This is the first of several posts covering topics discussed at the recent Lindau Meeting for Economic Sciences.
Several economists at the Lindau meeting were severely critical of central banks' conduct of monetary policy in the light of continuing depression in the US, Japan and much of Europe, and called for greater use of fiscal policy to bring about recovery. Among the most critical was Christopher Sims, who gave a trenchant presentation on “Inflation, Fear of Inflation and Public Debt”.
He started by announcing the death of the quantity theory of money, MV=PY. Due to interest on reserves and near-zero interest rates, “money” can no longer be clearly distinguished from other financial assets. This is a fundamental point which requires some explanation.
These days, nearly all forms of money bear interest, which makes them indistinguishable from interest-bearing assets. For Sims, the paying of interest on bank reserves, coupled with the decline of physical currency, all but eliminates the distinction between interest-bearing safe assets such as Treasury bills and what we traditionally call “money”. All assets can be regarded as “money” to a greater or lesser extent: the extent to which assets have “moneyness” is really a matter of liquidity.
To give an example: while I was in Lindau, I had very few Euros, no Euro bank account and no credit cards. I have sterling bank accounts, of course, but what use are those in a country that does not use sterling? Ten years ago such illiquidity would have meant I did not eat for a week. But not any more. Now, I can obtain Euros using my VISA debit card (issued by a British bank for a sterling bank account) in an ATM. Or – even better – I can use the same card to pay for a meal in Euros directly from my sterling account. Because of technological changes, what was formerly an illiquid asset in Germany (sterling) has become transparently liquid.
Sterling is, of course, a currency, so my example does not prove the fungibility of currency and interest-bearing assets (though of course my sterling bank account is an interest-bearing asset). But suppose that instead of a sterling bank account, a smartcard or a smartphone app enabled me to pay a bill in Euros directly from my holdings of UK gilts? This is not as unlikely as it sounds. It would actually be two transactions – a sale of gilts for sterling and a GBPEUR exchange. This pair of transactions in today's liquid markets could be done instantaneously. I would in effect have paid for my meal with UK government debt. It is ridiculous not to regard such a highly liquid asset as money. ANYTHING that can be used to settle transactions really should be regarded as money: and as technology increases liquidity in all asset classes, so they become easier to use for transaction settlement and therefore more money-like.
The pricing of money-like financial assets is something of a challenge. As assets become more liquid, the term premium becomes less relevant: after all, if you know you can sell the asset tomorrow, you are not locking up your money for extended periods of time even if the asset itself has a maturity date years into the future. Behavioural measures of liquidity suddenly become far more important than structural ones. This might partly explain why term premia have been falling over the last thirty years.
Credit risk does matter, of course. But we are amazingly good at ignoring it when it suits us, and liquidity trumps credit risk anyway: we like to believe that we can always get out of a risky investment if it is sufficiently liquid. So a highly liquid asset is likely to be regarded as low risk. Not that liquidity is always certain, of course. CDOs were highly liquid near-money assets prior to the financial crisis. Now, they are about as liquid as flies in amber, and worth considerably less.
When nearly everything is highly liquid, even long-duration assets won't pay much in the way of interest. The yield curve on an ultra-liquid asset is flat. If you want yield, you have to look for illiquidity – which these days can be hard to find. No wonder yield-hunters go for intrinsically illiquid assets such as property, or assets that carry significant credit risk such as junk bonds. They can't get any sort of return on safer and/or more liquid assets. And the more technology improves market liquidity – and the more regulators push for improvements to market liquidity – the lower returns will be across all classes of financial asset. Perhaps this might explain not just falling term premia, but falling interest rates generally for the last thirty years?
So in our ultra-liquid world, nearly everything bears interest, and almost nothing pays interest. “High powered money” - the forms of money issued directly by the central bank, which collectively make up the monetary base - is indistinguishable from any other sort of liquid asset. The traditional distinction between the monetary base, other monetary aggregates and other financial assets is spurious. They are all liquid, all fungible and most are either explicitly or implicitly backed by government*. This is unfortunate, because most monetary policy is based upon the idea that there is some fundamental distinction between the monetary base and all other forms of money.
A couple of things follow from this. Firstly, monetary policy based solely on changing the compositional mix of the various forms of “money” in the economy can only work to the extent that some forms are more liquid than others. On that basis, QE, which replaces government debt and certain private sector assets with new monetary base, should provide additional liquidity in markets where the substitutes are not liquid. Unfortunately the only market where the substitutes are not liquid is the real economy: in financial markets, liquid assets are a near-perfect substitute for cash. But as we know, QE's impact on the real economy is unclear. Market monetarists argue that the increase in liquidity is bound to have some effect through increased transactions (the “hot potato” effect), but others argue that swapping liquid assets for base money merely encourages unproductive investment. What is very clear though is that in the real economy, there is still a distinct shortage of liquidity despite years of monetary easing. For Sims – and for me - however it works, QE is a very weak policy.
Similarly, Operation Twist, which shortened the maturity profile of US government debt in circulation, had less effect than might have been expected. But the UK's Funding for Lending scheme, which improved the liquidity of bank balance sheets by swapping illiquid mortgage loans for liquid treasury bills, does appear to have revitalised the UK economy to some extent – though the relaxation of capital rules for new lending under the scheme might also have had something to do with it. UK banks, expensively and painfully recapitalised over the last few years, are understandably very wary of committing capital to new lending.....
Secondly, if “money” is indistinguishable from any other sort of asset, it is impossible to separate monetary and fiscal policy. After all, fiscal policy itself involves the issuance of money-like assets, and monetary policy acts upon those money-like fiscal instruments. Every monetary action therefore has fiscal implications, and vice versa. Attempting to enforce separation of monetary and fiscal policy renders the central bank powerless. We see this all too clearly in the Eurozone, where everything the ECB can do within its mandate to counteract the falling money supply is ineffective because interest rates are already on the floor and banks (on which the European economy is heavily dependent) take little notice of the ECB's policy guidance, and anything that might actually work founders on the prohibition of central bank financing of fiscal deficits and – above all – the incomplete fiscal union.
So it is not so much that the quantity theory of money is dead, as that it is incomplete. Sims pointed out in his lecture that M has already been redefined by most people to include the money created by banks when they lend: the "money multiplier" is no longer a meaningful concept. But now M needs to be redefined again to include all classes of asset that can be used as money. For fans of IS/LM charts, that doesn't leave much on the IS curve. If everything is money, how can anyone make any money simply by holding money? Money only makes money if it flows.....The only way of making money from money in an ultra-liquid world is to find something that isn't liquid and invest in it – the less liquid, the better the return. And there seem to be fewer and fewer illiquid investment opportunities in traditional asset classes.
Insurance companies, pension funds and private equity firms are already wise to the problems that ultra-liquidity poses for them. They are actively looking for new types of illiquid investments. Insurance companies are keen to find ways of investing in illiquid mortgage loans on bank balance sheets without incurring capital penalties, for example. And private equity firms are talking about long-term investments in infrastructure and technology, perhaps in partnership with the quasi-public sector institutions from which many governments are withdrawing funding in the interests of fiscal consolidation. There is clearly an opportunity here to open up new sources of capital investment to benefit both investors and the liquidity-starved real economy. For where investors invest, money flows.
But investors need to take care. Long-term investment projects can lose you a serious amount of money if you get the evaluation wrong – and the longer-term the project, the harder it is to estimate its real returns. And some illiquid investments simply aren't valuable. A barren lump of rock in an ocean is just a barren lump of rock: it produces nothing and is therefore intrinsically worthless. An asset that has lost its liquidity – such as a CDO – has most likely also lost its value.
And even long-term investors need some liquidity. We really don't need insurance companies, pension funds and the like becoming like the Ancient Mariner – surrounded by liquidity but dying of thirst. Liquid assets may not make any money, but they do make it possible for companies to meet their day-to-day obligations. It's a balance, really. Too much liquidity, you can't make any money. But too little liquidity, and you become like that fly embedded in the amber. Dead.
Above all, though, we need to rethink how we do monetary policy. We already know that post-crisis, monetary policy can no longer use reserve scarcity to create the illusion of money drought and force agents to pay more for liquidity. But actually central banks now lack the ability to influence the cost of liquidity much at all. Liquidity is becoming to all intents and purposes free. The question for policy makers now is how to influence the returns earned on illiquid investments such as property.
If ultra-liquidity is here to stay, as seems likely, we must either re-unify monetary and fiscal policy or resign ourselves to central bank impotence. There is no longer any justification for even the pretence of separation. Fiscal and monetary authorities together need to find new ways of transmitting policy to a world flooded not only with reserves, but with liquid assets of many kinds.
Moneyness - J.P Koning
Theory of Money Entanglement – Izabella Kaminska at FT Alphaville
The liquidity trap heralds fundamental change – Coppola Comment
Christopher Sims's lecture at Lindau Economics Meeting (video)
Image from the 8th International Congress of Dipterology, which was held in Potsdam, Germany, 10th-15th August 2014 - a week before the Lindau Economics Meeting.
* It should be noted that when the central bank actively buys, or even declares an intention to buy, private sector assets, they implictly gain a sovereign guarantee. Central banks are backed by the fiscal authority: no sovereign would fail to guarantee assets held by its central bank.