I say this because there is a possible scenario in which gilt yields fall even lower than they are now. The fact that yields on 10-year Swiss government bonds are now negative warns us that even long-term yields can fall below zero. This could happen if Bank rate turns negative and/or if a recession or new financial crisis triggers even greater demand for 'safe' assets: Bank of England chief economist Andy Haldane recently warned that 0.5 per cent is not the lower limit for Bank rate.
In such circumstances, index-linked yields could become very negative indeed. Some economists believe one possible solution to sustained low growth, or to the next recession, would be to raise the inflation target, say from 2 to 4 per cent. Doing so would require a big loosening of monetary policy - some combination of quantitative easing (QE) and a commitment to keep short rates at or around zero for a long time. That in itself would stimulate the economy. So, too, some economists believe, would the prospect of higher future inflation, as this would encourage customers to buy before prices rise.
Were this to happen, index-linked yields could fall from their current minus 1 per cent to minus 4 per cent: zeroish nominal yields because of QE and zero short-term rates, minus expected inflation of 4 per cent.
Of course, all this is only a risk. Another possibility is that a continued global economic upturn (which the eurozone seems to be joining) would eventually return yields to more normal levels. If there were no fears of secular stagnation or of a global savings glut, we'd ordinarily expect nominal yields to be around 4 per cent, a real yield of 2 per cent (roughly trend long-term real GDP growth) plus expected inflation of 2 per cent, the current target.
We should think of current yields as being the probability-weighted average of these two scenarios. An index-linked yield of minus 1 per cent represents a 50 per cent chance of minus 4 per cent yields, and a 50 per cent chance of a return to more normal 2 per cent yields.
Of course, this is a simplification, not least because there are countless intermediate scenarios. But it helps explain two facts.
One is why gilt yields have stayed low despite the economic recovery. It's because there is still a risk of a scenario in which yields become very low indeed.
The second thing it explains is the increased volatility of gilt returns. Small and reasonable changes in the probabilities investors attach to different scenarios can generate big changes in prices; this is why high equity volatility can be rational. And gilt volatility has been trending upwards for a long time. If we measure it by the average absolute weekly change in total returns over a 12-month period, it has recently been at its highest since the financial crisis of 2008, and double what it was in the early 2000s. This process is exacerbated by the mathematical fact that lower yields mean higher duration: a given change in yields means a bigger change in prices than it does when yields are high.
This higher volatility isn't necessarily a bad thing. Yes, it means bond holders could lose a lot if the fear of recession or secular stagnation were to recede. But in such circumstances, shares would probably rise - so equities hedge us against possible gilt losses, and vice versa.
What it does mean, though, is that low yields are not proof of a bond bubble. They might instead be just reasonable bets on the possibility of a scenario in which yields go much lower.
