Join our community of smart investors

What the gilt market knows

The gilt market is pricing in low yields for years to come. But the market is a poor predictor
July 23, 2020

The gilt market doesn’t believe in a V-shaped recovery. We know this simply because longer-term gilt yields are low, which implies that the market expects interest rates to stay low for a long time. And if it expects low rates, this is probably because it anticipates a weak economy.

We can quantify this. To see how, ask why anybody would want to hold a five-year gilt that currently yields minus 0.08 per cent when they could hold a 10-year one that yields 0.14 per cent. The answer is that they expect interest rates to rise, so that when the five-year gilt matures in 2025 they can reinvest their money at a better rate. Two consecutive five-year bonds should therefore have the same total return over 10 years as a single 10-year bond. (And for that matter the same as five two-year bonds or 10 one-year bonds and so on.)

This being the case, we can derive the market’s expectations for yields in five years’ time from current yields. Doing so tells us that the market reckons that five-year gilts will yield just over 0.2 per cent in 2025. Yes, this is a rise from current levels. But not by much. The market expects savers to face nugatory returns on safe assets for years to come. Which suggests it foresees a fragile economy for a long time.

But what does the market really know about the future? The answer is: not much.

My chart shows the point. It shows the actual five-year yield minus the yield that the market expected five years earlier – the latter being derived from 10- and five-year yields. So, for example, the current five-year yield is slightly negative, but in July 2015 the market was pricing in a yield today of 2.9 per cent, giving us a gap of almost three percentage points.

If the gilt market had always forecast future yields perfectly accurately, my chart would show simply a horizontal line at zero.

But it doesn’t. In fact, it was only in the mid-2000s that yields were for any length of time what the market had predicted five years previously.

Instead, for most of the time since the 1980s, yields have actually been below market expectations five years earlier: the main exception to this being that in the mid-1980s the market did not foresee the inflation and monetary tightening of the late 1980s.

Of course, there are good reasons why yields should occasionally be lower than expected. The market did not foresee the depth of the 2008-09 crisis or this year’s pandemic. Both caused yields to fall below expectations.

But this isn’t the whole story. For most of the past 30 years yields have been below expectations even in normal times. The market hasn’t just failed to foresee recessions, therefore. It has also failed to anticipate the strength of the long-term forces depressing yields, such as the savings glut (or investment dearth) and secular stagnation.

Which raises a possibility. Perhaps gilt investors make the same mistake equity investors do. I’m thinking here of the recency bias, the tendency to overestimate the durability of current events and therefore to underestimate future changes. One fact is consistent with this. It is that actual changes in yields over a five-year period have been greater than predicted ones. Since 1970 the five-year change in five-year yields has averaged 1.8 percentage points (ignoring the direction of change), but the average predicted change was only 1.1 percentage points.

We tend to think of gilt investors are being more serious and rational than equity investors – and especially more so than retail investors: most studies in behavioural finance look at stock rather than bond markets. But perhaps this is wrong, and gilt investors are as prone to systematic error as the rest of us.

You might wonder how all this can be reconciled with the fact that the yield curve has historically been a great predictor of recessions. 

Simple. For one thing, the yield curve works as a forecaster at a horizon of up to around three years, whereas I’m considering a longer horizon. And for another, the curve tells us more about the probability of recession than its depth. The curve was hugely inverted before the very mild 1991 recession, for example, but only slightly inverted before the much deeper 2009 recession. 

All this might seem very worrying. If the gilt market’s habit of over-predicting yields continues then yields in coming years will be even lower than it expects. That’s not just terrible for savers. It betokens a desperately weak economy and hence tough times for workers, business owners and equity investors, too.

I’m not sure, though, that we need draw quite so gloomy an inference. Perhaps the point is instead merely that the market doesn’t tell us very much. We know less about the future than we think. And in today’s climate, this alone might be good news.