Tuesday, December 08, 2009

Monetary Time Lag & Money Demand

I recently discussed the issue of a monetary time lag and why it exists. Anyone observing data on money supply and prices can see that the immediate empirical link between money supply changes and price changes at a certain point in time is weak to non-existent. That leaves us with two options: either the causal link doesn't exist or it exists with a time lag. As I and my good friend Daniel Halvarsson documented here and here, an empirical link seems to exist, but with a significant time lag of 1 to 2 years.

But why doesn't money supply immediately affect prices? Why does it take as long as 1 to 2 years?

Well, one reason is that many prices are relatively sticky in the short-term. Most retailers are for good reasons reluctant to change their prices up and down on a daily or even weekly or monthly basis because they know customers will resent price increases and because contracts with suppliers usually lasts for several months. Furthermore frequent price changes will create menu costs for businesses.

Another, more overlooked but arguably even more important factor is the role of fluctuations in money demand.

During cyclical downturns, investors often get more risk averse. And the latest downturn is a very good example of this, as the fear of a financial collapse led investors to avoid anything deemed risk and instead demand less risky assets. The demand for money in its various forms, not just physical notes and coins but also demand and svaings deposits and money market mutual fund (though the temporary loss in value for one money market mutual fund caused demand for it to go down temporarily until the Fed had restored confidence in it) holdings therefore increased dramatically.

So, during cyclical downturns the demand for money suddenly increases, something which both causes the price of money to increase (causing the price level of goods in terms of money to drop since money is now more expensive) while it also causes an increase in the supply of money by banks and other money producing financial institutions.

Because this increase in supply was largely (though usually not entirely since at this point in a business cycle, monetary authorities will try to produce "monetary stimulus") caused by this increase in money demand, it will usually not be associated with immediate price increases.

However, once people believe that the crisis has abated, people will reduce their demand for money and instead demand more risky assets. That reduction in money demand will at the same time reduce money supply growth and raise money prices of various goods. That is why money supply growth has been particularly strong in the end of downturns and beginnings of booms (like in 1983 and 2001) and then been more moderate even as the boom has continued. And that is also why booms usually have preceded even some time after money supply growth halted. 2008 differed somewhat from that pattern as a downturn followed and was combined with significant price deflation just a few months after money supply growth ended. But that was a result of the dramatic increase in money demand caused by the increased and very justified worries about the soundness of institutions like AIG, Fannie Mae, Freddie Mac and Lehman Brothers and securities associated with them.