An Autumn Bond Selloff?

Summary

• The spread between real yields and real GDP is at its widest ever.
• This reflects mainly stable inflation against expectations of continued large Fed bond purchases.
• The real yield/GDP spread could tighten in autumn as the taper conversation brings home that, with more active fiscal policy, less Fed support is needed to stabilize the economy.
• Germany’s September general election could reset fiscal policy in the euro area and beyond and pressure US nominal and real yields.

Market Implications

• Real and nominal yields to rise.

The Spread Between Yields and Nominal GDP Is Widening

Bond yields tend to track long-term nominal GDP yields (Chart 1). Yields have gradually fallen below GDP growth following the Volker recession of the early 1980s, possibly reflecting a more stable macro environment and lower risk premium.

As well as a stable macro backdrop, the (absolute) decline in the yield/GDP growth spread and the term premium post-crisis reflects a structural increase in demand for bonds. This demand stems mainly from the Fed but also MMFs, MFs and ETFs held mostly by households and institutional investors. By contrast, foreign holdings of Treasuries have remained broadly stable, and banks’ share of Treasury holdings has been too small to make much difference.

Real growth and yields account for most of the widening in the nominal growth/nominal yield spread since nominal yields tend to track inflation closely. Real yields are essentially a residual from the pricing of nominal yields and BEs, as shown for instance by the March 2020 spike. As a residual, they can take longer to discount changes in fundamentals than either BEs or nominal yields.

Real, and Therefore Nominal, Yields Are Mispriced

I believe real and nominal yields are mispriced. 10-year BEs around 2.4% align with the FOMC forecast of end-2021 core PCE around 2.2%, with which I agree. By contrast, even accounting for the wider post-GFC spread, the real yield/real GDP growth spread is too low. For instance, the past five-year average of that spread is about 2%; assuming real growth returns to its long-term trend of 2%, real yields should be about 0%, against currently -84bp. This gap reflects the volatility of nominal yields against stable inflation and BEs.

The downtrend in real rates started with the September 2018 equity selloff. In my view, concerns over decreased Fed support triggered the selloff. By end-2017, earnings growth had peaked (shown by EBITDA, as EPS reflected the impact of the 2017 tax cuts). This implied that further increases in equity prices would have to come from multiples.

Nominal 10-year yields peaked in November. The equity selloff and negative bond stock correlation, concerns over the trade war, and expectations of continued Fed support including asset purchases saw a steady decline from 3.2% in November 2018 to 1.5% in August 2019. Meanwhile, with unemployment sliding further to 3.5%, inflation and therefore BEs stabilized in a range of 1.6% to 1.8%.

Real yield movements are simply the by-product of these developments. Between December 2018 and August 2019, due to nominal yields falling faster than inflation and BEs, real yields fell into negative territory. Real yields recovered alongside nominal yields in Q4 2019, but then the pandemic hit them.

The large 2020 decline in real yields largely reflects the collapse of nominal yields caused by pandemic concerns and extraordinarily large Fed purchases. By contrast, inflation was more stable: core PCE remained above 1.4% except during April-August, and BE bottomed out at 55bp in mid-March. By end-Q3, however, BEs were back to their pre-pandemic levels of 1.75%. More recently, the decline in real yields is again the result of a decline in nominal yields against stable inflation and BEs.

An Autumn Bond Selloff?

Real yields have strayed so far from real GDP growth that they could be ripe for a correction. Over the next six months, two developments could catalyse one.

First, the realization that Fed LSAPs could be lower in future. This would not reflect a move to a higher-inflation regime (though such a move is likely within a few years). Rather, it would reflect a shift in the US policy mix towards greater fiscal activism. At the same time, I share the Fed’s view on benign inflation trends. This is due to a lack of changes in the structural factors driving low inflation, namely an imbalance between the market power of workers and employers.

The second development that could lift nominal and real yields is the September 2021 general elections in Germany. Current polls show the Greens polling higher than the CDU. Should the Greens become the main coalition partners, euro-area fiscal policy would undergo marked changes.