​The simultaneous sell-off in global government bond yields in January 2018 provided an opportunity to inject, in greater size, a core theme of ‘divergent economies and monetary policies’ into our portfolios. Specifically, we increased duration in our strategic bond portfolios via 10-year government bonds in three developed economies: Australia, Canada and Sweden from early February. In this note we set out the rationale for this positioning and how it may develop over time.

Private debt build up: classic late-cycle behaviour

The common feature of the Australian, Canadian and Swedish economies is how desynchronised they have been from the experience of consumer deleveraging seen in most developed economies in recent years. The resulting private sector debt build-up is stark (see chart 1) and piqued our interest some time ago, as it strongly suggests that these economies are in a late-cycle phase and as such are structurally more fragile to rising interest rates.

The work of economist Steve Keen provided our initial introduction to this theme and links with the work of Richard Koo in Japan on the impact of excessive debt in the private sector in driving booms and busts. Fuelled by explosive growth in their domestic banking systems, the property cycles in these countries appear to have simultaneously peaked in 2017. In each country, the fate of the property market had become a political and financial stability issue resulting in blunt policy tools, such as restricting mortgage regulation and additional property taxes, aimed at slowing or halting its advance.

Other signs of excess come from the banking systems of these economies as Absolute Strategy Research points out. In these three economies, the banking sector has swelled in size and now accounts for approximately 20% of market capitalisation – typically a warning sign in other property market bubbles (see chart 2).

Needless to say, timing multi-year cycles and/or long-term themes is fraught with danger. However, the government bond markets of Australia, Canada and Sweden were attractive not simply because of a property price correction but also based on three other observations:

1.It seemed clear that these economies were likely to suffer first, early and disproportionately, from a global rise in bond yields and interest rates (and hence mortgage rates)

2.The financial stability risks had prompted the regulators and central banks to use macro prudential tools such as mortgage availability to slow the trend, lessening the need for interest rate rises, and

3.These economies have low inflation and their governments are not participating in the kind of late cycle fiscal stimulus seen in the US.

Australia first, others later…

Australia has been the easiest of the three economies to pick for adding duration and we began in late 2016. This economy has been weighed down by below target inflation, stagnant wages and appears to have participated very little in the recent much-vaunted “synchronised global growth” of 2017. The front end of the yield curve has been remarkably stable and hence we have stuck with ‘four years and under’ government bonds throughout. This was vindicated in February when the central bank governor made a speech, in which he was keen to point out the differences in the Australian and US economies, as justification for why they would not blindly follow the US Federal Reserve in tightening monetary policy. At a portfolio level, we added 10-year exposure in February.

… and, in Canada…

The market’s perception of the rate hiking cycle in Canada seemed excessive with another 100 basis points (bp) of rate hikes priced in, in addition to 75bp of actual hikes over six months. Much tightening in financial conditions was already underway via a stronger currency, higher bond yields (mortgage rates) and a new mortgage regulation that came into effect on 1 January 2018.

… finally, Sweden the smallest of these economies

It has the same structural issues of household debt and a rampant property market, coupled with persistently below target inflation and wage inflation. The shape of the yield curve is relatively steep and the global sell-off in government bonds repriced Swedish bonds back to the top end of a yield range that has been in effect since 2015. Hedged back to sterling or the US dollar, this provided a more attractive yield given the interest rate differential.

Duration management: more about emotion than data
We have been on record as saying that, as bond investors, the opportunities around duration management are often driven by the emotion of investors (inflation or deflation hysteria) rather than actual changes in inflation or growth regimes, which are incredibly rare.

In this context we viewed the recent spike in government bond yields, driven by the US, as one such phenomenon in a classically late-cycle environment. Having used US interest rate futures (10- and 30-year) to reduce duration from the first week of January, by early February we shifted to lengthening the overall duration of the portfolios.

We view the likely uptick in US core inflation as a late-cycle cyclical phenomenon, rather than a structural move, and in other economies there are very few signs of any shift. In addition, the best of the global synchronised growth environment is now behind us as lead indicators, such as monetary aggregates, suggest a sharp slowdown in growth may well occur this year.

With that I’ll leave you with a chart of real M1 in the three economies discussed above courtesy of our Chief Economist, Simon Ward. This demonstrates the notable weakening of money supply in Canada and Australia and further strengthens our belief in the ‘divergent economies and monetary policies’ theme. As with all our themes, however, we monitor developments carefully and maintain a pragmatic approach as we seek to manage opportunities in the best interests of our investors.