The analyst in search of a property bubble: Arek Drozda

Since property prices started to rise again a key question that comes to mind is – what about the bubble? Are we there yet or far from it? Opinions are many but I prefer to deal with facts and figures so, let’s examine the evidence…

Firstly, we have to define the “price bubble”. As it turns out, even this simple task is not without some challenges due to a variety of perspectives on the issue. Therefore, let’s settle on a simple and succinct definition from Wikipedia which explains price bubble as “trade in high volumes at prices that are considerably at variance with intrinsic values".

So, according to this definition, the first precondition of a bubble is “trade in high volumes”. Since I do not have access to property sales data for the last few decades, the best proxy of this information is the number of mortgages issued by financial institutions, as published by the Australian Bureau of Statistics. This data series goes back to 1975 and is presented on the chart below. The red line on the chart is a 12 month moving average of monthly volumes of loans for purchase of established dwellings, new dwellings as well as for construction of dwellings. Loans classified as refinancing are excluded.

The spikes in volume can be clearly identified in 1984, 1988, 1994, 2003 and 2007. Volumes were also elevated in 2009 and are up in 2013 from a very low base but these appear so far to be “reactions” in a downtrend hence, rather unlikely candidates for a bubble.

 total_dwellings_financed

The second precondition for the existence of a bubble is that “prices are considerably at variance with intrinsic values”. The definition of an “intrinsic value” is a much bigger problem. So again, let’s opt for something simple and widely used by economists. In particular, the most common approach to determine the intrinsic value of properties is to calculate price to household disposable income ratio, or price to rent ratio, and compare present and historical values. The second measure is my preference since there are long term data series readily available which makes construction of this indicator much easier. However, both those measures should show a very similar result.

I set the ratio of house price to rent as 1 for 1975, which is the start of housing mortgage data but I also calculated the ratio back to 1950’s - to a point in time when modern and unregulated post World War II housing market began in Australia.

Read more over page

 

 


 House_prices_to_rents

Looking at the above chart we can immediately identify a problem – the line only rises and it never fell back to or below 1950 level. The ratio first jumped in early 1950’s, then in mid 1970’s, again in late 1980’s and in early 2000’s, peaking so far in 2010. In other words, it implies that prices are “at variance to intrinsic value” continuously for 60 years now… Yet, despite that implied long lasting exuberance, property prices in Australia fluctuate periodically – they advance and then they pause or fall slightly, then they advance again to a higher level.

It means that this popular measure of intrinsic value is rather insensitive to short term variations in market conditions and therefore should probably be only applied in analysis spanning many decades or even centuries. In other words, simple measures like price to rent ratio, are rather inadequate for a shorter and medium term analysis, relevant to a person’s lifespan.  A more responsive measure is needed in order to determine whether prices are exuberant or not. I will explore this further in my next article.

For now, the only conclusion that can be drawn is that, according to our original definition, if we accept that 1950 is an appropriate starting reference point and that prices to rents ratio is an appropriate measure of a variance to intrinsic value, Australian property is in a very long term bubble that peaked so far in 2007 (the point of the highest volume of loans and high ratio of prices to rents) and that this bubble is currently deflating.

However, there is more that should be said. In particular, it is an awfully long term bubble and it may take an awfully long time to deflate, if at all. It is not out of the question that what we observe on the second chart is a reflection of ongoing structural changes in the housing market – therefore, that the chart indicates what the current intrinsic value is and not the variance from a historic intrinsic value, as this information is commonly interpreted. If what we observe is indeed the result of long term structural changes in the market, this “bubble” could as well inflate to even bigger proportions, as it happened in 1950s, 1970s, 1980s and in 2000s.

I will leave you with an interesting question to ponder about - what would it take to completely deflate this bubble and bring ratio of prices to rents down, back to 1950’s level? It would involve a serious modelling exercise to come up with a range of scenarios that could deliver that outcome but, as an illustration, prices would have to halve from current levels and, at the same time, rents would have to double to get us close to that result in the next 20 years. How likely it is to happen without a catastrophic economic shock or a drastic government intervention is not difficult to guess.

 


Arek Drozda is an independent analyst who has worked in the public and private sectors for over 20 years in business development, data analysis and in building geographic information systems.

 


 

Community Discussion

Be the first one to comment on this article
What would you like to say about this project?