Economists expect the Federal Reserve to increase the fed funds rate by a quarter point later this month. This innocuous development hides an awkward fact – that economists don’t really know what interest rates should be.
Theory tells us that there should be a 'neutral' or 'equilibrium' or 'natural' rate of interest at which inflation is low and stable and the economy growing at its trend rate. The idea here dates back to the Swedish economist Knut Wicksell writing in 1898. He said that if interest rates were below the return on capital, then companies would borrow to expand. That would lead to booming employment and demand and hence to rising inflation. The natural interest rate, he thought, would be the rate that prevents both this and its opposite, declining demand and deflation.
The facts, however, suggest that actual US interest rates have not been below their neutral 'level'. Wage inflation has been stable around 2.5 per cent since mid-2016; price inflation (measured by the CPI excluding food and energy) is lower than it was last year; and GDP growth in the third quarter, at 3 per cent annualised, was actually below the average rate we saw in the 50 years to 2007. None of this is a sign that the fed funds rate has been below the natural rate.
With the fed funds rate negative in real terms, this suggests that the 'natural' real interest rate might be negative. This means it has fallen massively since the 1980s and 1990s; John Williams at the San Francisco Fed estimates that it was over 3 per cent then.
This fall is a symptom of what Larry Summers calls “secular stagnation”. A combination of lower trend GDP growth, an ageing population with more savers, companies’ reluctance to invest and increased demand for safe assets have reduced the 'natural' real interest rates not just in the US but across western economies.
Which brings us to the problem. Nobody knows for sure what the 'natural' rate should be. We can’t observe it directly, and estimates of it are, says Mr Williams, “highly imprecise”. This imprecision is evident in the Fed’s own thinking. The Federal Open Market Committee members’ estimates of what the appropriate fed funds rate should be in 2020 range between 1 and 4 per cent.
What’s more, even theory is vague. We know that lots of things can influence the 'natural' rate: all the determinants of the supply and demand for credit, willingness to hold stocks of different assets, the marginal product of capital (which itself is unobservable and perhaps theoretically incoherent) and so on. But reconciling these into a theory that fits the observed pattern of interest rates is tricky.
In fact, there might not even be a single 'natural' rate says M&G’s Eric Lonergan. Ultra-low rates, he says, might cause people to rein in their spending – because if we can’t increase our wealth by getting decent returns we must do so by saving more. This means low rates might be self-perpetuating; in weakening demand they cause central banks to keep rates low.
You might think all this shows that economics isn’t much use. Not so.
For one thing, we have the Brainard principle. This says that when faced with uncertainty, central bankers should move slowly. Uncertainty about the 'natural' rate thus justifies small and gradual moves in the funds rate. Which is what markets expect. Futures markets are pricing in a funds rate of only 1.8 per cent for December 2018, implying only two rises next year.
And for another, uncertainty about the current and future 'natural' rate means we cannot say with any confidence at all that there is a bond bubble. Quite the opposite. The very fact that inflation has stayed low in the negative real rates is a big clue that the 'natural' rate, if there is such a thing, is very low indeed. Which justifies low bond yields.
