At the time of writing (10 May), the Global Aggregate fixed income benchmark has returned -12.1% year to date, with many underlying fixed income markets cheapening by more. Neither did duration-hedged asset classes prove to be the panacea to Federal Reserve policy tightening that many had claimed, with EUR investment grade excess returns (that is, performance versus governments) year to date of -3.25% and EUR high yield -5.78%. Evidently, monetary policy tightening cycles that are rapid, that are not expected by the consensus and moreover involve balance sheet tightening, are generally less supportive for spread products than periods of low interest rate volatility and quantitative easing. Government bonds, despite their historically negative performance, are therefore among the least worst-performing sectors across the fixed income landscape. Still, these things are all relative.
But has the historic amount of cheapening reinjected value yet? To get fixed income duration right over time, you have to be right on two factors: the secular regime and the (economic and monetary policy) cycle. Many investors who underestimated the secular regime became too short duration in the period 2010-2020, particularly in Europe. On the other hand, many that only focused on the benign secular regime since the early 1980s got caught being long duration in the 1994 and 1999 hiking cycles. Both clearly matter.
Today, there are well-articulated divergent views in the marketplace on the secular question, and the outcome of the debate is far from settled: will the demographic factors and internationalized capital flows prevalent in the past few decades since the 1980s continue to offer a benign tail wind? Or might we be entering a more challenging inflation regime akin to the path from 1965-1981? With bond yields close to the top of their 40-year downward-trending channel – and close to the 2018 yield peaks of last cycle – the question is pertinent, but it may pay to keep an open mind. For the second question, on the cyclical outlook, we think the answer is starting to become clearer.
As Larry Summers and Alex Domash have argued, the chance of a recession over the next two years is arguably closer to likely than unlikely, given initial conditions of late-cycle unemployment (now at 3.6%) and the highest headline CPI in over forty years (currently running at 8.5%). This combination is usually hazardous for economic expansions over a two-year horizon, because of the inflationary pressure more often associated with very low unemployment rate environments, and the monetary policy tightening required given high inflation. As Milton Friedman and Anna Schwartz wrote in 1963, it is Federal Reserve tightening that typically ends US economic cycles. In the sixty years since, that hasn’t materially changed. Indeed, the Bank of England seems to side with Summers and Domash on their recent negative growth forecast for 2023.
A substantial number of hikes are now priced into the front ends of many global government bond curves. Some markets have nearly 250 bps priced in, if we include hikes just delivered (the US, Australia, New Zealand and Hungary for instance), some over 150 to 200 bps (South Korea, Norway, Sweden and even the Eurozone). The inclusion of the latter on this list is remarkable, given the ECB have not managed to hike rates once since 2011, and they are still not yet finished wrapping up quantitative easing.
With a substantial amount priced in, the million-dollar question (or the USD 63 trillion one, given that is the size of the Global Aggregate benchmark) is: at what point do higher borrowing costs create maximum pain for the economy? After all, if Summers and Domash are right, a recession by the end of 2024 needs to be considered. Caveats related to the uncertainty of secular regimes aside, we are focused on four factors:
The performance of front-end yields when yield curves invert and policy rates peak
The relationship between peaks in year-on-year inflation and peaks in bond yields
The debt structure of economies, particularly given their rising ‘interest rate intensity’
The performance of bond markets after the second-last hike of each cycle.
On the first, with the US 2s10s nominal spot curve having briefly inverted in early April, our analysis shows that 2-year yields are much more likely than not to end up below their 12-month forwards in one year’s time. In other words, a curve inversion is a medium to longer-term signal to buy the front end. A word of caution: inversions can be deeper and more persistent in periods of high inflation, as the 1970s showed, so there can be a cost to being too early.
Second, while inflation has continued to surprise the consensus by peaking later than forecast, and at higher levels, base effects suggest some moderation into the second half of this year, at least in the US. If the base effects overpower the evident upside pressures of wages and second-round effects, this metric suggests a potential yield peak some time this summer.
Third, most developed economies have become even more indebted in recent years. In 1970, US household and corporate debt each totaled around 45% of GDP. Today they are each around 80% of GDP. Government debt was also just below 40% of GDP in 1970 – it has now tripled to 120%! This matters for the calculus of when Federal Reserve hiking cycles come to an end, because the amount of interest-bearing debt that requires servicing is now substantially higher. In other words, the interest rate intensity of the economy has gone up. All else equal, that means the terminal rate will be substantially lower, given the higher level of debt.
Finally, as a rule of thumb, US bond yields tend to peak close to the second-to-last rate hike each cycle. Of course we only know when the second-last hike occurs after we’ve had the last one, and after the pause or first cuts that then follow! The ex ante uncertainty clouds the current view, but the interplay of the growth outlook with any progress on the inflation front, is likely to be key.
Because timing turns can be hazardous, in the meantime cross-market trades offer better risk-adjusted opportunities in our view. Markets where central banks have already tightened significantly relative to terminal rates in prior cycles, and have already inverted their forward (and in some cases their spot) curves, have less scope to sell off at the front end. For yield curve strategy, while it still looks too early to put on steepeners in many markets, the first step to getting there is to take profits on flatteners.
So as questions over the market cycle suggest a turn at an indeterminate time in due course, there’s alpha to be made in the meantime. After all, it’s all relative.