A ghost is haunting bond markets – that of 1994. Back then, the Federal Reserve raised interest rates, having cut them in response to a financial crisis. And although that move was widely expected, bond yields rose sharply. Ever since then, markets have been jumpy about the possibility of tighter monetary policy – as we saw, for example, in the “taper tantrum” of 2013 when bonds sold off in fear of the Fed ending quantitative easing.
We’re seeing a similar thing now: 10-year US Treasury yields have risen to 2.3 per cent, their highest level since March, and their UK counterparts have risen to the highest level since February.
One reason for this rise is that the Fed plans to begin reversing QE later this year – what it calls “balance sheet normalisation”. In effect, the proceeds from maturing bonds and interest payments won’t be fully reinvested into the bond market.
Just as QE reduced bond yields, so reverse QE should raise them. Because government bonds in developed countries are close substitutes for each other, any rise in US yields should mean rises in UK and European yields too.
By how much? Fed economists estimate that QE cut 10-year Treasury yields by 1 percentage point. A full reversal of QE should, therefore, raise them by this amount. The actual impact on yields, however, is likely to be less than this. For one thing, nobody expects a full reversal of the $3.6 trillion rise in the Fed’s balance sheet between 2008 and 2016. And for another, reversing QE is to some extent an alternative to raising the Fed funds rate: more of one entails less of the other. The damage done to bonds from reverse QE should therefore be mitigated by the path of interest rates being lower than it otherwise would be.
Gilts, however, have a second problem – that fiscal austerity might well be relaxed. This would raise yields not because markets would take fright at increased government debt, but simply because, for a given inflation target, looser fiscal policy means higher interest rates. Given that fiscal multipliers might be big and the response of output and inflation to interest rates is small, this might mean a significant rise in rates.
Because gilt yields should be equal to the expected path of short-term rates over the lifetime of the bond, this means higher yields. It's no accident that the sell-off in gilts has come at the same time as markets have started to price in higher rates. Futures markets now expect three-month interbank rates to be 0.9 per cent at the end of next year: a month ago, they were pricing in a rate of just 0.5 per cent.
There are, therefore, reasons to believe yields might rise. But not perhaps by much.
History tells us that yields don’t move much in a short time. Since 1986 the standard deviation of annual changes in 10-year gilt yields has been 0.9 percentage points. This implies that there’s only around a one-in-six chance of them being above 2.2 per cent this time next year. While this would entail heavy losses (because prices are very sensitive to moves in yields when yields are low) it would still leave yields extraordinarily low by historic standards.
What’s more, many of the factors that have caused negative real yields are still in place. Long-term economic growth in the west looks like remaining lower than it was before the financial crisis – not least because labour productivity growth is still low (and non-existent in the UK.) Companies are still reluctant to invest; in the UK, non-financial companies have been net savers for the past 15 years. And there’s not much sign of a diminution in the global savings glut or in the fundamental causes of the worldwide demand for safe assets.
In fact, one recent development should point to yields staying low. In both the UK and US, wage inflation has been surprisingly low despite falling unemployment: Phillips curves are flatter than central bankers expected them to be a few months ago. While this remains the case, there’s less danger of sustained inflation than we thought a few months ago. This should keep a lid on interest rate expectations and hence gilt yields.
So yes, it’s possible that gilt yields will rise. But it’s hard, for now, to see them rising very much. It’s not 1994.
