Once again, market chat is dominated by two subjects: Brexit and the threat of a US recession. Even now that we are approaching the re-set deadlines for Parliament to come up with a solution, it is not at all clear which of these is going to happen first. No doubt many are hoping that neither will ever occur. That’s certainly possible for the former event, but not for the second.
Even though economic cycles can rumble on for a long time – Australia hasn’t had a recession since 1991, for example – they certainly haven’t been abolished. The sheer longevity of the current US cycle, which was born out of the depths of the financial crisis in 2009, is making many investors nervous, especially as it promises to become the longest ever this summer.
One key early warning sign of a recession is when the yield curve inverts, and that has happened recently in the US. A yield curve inversion occurs when the yield on longer-dated government bonds (normally 10 year maturities) falls below the yield on shorter dated instruments (3-month Treasury bills or 2-year Treasury bonds gain the most attention).
An inverted yield curve implies that current interest rates are too high for the economy to bear and that cuts are ahead.
In normal circumstances investors demand a higher yield (or term premium) for bonds with a longer maturity to compensate for the greater uncertainty of investing with a more distant time horizon. An inverted yield curve implies that current interest rates are too high for the economy to bear and that cuts are ahead. That message is backed up in the US interest rate futures market, which is now pricing in a high probability of a quarter-point cut by the Federal Reserve later this year.
Even so, nothing is crystal clear. First of all, the yield curve is not totally inverted. The trough is at the 3-year maturity, where the bond yield is 2.25% relative to the 3-month bill yield of 2.39% and the 10-year yield of 2.44%.
It is also interesting that until a few months ago the focus was almost entirely on the 2-year/10-year part of the curve, but that steadfastly refused to invert, bottoming out at 11 basis points on December 19th last year. It currently stands at 14 basis points.
When the 3-month/10-year curve did invert on March 22nd, it was a much better story for the scaremongers to latch onto.
So, time to panic? Well, not necessarily. These sorts of indicators are notoriously open to interpretation. I have a range of notes from various investment bank strategists informing me of what might happen next. First, a little history. The correlation between the yield curve and the future risk of recession was first noted in the 1960s, but only entered mainstream financial analysis in the 1980s, and so it has only a relatively short record, and perhaps only during a certain type of economic environment and interest rate regime.
The yield curve canary fails to send any recession warning, it appears, in a structurally low inflation and interest rate environment.
It has been noted, for example, that the relationship between the Japanese yield curve and the economy broke down in the 1990s. This is important if Japan’s ageing, shrinking population provides a template for the future of other maturing economies. The yield curve canary fails to send any recession warning, it appears, in a structurally low inflation and interest rate environment. That said, the US is not Japan – it at least has positive interest rates and inflation is trending higher - and so there could still be some utility.
Another confusing factor is that any potential recession could be as far as two years away, which makes market timing even more difficult. One of the studies points out that following the last three instances of yield curve inversion (1989, 1998 and 2006), US equities rose on average a remarkable 32% on average. Even stripping out the Tech Boom (1998-2000) delivers 19% upside on average before peaking and entering a bear market. Thus to bail out of equities now could be to leave substantial gains on the table.
There are a number of other questions. Foremost, are longer dated bond yields being depressed by, for example, the effects of Quantitative Easing and/or the regulatory pressure on banks, insurance companies and pension funds? With German 10-year Bunds now sporting a negative yield again, the US equivalent yielding 2.44% looks relatively attractive, even allowing for currency hedging. And if a recession does unfold, what sort will it be? The Tech Bust and the financial crisis were very specific in nature, reflecting the previous excesses built up in those industries.
With German 10-year Bunds now sporting a negative yield again, the US equivalent yielding 2.44% looks relatively attractive
As I have written before, we do not currently observe similar excesses, and so do not expect to see similar market declines. That’s not to say there is nothing to worry about, but the second order effects of a setback in one industry are less likely to ripple through the whole economy. An example of this could be the sharp downturn in commodity-related sectors in 2015 (primarily Oil and Mining), which failed to trigger wider recessions.
The yield curve cat is certainly out of the bag, though, and so we had better get used to being scratched a few more times. As I have commented previously about the relationship between the High Yield credit spread and equities, there are a lot more computers and algorithms out there programmed to respond to correlating factors. This will make equity markets more prone to sharp short-term adjustments.
We could also invoke Goodhart’s Law at this point, which states that “When a measure becomes a target, it ceases to be a good measure”. Thus, all this talk about yield curves and recessions could actually help to create a recession! Our opinion is that the yield curve is one of a number of indicators that is worthy of monitoring, but it needs to be set in the context of other developments. For now we accept that a US recession is out there, which curtails our appetite for risk; but it is not sufficiently close to warrant excessive caution either.
