There are impassioned campaigns for replacing our current system of banking with various proposed alternatives termed “full reserve banking”.  Typically the aims of these proposals are to stabilize the banking system AND to prevent private banks from controlling the supply of money. This post is about how the first aim seems realistic and wise but the second concern is something perhaps much better addressed by other means.

This discussion needs to start from a clear understanding of what banks are (I’ve attempted a background overview of our monetary system on page 13 of this pdf). I also want to make the case that this is all about “maturity transformation”. Imagine lending without maturity transformation: Adele and Beyonce are neighbours who each own their own homes. Adele also has £100000 as paper cash in her mattress. Adele decides to sell her house to Beyonce and then live as a tenant, paying rent to Beyonce. Beyonce goes to a loan company to get a loan to buy Adele’s house with. The loan company grants a £100000, ten year, interest only, mortgage to Beyonce. Beyonce has to pay 4% (ie £4000) every year to the loan company and then after ten years pay back the £100000. The loan company funds the loan by selling a £100000 bond to Adele in return for the £100000 paper cash Adele had in her mattress. When Beyonce buys Adele’s house she pays Adele £100000 and Adele puts that as paper cash in her mattress. Adele receives 3% (ie £3000) interest each year from the bond and also gets the £100000 principle back at the end of the ten years. Adele is also an employee and shareholder of the loan company and the loan to Beyonce adds to Adele’s annual pay and dividends by £1000. Adele pays £4000 rent a year to Beyonce. At the end of the ten years Adele hopes to buy her house back with the £100000. The entire ten year long process simply uses Adele’s starting money and circulates it about leaving things at the end just as they started. During the period of the loan however there is extra asset in the system in the form of the bond owned by Adele. This is a corollary of the debt owed by Beyonce. This process is NOT bank lending. The loan company would not need a banking license to do this.

Imagine the loan company instead decides to fund the ten year loan to Beyonce by selling a one year bond to Adele. The loan company hopes that after the first year, it will be able to sell another one year bond to Adele or someone else and use the money from selling the second bond to pay the principle on the first. In that way, ten successive one year bonds fund the ten year loan. Under this scenario, the extra asset in the system is in the form of a one year bond rather than a ten year bond. Adele now has £100000 of paper cash in her mattress but also knows that she is due to receive £100000 in a year’s time. The loan company needs to find a buyer for each successive bond. This necessity to find a buyer is termed “liquidity risk”.  Nevertheless the loan company still would not need a banking license to do this.

In the third scenario the loan company decides to fund the loan by rolling over successive one week bonds. Adele now has £100000 in her mattress and also knows that by the end of the week she has another £100000 coming. For most practical purposes Adele has £200000. Nevertheless the loan company still would not need a banking license to do this.

In the final scenario, the loan is funded by a continuous succession of instantaneously maturing bonds. Now the loan company has become a bank and the loan is funded by bank deposits.  What we know as bank deposits are in effect a continuous succession of instantaneously maturing bonds. Whilst the bank deposit is with the bank, it is funding the bank’s loans. Adele would now have £100000 of paper cash in her mattress and £100000 in the bank. She would have £200000.

Note however that in practice banks typically would not have a large book of ten year loans funded over the entire ten year period by bank deposits. Banks avoid undue liquidity risk by reducing the extent of “maturity transformation” by ensuring much of the funding is in the form of longer term bonds or fixed term savings accounts. The bank constantly needs to fund the outstanding loans either by bank deposits or such longer term liabilities.

Of course we use bank deposits as the medium of exchange for most payments (see page 15 of this pdf for an overview of the payment and settlement process). Paper cash is much less significant. So the debt from bank loans is the predominant form of money in our economy. It seems to me that banks currently provide two completely separable functions. One is the payments system; that could potentially be a “full reserve system” just like PayPal; the other is the lending system. Where we run into trouble is that we mix them up. We have a system where the payments system uses the debt from lending as the medium of exchange. That entails all of the hazards of maturity transformation and liquidity risks. It means that the payment system (a vital utility for the entire economy) is held hostage by the lending system. That creates a slippery slope towards bailing out an irresponsible lending system so as to keep the payment system intact.

If the payment system was just by something such as PayPal and the lending system were separate; THEN it wouldn’t matter if systemic credit defaults led to defaults on the bonds sold to fund loans. That would not impact on the vital role of the payment system. It would simply be irresponsible lenders getting what they had coming to them.

Lending could potentially be by loan companies that did no maturity transformation and simply sold bonds with the same maturity profile as the loans made. Retail savers could hold savings in the form of ETFs that were made up of thousands of such bonds (just like existing bond ETFs). Customers could sell their ETF holding whenever they needed to draw down savings.

I think it is worth remembering that half of all the credit provided in the USA is already by shadow banks (ie lenders that don’t hold deposits) and we already have a non-bank payment system in the form of paypal. Transition from our current system to a system without maturity transformation need not be at all disruptive. There is already a vast stock of excess bank reserves in the UK banking system (the same is true in the USA and Japan). If banks were to now match the maturity profile of their liabilities to their existing loan book, then that would entail selling of bonds. Those bond sales would drain the bank deposits and so replace much of the current stock of bank deposits with bonds. The remaining bank deposits automatically would be those entirely matched by bank reserves. At that point full reserve banking would have been achieved. In effect the banks would have simply been converted into being the functional equivalents of PayPal together with a loan company that could be separated off as a distinct institution. A bank run would be an accounting impossibility for the payment system. Even if all of the customers withdrew all of their money, the payment system would suffer no liquidity stress. Obviously such accounts would purely be provided as a service. The payment system would no longer have any reason to seek account holders custom except so as to charge a service charge. That seems no great problem to me. Mobile telephone services are provided efficiently and providers of that service charge their customers. The payment system could similarly be such a mundane, low cost, utility. We now indirectly pay banks to administer the payment system. We would simply be paying for it in a more direct manner.

If people wanted to receive interest on their savings, then they could hold their savings in the form of a bond ETF. Various bond ETFs could be offered each with a different maturity profile. An ETF with bonds of less than one year duration would have minimal price fluctuations. An ETF with bonds of over 15 years’ duration would have considerable price fluctuations. In either case however any given saver could sell their holding on the market, at any time, with a click of the mouse, just as with current ETFs. The demand from savers to hold such interest bearing bonds would be what set lending rates. A key aspect of what I’m proposing is that there would be a regulatory ban on maturity transformation. Loans of each time to maturity would be made available at the price the debt was actually sell-able for. I consider maturity maturation to not only be a key cause of instability but also to be a distorting source of mis-pricing that causes more robust equity financing to be displaced by fragile debt financing.

It is important to contrast this proposal with maturity matched peer to peer lending systems such as Zopa and Funding Circle. Zopa and Funding Circle simply offer savers a stake in aggregated maturity matched loans without any shielding from losses due to defaults. Zopa and Funding Circle simply administer the system and do not bear the credit risk. I’m proposing that loan companies would have to themselves hold capital that would stand first in line to take losses in the case of defaults. If losses were so great that a loan company had insufficient capital, then the bonds sold would convert to shares in the loan company until the loan company were adequately capitalised. Loan companies would also have to keep any loans made on the books until they matured or were written down. It would thus be the responsibility of the loan company to assess the credit risk of the borrowers.

This proposed system is starkly different from currently prominent “full reserve banking” ideas. The key difference is that it makes no attempt to prevent debt from taking on “money-like” characteristics. Savers would presumably view their holdings of my proposed ETFs as being “money” they had available and would act accordingly. This “money-like” quality would be especially true because the loan companies would be well capitalized, would not be subject to liquidity risk and (because they kept loans on their books) would be cautious about credit risk.

The reason why most “full reserve banking” proponents wish to prevent debt from taking on “money-like” characteristics is because they consider that only the state ought to have the role of deciding a country’s money supply. They consider it unfortunate that commercial provision of credit confers that power to private companies. The “positive money” campaign hopes that debt would be prevented from slipping into being “money-like” by insisting that it is held as fixed term saving accounts. The intention is that savers would then not be able to access their savings for several years and so would consider such accounts as being money that they were due to receive rather than being money they actually hold. In practice I don’t believe it would work. People who needed access to money in the interim would simply take out a shorter term loan knowing that they would be able to pay it off when their savings account matured. Credit providers would also take account of what savings accounts would be borrowers held. On a more general note, I think any such attempt at placing inconvenient obstacles would fail to erase the “money-like” properties of debt. I consider that the government does have an important role to play in controlling the supply of credit but I consider that the tool to use is price. Furthermore the way to put a price constraint on credit formation is to replace existing taxes with a tax on all gross assets (this tax idea is detailed in this pdf).

The “full reserve banking” proposal of Benes and Kumhof takes the approach of insisting that all debt from lending is transferred to government ownership so as to eliminate “debt as money” altogether. The government then becomes the entity behind decisions as to who is lent money. The country as a whole pays the consequence if debts default. Perhaps more worryingly, the political system becomes embroiled in the minutiae of the entire economy. To me it seems a recipe for cronyism and gross inefficiency.

Edited 29April2013, added: I also want to make the case that this is all about “maturity transformation”. Imagine lending without maturity transformation (In response to Ralph Musgrave’s comment)

Related stuff on the web:

Toward a run-free financial system -John H Colhrane (ht Nick Edmonds) (link added 30April2014)

100% Reserve Banking-The History -House of Debt (link added 30April2014)


Money creation in the modern economy -Bank of England-(link added 19March2014 ht JKH)

‘Loans create deposits’- in context

What is a bank loan- interfluidity

What is money and how is it created- Bank of England (see page 377)

Hitch-hiker’s guide to monetary infrastructure -FT

Positive Money

Banking in the Abstract -The ‘Chicago Plan’ -Monetary Realism

Proposals for the Banking System- Warren Mosler

Bagehot was a shadow banker

More on M-pesa and e-money -FT

When safe assets return -FT (link added 26May2013 ht Mike Sankowski)

Endogenous Money and “Out of Thin Air” Money – Nick Edmonds (link added 14July2013)

Repeat after me: banks cannot and do not “lend out” reserves -Standard and Poors (link added 15Aug2013 ht Mike Sankowski)

Comercial Banks as Creators of “Money” – James Tobin 1963 (link added 20Aug2013 ht Ramanan)

Banks, Non-banks and the Medium of Exchange – Nick Edmonds (link added 06Sep2013)

Financing Investment In A World Without Maturity Transformation -Ashwin Parameswaran (link added 30nov2013)