Thursday, October 1, 2009

No one understands quantitative easing.


Or at least the standard explanation as to how quantitative easing works is flawed.


Summary. The standard explanation as to why quantitative easing (QE) works ignores the crucial difference between QE given constant national debt, and QE given rising national debt. The standard explanation seems to assume the former (not relevant in 2009).

Given constant national debt, obviously QE puts liquidity into the hands of banks, pension funds etc., plus it reduces interest rates. In contrast, there is QE accompanied by Keynsian “borrow and spend” which is what has actually taken place in 2009. In this circumstance, trying to work out the effect of QE alone is pointless because all QE does is to reverse two of the elements of borrow and spend. These two elements are, 1, “cash moves from private sector to government”, 2 “government issues gilts to the suppliers of the cash”. The NET effect of borrow and spend plus QE is simply “government prints money and spends it”.

The latter certainly should be stimulatory for the economy as a whole, but the stimulation will NOT come from increased lending by banks, which is supposed to be the effect of QE. The simulation comes from increased spending by government and consumers.

_____________________


Terminology
. First, a few words about the terminology used below are required. When referring to government debt, I’ll use UK parlance, i.e. “gilts”. The equivalent in the US is “treasuries”. Also, QE can consist of buying private sector bonds as well as public sector bonds (i.e. gilts). In practice, in 2009 the Bank of England’s QE has consisted almost exclusively of buying gilts. So to keep things simple, I’ll refer just to gilts and won’t mention private sector bonds from now on.

The standard explanation of QE runs approximately as follows. Central bank prints money and uses it to buy gilts from those in the habit of holding same (banks, pension funds, wealthy individuals, etc). This raises demand for bonds, which reduces interest rates, which should boost aggregate demand. Also those in receipt of this cash will tend to buy other assets, e.g. houses or shares which boosts the housing market and stock exchange. Net result: a boost for the economy.

The problem with this explanation is that it is only a partial view of what has happened in 2009 (certainly as far as the UK is concerned). The whole picture is thus (in bold italics).

In 2009 some countries, including the UK, have substantially increased national debts at the same time engaging in QE.

This is a different kettle of fish to QE alone. With a view to gleaning the net effect of this exercise, let’s look at the various movements of money, government bonds and so on with a view to working out the crucial factor: the NET effect of the above (“increase national debt plus QE”).

Take the “increase national debt” first. Governments go into debt to acquire funds to spend on health, education and the usual public sector items. Or they go into debt because they want to maintain spending despite a fall in tax revenue (e.g. because of a recession).

Increasing national debt consists of the following three movements of cash and gilts. 1. government / central bank machine (gcbm) sends gilts to those willing to purchase same. 2, the latter send cash in return to pay for the gilts. 3, gcbm then spends the cash on education, health etc, thus the money ends up back in the private sector.

As for QE, this simply consists of reversing items 1, and 2 above. Hey presto, the NET effect of “increased national debt plus QE” is just item 3: gcbm printing money and spending it on education , health, etc. At least this is the case to the extent that the total amount of QE is is the same as the rise in the national debt, which in the case of the UK it is approximately (see second graph here).

To repeat, gcbm printing money and spending it on health, education etc is the only net effect. So why do we hear so much technical “City of London” stuff about QE making markets more liquid, reducing interest rates, when the only net effect is “print money and spend it”?

Moreover, QE has had less effect than anticipated: banks who have been in receipt of the proceeds of QE have not spectacularly raised the amounts they are willing to lend. Hardly surprising ! This NET effect does not have an obvious effect on bank behaviour.

The net effect of Keynsian “borrow and spend” i.e. the effect of a rising national debt is to place additional gilts into the hands of those in the normal habit of holding gilts (pension funds, banks etc). In other words the effect of “borrow and spend” is amongst other things to induce banks etc to lend to government. Having done that, banks will not then be falling over themselves to lend even more!

Thus the main effect of QE is arguably just to reverse the crowding out effect. It doesn’t give banks a big incentive to increase lending to businesses.

If the above argument is correct (a very moot point), then it is hardly surprising that the effect of QE has been not as anticipated. Again, on the assumption that the above argument is correct, QE plus "Borrow and spend" WILL have reduced the severity of the recession, but it will have done it via a general stimulation of the economy, including putting more cash into consumers' pockets. It will NOT have done it, and clearly has NOT done it via more bank lending.

No comments:

Post a Comment