INTEREST rates will eventually rise in the developed world, perhaps in early 2015. The markets are braced for such a possibility. But the more difficult question is how far they will rise: what is the normal level of rates in the post-crisis world?

This is a crucial question. The Federal Reserve’s benchmark rate averaged 2.96% in the first decade of the 21st century but 5.15% in the 1990s. Imagine the effect on the borrowing costs of mortgage-holders or small businesses if rates moved back to the latter level.

The general view is that rates will remain low by historical standards. The chart shows the implied yield on five-year government bonds five years from now (generated by comparing the yield on five-year bonds with that on ten-year issues). This yield has fallen in 2014 to around 3.25%. Given that short-term rates are normally below five-year yields, that suggests investors do not anticipate either a return to the robust economic growth of the 1990s or a resurgence in inflation.

Economists talk about the “neutral level of interest rates, where the central bank is neither trying to stimulate the economy nor to slow it down. This is normally expressed in real (stripping out inflation) rather than nominal terms. As Stephen King, chief economist at HSBC, points out, real rates have varied a lot since the second world war.

In the 1950s and 1960s, real rates were quite low. Capital controls kept investors from moving their money in search of higher returns and governments held rates down as a way of reducing their debt burdens, a technique known as financial repression. In the 1970s a sudden surge in inflation caught investors by surprise, pushing real rates into negative territory. The bond vigilantes responded in the 1980s by forcing bond yields higher. As inflation declined, real rates rose sharply. Finally, the savings glut of the early 2000s, followed by the monetary response to the 2008 crisis, meant that rates plunged once more.

All this suggests that the neutral level of rates is far from set in stone and that the past may be a poor guide to the future. The rate of real economic growth is probably the most important determinant and in that respect many people are gloomy: the trend rate of growth in America may have dropped to 2% or below while pessimists put the European rate at 1%. Demographic trends mean the workforce is likely to stagnate or shrink, leaving productivity to do all the work expanding the economy. Other things being equal, a slower growth rate will mean a lower real interest rate.

But will other things be equal? An ageing population means the elderly will start to draw down their savings in retirement. Real rates in greying Japan have been positive in recent years (thanks to deflation) despite the moribund economy. And some investors may conclude that the debt burden in the developed world is so great that the eventual result must be either default or inflation, and will demand a higher rate to compensate for the risk.

A further factor is globalisation. Emerging markets are growing faster than the developed world and will thus attract investors seeking higher returns. In theory, this should cause real rates to equalise around the world at a higher rate than if the developed world were the sole determinant. In practice, real rates may still be held down by a modern version of financial repression: pension funds and insurance companies have been forced, by regulations and accounting, into owning more bonds, while central banks have bought huge amounts of government debt and show no sign of selling it.

On balance, then, the neutral level of rates may be even lower than markets expect. PIMCO, a fund-management group, suggests the level of neutral rates is likely to be close to zero in real terms (so perhaps 2-3% in nominal terms if inflation targets of 2% are met).

Before homeowners and small business breathe too big a sigh of relief, remember that this discussion has concerned the neutral level of rates. Rates could go higher if central banks decide they need to rein back the economy, perhaps because inflation returns. Richard Barwell of Royal Bank of Scotland says there may be too great a belief in “the miracle of immaculate monetary exit”. This would require monetary stimulus to be withdrawn before the economy recovers, not after, and for central banks’ economic forecasts to be unerringly accurate.

Since central banks failed to anticipate the debt crisis of 2007-08, this is an attitude of the purest optimism. Given the debt burden still facing the rich world, the risks of policy failure are enormous.