The chart below depicts the inverse relationship of the index with the unemployment rate. Note SVI is lagged six months, meaning the levels indicated for unemployment, say in the last observation, October 2008 is actually matched with the value for SVI back in April (hence the predictive potential of the indicator).
The reason for this sequential relationship? Labour economics tells us that the costs of acquiring skilled labour cause employers to smooth their hiring decisions during peaks and troughs in the economic cycle (see Walter Oi's article, Labor as a Quasi-fixed Factor, in the Journal of Political Economy, Chicago Press, 1962). In a down cycle, they will maintain their skilled workers for as long as possible instead of laying them off. During upswings they do not necessarily start hiring as a certain amount of "slack" has been pent up from the low season. So the skilled vacancy reports can be used as an early warning device for predicting future employment conditions.
The problem? The most recent reports show the index diving to its lowest levels in recent times. The early November report has SVI at 65.6 where dipping below the base (of 100) signals a decline. This is a bad omen especially since the downward trend has yet to hit a bottom. Even if we are to assume fiscal stimulus and monetary easing occurs this December, the sheer momentum of the current mood will see unemployment inching back up from 4.3 per cent in October to about 6 in May next year (according to our estimates).
There are of course other factors such as interest rates and industrial relations policy that need to be considered, but all other things being equal, we are in for a bumpy ride ahead. (For those who would like to read a more detailed forecast of where it is headed, please send me an email, and I can provide you with a more technical presentation of the model that underpins the analysis).