Explaining why a crash can be contained to new development areas

I just thought I'd post some ideas on how a crash can be contained. We often hear, for example, that major post-bubble markets crashed between 30-60%. In Dublin it was about 50%. I wanted to point out, however, that this 50% was very uneven. I don't have a good direct comparison of suburbs, but let's illustrate it this way:

Say you have a city with 10 suburbs. Each contains 1000 homes, and typical fundamentals-based demand sees 5% capital growth annually (driven by for example population change). One suburb is the CBD. One is the fringe (the "rapid" zones). The rest we'll call "normal" suburbs.

Now suppose the city goes through a bit of a boom for a few years, bringing the median price up to an exciting $100,000. Boom turns to bubble, and this sparks a construction frenzy.

In one of these bubble years, 1000 homes are added to the CBD and 1000 to the fringe (we'll call these, collectively, the "rapid" growth suburbs. Many are bought mostly by excited recent investors, many are bought by builders, some are bought by families. The extreme demand sees prices in these two suburbs jump 20%.

Sales continue in the "normal" suburbs, but at a much slower rate of 5% (on average). Suppose we sell 500 homes between all the "normal" suburbs (assuming normal demand means a sales rate of between 30-70 homes depending on suburb annually, which is your fundamentals-based demand).

Our median for the "rapid" areas is now $120,000. Our median for the "normals" is $105,000, but our median for the whole city is $117,000. (This is why it can be risky to look at whole market trends). Remember the median is calculated based on the total number of sales, so the fact development has been extreme in those "rapid" areas skews the median. It has been distorted by all that construction.

The mania continues, and offshore or interstate investors (especially where inexperienced) may see the ImaginaryTown median and think holy moly, I need to buy in right now. Suppose a further 1000 homes are added to each of the rapid suburbs, again spurring 20% growth, and 500 sales again occur in the "normal" suburbs (supply there being tight). This causes a 10% jump in the "normals", a contagion effect, as investors want to buy there too.

Our new medians are $115,500 for the "normals" (an average, of course), and $144,000 for the "rapids". But here's the rub: the median for ImaginaryTown as a whole has ballooned to a smidge over $140,000. The vast majority of the sales have occurred in the "rapids", but the median has risen everywhere.

The above is hopefully a good insight into how bubbles become self feeding. People looking at whole market trends can be stunned into thinking they need to buy in the most rapid, frenzied areas, and in some cases prices can rise by a sort of contagion effect (the bigger a bubble gets, the more likely this distortion is). This has definitely been the case in Sydney with some suburbs rising into the multi-millions, but arguably it hasn't been the case (at least not to a significant extent) everywhere and in all cities outside of those rapid development areas.

Now, suppose the bubble bursts. If you own a "normal" home in ImaginaryTown, you might have seen some out-of-the-ordinary growth, sure. A total of 15% is your jump (varying by your particular suburb obviously), maybe 5% faster than you would have seen based on fundamentals. If the bubble bursts, this is likely to retrace a bit. You could also see an overcorrection, but regardless the "normals" are suburbs that will always see strong demand from families and owners, and from savvy investors looking for cashflow. If there's been a big distortive effect (like in Sydney) the risk may be a little higher, but it's also a big, economically diverse city, and demand always has a floor. People need homes.

If you own in the "rapids" then you've got a big, big problem. Firstly, 1000 homes annually is faaaar in excess of fundamentals-based demand, and this oversupply is cumulative. About 80% of your stock is probably in excess of any actual needs. Yes there's been demand but it's all been speculative, sentiment-based demand, at the peak of a bubble, and many of the buyers are probably recent investors.

You've also got the issue of a lot of developers having money on the line, and few recent buyers having much in the way of equity. Those buyers are more likely to have low incomes, and thus be at risk in a recession. You may also lack infrastructure on the fringe, and your tenants might only be living there because they are struggle to compete with the temporary jump in demand for "normal" suburbs.

If you see a crash in these areas, hopefully you can see how the downward spiral could be very steep. You could easily get a "40% crash" in ImaginaryTown, yet the assets losing out are mostly contained to those "rapid" pockets. The median of ImaginaryTown will probably hit the floor, but few suburbs actually grew "20% per year".

Obviously this is just an example but it's a good illustration, I hope, of what I saw happen in Dublin. Fringe suburbs (later called "ghost estates" were rapid development areas and when the music stopped, many of them became next to worthless. "Normal" suburbs though held up reasonably well, at least beyond the initial panic, and attracted strong rental growth because the supply excess was also swallowed up as people shunned new development zones. They wanted to live in good suburbs).

I hope this also shows why it's critical to look for cash flow as a big part of your strategy. And why it's a terrible idea to attribute city-level median trends to your own portfolio. House price indices can only measure very broad trends, and they are easily distorted. Look for suburb-level trends.

 

Same can be said for investing in mining towns where there was huge cashflow on the back of rising asset prices and strong temporary demand.

 

Yes, definitely. A lot of regional towns, too. In those areas fundamentals-based demand also has a bigger effect because that can see such big shifts. In say a seaside town, you lose a lot of demand if tourism goes bad, and so on.

 

I've been seeing this effect at work in Armstrong Creek and thereabouts.

There are rental opens for places that haven't finished final construction. You walk in to this brand-new house (the ground is still mud, the whole street is a construction site) and see no oven, no central heating/cooling system.

They're trying to get a tenant in place before they put those things in.

You leave, not for you perhaps and then an amazing thing happens... a rental agent calls you. Are you interested? Want to put in an application?

Astounding stuff... normally when you rent agents DGAF. But now they're chasing tenants to live in these muddy streets in new estates.

I've been watching it for a while... I see houses go up for rent, no tenant... and then they go up for sale.

Or there is a sale and then they immediately hit the rental market... and sit there because why would anyone pay $400 p/w to live in Armstrong Creek when you can pay the same and live five minutes from the CBD in a street that isn't mud?

There is so much construction going on, all these investors pouring money into these estates and you can see prices rising as the money froths in. People who live there are seeing capital gain that will inevitably reverse once oversupply hits.

The rental market always shows the truth of what is happening. When you have agents chasing tenants... that's a blaring warning of oversupply and desperation.