In trying to recover from the recession of 2001 the Fed expanded money supply rapidly, setting the stage for the surge in housing prices and the Great Recession of 2007-09.

The Fed's problem was that it did not know when to stop the monetary expansion and the reason it did not know was that it had no accurate measure of money. The graph of M1 below from 2001 to 2011 shows how useless a measure it is. During a period which saw the worst money crunch since the Great Depression, M1 does not fall at all. In fact during the severest period of the crunch it shoots up drastically.

The M1 graph on the St Louis Fed web site marks off the bounds of the recession and shows the contradiction more clearly. If M1 were to be mirrored in the x axis the reflection would be a better indicator of the economy at the time. The measure of money, in other words, behaves exactly the opposite of how we would expect it to behave. In the absence of an accurate money measure the Fed was in essence shooting at a target in the dark, with no way of judging whether its shots were approaching anywhere near the target.

Does the Fed know any better now? The answer alas can only be negative. There is every reason to believe that Bernanke is following in Greenspan's footsteps, simply because the Fed has developed no better measure of money.

In What is money supply? written in August 2010 I argued the rationale for a new measure of money, Corrected Money Supply or M_c. I showed there that a good measure of M_c for any month was M1 for that month minus the sum of savings for that month and the 9 previous months (i.e. 10 months in all). Shortly after that article was written the values of savings (statistics available on the St Louis Fed web site) were drastically revised not just for two or three previous months but for more than a year before. The graphs in that article are therefore no longer accurate. But the logic remains valid.

Using that logic with the revised figures it is possible to show that M_c is roughly equal to M1 minus 12 months savings. The resulting graph below accurately captures events from 2001 to 2010, which M1 does not. More frighteningly it shows that M_c at the beginning of 2011 is approaching a figure not much lower than its peak in 2006.

On 1 March 2011, in presenting the Fed's semiannual monetary policy report to Congress, Bernanke said that "last quarter, for the first time in this expansion, our nation's real gross domestic product (GDP) matched its pre-crisis peak." Given that M_c can be calculated only after a lag of a couple of months (because savings figures are available only with that time lag) it is probable that the US economy is right now in exactly the situation it was in 2006 before the housing market collapsed.