At this time of year, it is traditional for market economists to release detailed macro forecasts for the 12 months ahead. I am alarmed to find that I am now completing my fifth decade of performing this difficult exercise, with much less than perfect success so far. Important economic and market dislocations do not cram themselves neatly into a calendar year.
This time, I will turn to Fulcrum’s suite of econometric models, built in the past decade, for guidance on possible economic and market scenarios for next year and beyond. It is useful to know what the models — or algorithms — are saying, if only to know where measurable “facts” end and human judgment begins.
Where global economic activity is now
The main feature of 2019 has been weakening global activity, especially in the manufacturing sector of the advanced economies, with consensus gross domestic product growth forecasts for the calendar year being repeatedly downgraded. The eurozone and UK have been particularly affected and have only narrowly avoided technical recessions (defined as two quarters of contraction).
Pessimism about Chinese growth has been rife throughout the year, but an easing in the country’s domestic policy has cushioned the blow from the trade wars, and growth has remained close to target.
Since mid-2019, global manufacturing surveys have started to recover, but service sectors have still weakened. Overall, business conditions have stabilised but not yet decisively improved. The big question is the direction of the next change in growth rates.
The Fulcrum nowcast models — designed to assess activity in exactly this type of uncertain situation — show annual global growth is still hovering at about 3-3.5 per cent, about half a percentage point below trend.
The latest projections from these models show the US gross domestic product growing at a disappointing annualised rate of 1.5-2 per cent in 2019 Q4 and 2020 Q1. China is still defying the pessimists with healthy 6-6.5 per cent growth. And eurozone growth has recently stabilised after a worrying summer, though it is still growing at only around 0.3-0.6 per cent.
The outlook from the models shows global growth rates rising next year, returning roughly to trend rates. Recession risks are deemed to be low, currently standing about 5 per cent for the US and 15 per cent for the eurozone.
Alternative models, such as those from Jan Hatzius’ team at Goldman Sachs, include other variables, including the yield curve, credit spreads and private-sector financial imbalances. The bank’s model for the US suggests a recession risk of 24 per cent in the US next year — significantly higher than suggested by the nowcast model but still fairly low.
All the major advanced economies have had a further extension of low inflation in 2019. Consensus forecasts and expectations from the inflation swaps markets have dropped almost 0.5 per cent since late 2018.
Inflation snapshots from the models continue to report underlying inflation running stubbornly below target in the eurozone (at 1.5 per cent) and Japan (at 0.4 per cent). In the US, where capacity utilisation in the economy is extremely high by historic standards, the same snapshot has nevertheless fallen slightly in recent months and is now at only 1.6 per cent.
The Phillips curve relationship between declining unemployment and rising price inflation therefore remains inoperative in the advanced economies, probably because the credibility of central bank inflation targets remains high. This implies that unemployment can fall significantly without causing any immediate rise in reported inflation.
Forecasts from the Fulcrum models clearly indicate that core and headline price inflation in the advanced economies will remain well below target in 2020. However, wage inflation is beginning to rise in the US and Europe. This suggests profit margins in these economies will fall, with corporate earnings growth coming under intensified downward pressure.
Asset return forecasts
The Fulcrum outlook for US equity returns is much lower than the historic real return of 11 per cent per annum in the past decade, and below those expected in the eurozone and emerging markets.
The main features of the forecasts, for dollar-based returns, are as follows:
- US equity returns are expected to be below average at 3.9 per cent a year in nominal terms, or 2.5 per cent in inflation-adjusted terms, in the next three years.
- Although still positive, these expected returns are lower than any previous projection from the model since 2007, reflecting relatively expensive equity valuations and expectations of slowing growth in US corporate dividend payments.
- The dollar-based returns in eurozone and emerging market equities are higher than those in the US, reflecting higher dividend yields and higher expected dividend growth.
- Government 10-year bond returns are expected to be extremely low or negative over the next one to three years. Long-term yields are now abnormally low relative to policy rates. Over the forecast horizon, these “term premia” are predicted to normalise, taking long-term bond yields higher.
- The dollar is overvalued, according to the models, and is expected to fall by 2.6 per cent per annum over the next three years against the euro.
In summary, the models expect recovering GDP growth, extremely low inflation, exceptionally low bond returns, and low but positive equity returns in 2020-22. If the models are wrong, please do not shoot the messenger!
Key model results for the global economy and asset markets
Global activity has rebounded slightly in the past couple of months, but it still remains below trend:
Underlying price inflation in the advanced economies shows no sign of picking up:
The models for one-year, three-year and seven-year asset returns are explained in this technical paper, with the latest results for dollar-, sterling- and euro-based investors shown in its tables. Real bond returns may be generally very low and outpaced by real equity returns, although these may also be below average: