As many of you know, I have spent much of the last seven years explaining to anyone who will listen that banks do not “lend out” deposits or reserves. Rather, they create both loan assets and matching deposit liabilities “from nothing” by means of double entry accounting entries. Creating money with a stroke of the pen (or a few taps on a computer keyboard) is what banks do.
But this does not mean that the money that banks create comes from nowhere. It doesn’t. It is only created when they lend (or when they purchase assets, which is equivalent to lending). As Pontus Rendahl explains in a comment on my previous blogpost, what banks do is liquidity transformation – exchanging long-term illiquid assets for short-term liquid ones:
How do private banks create money? They create a deposit. A deposit is a Barclays-pound/Bank of America-dollar, or what not, that is traded and accepted as a means of payment at a one-to-one exchange rate with the underlying national currency. But as with currency pegs, such exchange rates can only be maintained if the bank has a healthy asset side on the balance sheet, and sufficient reserves.
But did the bank create this deposit out of nothing? No, it created the deposit out of an asset; a loan. That is not “out of nothing”. In fact, it’s very far from it; a more accurate description is that the bank converted an illiquid asset (the debtor’s future ability to repay) into a liquid one. If the bank did not create the deposit out of an asset, it would end up in a asset<liabilities situation. It would be insolvent.
All money that banks create is their own liability towards a third party – in other words, it is debt. This debt money is always backed by an equivalent claim on the income and, as a last resort, goods of a third party. The problem for banks is that the debt they owe to their customers as a consequence of lending is short-term and highly liquid, whereas the debt their customers owe to them as a consequence of lending is long-term and illiquid.
When the customer draws down a loan – or, for that matter, when a customer draws any money from their deposit account, whether or not it is created through lending – the bank must have sufficient cash, or liquid assets readily exchangeable for cash, to pay them. But banks don’t keep much in the way of liquid assets: after all, if they had to hold sufficient liquid assets to enable everyone to draw down the contents of their deposit accounts on demand, they wouldn’t be able to do much in the way of long-term illiquid lending. Therefore, although banks create money when they lend, they still need to borrow money when they make payments. It is bad accounting to assert that “banks create money when they lend” without recognising the funding problem caused by liquidity transformation. Yes, lending creates deposits, but deposit drawdowns need to be funded with liquid assets – which for banks are by definition scarce.
So although we often say banks create money “from nothing”, what we really mean is that they create money from lending. And the purpose of the lending is completely irrelevant. Money created for purposes that Zoe Williams in the Guardian dubs “unproductive”, such as mortgage lending, is just as much money in circulation as money created for what she calls “productive” lending to corporations. Who are we to judge what is “productive” and “unproductive”? People who buy houses don’t just create work for lawyers and estate agents, they spend a lot of money on doing up their houses – money that people who rent often either can’t or don’t want to spend. They also free up money for other people to spend. If what you want is demand, mortgage lending sounds like a good bet.
This is not to say that the dominance of real estate lending is necessarily a good thing. It has unfortunate distributive consequences, since it pushes up house prices, pricing younger and poorer people out of the market. But it is more than slightly unreasonable to bash banks for expanding mortgage lending when the prevailing political belief is that everyone wants to buy their own house, media constantly talk up the market (when did you ever see news of a fall in house prices portrayed as a good thing?) and government actively intervenes to help people buy houses they can’t afford on their own. If government really wants banks to lend to businesses, it is sending entirely the wrong messages.
To my mind, the bigger problem is the extreme illiquidity of mortgage portfolios on bank balance sheets. Prior to the financial crisis, banks thought they had solved this problem, since securitising mortgages is another form of liquidity transformation. But the securitisation engine failed in 2007-8 and has never really revved itself up again. So banks are now stuck with large amounts of highly illiquid, long-dated assets on their balance sheets.
Of course, since the crisis regulators have worked hard to reduce the crippling illiquidity and constant risk of insolvency that arise from the nature of modern commercial banking. These days, banks have to keep more of their assets liquid, thus reducing the likelihood that a sudden run on their liabilities will force them either to tap central banks for emergency liquidity or sell assets at heavily discounted prices (fire sales) to raise liquid funds. And they also have to keep a larger gap between their assets and their liabilities, to reduce the likelihood of insolvency if the aforementioned fire sales, or a nasty drop in market prices, renders the market value of their assets insufficient to meet all the claims upon them. This gap is the “equity cushion” that is so often misreported in the media as “holding capital” or “setting money aside”. It is nothing of the kind. Bank capital is “own funds” – shareholders’ equity, retained earnings, debt convertible to equity. The whole idea is that the bank should be able to absorb a significant fall in asset values without defaulting on its obligations to its creditors. Whether these measures are sufficient to mitigate the extreme illiquidity and high insolvency risk of banks whose main activity is mortgage lending remains to be seen.
The central bank does for commercial banks what commercial banks do for their customers. It exchanges illiquid loans for newly created, highly liquid, risk-free money, which banks then use to pay their customers. It also acts as deposit-taker, accepting excess cash from banks that have more than they need to meet short-term payment obligations, and paying them interest on that money. And it is responsible for final settlement of payments cleared through commercial banks.
But if commercial banks can’t create money unless it is backed by assets, can central banks? This is a matter of fierce debate.
At present, they don’t. A central bank that creates money through open market operations is creating asset-backed money, since it is exchanging existing assets held by the private sector for newly created money. Similarly, the money created in QE is exchanged for existing assets held by the private sector, and is therefore also asset-backed. If a bank borrows reserves from the central bank, the accounting at the central bank is the same as if a commercial bank lends to a customer: new deposit in the commercial bank’s reserve account balanced by a new loan asset to be repaid at an agreed date, with interest. Assets that the commercial bank pledges in support of the loan do not appear on the central bank’s balance sheet, for exactly the same reason that the houses mortgagees pledge in support of their mortgages are off the commercial bank’s balance sheet: they are not owned by the bank, but the bank has a first claim upon them in the event of default.
But in fact, a central bank can create money that is not explicitly backed by anything. Unlike commercial banks, they can genuinely create money “ex nihilo”. The value of the money they create is maintained entirely by the credibility of the central bank/government combination.
Suppose our central bank decides to do “helicopter money”, which is a simple transfer of newly-created funds to the private sector. No assets are purchased, and because this is a central bank transfer, no debt is created. When a currency is not asset-backed (as in a gold standard), the money created by the central bank is redeemable only as more of itself, and is therefore not “debt” in any ordinary sense of the word. So the balance sheet entries for a helicopter money transaction would look very odd: huge quantities of zero-coupon irredeemable liabilities, and no assets. The central bank would be technically insolvent. Would this matter?
Arguably, no. The implied asset backing for helicopter money is the net present value of future expected tax receipts (for the Eurozone, read this as the collective future tax receipts of all current and future Eurozone governments.) This doesn’t mean that governments must always run primary surpluses: as central bank money is perpetual and irredeemable, and there is no interest to pay, the government could run sustained fiscal deficits far into the future without destroying the central bank’s implied assets. As long as the government is credible, the central bank should be able to do helicopter drops without destroying the value of the money it is dropping.
The risk arising from a central bank doing helicopter money while the government runs sustained fiscal deficits is inflation. Irresponsible money printing by an irresponsible central bank at the behest of an irresponsible government is rightly feared as a cause of hyperinflation. But there is a huge difference between an irresponsible fiscal authority mismanaging its finances and dominating the central bank, and a responsible fiscal authority and central bank that together choose a reflationary path that combines helicopter money and deficit funded investment. Of course, even if done by a responsible central bank, helicopter money should be inflationary, especially if accompanied by fiscal deficits: but that would be the whole point of doing it!
The combination of helicopter money with investment financed by large fiscal deficits should be extremely powerful. The helicopter money would provide a direct demand stimulus to the real economy, with money going to those who are most likely to spend it rather than to those who are rich enough to own assets, while deficit-funded fiscal investment would kickstart the supply side of the economy and prime the pump for private sector investors. Such an approach would force the central bank and government to cooperate while still preserving the distinctive remits of the central bank (demand management) and fiscal authority (supply side and distribution).
So why didn’t we do this after the 2008 crash? Simple. Fear of inflation. We preferred to court deflation than risk resurgence of inflation. Because of this fear, we did not provide the really powerful stimulus that the economy needed after the crash. We provided a much weaker stimulus in the form of QE, and then weakened its effect still further by embarking on premature and ill-considered fiscal consolidation. Instead of allowing the public sector to plug the gap left by the withdrawal of a wounded private sector, we tried to kick damaged banks into lending and damaged businesses and households into borrowing. We completely failed to deal with the widening inequality that arose from the combination of QE with fiscal austerity, and the destruction of hope that arose from ten years of stagnation, in some countries accompanied by very high unemployment, especially among the young. The price we are now paying for this abject failure of public policy is widespread political unrest. Some might say it is surprising it has taken as long as ten years for people to lose patience with the broken political establishment.
I have proposed here that helicopter money should be the central bank instrument of choice after a severe financial crisis. But that does not mean that all money should be delivered by helicopter. Those who propose “sovereign money” to replace money creation through bank lending appear to be driven by an irrational, though perhaps understandable, fear of debt. And they also, to my mind, place far too much faith in central planners, as this comment from Zoe Williams shows:
The nature of centrally created money should itself be opened up for debate, whose starting point is: if we agree that commercially created money is skewing the economy, can we then agree that it should be created by a public authority, even if we don’t yet know what that authority would look like.
No, Zoe, we can’t agree on that. Firstly, although I might agree that commercial bank lending can skew the economy (though you totally ignore the fact that government actively fosters this distortion) I don’t agree that the associated money creation necessarily skews anything. To my mind, QE, which gives money to the rich to spend on things that only the rich want, is far more distortionary – and QE is publicly created money.
Secondly, I don’t believe that any central authority is capable of deciding how much money the economy needs. Far better, in my view, to allow the amount of money in circulation to respond to private sector demand. The job of “central planners” is to regulate the bipolar swings characteristic of private sector money demand.
And finally – a warning. The “money tree” exists, but its fruit is not cost-free, no matter who creates it. We should be cautious about advocating completely unrestricted money creation by any institution, private or public.
Money creation in the modern economy – Bank of England
Money tree image from the BBC.