UPDATE: Today I came across this old post that I started a year ago and never quite finished. I’ll finally post it now, but I’ll try to put it in at the date when I originally wrote it.
Suppose everyone in the world burns up 10% of the cash that they possess in a temporary fit of insanity. That destruction of money would decrease the money supply, but we would have no way of measuring that a decrease had happened. The monetary authorities (like the Fed) would go on assuming that the supply of money is holding steady with historical trends as they always do. But the decrease in money would have an effect on the economy. GDP would likely decrease because it would be harder to make transactions and interest rates would tend to rise as people tried to take more cash out of banks. The Fed would observe a decrease in output and assume that it was due to a decrease in the velocity of money instead of a decrease in the money supply.
One of the fundamental problems in macroeconomics is the difficulty in measuring basic variables. It is impossible to directly measure V at all. The Fed calculates velocity based on the Fed’s imprecise measurements of nominal GDP (PY) and the quantity of money (M). Nominal GDP is PY, where P=price level, and Y= real GDP. M = the quantity of money. The Fed then calculates velocity (V) using the formula V=PY/M.
In reality, when 10% of the cash disappears, the real velocity of money tends to increase as people are forced to make a smaller pool of money work harder to generate the flow of goods that they want. But when the Fed measures a decrease in PY, the Fed cannot measure the decrease in M and erroneously assumes that V must have decreased, when in fact it increased.
Burning cash might seem like an unrealistic scenario, but a similar thing actually happened during the Great Depression when ordinary people lost faith in banks and withdrew their cash. But instead of burning it, they literally buried it. They stuffed cash in jars and buried them in the yard or stuffed cash in mattresses and sewed them up. This semi-permanent sequestration of cash is just as much a decrease in the money supply as if people had burned the cash except that the buried cash could eventually be unburied again (if it were not lost or otherwise destroyed).
When the pool of actively used money shrank out of the economy, the economy shrank for the reasons mentioned above. Anna Swartz and Milton Freidman’s book, A Monetary History of the United States blamed the Great Depression on the Fed’s failure to keep the money supply from shrinking. Swartz and Friedman based their book on new measurements of the money supply that economists had been ignoring. Monetarism could never have been born without better measures of the money supply. Today we are in a similar situation to the Great Depression because the money supply is being missmeasured and that is leading to misunderstanding about the macroeconomy.
Recessions Happen Because of Hoarding
The Keynesian-Monetarist explanation for recessions is when there is too much savings. Economists typically say that savings (S) equal investment (I) in an economy. The only way this could be true is if economists define investment as borrowed money because financial savings is always money that is lent out. But that is wrong. Savings (S) does not always equal investment (I) because sometimes money is hoarded. Hoarding is when people don’t spend money AND they don’t lend it to borrowers who would spend it on investment (I). In the Great Depression, people hoarded their money in jars in the yard.
Hoarding is the equivalent to a decrease in the true money supply. The Keynesian explanation for a recession is that people increase savings (S) and/or decrease borrowing (I) at the very same time, but that is impossible if S=I. What really happens in a recession is that there is an increase in hoarding. This hoarding is the amount of increased savings that are not lent to anyone. It is simply money that is taken out of circulation by sewing it inside mattress stuffing or secretly burying it in a jar in the yard. Usually when someone saves money, they lend it to a bank who lends it to businesses who invest it in building something useful. But this virtuous cycle breaks down during a recession. Some saved money ceases to be lent out. It is hoarded instead and hoarded money might as well be burned into ashes for all the good it does anyone during the years it is being hoarded. The hoarded money only becomes useful again if it is resurrected from whatever dark pit it had been sequestered in.
Unfortunately, it is impossible to measure the amount of money that individuals sequester in pits in their yard, but that is rarely a big problem. Ordinary individuals only hoard cash in pits under conditions like the Great Depression when there was deflation and fear of bank failure. Deflation makes the real value of cash rise and encourages people to hoard it and bank failures make people withdraw their cash from banks which they see as being less safe than a pit in the back yard.
So if the US isn’t having problems with people hoarding money in pits in their yards, where is the money been hoarded during this recession? This is one of the great empirical flaws of Keynesian-monetarist theory. We don’t have a way to measure hoarded money so we still say that savings equal borrowings (S=I).
During a liquidity trap when private individuals and for-profit corporations lend their money for zero interest, it is obvious where the hoardings are going. Any savings that are lent for zero interest rate are being hoarded because the money earns the same ROI as cash and it is an attempt by savers to lock up the money for a period of time. For example, a 5-year bond that is bought with a zero interest rate is the same thing as cash that is buried in a time capsule that cannot be opened for five years. This is where some of the hoarded money goes. Instead of going to people who will spend it on investments (I) or consumption (C), money flows away from risk-taking businesses towards ultra-safe government bonds even though the government does not spend more money when its borrowing costs decrease. Almost anyone else would spend more money when their borrowing costs declined to a negative real interest rate, but government spending is hardly influenced by low interest rates. Private businesses tend to spend more when interest rates decline, but politicians defy market incentives and common sense. They have actually been spending less than usual over the past four years during which private savers have been willing to pay the government (a negative real interest rate) to take the money.
Another place where money is being hoarded is in the excess reserves of the banking system. This has been ignored by most economists because it is a new phenomenon and they don’t seem to know what to make of it yet. Excess reserves are moneys that banks are hoarding instead of lending out. Bank reserves have the same effect on the economy as money that is secretly sequestered in backyard pits. It is just as sequestered in the vaults of the banks as it would be if it were buried in private pits.
Banks are special because they create money when they lend it out, but we don’t really know how much money is being created by bank lending (or ‘leverage’) because part of the banking system is “shadow banking” done by financial institutions that lend out money (and thereby create money) while avoiding bank regulations by not entering the formal banking system. Note that this is one of the big potential problems for Admati and Hellwig’s recommendations to raise reserves requirements or proposals for full-reserve banking. Bigger reserve requirements increase the gains of shadow banking.
Bank reserves should not be counted as part of the money supply and have no effect on inflation because bank reserves have absolutely no more effect on MV=PY than cash that is burnt. Normally bank reserves don’t matter much for economic analysis because they are usually fairly constant. When reserves don’t change, they don’t change their effect upon monetary policy. But reserves changed dramatically during the 2008 financial crisis. They exploded. This caused many people to predict runaway inflation because they looked at the monetary base (MB) which mostly measures bank reserves and the MB showed what looked like a dramatic expansion of the money supply. For example, the NY Fed said that “Meltzer (2009), …worries that “the enormous increase in bank reserves — caused by the Fed’s purchases of bonds and mortgages — will surely bring on severe inflation if allowed to remain.” That is ridiculous. Bank reserves are irrelevant because they are not money and inflation is a monetary phenomenon. Reserves could only become inflationary if they were first converted into money that enters the real economy. That didn’t happen as the following graph demonstrates.
Many naïve economists like Meltzer think that the doubling of an official money measure (the Monetary Base in blue above) should cause a big increase in prices (P), because of the equation MV=PY. GDP (denoted as Y in the equation) and velocity (V) normally don’t change that much in one year, so the only other alternative is an increase in prices, but inflation didn’t happen. If people had been looking at better measures of the money supply like MZM (the green line) instead of an irrelevant measure of phantom dollars hoarded in banking vaults (like MB), they might have worried about the deflation and recession like what actually happened in 2008.
MZM (Money of Zero Maturity) attempts to measure all financial assets that have “zero maturity” which means that they are about as liquid as cash because they can be used as a medium of exchange at any time with zero advance notice. A Treasury bond that matures in 3-months cannot be used as a medium of exchange unless it is first sold in exchange for money. It has a 3-month maturity, not a zero maturity. Any financial asset that cannot be immediately exchanged for goods and services is not money and any financial asset that imposes a penalty for early withdrawal (before maturity) is not money. MZM attempts to measure everything that can be immediately used for transactions buying goods and services.
The conventional rationalization for the failure of MV=PY to predict that the MB expansion would increase inflation (P) or output (Y) was that there was a collapse of the velocity of money (V). But this is contrary to the conventional wisdom about V which says it should not change rapidly and unpredictably. The conventional wisdom makes sense. Given official measures, velocity of MZM plummeted by almost half (in total) since the recession began, but nobody has come up with an explanation for why the true velocity of transactions could have slowed to half its ante-recession speed. Why would people become permanently so much slower to pay than they had been before the recession?
A better explanation is that MZM did not really explode and its velocity did not plummet. MZM is just poorly measured. I suspect that the apparent rise in the money supply is miss-measurement. MZM abruptly rose when bank reserves skyrocketed in fall of 2008, but it is hard to believe that the real money supply rose right at a time when velocity was plummeting. Most money is created by lending and why would lending increase when people aren’t using money as fast? What really happened was an increase in hoarding reduced the real money supply. I suspect that bank hoardings (reserves), also artificially inflated the official money measures somehow.
Bank reserves are far from money. Reserves are never directly exchanged for goods and services and nobody who wants to buy goods and services has any direct claim on reserves. Nevertheless, many economists think the MB, which mainly measures reserves, is important for macroeconomic policy. Perhaps this misconception is partly due to the Streetlight Fallacy. This is the fallacy of focusing too much on things that are easy to measure (well illuminated by the streetlight) even though the true answer is obviously elsewhere (in darker places). Bank reserves have always been the most easily measured kind of “money” since they are directly held by the very authorities that are responsible for measuring money. But bank reserves are no more a form of money that can be used for transactions than a government bond is. Both must be converted into money before they can be used for spending on goods and services.