What is now commonly dubbed the ‘US-China trade war’ is a series of tariff-imposition threats and actions started by the US on the initiative of President Trump against China at the beginning of 2018 and responded to by way of similar retaliatory measures by China. The sequence of events is seen as one reason for what looks like a global economic downturn, not only by business leaders and the press but also by major institutions with significant academic backing such as the IMF. In any case, this trade war is drawing significant attention which, inter alia, shows in more than 20 million Google search hits by October 2019.
At this point, sound statistical and holistic economic analysis of the consequences of this trade dispute between two of the largest economies on the globe is difficult. The reason is that few countries provide revised data on all the relevant economic outcomes that this dispute may have affected. Systematic analyses of the effects of the events are currently mostly available for the US, where researchers have documented that the terms of trade of the US have not improved and consumer prices have risen on average so that US consumers bear a cost of the US-China trade war (Amiti et al. 2019). Moreover, the county constituencies with a strong Republican majority have been most negatively affected by the trade war (Fajgelbaum et al. 2019).
In a recent study (Egger and Zhu 2019), we scrutinised stock market responses all over the world in assessing the trade war events. Stock-market data have two particular merits for analysing effects related to the US-China trade war: (i) they are available on a daily basis, and (ii) they are available for very recent episodes.
The availability of daily data is important because using quarterly or yearly data aggregates bears the risk of misattributing effects to the trade war that actually originated in pre-trade war shocks. As the trade war measures of interest were implemented on various days during the 2018 and 2019 calendar years, stock market data allow us to analyse responses immediately after new events (threats or actual actions taken). The fact that stock market data are not administrative data means that they are almost immediately available and cover very recent time spans, and that they are not ‘filtered’ or adjusted by any statistical authority prior to their release. Hence, they reflect immediate market responses to news.
We use the data to first determine the standard market expectations within a period of about one and a half years (250 trading days) prior to 31 December 2017. We then predict market outcome for each stock between March 2018 and May 2019, during which period we focus on 19 US-China trade war events. In line with the related literature, we dub as ‘abnormal returns’ any difference between observed and predicted stock-market prices around the event dates of interest.
We accumulate these abnormal returns, from the period of one day before a tariff-inception announcement event by the US or China until up to ten days after each event, into ‘cumulative abnormal returns’.
Finally, we determine by how much a 1% tariff increase changes cumulative abnormal returns
- in the same sector at home (the US or China),
- in the same sector abroad (the US or China), and
- in other sectors at home, abroad, or in 38 ‘third countries and territories’ (other major economies on the globe).
We refer to the first two of these as ‘direct’ effects and the last type as ‘indirect’.
Regarding indirect effects, we use global input-output tables to establish links between countries and sectors for the following reason. Placing a tariff on the products of a sector is equivalent to raising the customer price for these products. With the cross-border organisation of value chains, the business activities of virtually all sectors and countries are linked to each other (Baldwin and Lopez-Gonzalez 2015). This means that tariff-related price increases from the trade war will travel to other sectors and countries through value chains and will have ripple effects on the same sectors abroad as well as in other sectors at home and abroad.
Based on our analysis, we arrived at three observations:
- Observation 1: On average, the trade war tariffs of the US and China directly hurt targeted firms/sectors abroad as intended (i.e. US tariffs hurt Chinese firms and vice versa) but they also hurt firms at home (i.e. US tariffs hurt US firms in the same sector and similarly for China).
- Observation 2: The actions of both the US and China indirectly affect stock prices through global value chain linkages in the US, China, and in third economies which do not directly participate in the trade war. Such indirect effects can be positive or negative, depending on a sector’s and economy’s position in the global value chain.
- Observation 3: On average, the direct effects on US firms are the largest, and the indirect effects induced by US tariff changes (through global value chain relationships) are much larger than those of China.
These three observations are demonstrated in the following two figures, which show the overall effects of the tariff changes (Figure 1 for direct effects and Figure 2 for indirect effects). They indicate that the scale of the direct effects tends to be much larger than that of the indirect effects.
The direct effects vary between about -4% and +9%. The effect of the US on US firms tends to be more skewed towards the positive and that of the US on Chinese firms more to the negative. However, there is a significant amount of unintended negative effects of the US on US firms and positive ones of the US on Chinese firms. Also, China’s retaliatory tariffs apparently had unintended effects at a nontrivial frequency.
Figure 1 Estimated direct trade war tariff effects on firms’ stock market prices (%)
Notes: The panels show the overall direct effects of tariff changes, in percent, defined as the coefficient estimates per sector multiplied by the tariff changes by sector. These represent density estimates. A higher elevation suggests that a larger share of sector effects occurs in the neighbourhood of percentage effects recorded on the horizontal axis.
Not surprisingly, the indirect effects are much smaller than the direct ones (about 1/20th the size). They are largely skewed towards the negative, and they tend to be larger in the extremes for tariffs set by the US than for those set by China.
Figure 2 Estimated indirect trade war tariff effects on firms’ stock market prices (%)
Notes: The figure shows the overall indirect effects of tariff changes, in percent, defined as the coefficient estimates per sector multiplied by the tariff changes by sector. These represent density estimates. A higher elevation suggests that a larger share of sector effects occurs in the neighbourhood of percentage effects recorded on the horizontal axis.
Overall, our analysis shows that there are unintended effects of the trade war on the US and China, as well as on third parties, mediated by global value chain interdependencies. The results suggest that there is an irony to the US-China trade war in that it appears to have resulted partly in the exact opposite of what was intended.
Firms in the modern world are organised in complex ways across the boundaries of both sectors and economies. Instituting well-targeted protectionist tariffs in such a world is not easy and eventually, as in the example of the US-China trade war, hurts those they are meant to protect.
As with Brexit, investors are much more concerned about the economic dismay that recent deliberalisation attempts and protectionist policy announcements may trigger than the politicians proposing them appear to be.
Amiti, M, S J Redding and D Weinstein (2019), “The impact of the 2018 trade war on US prices and welfare”, NBER Working Paper 25672.
Baldwin, R, and J Lopez-Gonzalez (2015), “Supply-chain trade: A portrait of global patterns and several testable hypotheses”, The World Economy 38(11): 1682–721.
Egger, P H, and J Zhu (2019), “The US-China ‘trade war’: An event study of stock-market responses”, working paper.
Fajgelbaum, P, P K Goldberg, P J Kennedy and A K Khandelwal (2019), “The return to protectionism”, NBER Working Paper 25638.