Currently new dwelling permissions is well outpacing completions and starts. The former is much more than a leading indicator of the latter, it has on occasion wildly risen above it. Most notably in the period 2005-2007, just before the crash, and was a sign of trouble.
What happened was that interest rates nudges upwards taking the heqt off demand just at the point the Brown Government was turning up the heat on housing in the light of the Barker review etc. As a result developers built up their landbanks and maximised densities on sites in their landbanks (providing that supply is not fixed and does not vary according to a crude neoclassical supply-demand formula), at the point when demand was slackening, leading to a severe overhang of supply when credit dried up in the credit crunch.
The same could easily happen again. Which is why I think Carny is reluctant to raise interest rates as that sucks effective demand out of the economy as a whole whereas macroprudential measures do not. But beware an excessive continued uplift in permissions, especially if developers leverage off the rising value of their landbanks, could trigger another crash. It is not simply an issue of too little housebuilding, but developers building up excess inventory in comparison with effective demand. If effective demand falls for any reason developers could be left with too much inventory and cease building, in these cases the macroprudential measures to take is to restrict housebuilding.
In all cases of housing booms and bust whilst shortage of housing supply triggered the speculation, it was excess resultant supply in comparison to lowered demand that caused a glut triggering the crash.