Questioning Policy Effectiveness: Quantitative Easing

Quantitative Easing is Pushing on a String

They’re at it again. Last October the US Federal Open Market Committee announced it “is prepared to provide additional accommodation if needed”. This means ‘quantitative easing’, or QE. Indeed, under QE2, the US Federal Reserve committed to buying $600 billion of Treasuries between November 2010 and the end of this month. So far, the QE2 program positively disappointed and the markets have been buzzing with speculations about ‘QE3’. In these circumstances, it is news that in Tuesday’s speech on the U.S. economic outlook, Mr Bernanke did not mention QE3. But many expect he may still do.
The QE idea is simple. With deflationary fears building and the federal funds rate between zero and a quarter of a per cent, not much more can be done with the price of money. So why not try influencing the quantity directly? The Federal Reserve would buy financial assets off banks, and in paying the banks is thought to put money into the economy, boosting spending and (hopefully) avoiding deflation. During the latest QE experiment, in Britain from March 2009 to February 2010, Charlie Bean, Deputy Governor of the Bank of England, explained on the Bank of England web site that “quantitative easing aims to increase money spending”.
It will not work. Despite its popularity, QE is based on a misunderstanding. Banks are thought to multiply their central bank reserves into loans to the economy. In this causal scheme, expanding banks’ deposit with the central bank (their reserves) will lead to growth of the public’s deposits with banks. This ’fractional reserve banking’ theory seems to be supported by the fact that in normal times, bank loans and reserves tend to move together, so that the ratio is indeed fairly stable.
The problem is that the loan-reserves ratio is a number that can always be constructed ex post, but not therefore a causal determinant of bank lending. Increases in reserves are the result of increases in bank lending, not the other way round. Any bank short of reserves because it increased its bank lending will seek to buy reserves from another bank. If the banks in aggregate are short, the central bank will accommodate the demand. To start with increasing reserves is to reverse the sequence, which is not possible. QE is akin to pushing on a string.
QE is a metaphor for the central bank increasing the banks’ account balances with the Fed. However, what needs to happen in order to put money into the economy is that banks increase the public’s bank account balances - banks need to increase their lending, and the public must be willing to borrow more. Academic research has long recognized that money growth happens endogenously within the economy, between banks and the public - not exogenously, by policy interventions from outside. There is nothing to stop the central bank from multiplying reserves, but this by itself will not multiply lending.
And this is precisely what we observe.. In recent research*, we found that British loans and reserves moved together in the years prior to the 2009-2010 QE experiment.. But when British reserve balances quadrupled from £39.4 bn in February 2009 to £156.4 bn in February 2010, the total amount of outstanding loans remained flat (at £2,110 bn at the start of QE in March 2009 and £2,233 bn at its end). QE failed to increase lending in Britain in 2009-2010 just as it had failed in Japan in 2001-2006. Although this was clear from the figures already last March, the Bank of England admitted the failure only implicitly last week, when Mr Bean went on television to implore savers to spend more. This should not have been necessary, had QE worked and increased money spending as Mr Bean expected in 2009.
Instead, what we see is ‘Goodhart’s Law’ in action - the principle proposed by young Charles Goodhart in 1975 that ‘any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes’. QE proponents confuse correlation with causality. Loans and reserves tend to develop proportionally—until policymakers mistake this statistical regularity for causality, and purposely boost reserves. Then it is revealed that reserves do not and never did cause loans in a systematic fashion.
What are the lessons? Banks do not lend Central Bank money to the public, but their own money, by creating a liability on themselves. Money is not a thing to be pumped into the system from outside, in order to unclog the credit pipelines. Money consists of accounting relationships, and they come in different sorts that are not simply interchangeable. Reserve balances, for instance, are not money balances. Unless we want to have a realistic look at the accounting relationships that make up our financial system, we will remain clueless about effective monetary policies.

Dirk Bezemer is a fellow at the economics and business department of the University of Groningen in the Netherlands. Geoffrey Gardiner is a Former Manager in the Financial Services Division of the Barclays Group.

* ‘Innocent Frauds Meet Goodhart’s Law in Monetary Policy’, Levy Institute Working Paper.