With the Oz stock market falling nearly 5 per cent on Thursday, and down nearly 15 per cent in the last few weeks, it’s a good time whether this will have real effects beyond the value of our superannuation. The immediate starting point of the current disruptions was evidence that defaults on US mortgage markets were worse than expected. An obvious question is whether this underlying shock is large enough to have substantial effects in itself, or whether the problem is mainly one of liquidity and confidence.
Ben Bernanke has estimated possible losses from subprime mortages at between $50 billion and $100 billion. As this article notes, that figure may sound large, but it would represent a tiny fraction of the $56.2 trillion in U.S. household net worth. To be precise, it’s somewhere between 0.1 per cent and 0.2 per cent of net worth.
Arguably, to the extent that real resources are dissipated, as commonly happens in a process of default the relevant comparator is something like annual output. Expressed relative to annual output, a loss of $50 billion to $100 billion bit larger (between 0.5 and 1 per cent) but still not really scary. We could get to a similar number by supposing that house values will decline modestly in response to the crisis.
It appears that there is around a $1 trillion dollars in sub-prime debt outstanding, and much of this is likely to go into default this year, as house prices fall and interest rates (initially at low teaser rates) reset. The loss from default is confined to the costs of foreclosure and the shortfall in recovery when the house is sold, but it seems likely to be at or above the high end of Bernanke’s range.
The real problems emerge if there are more bad loans elsewhere in the economy. It’s already become apparent that housing defaults go beyond the sub-prime sector to the Alt-A sector (low-doc loans to borrowers with good credit) and maybe beyond. More generally, a lot of recent financial market activity has been premised on the assumption of cheap debt or, more precisely, a low price for risk. A higher price of risk implies that there must be quite a few heavily geared assets that are worth less then the debt they secure.
We’re still waiting to see how large the losses are. As a rough guess, I’d say that a quick response from the central banks can deal with liquidity problems if the total losses are less than about $500 billion. After that, it’s anyone’s guess.