established that:

1. Proximity to certain facilities increases the appeal of a site and

2. The rise in value caused by a particular facility is not necessarily linearly related to the site's distance from it

I then considered whether it was possible to develop a formula for assessing the likely site value of a particular residential location. Note the stress on 'residential'; commercial or industrial land values are influenced by other factors not discussed here.

Such a formula might include the following factors:

a. A starting point (at least in urban areas, no land is worth nothing)

b. Particular qualities/facilities that attract or detract from the land value

These attractors were discussed earlier and include:

1. Proximity to CBD
2. Proximity to (good) schools and other educational facilities
3. Proximity to transport
4. Proximity to shops etc
5. Proximity to beach or river
6. Proximity to recreational facilities (which may include 5 above)
7. Proxmity to hospitals/health facilities

c. Some form of weighting for the above (as not all are equal value). I will call this an 'Attraction Value Weighting'.

d. Some way to compensate for the non-linear distance/value influence discussed earlier. I will call this a 'Proximity FA

a. STARTING POINT

The starting point could either be (i) the minimum land value price with other factors only adding to it, or (ii) a metropolitan average, with other factors adding or detracting from it.

A benchmark for (i) is hard to agree on. It could be a block a long way away from the city and other facilities. But how far out should one go. Alternatively it could be the cheapest standard house block from the cheapest suburban fringe estate available at the time.

(ii) could be the more straightforward median established house sale price. Though it should really be the land component of this this is not essential if we are less scientific and agree to discuss house+land (and not just land) prices.

Note that choosing the less rigorous approach would cause inaccuracies. This is of highest risk if assessing properties a long way away from the median (eg small flats and outer suburban mansions both with low land components, and fibro shacks in 'good' suburbs with high land components).

Because people like to compare suburbs with one another, I will go for (ii) and allow negative points to be given for a poorer than average location.

b. PARTICULAR FACILITIES

These are identified in previous postings.

c. ATTRACTION IMPORTANCE WEIGHTING

Not all attractions are of equal value. You will need to weigh up the merits of various different characteristics or facilities. For example a beachfront location may have limited shopping facilities and poor public transport. However its prime location may still be very highly prized and deserves a higher weighting than other attractors. I will call this an 'Attractor Importance Weighting', later on. Of course different buyers will have different ideas of what they value this most so none of this can be accurate.

This is a number that sets a ceiling on how important each attractor can be.

d. ATTRACTOR QUALITY WEIGHTING (of specific facilities)

Not all facilities are of equal value. Hence there is a need to weight the value they are likely to add to site value. For instance some properties might be near 'good' schools, but another might be near a 'prestige' school and attract a higher weighting. Similarly, some properties might have the advantage of a nearby railway station, but another might be near a major junction (eg South Yarra) where services are much more frequent and therefore better quality.

This is a score out of 10.

e. NON-LINEAR DISTANCE/VALUE FACTOR

This is the factor that puts in numerical form my initial posting; ie being opposite a particular school's gates ('good' or otherwise) is bad, 5 minutes walk good, 10 minutes walk equally good, but 40 minutes walk not so good.

For beaches beachfront could be worth at least double than 100m back, but this is still more prized than 3km back.

This is a score between -10 and +10, as suburbs a longer than average distance from key facilities will get a negative score.

CALCULATION

The assessor will need to devise a table for each of the seven key facilities listed above.

The overall formula could be:

Value = Metro Median + (Prox CBD + Prox Schools + Prox Transport + Prox Shops + Prox Beach/River + Prox Recreation + Prox Health)

Note that each proximity item can be negative or positive and includes weighting and the distance/value factor. The idea is to try to assign a value (probably in $5 or 10 000 lots) and add or subtract these from the median. You could even add positive or negative characteristics of the building, this should not be done if we are just assessing site value.

Each proximity item might include:

(i) Attractor Importance Weighting (overall multiplier factor)
(ii) Attraction Value Weighting (score out of 10)
(iii) Non-Linear Distance/value weighting (score between -10 and 10)

This should be done seven times, once for each attractors. i through to iii are multiplied with one another for each attractor (schools, transport, shops, beach, etc). Note that negative values are given when a location is more distant than average from desired facilities.

These seven totals are added to the median price to provide a 'guesstimate'. Work would need to be done with existing suburbs and prices to provide some realistic figures for the so far undefined Attractor Importance Weighting multiplier for each of the seven factors (i above).

LIMITATIONS AND CONCLUSION

Of course all the above assumes that there is some intrinsic value that can be determined by some formula. I haven't included factors such as architectual tastes that may make a particular precinct more highly sought after than numbers might suggest.

Mark's previous point about beaches was also extremely important because it hightlighted social factors that are more to do with culture, myths, reputation and snobbery than any intrinsic value of Bondi Beach.

Such factors are unlikely to be captured in the above unless we add an eighth factor, ie a 'suburb reputation index'. However since reputation is built into existing property prices, this might not be very meaningful, and could be counter productive if we are using exercises such as this to calculate 'undervalued' suburbs that though rich with facilities are not loved by the market and may be set to boom in the future.

Regards, Peter
 
Peter,

Are you including the irrational factor. A property may be more valuable to one potential buyer than another due to personal factors.

Cheers,

Aceyducey
 
Aceyducey said:
Are you including the irrational factor. A property may be more valuable to one potential buyer than another due to personal factors.

This is true if you are looking for a PPOR, but shouldn't come into it if you want to buy for investment purposes.

If some irrational factor causes an owner-occupier to offer more than you as an investor think it's worth, then let them have it ;)

The individual factor would be most applicable for (i) tenantability (especially if vacancy is a problem in the area and/or you want to charge high rents) and (ii) saleability. In this case the onus is on you the seller to prepare the property so it appeals to what buyers in the area want and for your agent to find the buyer who will pay most for it.

I suppose individual and cultural factors are part of the 'irrational' element that the investor shouldn't pay too much for but which should be known so they can be exploited when selling or letting.

Rgds, Peter
 
beech said:
Acey it sounds difficult enough without you adding another factor. :D

I'm about to add a couple more :D

After graphing average suburban prices along particular radial corridors extending from the CBD I can confirm that average prices do indeed fall off with distance.

However the rate of fall-off differs with what side of the city it's on.

For instance, a given western suburb 20 km from the CBD will almost certianly be cheaper than an eastern or southern suburb 20km out. Part of this could be due to the higher prestige of the eastern suburbs. Also in the east there are very major facilities in the 10-20km band (eg Monash Uni x2 and hospitals) whereas in the west they might be closer in (eg around Footscray).

Another factor is that the rural/urban fringe might be 40km or so in the east but only 20-25km in the west, so the eastern suburb qualifies as 'middle distance' whereas the western suburb might be considered 'outer'.

Things like the Westgate Bridge challenge this, but the population centre of Melbourne remains in the east (Oakleigh). Thus people are still quite happy to settle in the east even if they're 30-40km from the CBD as they'll be a much lesser distance from many friends and facilities.

Another factor is that people tend to be familiar with their own suburban corridors only and if they (or their children) need to move generally move along them. For instance new residents of Narre Warren many come from Dandenong, Springvale or Warragul, but seldom Werribee or Altona. Hence demand for a suburb may be higher if it is in line with a growth corridor, eg south east, west or north. Conversely the same factor may work against suburbs, and encourage emptying and economic stagnation of some suburbs (eg Dandenong) in favour of newer outer areas.

If this is the case, it could well be identified by a donut effect, where property prices dip around a particular centre, only to rise somewhat in suburbs further out (which bucks the 'normal' pattern of a gradual fall).

Rgds, Peter
 
Spiderman said:
a. STARTING POINT

The starting point could either be (i) the minimum land value price with other factors only adding to it, or (ii) a metropolitan average, with other factors adding or detracting from it.

A benchmark for (i) is hard to agree on. It could be a block a long way away from the city and other facilities. But how far out should one go. Alternatively it could be the cheapest standard house block from the cheapest suburban fringe estate available at the time.

(ii) could be the more straightforward median established house sale price. Though it should really be the land component of this this is not essential if we are less scientific and agree to discuss house+land (and not just land) prices.

Apologies in advance if I've missed it in your posts, but have you taken into account what sort of properties contribute to a median price in the suburb in question? For example, you may be considering the purchase of a 3 bedroom bungalow in a suburb that is 60% apartments and townhouses, say. The median house price would thus be skewed to some extent toward the value of the townhouses and apartments.

Perhaps an RP Data report or similar could be used, where the recent sales prices of similar properties could be used to form some sort of rough median house price on which to start on. Or perhaps even the asking prices of similar houses, in the Saturday paper, (this may cause more unwanted inaccuracies, though).

Just some thoughts. ;)

Thanks, and great thread by the way. I like these threads on mathematical analyses. :)
 
It's the "Old Sol" factor !!!

For instance, a given western suburb 20 km from the CBD will almost certianly be cheaper than an eastern or southern suburb 20km out.
Yeah, and "Old Sol" is to blaim for that (in my opinion).

I've lived in Western suburbs in the past - and having the Sun in your eyes when driving to AND from work is no joke. (Mornings seem to be particularly bad, with foggy windscreens, etc).

The anomaly to this might be Perth (and other Western towns/cities) where you get closer to the water as you go further West......

Regards,
 
Les said:
Yeah, and "Old Sol" is to blaim for that (in my opinion).

If this was true, Penrith and Ipswich would have overtaken Sydney and Brisbane CBDs as THE place to do business by now. And no one would want to commute to a workplace anywhere along the east coast.

If reduced to another Sydney-Melbourne thing - more of Melbourne lives in the east, whereas most of Sydney in the west, so Melbourne wins here!

Rgds, Peter
 
G'day Spider,

Whoaa! You're about 180 degrees out of phase with my statement here. I'm SUPPORTING your prognosis. The Western suburbs are the ones MOST affected by Old Sol - as such are LEAST desirable. Hence why Penrith hasn't taken over..... :D

As far as
Spiderman said:
whereas most of Sydney in the west
you can't go too far East of Sydney without water-wings ...... Thus, Sydney almost becomes a "special case". My thoughts would be that North or South of Sydney (compared to West) would be inherently more desirable because of old Sol. And, of course, what little of Sydney IS East, is EXTREMELY favourable.

Edited later: Ahhh! I SEE where you're coming from now, Spiderman - that having businesses to the WEST should BECOME favourable if my "Old Sol" comment were valid !!! Right??? OK - a VERY good comment. (You "out of the square" thinker, you).

Maybe in 50 years time, that could be the case. But the world in which I grew up, YOU went to WORK (they didn't come to you) - and, since most businesses seem to have become esconced around most CBD's, it might take a generation or two to sort this out. By 2100, the West might well be more desirable - unless the computer generation dictates that we DON'T NEED to travel to work - in which case, all bets are off !! :D

Regards,
 
Last edited:
Les said:
Yeah, and "Old Sol" is to blaim for that (in my opinion).

I've lived in Western suburbs in the past - and having the Sun in your eyes when driving to AND from work is no joke. (Mornings seem to be particularly bad, with foggy windscreens, etc).

The anomaly to this might be Perth (and other Western towns/cities) where you get closer to the water as you go further West......

Regards,

That's an interesting point Les. I'm not sure if it's true, but someone once told me that the reason why the "poor" live in the western suburbs and the "rich" live in the eastern suburbs is because of this very reason. The "rich" have the sun behind them when they travel to and from work and vice versa for the "poor".

This is mostly true for most cities around the world except for Christchurch where i visited recently and found the more affluent suburbs were in the West and the poorer suburbs were in the East.
 
Les said:
G'day Spider,

Whoaa! You're about 180 degrees out of phase with my statement here. I'm SUPPORTING your prognosis. The Western suburbs are the ones MOST affected by Old Sol - as such are LEAST desirable.

Hence it's better to buy business/commercial/industrial/retail property (ie workplaces) in the west and homes in the east, so that you can promise all tenants that sun won't get in their eyes (assuming standard working hours).

Rgds, Peter
 
G'day Peter (Spiderman),

Yeah, it took me a few minutes, but I now see where you're coming from. See my original retort, as I added more to it (after the light went on).

Your words might well become prophetic ...... And commercial rents MUST be substantially cheaper in the West anyway - an overall WIN/WIN !!!!

Good point,

Regards,
 
Mary said:
This is mostly true for most cities around the world except for Christchurch where i visited recently and found the more affluent suburbs were in the West and the poorer suburbs were in the East.

I think it's more to do with a mixture of natural features (eg the coast and hills) and history. Eastern suburbs of some cities (eg Perth, Adelaide and Melbourne) abut a mountain range hence get more rain and are greener than the more arid western plains. Although in the Victorian town of Seymour the CBD and the 'better' area is flat and featureless, while the housing department area to the south east is undulating with valley views.

Perth's wealthiest areas are west. Ditto for Geraldton and other west coast cities. Ballarat's best area is around the lake just west of town but its worst is a bit more west. I wouldn't refuse a house in Adelaide's beachside west, though their poorer areas are more north than east, which is also true of Geelong. And Melbourne's Yarraville and Williamstown, though west, aren't exactly struggletown these days either.

I blame Captain Cook for it all!

Rgds, Peter
 
Back
Top