The velocity of money is stable
This article has been superseded by The velocity of money is a function of interest rates
The income velocity of money (which I shall hereafter refer to as the velocity of money) is calculated by dividing a measure of income, say GDP, by a measure of money, say M1.
In a speech on June 11, 2011, Economics instruction and the brave new world of monetary policy, John C. Williams, president of the San Francisco Federal Reserve Bank, said: "Let's take a closer look at the classic quantity theory of money: MV = PY. It becomes very tenuous when traditional measures of M make up a smaller and smaller fraction of the value of transactions. For example, the velocity of M1 was around three or four in the 1950s. Now it is about eight-and that's down from a peak of about 10½ a few years ago. Today's economy uses cash and checking accounts much more efficiently."
Williams's speech can be taken as typical of the general consensus among economists that the velocity of money is unstable and of little use as an economic variable.
The figures certainly bear him out, as the graph below shows.
And yet, when one views matters at a micro level, the idea that the velocity of money has become unstable because of financial innovations is difficult to credit. If I spend $1000 a month, it matters little whether I keep the money in the bank and pay by check or spend the entire amount by paying it out as currency. Both currency and demand deposits are part of M1, so either method shouldn't change the velocity of money. The same holds if I use the internet to make an instant payment. Making the payment through a debit card too shouldn't alter the velocity of money. If I pay through a credit card, the bank is effectively extending me a loan. This increases M1 and should therefore reduce the velocity of money, not increase it. In other words, financial innovations that do not change the quantity of money shouldn't change the velocity of money.
That leaves only one other explanation for unstable velocity: we are measuring money wrongly.
In the graph above, the velocity of money stays between the limits of 6 and 8 from around 1976 to around 1996 and thereafter stays between around 8 and 10. This sudden increase is attributed to financial innovations that allow the more efficient use of money. But there is another explanation. In 1994 the Federal Reserve permitted commercial banks to use a new type of software that dynamically reclassifies balances in their customer accounts from transaction deposits to a type of personal-saving deposit, the money market deposit account. [Retail Sweep Programs and Bank Reserves, 1994-99 by Richard D. Anderson and Robert H. Rasche]
When sweeps are added to M1 to get a monetary aggregate continuous with the earlier M1 series, the graph of the velocity of money is as below.
From around 1984 to 2011 (or more than a quarter of a century), the velocity of money stays between 6 and 7, which I am sure any economist will agree is remarkably stable.
In other words, the perceived instability of velocity is only an illusion. The real reason is that we have been inconsistently measuring money, because of various changes introduced by the Fed at various points of time. I am not sure I can explain the rise in velocity before the eighties; any ideas are welcome.
The other point to be made is that there is no reason to expect the velocity of money to be absolutely constant.
V = GDP/M1 (1)
The numerator GDP measures the money value of real goods and services produced by the economy. But the denominator M1 is not used to purchase only real goods and services; it is also used to purchase financial assets like stocks and bonds, which do not appear in GDP.
So, modifying (1) we get
V = GDP/(M1r + M1f) (2)
where M1r is the money used to buy real goods and services, and M1f is the money used to buy financial assets. At the height of a financial mania, we would expect M1f to be very high and therefore V to be at its lowest. And when financial markets are at their gloomiest we would expect V to be at its highest.
The graph below, an enlarged view of Velocity (M1 plus sweeps) for the period 2001 to 2011, shows exactly that.
Using the monetary aggregate I have developed (called Corrected Money Supply) yields the graph below.
In conclusion, the velocity of money is not only extremely stable but also very useful. Looking at it helps us to know when speculation is rising to unacceptably dangerous levels.