The Bank of England (BoE) publishes both M0 and M4 money supply figures (more accurately these could be called money quantity figures). Broadly M0 is central bank money or 'narrow' money and M4 is that plus all of the money created through mechanisms such as fractional reserve banking, whereby money deposited in a bank is lent back out, spent and deposited in a second bank account thus becoming double counted. In fact there is no defined limit to the number of times money can be lent back out, but the reserve requirement on banks results in an upper limit to the overall multiple that can occur of 1/reserve_proportion. So a reserve requirement of 10% results in a maximum money multiplier in the system as a whole of 1/0.1 = 10. Thus 1GBP of central bank money could become as much as 10GBP. Actual reserve requirements can be a tad murky, e.g. the UK currently has no hard defined requirement ( Reserve requirement), with the intention being that other regulations serve to throttle risky behaviour without having to define an arbitrary limit.
One of the central ideas behind a fractional reserve system is that it allows the money supply to expand naturally with the economy. Money is created to fund new enterprise as and when there is appetite for that enterprise and the risk-reward ratio of investment opportunities look good. Think of the economy as an engine requiring oil to keep it lubricated, as the economy grows in size we need more oil to keep it moving. Also note that this growth in money through re-lending (money multiplier) is independent from the form of money in use, e.g. a strict gold standard would prevent the BoE from issuing new central bank money ('printing' money) but would still be subject to the money multiplier, whereas a fiat money system where fractional reserve banking was banned would not.
The banks therefore are a critical part of the system, not just because they perform the mechanics of re-lending and thus creating new money, but crucially they perform the risk-benefit analysis that determines what loans they issue. Good loans result in an increase in production capacity of the nation (e.g. by paying for a new widget making factory) and thus grow the economy. Bad loans divert labour away from productive enterprise. Further, a failed factory cannot repay the loan, therefore a loan default occurs which has the effect of destroying the new money that was created - the new money is destroyed in tandem with the new factory failing.
Note that the new money from the good loan is also destroyed as the loan is paid back, however, the interest payment boosts the bank's coffers over time and ultimately allow it to make further loans with a larger overall amount. The money to cover the interest payments comes from the profits of the widget factory, which in turn are obtained from the economy as a whole - the widget purchasers and users who in turn have their own enterprises and thus money creating loans. In effect it's a big game of musical chairs in which the number of people in the game increases continuously and we hope the music never stops. This is similar to a Ponzi scheme with the distinction that there is genuine investment in productive capacity occurring.
So money lending by commercial banks increases M4 and, if M0 is constant, we can easily track overall lending amounts by observing changes in M4 - otherwise we can fairly trivially compensate for changes in M0. Currently it looks like M4 is shrinking (see Provisional estimates of broad money (M4) and credit (M4 lending): August 2010), that is, net loan amounts are negative. As such the economy is probably shrinking right now, although it is still up for the last 12 months as a whole.
Quantitative easing is the increase of M0. In a nutshell the BoE creates central bank money out of nothing and buys stuff with it - in the UK this is mostly government bonds and some commercial bonds. The government issues a bond, the BoE buys it with the new money and the government now gets to spend the money. Commercial bonds work the same way, it's just an alternative form of obtaining a loan.
There are several motivations for QE:
(1) It creates an artificial demand for government bonds and thus provides a cheap supply of money for the government, it also suppresses government interest payments for bonds not bought by the BoE. Government spending can then try to fill in the gap caused by the slowing of the economy. This is often considered to be flawed as public spending nearly always has poor productivity in comparison to the private sector, plus it tends to compete with the private sector thus crowding out much needed healthy private enterprise that is ultimately required to pull the economy out of recession.
(2) Low bond yields cause money to flow away from bonds into other asset classes, boosting prices elsewhere. The hope here is that bank assets increase in price, shoring up bank reserves and promoting bank lending. The problem here is that bank reserves aren't the main barrier to new loans being issued - the banks don't want to lend and people/companies don't want to borrow as much in uncertain times. There's also a question mark over the permanence of asset prices that have been boosted by QE, or rather, we all know it's a temporary effect and therefore tend to compensate for the price changes when performing capital reserve calculations.
(3) Inflation targets. The absence of new loans is naturally shrinking M4. Less money chasing the same amount of stuff means falling prices (deflation), which doesn't sound too bad on the face of it. The argument is that falling prices cause people to hold off on purchases because they can save now and buy stuff for less in the future. IMO this is a slight mis-framing of the situation, people will tend to save when the true savings rate is positive and will spend when the true savings rate is negative. By true savings rate I mean the savings rate you get on your money (the nominal rate) minus the inflation rate. If prices are rising (inflation) at the same rate as your savings then you're walking the wrong way up the escalator and going nowhere. Framing the savings issue in terms of inflation alone is thus misleading to some extent. The BoE reduced the base rate to 0.5% to discourage saving and encourage spending, to further encourage spending they use QE to increase inflation - buy now before the price goes up.
The problem with using QE to boost inflation is that you're swimming against the tide. M0 may be expanding but M4 is shrinking, and because of the credit splurge of the last 10 years M4 can probably continue shrinking for much longer than you (the BoE) feel comfortable increasing M0.
One of the biggest causes of M4 expansion in recent years has been new mortgages, in effect new money in the economy was being largely directed at one asset class rather than being spread evenly. This new money creation span out of control as it forced up prices which in turn increased the amounts of new mortgages as prices rose, which resulted in greater amounts of money creation, and so on. The banks failed at their key role of analysing the risk-reward of each loan/mortgage - they couldn't see that in order for *all* the mortgages to be paid back with interest there would need to be a source of new money and therefore new productive enterprise. IMHO M4 shrinkage is likely to dwarf the BoE's appetite for counteracting it with QE, in fact they've been uncertain about QE from day one and appear to be extremely hesitant at each new round.
Understanding bank reserves is key and is perhaps best explained with an example. Say there is a new bank and the legal minimum reserve requirement is 10%. You deposit 1 million GBP in a new account, the bank loans out 900K and keeps 100K as reserve (the absolute minimum reserve requirement). At this point if you ask for even a single pound of your money back the bank can't do it - it would break the reserve limit. You could however transfer money between your account and another account at the same bank, e.g. paying a local Builder who happened to use the same bank, because moving money within the bank doesn't draw on its reserves.
Ok so basically the bank can't operate with just the minimum reserve so they operate a 'buffer', that is, they hold reserves in excess of the minimum requirement, let's say 20%. So now they've lent out 800k of your 1 million and kept 200K. If you now write a cheque for 100K to an account at another bank the cheque can be honoured and the money paid, anything over 100K and your bank once again would have less than the minimum requirement. The main point here being that the total amount of money deposited by a bank's customers is not available for withdrawal at any one time, in that respect there's some degree of smoke and mirrors going on. The system works because generally not everyone wants to withdraw their money at the same time, the excess bank reserves cover the fluctuations in the reserve level as customers deposit and withdrdaw funds. With a large number of customers the fluctuations tend to balance out nicely - on a given day the bank's customers make and receives payments that will tend to mostly balance out. Occasionally something may spook the bank's customers and cause more than usual to withdraw money, causing a bank run - a rush to access the bank's reserves before they run out. If the bank had simply held the deposited money instead of loaning some of it out then there would (A) be no risk of losing money from loan defaults and (B) No bank runs (all of the deposited money is always available for withdrawal). The downside to this would be no interest payment on your money, in fact the bank would have no source of income and would therefore have to make a small charge for holding your money to cover running costs. Your money would also not be keeping up with inflation (currency devaluation), although this works in reverse in deflationary environments.
Colin, September 2010