Saturday, 1 March 2014

Money and the Structure of Financial Balances

There's no shortage of theories of money, but I'd thought I'd add mine.  Actually, this isn't really a theory of money, it's just something I've been thinking about, to do with what money is and how it fits in with other things.  I think it makes sense.

There's nothing really ground-breaking here, but perhaps the logical order is unconventional.  Rather than seeing money as a fundamental base on which the broader structure of financial assets and liabilities is built, I go the other way.  So I start with the observation that an economy needs financial assets generally.  I then look at how these are most effectively used as an exchange medium.

The Role of Financial Assets

So the first step is to recognise the benefits of financial assets.  As a starting point, we can take it that people sometimes want to save, consuming less than they earn, which implies that they will be holding some form of asset for a period of time.  They could hold real assets, such as land or machinery.  However, people who wish to save for the purposes of deferring consumption are not necessarily interested in actually using the physical attributes of the real asset.  There is therefore a benefit in having a different person hold the real asset whilst giving the saver some kind of claim on future resources.

At this point, we don't need to worry about the fact that we haven't provided for any unit of account, because it doesn't matter if these contracts are for the delivery of specified real goods.  Even if they were, they would still constitute a form of financial asset.

The existence of such contracts as a means of saving also allows some agents to have a negative net worth.  For example, someone might own no assets, but have contracted to deliver something in the future.  This is not possible where the only assets are real assets.  The state is the main example of an entity that is likely to have a significant negative net worth.

Creating a Homogenous Exchange Medium

If these financial claims are transferable, they may themselves provide a useful medium of exchange, as they are portable and non-perishable.  However, the efficiency can be improved by some kind of pooling.

This is where a large number of claims are held by a central depository (the "bank").  Savers ("depositors") then hold deposits entitling them to a share of the pool.  Rather than transferring underlying claims between each other, depositors would transfer their deposits, with the bank maintaining a record of each depositor's balance.

The main advantage of this is that provides for an homogenous medium of exchange.  Where the underlying claims ("loans") are being used as the medium, there is the problem that they will all be different.  In particular, the "borrowers" under the loans will be different, so each time a party received a "payment", it would need to make a separate credit assessment.  With deposits, what they receive is identical each time.


Pooling also means that the deposits can have an indefinite life, even when the loans have a finite term.  This allows some further features.  In particular:

1. The deposits can be non-convertible.  This means that the only way they can be used is by transferring them to someone else (or by settling a loan - see point 2).[1] 

2. The loans no longer need to be for delivery of goods or real assets.  Instead, they can be specified to be settled only by cancellation against deposits.  This also allows loans held directly, i.e. not by the bank, to be settled by transfer of deposits.

These two steps then mean that loans and deposits do not need to be pegged in some way to the value of any particular good or real asset.  They can form their own unit of account.  Their exchange value against real goods is ultimately determined by demand and supply.   Note that this needs to match demand and supply for all assets denominated this way, not just deposits.

Removing convertibility also means that the (nominal) rate of interest can be set by the bank.  It is no longer dictated by the demand and supply of funds.

What we have now is a model of modern banking, provided that we interpret the bank as representing the entire banking sector, including the central bank. 


So far, I have said nothing about liquidity.  This is because I have only looked at a single bank, and when we have a banking monopoly with no convertibility, liquidity is not an issue.  With a monopoly, the bank is indifferent to payments between depositors, because the deposits cannot go anywhere else.  The bank can afford to allow any depositor to pay away the entire amount of its deposit.  The bank therefore does not need to apply any restrictions on payments and the entire deposit base can constitute the medium of exchange.

With competition, payments between depositors may involve the liability side of the deposit moving between banks.  Banks can then no longer afford to allow depositors to pay away their entire balance.  Some deposit amounts therefore need to be locked in for minimum periods.  This gives us sight deposits and time deposits.  The split depends on the requirements of banks and their depositors and represents a kind of market in liquidity, which extends into the willingness of banks to provide short term credit.

The above is not intended to be historical, in that I'm not suggesting this is how money actually evolved.  It's just meant to set out what I see as the logical order.  Money rests on the back of the wider network of financial claims, rather than the other way round.  Financial assets serve an important role in an economy.  Once we accept that role, it is then natural to use them (or at least a subset of them) as an exchange medium.  We need to understand the function of financial assets before we can understand how money fits in.

[1] We can allow for depositors to exchange some or all of their registered deposits for bearer notes, or back again, but the total balance of registered and bearer would remain the same.  These notes would then function as currency.


  1. Nick,

    I think the State may really not have a significant negative net worth. As per the SNA, it is just about minus £300 bn for the UK public sector at the end of 2012 as per the Blue Book. The SNA only includes nonfinancial assets which can be immediately exploited for production and the State actually has much more than that because of assets which cannot be immediately used.

    1. £300bn would certainly be significant, but I think you're right that it may not be a true reflection. To be honest, I just threw that comment in as a bit of an afterthought.

  2. I tend to emphasise the unit of account property. It is exceedingly difficult to do commercial operations without a well understood numeraire. I play a bit of a trading card game, and although you do see some simple 1:1 or 2:1 trades, people use their cell phones to check dollar prices to see if trades are "fair". (Before the prices were available on cell phones, people used price lists in magazines to assess value, which created a lot more opportunity for traders who could assess how card values would shift over time.)

    Also, debt repayment (and the payment of taxes) without a common unit of account would be largely impossible, since it makes settlement by assignment very difficult to arrange.

    If you start with a unit of account definition, you could end up with a logical progression different than yours - banks are not particularly special, and would be just a final step on top of a pre-existing shadow banking system.

    I also think the negative net worth of the State is debatable. What monetary value on a balance sheet would you assign to the Army? Since the modern State also has first claim on wage incomes and profits, those values should be capitalised on its balance sheet. But when GDP growth rates are close to the discount rate, those values (as well the discounted values of expected spending) are infinite.

    1. Brian,

      I very much didn't want to start with a unit of account, because:

      a) I wanted to talk about how there is still a benefit in having "financial" assets even in a barter economy. By "financial" assets I mean something a bit wider than normally understood, because obviously financial assets usually reference the numeraire. So I include arrangements like a loan of an enduring real asset like land, where the borrower has use of the asset in the interim in return for agreeing to deliver additional resources at the end, but it could also be something like 100 sacks of corn now for 120 sacks in 5 years, where the asset is perishable.

      b) I wanted to mention the structural changes that are required to introduce a numeraire and the implications that follow from switching to one. I only really touched on this, but I think there's a lot more that can be said here.

      Obviously, there is a separate line that says that a numeraire would be beneficial even in an economy where there were no savings. So, you can take a different route.

      Also, what I have written here probably makes it sound like banks are more special than I actually believe. I've written before about how I think that there is something special about banks in terms of the asymmetric relationship with the rest of the financial system and the implications for interest rates, but on the whole I think the specialness of banks is overrated.

      I think there's actually quite a lot of interesting stuff in this, but I like to keep my posts short. I'm thinking of doing more on this, but I'm just resolving a few points in my mind, so I appreciate the feedback.

      The point about the net worth of the state probably opens up a whole range of questions. I probably shouldn't have mentioned it as it's not important to what I'm saying.

  3. It sounds like your description is money as a "CLO" (but with the return being a liquidity yield). In fact this is the argument given by the safe assets and proponents of securitization. It seems to make sense since, at least for saving for retirement, what we really want to do is purchase a basket of goods and services for future delivery, but are limited in how we can do this.

    1. Maybe a bit, but I wouldn't take that comparison too far. In the de-pegged version, the deposits are non-convertible, so effectively irredeemable. Also, I've tried to avoid saying what I think money actually "is" and have just talked about how financial assets can be made into an effective payments medium. If my bank deposits look a bit like CLOs it's partly just because I've assumed a very simple bank.

  4. I like it and I think this is a useful backdrop for thinking money in more realistic setup.

    But is the "CLO" property at odds with MM's view?

    1. I'm sorry: MM = market monetarism.

      Just to make sure, I'm not trying to label you as a market monetarist but just to wonder if you think that this type of logic/story is at odds with their view. I mean in your story the money was only based on the pool of assets, not on the base money.

      Btw. isn't your story pretty much what is going on in the shadow banking?

    2. I don't think this fits very well with a market monetarist view. Base money is almost an afterthought in this analysis. that said, i don't dispute the power the central bank can weild through the way it uses its balance sheet.

      Maybe the first bit is a little like shadow banking, but not the whole thing. Shadow bank liabilities aren't at the stage where they're homogenous enough to be able to form an exchange medium.