After several interesting and, on other blogs at least, long discussions on the nature of money, my view on what money is and why it matters has changed and, in some aspects at least, hardened.
What I remain agnostic on is the need for commodity money - returning to the gold standard or some other commodity. The efficiency of fiat money appeals to me - there is no need to confuse money with gold (or silver), nor is there any need to store anything against the possibility that people will want to redeem their notes and coin for the commodity.
The upside of commodity money is the discipline - the need for the currency issuer to ensure that they will be able to meet any call. This imposes at least some limits on how much they should actually issue. Without it being fully backed, though, there is a long history of governments, at least, cheating on this. The history extends from the collapse of Bretton Woods in the early 1970s to the over-issue of assignats in revolutionary France and past that in many places.
The point to note here is that, as far as I can see, this sort of cheating has been largely restricted to government issuers - but this may just be because private issuers have been rare over the last few centuries.
In the rest of this post, though, I am interested only in the current position - not anyone’s ideal of where we need to get to.
Monetarism
Perhaps a good starting point for a consideration of money is the economics school that focused (or obsessed) on it. A crude monetarist position runs from the old MV=PQ1 equation - the theory being that since V and Q only changed slowly and predictably, changes in the money supply would directly feed through the general price level. Clear enough - if you know what the money supply is.
Many monetarists took M3 to be the most reliable broad measure. This is normally defined as being:
- M0: The total of all physical currency, plus accounts at the central bank that can be exchanged for physical currency.
- M1: M0 - those portions of M0 held as reserves or vault cash + the amount in demand accounts (”checking” or “current” accounts).
- M2: M1 + most savings accounts, money market accounts, and small denomination time deposits (certificates of deposit of under $100,000).
- M3: M2 + all other CDs, deposits of eurodollars and repurchase agreements. (Thanks Wikipedia)
Other publications
As “Jim” noted in comments on a previous post on this topic, most, if not all, of the textbooks on this topic are adamant that banks create money - he was right in this - but the question is whether they are right, and whether it matters.
By the traditional calculation of the money supply above, he and they are right. Any simpleton can see that, if a bank accepts a demand deposit and then makes a loan on the back of it, the money supply, as calculated above, has increased. The deposit is still money, the loan out (once drawn down) is money, ergo money has been created.
The issue
The problem I have with this is a serious one - are all bank deposits truly “money”? There are probably as many definitions of what money is as there are people - but let’s use this one for the moment:
Money is anything, which has reached such a degree of acceptability, no matter of what it is made, or why people want it, no one will refuse it in exchange for his goods if he is a willing seller. (Professor Walker - Money, Trade and Industry) (thanks to Jim again)
Now, look at bank deposits and some of the other things in M3. Would someone seriously take my bank deposit or CD in exchange for his goods? No. He or she will normally accept cleared funds of one description or another, but not my bank account - there is no real way I can transfer that, other than changing the name on the account, which would be silly. They may accept a cheque, but if it is not met at the bank then I still owe them money (and the bank some additional fees).
The M3 calculation also fails to take into account banking practice, or even the transaction processing ability of banks. If all of the bank deposits, CDs, eurodollars were suddenly to be used for transactions the system would simply collapse - not only because there is not enough physical cash to redeem them all but also because the system would not be able to process that much at once - or even over a few days.
The practice of banks of borrowing short and lending long also, to me at least, makes the bulk of the amount in deposits less than fully “money”. Any given depositor can normally withdraw all of his or her funds, but, as we saw with Northern Rock, if all the depositors appear at once for their “money” there are serious problems.
If much of the amount in the banks cannot be used for transactions, are they still money? Should we only regard the amount held in banks as liquid reserves - or even only the expected daily funds usage as money?
Why does it matter?
Application
In some ways it does matter - and in other ways not. From an individual’s point of view it does not matter - and this was my previous point of view. In practice, I can go and get “my money” from the bank and spend it - using any one of several differing methods. It does not matter if some economist somewhere thinks of it as money or not.
From an bank’s point of view it probably matters only as far as the ALM2 function cares - can we meet expected or stressed withdrawals?
From a central bank’s point of view, though, I was wrong - it may be important. Government central banks are now typically charged with maintaining the general price level (remember P above?). The clear result of the obvious disconnect between all of the monetary measures (including M3) and the general price level has resulted in the dropping of monetary targeting and the use of (a measure of) inflation targeting using interest rates to achieve this.
The problem is that the measure of inflation being used is never going to be able to encompass all (or even much) of the detail of a modern economy. Typically, they also ignore asset prices as they are difficult to incorporate. As a result, P is difficult to calculate, much less target. Consumer Price Inflation is typically used as a proxy
Hypothesis
One possible reason for much of the disconnect between the calculated money supply and the (much lower) growth in measured inflation is due to the steadily dropping reserve ratios of banks - and therefore their increased lending ratios. The lower reserve ratios have been driven by improved ALM within the banks - the assets of the banks are sweating more and the amount of needed as liquid reserves is reducing. Deposits, though, are still needed to allow the lending. This means that, using M3 as a measure, the money supply is bounding ahead - but inflation is comparatively static.
If this hypothesis is correct, a better way to calculate the money supply would be to include only physical currency not currently in the hands of the banking system and the amount held at banks and available for either deposits or withdrawals - these being the only amounts that are truly liquid within the system and therefore the only components of M that are actually able to be used for transactions.
Is this useful? Probably not. But I have enjoyed thinking about it.
1. Money supply times Velocity equals the general Price level times the Quantity of transactions. I know the arguments were much more detailed than this, but this is a blog post, not a textbook.
2. ALM - Asset and Liability Management - the function of the bank that is primarily charged with ensuring the bank can continue to pay out what it owes and making sure the assets of the bank make enough profit to keep the res of the show going.
5 comments
Comments feed for this article
6 November, 2007 at 1:15 pm
André Levy
I agree with you, Andrew: bank deposits are not money. But the claim on deposits is. And, yes, you can transfer that claim at the press of a button. Your point on the availability of deposits is well taken, and that is reflected on their value. So, no, bank deposits are not money at face value, but the piece of paper saying that that money sitting their in the vault is yours, is.
Obvisouly, as you point out, the money is not sitting idle in the vault. It has been lent out to some better use, while you don’t need it. And that how banks create money, through fractional reserve banking. This is captured in the definition of ‘broad money’. Here is how the RBA defines it: http://www.rba.gov.au/Glossary/detail.asp?offset=40&term=broad%20money
But not only banks and governments create money. Anyone collecting rights against its obligations, i.e assets against liabilities, in essence, creates money. Companies, when issueing stock to fund their investments, for instance, create money. And who’s to say stocks don’t work as money? People get paid in stock! And stock options!! BHP share certificates are no less money than the RBA’s money bills. As long as they can both be exchanged for something else, they work as reserve of value and means of exchange. The difference, of course, as you rightly point out, may be in liquidity. But information technology has made the transfer and trade of practically any asset just as seamless as that of money.
Is this useful? I think so. When leverage is higher than growth, i.e. when we promise more than we can deliver, those promises, those IOUs we call money, become cheap. And it’s when we all start failing to deliver on our promises that things begin rolling downhill. Further, whoever laughed last, laughs least! Cheap money fuel dear assets, i.e. asset inflation. Those relying on salary, those company IOUs, to make ends meet will see everything around them going up - property, stocks - and will rush to save and jump on the get-rich-fast bandwagon. Who can blame them? If they don’t, they are left behind, everything went up but theirs, and the alternative to catching-up-to-the-joneses turned out to be the get-poor-even-faster. The trouble is that by the time they manage to get in, those that were already there from the beginning move out to make room, selling them their seat for something a little less fleeting (gold, anyone?).
11 January, 2008 at 12:58 am
Joanne Nova
Andrew,
Isn’t it true that since I can use my EFTPOS card to pay for bread at Woolies, then numbers in my cash account are effectively the same as cash in my hand? Albeit, I can only use these numbers to buy from people with EFTPOS machines. But for most purchases, especially large ones, this is true.
A term deposit is different. I can’t buy bread directly with it. But I can list term deposits as assets when asking for a loan, and mentally, when I’m walking past the eight pair of jeans I don’t need, somewhere in the back of my head I’m including that term deposit in my general sense of wealth. At Harvey Norman, I can ’spend’ that term deposit now, knowing that I have 2 years to front up with the cash.
My credit card even allows me to spend money I don’t have, that presumably gets created the minute I press ‘OK’. Therefore, my credit limit is arguably a part of the money supply.
Money that’s locked away still matters. If people have ‘money in the bank’ or a large credit limit at their disposal, it influences their consumer behavior. This ‘money’ means people spend more, with an inevitable effect on inflation.
And as far as the disconnect between Money Supply and Inflation goes, there is another explanation. Not only does CPI ignore assets - a huge soak for money, but hedonic effects, geometric calculations, and substitution (the fixed basket that is not fixed) have changed the way the CPI is calculated. They all reduce the CPI result.
John Williams at Shadow Statistics has done the detailed-number-crunching job of recalculating today’s US CPI rate according to methods used in the pre-Clinton era. If we ignore all the CPI method changes since 1992, current CPI is over 7%. http://www.shadowstats.com/
This article sums up why the CPI figures don’t ring true to the money supply figures any more.
http://www.weedenco.com/welling/Downloads/2006/0804welling022106.pdf
Joanne
11 January, 2008 at 2:12 pm
Andrew
Joanne,
A big difference between cash in hand and bank deposits is the question about who actually owns the money? At law, the answer is simple - the bank does; and the term they are deposited for does not matter. All the bank has is a contractual obligation to deliver cleared funds when you ask them to do so. They can refuse to do so, provided the refusal is in compliance with their contract with you. Cash in your hand is yours.
From a monetary point of view, I am still proudly uncertain as to how to measure the money supply - or even if it can be measured.
If, as you put it, you have funds on 3 month term, but you spend it today by using your credit card in the knowledge that you are able to pay it back once the term deposit matures, you have effectively transformed a term deposit into a call deposit. Does this ability to transform the “term” deposit in to a “call” deposit render it “money”? Or is the credit limit money as you are able to spend it? If the term deposit can be transformed in this way, does the ability to re-draw on the equity in your home render that into money - or does that occur when you get a new mortgage or only when you spend that mortgage?
If the bank does not have the ability to deliver on all the credit limits as once (as few, if any, do) just as they do not have the ability to deliver on all deposits at once affect this?
The point I am trying to make here is that trying to work out what exactly the money supply is may be impossible. A definition of money may be an easy thing to give, but actually applying it to come to a figure for the money supply may be impossible in a modern, diversified financial system.
I would agree that the crucial difference is the wealth effect - how it actually changes the spending habits. You captured this here:
I have also mentioned it elsewhere on this blog.
This, to me, is the crucial question - how does all this affect the behaviour of the actors in our economic framework? To me, getting an exact number for the money supply1 is far less important than working this out.
Perhaps I should have titled this post “What is the Money Supply and Why does it matter?”, to which my answer, at the moment, would be “We have a vague idea” and “Only in how it affects behaviour”.
.
1. Even if possible - and I would propose it is not as by the time you calculate it, it will have changed not only in quantum, but in definition.
5 July, 2008 at 1:13 am
Joanne Nova
Sorry about the delay. Thanks for your considered response.
I agree with you that a definition of money is intrinsically impossible, and that there is no such thing as an accurate, exact ‘Money supply’ figure. But the point of money is to be a store of value, and the point of inflation is a measure of how well that store of value works over time. So as you say, the crucial question, is about how money supply affects behaviour (which inconveniently is also impossible to define or measure), and whether it causes inflation.
So we’ll never get an answer to the question (exactly how big is our money supply), but given that money is our universal measure of equity, reward and justice, isn’t that enough to just declare, that even an imperfect estimation of money supply tells us something useful?
My problem then, is not whether all bank deposits fit any arbitrarily chosen definition of ‘money’, but whether people behave as though those deposits have monetary value, and spend accordingly. If they do, I find it hard to see how increasing money supply doesn’t automatically (with some lag) cause inflation. Indeed I find the Austrian’s arguments that increasing money supply IS itself inflation quite convincing.
And BTW M3 here does not include repo’s or most ‘eurodollars’ (foreign held $A), which IMO makes it more like the US M2. Not that that’s critical, but the RBA does do things it’s own way. We might as well know what we are comparing (and it’s unfortunately not easy to find out that info).
You mention the velocity of money. Do you feel that explains the disconnect between M3 and the CPI?
7 July, 2008 at 12:36 am
Andrew
Joanne,
I would start off a response with a couple of observations - firstly that it is only certain of the Austrians that argue that an increase in the money supply is, in and of itself, inflation - notably Rothbard. I would disagree, based in part on my second observation: if we do not know what the money supply is, how can we say it has increased or decreased?
If we look then at your fourth paragraph, particularly this sentence “If they do, I find it hard to see how increasing money supply doesn’t automatically (with some lag) cause inflation we need to substitute “money supply” for “bank call deposits” (as we do not know what the money supply is) we need to look at what would happen if people did not have call deposits - as this is where the “money supply” is coming from. Personally, if I were barred from having call deposits I would deposit my funds on term, but arrange a credit card with back to back funding from my term deposit - or I would just keep the credit card in credit the whole time to avoid the situation where I have a lot int he bank and no way to get to it.
The end result, if you look at the cash flows, is just the same.
To my way of thinking this means that bank deposits, and the propensity to lend them, cannot be the source of inflation - for the simple reason that, left to themselves, this behaviour would be just the same with or without call deposits. Even the radical step of banning them would have no real effect beyond everyone needing more bank accounts.
I will give it some more thought because, as I have said, there is a lot here I do not yet pretend to understand.
On your last question - FWIW I believe the velocity to be unstable, but this is not enough to explain the difference. It is because of this evident disconnect that I first started thinking about the problems of calculating the money supply (brought up, as I was, in the monetarist school) and brought me to the position I now find myself.