Preparing for the next recession
There are two certainties in economics. The first is that economic cycles exist: recessions eventually happen. The second is that economists are terrible at forecasting them. A recent study found that the IMF was able to forecast 5 out of nearly 470 past economic downturns.
There are many reasons for this shoddy track record. It’s in part because economic data are never exact, and typically arrive with a lag. In part, perhaps, it is because humans (economists included!) are hard-wired to look at the bright side of things. In part, it is because the task is very difficult. Many analysts develop elaborate models to forecast economic performance, inspired by the methodologies of sciences like physics. But as one economist, Friedrich von Hayek, whose work profoundly influenced my own thinking, famously put it: "The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design." The problem is that economics is not an exact science. Economic activity is the aggregate outcome of individual human behaviour.
I’m now expecting that the US economy will head into a recession next year. There’s a good chance that I’m wrong. Here’s why:
A segment of the US yield curve has been flattening for several months and has inverted between 2 and 5 years. Many would argue that yield curve inversion typically presages a recession. However, the relationship between the shape of the yield curve and growth isn’t one with solid foundations in economic theory. This thesis was first posited by a finance professor, Campbell Harvey in the 1980s, based on observations of past cyclical behaviour.
Many analysts will make heroic attempts at pinning down the predictive power of the yield curve by analysing the time lags between inversion and the start of a recession. They would argue whether what matters is the difference between the 3 year and 5 year yield, or 2 year and 10 year yield. This is missing the forest for the trees. The relationship between the yield curve and economic activity seems to be only robust in the US. It has shown no meaningful predictive power in Germany, Japan, and many other countries.
I will not attempt to be precise with the US curve. Rather, I would argue that rational economic agents respond to economic incentives. When shorter term yields are higher than longer term yields, the incentive structure for savers and investors is tilted towards taking shorter duration risks. This disincentivises longer-term investments which in turn can take a toll on growth. My argument, therefore, is not that the yield curve “predicts” a recession, but that it can be a factor in dampening growth.
US economic data have not turned sour, but they do not paint a uniformly positive picture either. The tax cuts enacted last year did not spur investment growth as expected. Rather, the earnings windfalls were spent by companies on share buybacks. Consumers are benefiting from modest wage growth but are squeezed elsewhere. The average interest rate on credit card debt is at its highest level since the 1990s. The demand for new car and consumer loans is shrinking. While interest payments as a share of household income is not especially high, it is increasing at an alarming annual rate of 15%. As a result, delinquency rates on many consumer loan segments are rising.
An important question to my mind is not whether a recession will happen in early or late 2020, or even 2021, but how long will it last? And what are the likely policy responses? Unlike most previous recessions, the next recession will happen at a time when the fiscal deficit is already very high. The fiscal levers – increasing government spending and/or reducing taxes – that tend to be favoured by politicians in every downturn are constrained by the fact that the US fiscal deficit is already about 5% of GDP.
The room for monetary policy is also constrained. The Fed has paused raising rates at a level that is considered “neutral”. In other words, unlike previous recessions, when monetary policy was considered “tight” and there was a lot of room for rate cuts, the rate cuts that can now be expected are much more limited.
What are the likely policy responses of policy makers in such circumstances? The most powerful policy tools will have to be structural in nature. Reasonable responses will include deregulation. But such measures take long to formulate and for the benefits to become apparent. Such measures also require sufficient legislative support, and it is not clear whether such support will be available given the current political climate. These limitations imply that the next recession may not be deep but may be long.
While the US domestic economic cycle is well past its’ peak, it is difficult to find drivers for growth emanating from the rest of the world. Parts of Europe are in recession (Italy), while others are on the verge of recession (UK). Germany’s economy is spluttering amidst the slowdown in global trade. China’s economic slowdown is expected to continue, impacting much of the trade-sensitive economies of Asia, such as Japan and Korea.
What does this imply for investors? As asset allocators we must expect that earnings growth, which is moderating this year, will decline further next year. This implies that US equities, which have already seen a strong performance this year are beginning to look far less attractive than they did earlier this year when we increased allocations. Equities in the rest of the world have relatively low valuations, but absent a catalyst for earnings growth, they are not particularly attractive either.
Within fixed income, the outlook for earnings, early signs of increasing delinquency rates at the consumer level, and the historically low level of credit spreads leave us uncomfortable with credit risk. High yield bonds are unattractive. The lowest tier of investment grade bonds (BBB) are also risky given the potential for downgrade to junk.
The evolution of the yield curve will be an important driver of returns. There’s a strong case to be made that heightened growth concerns argue for increasing positions in long term treasury bonds and high-quality credits. If the next move by the Fed is to reduce short term interest rates, short term fixed income investment will not have a positive inflation-adjusted return.
Having said all of, I’m just an economist and the track record of my profession suggests I could be very wrong. I would rather limit exposure to the next market downturn than miss some modest market upside.
