article

Central banks: crossroads or dead-end?

Central banks are at a crossroads, and if they carry on, maybe a dead-end. The good news, though, is twofold: central bankers recognise the need to reassess their frameworks and they are coming at it from a point of strength: unemployment at multi-decade lows in most of the developed world.

The problems have become readily apparent. First, the conventional framework in place today – flexible inflation targeting – aims to target a variable over which central banks very evidently have limited control. Japan provides the most extreme example. The BoJ has expanded the monetary base by over Y400 trillion (or c.US$4 trillion) in the past 7 years, equivalent to 100% of GDP. Rather than prices surging, the result has been an average inflation rate of just 0.7% p.a. (0.3% ex food and energy), consistently well below the stated objective of 2%.

Would the outcome have been meaningfully different if the dosage had been Y200trn, Y400trn or Y800trn? It is worth thinking about.

Either central banks have been far too timid in their approach, or other factors out of their control exert at least as much influence over inflation. The roles of technology, demographics and the nature of competition come to mind; in which case, a precisely specified level of inflation is the wrong target.

Second, the conventional economic framework used by central banks isn’t working. In one of her last (and most fascinating) speeches as Federal Reserve chair, Janet Yellen indicated as much. The model essentially relies on the concept of an output gap as defined by a measure of full employment and the role of inflation expectations in determining future inflation. On the concept of the level of unemployment that generates higher future inflation, the Fed has been completely wrong, as current chair Jerome Powell admitted to congresswoman Ocasio-Cortez only this month. Has the Fed been wrong about the calibration – or about the process by which inflation is generated?

Similarly, in today’s economy the role of inflation expectations – as typically analysed – appears confused and outdated, a relic of models designed in the 1970s and ‘80s when inflation was in the double-digits. Bank of England MPC member Silvia Tenreyro remarked in a recent speech that “if households expect prices to rise more, that increases their incentive to spend today rather than saving”. Is that really how households behave? It seems very unlikely at today’s low levels of inflation, even if it was the case 40 years ago.

We are lucky in most developed markets today. Greenspan’s definition of price stability has been reached – inflation is too low to be noticed and most households (or indeed businesses) make spending and savings decisions with little regard to it. People respond to relative price changes between goods and services, which in a low inflation regime dwarf changes in the aggregate price level. In the modern developed world, expected inflation is noise.

What about inflation as expected by professional forecasters and financial markets. Surveys of economists may be useful but they bear little relation to the wider population. Measures of expected inflation in financial markets matter to market prices and traders’ P&L accounts, but fluctuate far too much to be reliable estimates of future inflation many years out. Why would an apparent estimate of inflation over the next 30 years fluctuate with spot moves in the oil price? Much of the shifts appear to be related to fluctuating risk premia, which central bankers thankfully acknowledge.

So where does this leave us and what is to be done?

Central banks need some fresh thinking and a new mindset. They must recognise that their models are not working, try to understand why and revise them accordingly (something which should be happening anyway). They must also recognise that a precise inflation target is misguided. Their goals should be more modest, but pursued more aggressively.

As a starter, I propose the following goals in order of emphasis: (i) to promote financial stability and prevent collapses of the financial system; (ii) to promote low inflation as measured by a range of 1-3% over the medium term; (iii) to facilitate economic growth.

By far the most critical function of the central bank is its role of lender of last resort. Preventing economic catastrophes matters far more than tinkering with interest rates.

The other goals of maintaining low inflation and encouraging growth are valuable, but imprecise. Alongside, the revised mandate should come a change in mindset: do less more often, but be bold when required. Small changes in interest rates are usually irrelevant. But central banks should act much more boldly when they look like missing those wider targets.

For this, they need new and revised tools, as my colleague, Eric Lonergan, has argued. More QE isn’t the right medicine, as the Japanese experience suggests. Wider adoption of targeted lending (TLTROs) at dual interest rates, greater powers to buy equities and credit to reduce private sector risk premia during recessions and the capacity to make direct transfers to the private sector would, however, be game changers.

Central banks need a radical rethink, while conditions are good. More of the same is simply not good enough.


The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested. Past performance is not a guide to future performance.