Not PC has a very interesting post on identifying recessions. Using a special measure of the money supply as a leading indicator he shows that a large fall in this measure usually implies a recession (*).
This makes sense as a recession is a time when money demand falls rapidly. As the stock of money is determined by an exogenously set price and a money demand curve we would expect “money supply” in this stock sense to fall rapidly.
I disagree with his articles conclusion in the current monetary policy context (“the expansion and contraction of the money supply is the single greatest factor in causing booms and busts“) – but I do think that the method he points out is extremely clever. It is also a welcome reminder that monetary aggregates still give us some information – even if the information is all mixed up by definitional errors and technological change 🙂