Recent Posts

Tuesday, April 29, 2014

Are Banks Special? Yes And No.

One of the underlying questions of macroeconomics is whether banks are distinct from the rest of the financial sector. My answer is "yes" and "no". I am not saying that just because I enjoy impersonating a two-handed economist, rather it is because the answer depends on the objective of your analysis. For example, if you are looking at the operations of government finance, the government and banks form a system with circular cash flows. This has the not completely obvious implication that governments with "currency sovereignty"  are largely immune to default for financial reasons. But if the objective is to look at the behaviour of the overall economy under normal conditions, the banking system is not really special when compared to the "shadow banking" sector - or even "non-financial" businesses.

The role of banks and money has been grounds for debate for a long time. Recently, there has been a flare-up of interest in making banks hold 100% reserves against deposits, triggered by Martin Wolf's endorsement of the idea in the Financial Times. This article on Mike Norman Economics links to a couple articles, including one by Warren Mosler (which I largely agree with). Additionally, Nick Edmonds discusses some of the other articles on the topic.

I do not want to attempt to cover the debate about holding 100% reserves right now, rather I want to cover a more fundamental question about the importance of the formal banking system relative to the "shadow banking" system. This article also acts as a longer response to some of the criticism on my article about C&I loans on Seeking Alpha. However, if banks are not particularly special during normal economic times, reforms that are focussed on just the formal banking system are unlikely to be effective.

Banks Are Special: Arguments In Favour

It is clear that banks are privileged when compared to other businesses. This helps make them the target of widespread disapproval from across the political spectrum, as privilege is not a status that fits within the democratic ethos. (Privilege comes from Latin roots, and literally means "private law".) On the Right, Austrians attack banks as they argue that "fractional reserve lending" is fraud. On the Left, many progressives dislike the conservatism of bankers, as well as the relationships between banks and their regulators. The fact that banks helped blow up the global economy a few years ago did not help matters.

Many schools of thought within economics impute considerable power to "money"; this tendency hit its peak with Monetarism. This focus on money means that if banks are special, then bank deposits (which are included in almost any definition of money other than the monetary base) are of particular importance. Therefore banks remain at the centre of economic controversy.

I will not attempt to summarise those controversies here. I will instead look at the main factors which make banks special:
  1. A clearing system, where bank liabilities clear at par.
  2. Privileged legal status for settlement of debts. For example, you can pay your taxes with a bank cheque.
  3. Access to the central bank for lender-of-last-resort operations.
  4. The combination of the banking system and government finance operations creates closed circular flows.
The statement that "bank liabilities clear at par" is something that is not often thought now, but it is important. Before the advent of the Federal Reserve system, U.S. bank cheques ("checks" in American) would not clear at par. Banks would not cash cheques drawn on weaker or unfamiliar banks at their stated value, rather they would apply a discount in case there was a default during settlement. Although Austrians look fondly upon the pre-Fed era, such a situation is extremely complex and dangerous for the economy. Banking panics were all too common. (Other developed countries tended to have large centralised banking systems, with central banks, and thus these problems did not appear.) 

But when we look at points 1-3, they do not have too much of an effect on the economy under normal circumstances. A properly functioning clearing and legal system is a necessary component of a smoothly running economy, but this just allows other non-bank entities to carry on with their normal business. Access to lender-of-last-resort operations is important during a crisis, but it does not tell us much about an economy otherwise.

From the point of view of the analysis of the economy under normal circumstances, the circular flow property of the banking system is the most important. This is what allows the banking system to create debt/money "out of thin air", allowing nominal GDP to expand.

For the purposes of government finance, this is exceedingly important. For example, the circular flows between the government and banking system provide the basis for my view that anyone expecting a default in Japan is severely misguided. For a further discussion, see my article on "government bonds as a reserve drain".

But as a final point which is more sympathetic to bankers - their privileges are counter-balanced by the reality that their industry is highly regulated. Given centuries of bad experiences, societies know how to regulate banking systems. If there are problems, it is mainly a result of a poorly thought reforms. A migration towards shadow banking means that regulators will have a lesser grasp of what is happening within the financial system.

The real problems in the financial crisis within North America mainly originated within the shadow banking system (although the capital markets arms of the banks were major actors within the shadow banking system). Europe had more problems within their formal banking system, but this was only a reflection of the total collapse of the regulatory framework (national banking regulators facing international banks) coupled with a fixed currency peg system in the absence of fiscal cooperation. As long as countries avoid entering into botched fixed exchange rate systems, banking system problems can be contained by sensible regulation.

Banks Are Special: Arguments Against

Anyone can create money. The "trick" is to get it accepted.
Hyman Minsky 

Although I agree that banks are special for some forms of analysis, I generally do not view them as being particularly special from the point of view of analysing the economy during a normal growth phase. As the Hyman Minsky quote above states, money creation is not a monopoly of the banks and the government. During tranquil times, pretty well all businesses act as banks.

I will split up the discussion for the shadow banking and the "non-financial" parts of the business sector.

Shadow Banking

The shadow banking system is all the financing activities of actors within the financial system that is not done as conventional bank lending. Complicating matters is that regulators have allowed corporations that are banks to operate within the shadow banking system. This was probably a policy error. But for my discussion here, I am treating those activities as being outside the formal banking system.

Although shadow banking system entities cannot create "bank money", they are in a situation very similar to individual banks. For example, if customers re-allocate their "cash" from bank deposits to money market funds, the banking system will have no choice but to fund their assets in some other way than deposits. 

As a simple example, assume that a customer of Bank A moves a $1000 deposit to Money Market Fund B, which we assume banks at Bank C. Bank A will now be short $1000 worth of funding, while Money Market Fund B has $1000 to "put to work" in the money markets. Although there are many ways of resolving the situation, one possibility is that Money Market Fund purchases commercial paper with a market value of $1000 from Bank A. The end result is that the initial deposit has disappeared (since the seller of the commercial paper - Bank A - cannot hold deposits at bank A), and is now replaced with commercial paper.

More generally, large corporations can fund their working capital needs in the commercial paper market. They also simultaneously hold money market instruments, possibly via a money market fund intermediary, for their liquidity buffer. These corporations can expand their balance sheets via issuing commercial paper, and then they will have the capacity to absorb the paper. Even though bank credit may need to be created temporarily as part of the clearing process, at final settlement this balance sheet expansion can occur without any permanent expansion of bank balance sheets.

Minsky argued that in an expansion, risk appetites rise. The formal banking system is highly regulated, and can only loosen lending standards so far. This means that the shadow banking system will become increasingly prominent, and assets will be re-allocated towards it from the formal banking system. As a result, you need to incorporate the shadow banking system if you want a handle on aggregate credit growth.

The end result is that banks end up looking like just another component of the wider financial system, and their privileges do not cause them to stand out in terms of their ability to fund an expansion of private sector balance sheets.

"Non-Financial" Corporations

There are very few large truly non-financial corporations. This point is not obvious to modern consumers. Only a few retail (non-financial) small businesses extend credit directly to their clients any more. Although consumers use credit for a good portion of their transactions, they either use credit cards, or else small businesses get financial services companies to act as intermediaries for them. Only a few of the larger businesses have in-house lending arms - auto manufacturers being a well known case. Historically, this was not the case. In tight knit communities, small businesses like general stores extended credit to their customers, and businesses lived or died on their ability to judge the creditworthiness of their customers.

This is not true within "business-to-business" transactions. I am unaware of any industry that uses "cash on the barrel head" settlement; instead transactions are done via accounts payable or receivable. And as was seen during the technology bubble, "vendor finance" is even used to provide long-term financing.

Payment terms are one of the parameters adjusted by sales forces in order to clinch deals. And those accounts receivable can be re-discounted, which means they are effectively money market instruments. For these sorts of reasons, Minsky often analysed all businesses as if they were banks.

This means that the expansion and contraction of credit just within the "non-financial" sector is also an important source of cyclical variation. And this expansion is done entirely outside the banking system, at least until the clearing process is reached.

Banks Versus The Banking System

There is also a distinction that needs to be made between the position of an individual bank versus the banking system in aggregate.

If there was only one bank in the economy - for example, a combination "central bank"/postal bank - it could largely act without any constraints. (Since the central bank has "asymmetric redeemability" of its liabilities, central banks are somewhat free of constraints. This was also a topic of recent discussion; see here.) It can expand its balance sheet without any worries about funding its assets - deposits cannot flee it, only be rearranged. If a shadow banking system exists, the bank would not have total control over lending, but it could at least set the lower bound for the size of financial balance sheets.

However, even the Soviet Union was not that sovietised. Such an organisation would not work in practice, as the bureaucrats running it would be essentially unaccountable for their errors. ("A rolling loan gathers no loss.") Instead, we need a decentralised system where multiple banks allocate credit and inevitably make mistakes - but at least the errors are smaller and self-correcting in the long term (the banks are forced to eat their losses at some point).

In a multi-bank system, there is a de facto limit on how fast an individual bank can allocate credit. If a bank expands lending faster than its competitors, it will probably lose deposits are the proceeds of loans are spent. (A bank that is gaining market share in deposits can have a higher growth rate on loans, but this means that other banks have grow lending slower than average.) In the short term, the bank can borrow in the interbank market. But banks can only do this for so long (as some discovered the hard way during the financial crisis) - eventually, expanding bank assets have to be supported by either locking up long-term funding or equity issuance.

Therefore, from a micro level, bank officers cannot expand lending arbitrarily. And if the individual banks are constrained, one can argue that the whole system is constrained. Although he used an unfortunately incorrect phrasing, I believe that this logic was part of what Paul Krugman meant to write during his internet debate about exogenous money with Steve Keen. Although this point appears correct, one has to keep in mind that the constraint is lending relative to the average of all banks. During an expansion, all banks will tend to follow the herd, and so the average growth rate can accelerate as most individual banks decide to become "above average".

Concluding Remarks

Whether banks are "special" is a fundamental theoretical question, and what side economists are on does not seem to be driven by politics or the mainstream/heterodox divide. The fact that the answer is somewhat ambiguous allows this division to remain.

I will discuss the implications in later articles. But it is clear that reforms need to take into account the fact that the formal banking system is not particularly special in terms of its role during an expansion. Aggressive regulation of the banking system, such as a 100% reserve requirement, would just drive activity into the shadow banking activities undertaken by both financial and "nonfinancial" firms. Such an outcome would probably make the inevitable future financial crises even worse.

See Also:

The book the Hyman Minsky quote is from: Stabilizing an Unstable Economy - Hyman Minsky (affiliate link).

The paper: The Hierarchy of Money, by Stephanie Bell. Not purely on this topic, but covers a lot of fundamentals regarding money.

(update) My post "No, Banks Do Not Lend Reserves", which gives an example how the average growth rate of lending in the system restricts bank lending.

(c) Brian Romanchuk 2014


  1. Great Post and summary Brian.

    If we were to implement a Positive Money type 100% reserve backing of loans and deposits, where would you put the number of currently outstanding bank loans in need of matching reserves?

    In other words, how many trillions of dollars of bank loans are outstanding? Not total credit market instruments like corporate bonds and money markets etc, just bank loans in need of 100% reserves?

    1. Everything would depend on the implementation. From what I have seen of the proposals, banks would not want to have such deposits, unless they charge a fee. They would effectively be Treasury Bill funds, and would be managed as such - they would charge for services, and skim off a percentage as a management fee.

      I think the banks would skirt the spirit of the law and create accounts that do not require reserves, but add features that let them emulate demand accounts. For example, if there is a 30-day lockup period on the non-reserved account, just tie the account to a bank line of credit. They would make the accounts that require reserves extremely unattractive in contrast.

      As a result, I think most of the outstanding bank deposit balances would migrate to accounts that do not require 100% reserves, or else they would just pack up and go into the shadow banking system. Since this is a very hypothetical exercise, I do not know what percentage of deposits that leave: probably 60-90%? (I would have to look up the data i the Flow of Funds to get some numbers on that; unfortunately I am stuck with some other things right now.)

      I assume that the reformers believe that the non-reserved accounts would not get deposit guarantees. Based on the behaviour during the last crisis, that status would only last about 5 days into another financial crisis. Everyone knows the only way to stabilise the financial system is to extend deposit guarantees, and policy makers would do so. Doing otherwise would just be liquidationism.

  2. A favourite topic of mine, this and you've got some great points here.

    "...the formal banking system is not particularly special in terms of its role during an expansion."

    I think this is the really important one, that "banks are special" advocates do not pay enough attention to.

    Tying down where banks are special is quite a tricky one. An interesting comparison is this: "How small can non-bank lending be compared to the size of the banking sector" and "How small can the banking sector be compared to the size of non-bank lending". I think there are particular problems that arise as more and more credit is provided outside of traditional banking, that don't apply in reverse.

    And I think this is primarily to do with liquidity provision. Banks have a network of credit lines, whereby they effectively provide liquidity insurance to each other and to non-banks. Other FIs (and shadow banking entities in particular) are not in a position to do this and rely for liquidity on the marketability of their assets. When the sector gets big this can be much more fragile.

    That is not to say that bank liquidity is immune, of course. Part of the crisis involved big cuts in interbank credit lines, at the same time as the sector was having to provide massive liquidity support to shadow banking entities. This is where the central bank comes in, so the particular functional relationship that banks have with the central bank also makes them special.

    1. Thanks.

      For reasons of space, I did not want to go too much deeper into the relationship between the shadow banking and formal banking sector. Although I hinted at it, the shadow banking sector needs the banking sector as a backstop, as you point out.

      I think it was in "Lombard Street", but I have seen arguments that the international bond market developed in London and not Paris because the British had a more centralised banking system. A bond market can be seen as an extension of the banking system; a good portion of the activity is leveraged operators who use the banking system for funding. Those leveraged operators provide the equity, and the banks provide the bulk funding.

      When I track down the reference, I would probably cover that angle more explicitly.


Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.