The financial markets have become extremely focused on the twists and turns in the tariff negotiations involving the US and China and, more recently, on the likelihood of a technology war centred around US sanctions against Huawei.
Trade wars first assumed a central role in market psychology in the final quarter of 2018, when a deterioration in US-China trade relations coincided with a collapse in stock markets across the world. In recent weeks, there have been signs that the same might happen again.
Former US Treasury secretary Lawrence Summers has commented that this is a puzzle, because macroeconomic models indicate that the direct economic effects of tariff increases are likely to be rather small and certainly not enough to explain such large gyrations in equity markets.
His explanation is that the markets are not focused solely on tariff effects, but on the rising probability of a big economic and strategic confrontation between the world’s two largest economies, which could eventually extend well beyond trade in manufactured goods. Many of the benefits to global growth that have stemmed from decades of globalisation might be reversed in the worst scenarios.
I have some sympathy with Prof Summers’ suggestion. Markets seem to be reacting to changes in tail risks, rather than central scenarios.
However, there is an alternative explanation, which is that collapsing US markets late last year were not reacting entirely, or even mainly, to trade war shocks, but to other unexpected economic developments. This alternative is supported by the fact that the latest, and most serious, exacerbation of trade wars in early May has been accompanied by a moderate decline of about 6 per cent in US stocks, much smaller, so far, than the meltdown last December.
Measuring trade war uncertainty
Fulcrum economists have calculated a weekly index of tariff and trade war shocks that is derived simply from the number of Google searches for these terms since 2016. Although crude, this measure is appropriate, provided that the index is correlated with the true (unobservable) measure of trade uncertainty (see box).
There have been five large spikes in trade uncertainty since March 2018. Surprisingly, our initial econometric tests have failed to support any systematic relationship between the trade uncertainty index and changes in market prices. This suggests that investors may be paying too much attention to trade wars, but we need to do more work to be certain about that. Watch this space.
The deterioration in US/China relations in May has caused the largest spike so far in the trade uncertainty index. This makes sense, since the latest episode has involved threats of much larger US tariffs on Chinese and Mexican imports than seen last year, and also a widening in the conflict to include direct sanctions against Huawei, the largest Chinese IT supplier.
In response, Chinese negotiators appear to have adopted a more hawkish tone than before, including veiled threats about sales of US Treasuries and restrictions on the supply of rare earth minerals to the US.
Trade v other shocks
Another Fulcrum model (explained here) enables us to examine the shocks that have been driving the S&P 500 index since last October. The key point is that the US market plunge in 2019 Q4 was driven by a combination of several different economic shocks, while the trade event of May 2019 has not been accompanied by other adverse shocks.
The model identifies major negative shocks in the fourth quarter of 2018 from US domestic demand, foreign demand, risk aversion and, importantly, a tightening in monetary policy. Some of these variables may have been partly connected to trade war shocks but this is very difficult to prove.
The most important variable — tighter monetary policy — occurred because the US Federal Reserve persisted with hawkish guidance about increases in interest rates, despite deteriorating economic conditions, and falling markets.
The FOMC seems to have learnt a painful lesson from this error of judgment and communication, and has quickly softened its forward guidance on its policy stance as its perceptions of downside risks have changed.
There are three conclusions to draw from this:
- Markets are reacting to changing tail risks of extreme outcomes not to changes in central forecasts. The reaction of markets to trade wars is not consistent through time.
- The latest episode may have increased tail risks more than last year’s outbreaks of trade wars, but developed markets have reacted less than before.
- This is mainly because there has been no negative monetary shock from the Federal Reserve this time.
Provided the Fed remains dovish, developed markets should weather the trade storm better than they did last year.
Trade wars and financial markets — some model results
The index of trade uncertainty calculated at Fulcrum from Google search data indicates that the latest shock, since President Donald Trump’s unexpectedly hawkish tweet on May 4, has been larger than the four preceding trade shocks in 2018:
Although we have failed to find any clear statistical relationship between the trade uncertainty index and any global equity market, there have clearly been some occasions where the connection seems obvious. This includes the latest event, in which the response in Chinese equities has been much larger than in US equities:
Fulcrum also has a model that allocates US market movements to underlying economic “shocks”. This model suggests that the big market sell-off in the fourth quarter of 2018 can be attributed to a series of major shocks all working in a negative direction at the same time.
In the latest outbreak of trade concerns, these shocks have been much smaller, or entirely absent. This may explain the smaller apparent reaction of US equities, so far, to recent bad news on trade: