Several MPC members not previously known for their devotion to monetary analysis have cited weak broad money supply growth as a reason for expecting inflation to fall below target over the medium term, a prospect warranting consideration of "QE2" asset purchases. The implications, however, of a given rate of monetary expansion for the real economy and inflation depend on the velocity of circulation. The claim that money growth is "too weak" assumes that velocity will be stable or decline but it has risen strongly over the last year and there are grounds for believing that this pick-up will continue.
 
Velocity is defined as current-price GDP divided by the stock of money; it represents the flow of income supported by each unit of cash. Since the start of quarterly data in 1963, broad money* velocity has fallen by 0.5% per annum (pa) on average – see chart. To support sustainable real economic growth of about 2.5% pa with 2% inflation, monetary expansion and the change in velocity must sum to about 4.5% pa. If velocity were to decline at its historical rate of 0.5% pa, this would require an increase in broad money of 5% pa. Put differently, sustained growth at the recent slow pace – 1.6% in the year to August – would support a rise in current-price GDP of only about 1% pa, suggesting renewed economic contraction or a big inflation undershoot.
 
Velocity movements, however, are not fixed or "exogenous" but vary depending on the relative attraction of money as a store of savings. When interest rates on bank deposits fall beneath inflation, as at present, consumers and companies have a strong incentive to economise on cash holdings. This boosts current-price GDP both directly as part of the monetary "excess" is spent on goods and services and indirectly as purchases of other assets push up prices, leading to expansionary wealth and confidence effects. The combination of higher GDP and lower money holdings, of course, is reflected in a rise in velocity.
 
Recent economic and financial trends are consistent with such a process being under way. Current-price GDP rose by a stronger-than-expected 5.7% in the year to the second quarter; with broad money up by only 1.4% in the year to June, velocity surged by 4.3% – the largest annual gain since 1980. Other evidence of a “dash from cash” includes record retail sales of mutual funds, a decline in the "liquidity ratio" of insurance companies and pension funds (i.e. the proportion of portfolios held in money and short-term securities) and generalised strength in asset prices, with equities, bonds, commodities, houses and commercial property all appreciating over the last year.
 
The view that current monetary growth is "too weak" implicitly assumes that the rise in velocity will slow sharply or reverse but it is more likely that the financial shift is still at an early stage, with many consumers and institutions yet to take on board the MPC's message – delivered most recently by Deputy Governor Bean – that monetary savers should expect to suffer a sustained depreciation of their real wealth. When real interest rates were last significantly negative in the 1970s, broad money velocity rose by 39% over six years, or 5.6% pa – see chart. If such an increase were repeated now, money growth of 1-2% pa would deliver a large inflation overshoot.
 
The MPC’s born-again "monetarists" talking up QE2 are playing a dangerous game. Broad money has been rising faster recently – by 4.5% annualised in the three months to August. With banks in better shape, asset purchases could have a much larger monetary impact than in 2009, when cash injections were partly absorbed by capital issues. Combined with the rising trend in velocity, this suggests that QE2 on any significant scale would lead to a further acceleration of current-price GDP expansion, entrenching and possibly extending the recent inflation overshoot.

* "Broad money" here refers to the Bank of England's preferred aggregate M4ex (i.e. excluding money holdings of "intermediate other financial corporations") from its inception in 1998 and M4 for earlier years; quarterly growth rates were chain-linked to derive a break-adjusted level series.