Friday, April 26, 2013

Watts' model of cascading network failure

I have written in the past about how social and economic networks are a necessary ingredient for a proper understanding of economic patterns. The rise of social network platforms like Facebook and Twitter has allowed a thorough analysis of empirical regularities seen in networks in the social domain. Stephen Wolfram has a great blog about the regularities observed in Facebook data scraped using WolframAlpha.

One of the more interesting networks models is Duncan Watts' model of cascading network failure. Simply, each node in the network has a specific tolerance to failure, and it the share of adjoining nodes that have failed exceeds this tolerance, that node will also fail. For example, a node could have a tolerance of 0.5, so if more than half of its neighbours have failed, it will also fail, leading to a cascade of failures of its other neighbours.

This simple model is quite versatile. It suggests underlying mechanisms behind how fashion fads arise, or why investors tend to go with the herd, or why some industries produce superstars even though no one can objectively tell the difference in the quality of their skills

The methodological individualism so fondly embraced by the economics crowd has at its core the concept of utility, but stops short of answering the far more important question – where does our utility function come from if not our environment and our interactions with others? A model of networks can help explain the source of utility and begin to give a picture of how unique cultures and customs arise. 

In any case, I have generated an animated version of the model that simulates over a random network, with 5 random nodes ‘shocked’ to initiate the model. The histogram shows how many of the 20 different shocks have led to cascades of failure of a particular number of nodes. 

Enjoy. Follow me on Twitter. And please share.


Thursday, April 25, 2013

Timing the residential property cycle

One interpretation of recent data is that investors seem happy to jump back into Australian residential property markets. Perhaps due to a search for yield. Perhaps from foreign cash seeking a safe harbour. Or perhaps it’s simply time for the Aussie love affair to be rekindled. Holes are over. It’s houses turn.

With these winds of change in the air maybe it is time to take a step back and look at the long term property cycle itself.

Property industry types talk about the cycle like a mythical being - unless you have witnessed it yourself you won’t know how aggressive the beast can be to your leveraged finances.

Long term regularity of asset price cycles is an intriguing proposition. Is the 18 year cycle really a good rule of thumb? If so, why don’t investors expect the cycle, and remove it through their anticipatory actions?

A simple answer might be that investors would anticipate the cycle if credit markets would allow it. But the banks supplying credit are themselves constrained by previous movements of the market. Thus the interaction of prices and the willingness to supply credit seems to be pretty decent explanation of the peculiar regularity of long term cycles. Thanks Minksy.

One way to think about the nature of the cycle is in terms of returns from yield compared to capital growth. At the bottom of the cycle equities, including property, are seen as risky places to preserve capital. During the boom expectations of capital growth return, and equities become the assets to hold.

If this truly reflects some fundamental emergent dynamic in the economy, a simple rule of thumb is to buy the high yields at the bottom of the cycle, and sell capital gains at the top.

But how do we know when yields are high? We need a relative measure rather than an absolute measure.

In the past I have used the mortgage rate divided by gross yield as a measure of the relative value of residential property. The theoretical picture is the the mortgage rate is a good proxy for the yields, net of capital growth, available in the economic generally. Gains above this rate typically arise from capital gains.

When the gross yield is close to the mortgage rate, theory says that the price is reflecting an expectation of low capital gains. But that would be wrong, given that it is the same theory predicts and equilibrium asset price.

In reality this would be a good time to buy.

The theoretical explanation is that these low growth expectations arise from recent experience of low growth - the same feedback that feeds the cycle upswing when high prices feed into expectation. If markets are myopic, you can forget about finding anything useful about expectations in the prices themselves.

So where’s my evidence for this? The graph below is an update from a previous post. With recent rental growth, price falls, and falling interest rates, this simple measure is showing that now is a good time to buy.

I have also created a second measure - the mortgage payment per dollar of a 30 year loan divided by the yield. The second measure adjusts for the fact that the cost of buying asset, in addition to the cost of interest, is a higher portion of the total cost at lower interest rates.


What is more surprising is the regularity of a head and shoulders-type pattern - similar tops and bottoms, and a similar period to the cycle, in this case 15 years. Not too far from the 18 year rule of thumb. And not too far off the stylised asset price cycle seen so regularly when discussing the latest housing boom

Given this regularity, and the strong buy signal, my internal model of the market suggests two possible future paths.

1. Renewed cycle

A great time to buy in most capital cities was around 1998. This year preceded a boom in Sydney, that cascaded across the country for the next 8 years. My chart shows the cycle at around 15 years, meaning this year is a good time to buy. The 18 year rule of thumb is then 2016 - just three years time.

Given the expected resources shock in the second half of this year and early next, I would not be in a rush. Also it may be wise to get better signals about the direction of international markets, particularly the US before leveraging into Australian housing.

I expect to be on the lookout for well located land in about two years time unless I get strong signals that the second path is playing out.

2. Stagnation

Given the weight of private debt, the already low interest rates across most of the developed economies, and a general reluctance for increased public spending to maintain employment and stimulate private investment, could we be heading to a long credit-constrained stagnation that requires major price adjustments in wages, rents, and currencies.

I have no good reason to believe one way or another. Political outcomes in Europe, China and Japan will be critical, as will our domestic adjustment following the mining investment peak.

My gut says that the fundamentals of continued current account deficits, which reflect inflow of foreign asset demand, and scope for much lower mortgage rates will probably allow for another cycle to ramp up by 2017.

I don’t expect it to be as severe as the last cycle for a few reasons.

  • Inflation will be low unless the AUD falls significantly. Thus real gains could be high without such dramatic nominal gains.
  • Mortgage rates still have scope to fall to around 4% in the next two years.
  • But I expect the memory of the financial crisis and a stricter regulatory environment will mean tighter bank lending
  • The demographic shift of baby boomers seeking to get out of negatively geared residential property will dampen capital gains

If most investors are myopic, those who consider the long term will have an advantage in any market. And what we have seen here is that we seem to be in a relatively attractive period for buying residential property assets. Just remember to consider all the other macro and political factors in your own assessment of the market.

Tuesday, April 23, 2013

Australian age-dependency ratios

Everywhere you turn it seems that higher rates of population growth are seen as a 'solution' to an ageing population.  Here's one recent example.  My general views on this matter are found here.

At the very least there should be a publicly accessible model of population growth to verify the claims being made in this debate.  The productivity commission has modelled population growth for this purpose, although the intricacies of the model are not at all clear or public.

My first step towards this is to actually look at the historical demographic record in Australia.  As you can see from the interactive chart below, the country's age dependency ratio has been steadily increasing since at least the 1970s.

Offsetting this age dependency has been a quite dramatic decline in youth dependency as fertility rates fall. The total dependency, or number of children and retirement age people per working age person is at record lows.

The next step is the add some features to this model to allow a choice of assumptions about immigration rates (and ages) as well as birth and death rates, to see exactly what how immigration policy is affecting demographics and whether there are some circumstances in which the 'immigration solves ageing' slogan may hold.

Enjoy.