Stress Scenario - Tariffs and US-China trade conflict

In recent weeks, the US administration has implemented worldwide tariffs that have significantly impacted global trade dynamics. On April 2^nd, so called reciprocal tariffs were introduced, including:

• 10% universal tariff,
• 20% tariff on EU goods,
• 34% tariff on Chinese goods

In response, China reciprocated tariffs on US products and implemented export controls, sparking the trade conflict. Markets reacted negatively, with US treasury bonds experiencing a significant sell-off as investor concerns over escalating tensions grew. To facilitate trading negotiations, President Trump announced a 90-day pause on non-Chinese tariffs, although the universal tariff has remained in place. Tariffs on Chinese products were raised to 145% while China increased its retaliatory tariffs on US imports to 125%. Even though a trade truce was reached in May and the tariffs were dropped accordingly to 30% and 10%, the situation remains tense. Both Washington and Beijing argue that the agreement was violated by not removing the non-tariff barriers as agreed under the deal and threaten with retaliatory actions.

We are cautiously monitoring how these tariffs and in particular the US-China trade dispute could affect the global economy. The S&P economic research team has revised the growth forecast downward, noting that the introduced import taxes have negatively affected business confidence and investment. In advent of Liberation Day our economists published an updated macroeconomic outlook (Fig. 1) on which we have based our scenario considerations.

Design of the Stress Scenario

Our scenario modelling process begins with establishing a coherent narrative for the macroeconomic forecast.  With this narrative in mind, we try to identify the most plausible economic picture out of a range of variations supplied by S&P Global’s economic research. Subsequently, we focus on the change in economic growth linked to the narrative-consistent prediction, as shown in Fig. 2 for the Eurozone. 

Using the macroeconomic data as a foundation and considering recent developments in the tariff conflict, we propose two stress-test scenario assumptions:

1. Scenario 1 – Trade deal
   • Large economies agree on a trade deal with the US
   • Universal tariff rate of 10%
   • Reduction in exports to the US
   • Stifled economic growth in the Eurozone
2. Scenario 2 – US-China trade war
   • China and its allies respond with counteractions: supply chain disruptions, US treasury bond mass sell-off, avoiding tariffs through offshoring manufacturing
   • US treasury bonds and US currency come under pressure
   • Large economies experience substantial GDP fall
   • US and developed economies reach tariff agreement

As next, we project these assumptions onto financial markets. Recognizing that macroeconomic factors alone do not capture the full dynamics of the envisaged scenarios, we strengthen our analysis by supplementing our macroeconomic predictions with scenario-specific adjustments. Subsequently we determine the parsimonious set of risk factor shocks that clearly and unambiguously represent the outlined scenarios and divide them by risk class.

Selection of the Estimation Window

To ensure that the correlation matrix used in the propagation step is compatible with the set of shocks on the selected independent risk factors, we search for an estimation window that resembles the market environment envisaged for our tariff scenario.

We find out that the period between 19.04.2022 and 21.06.2022 is characterized by substantial losses in global equities accompanied by a sharp rise in US treasury bond yields which is consistent with our scenario narrative.

To design a robust shock propagation process, we need historical correlations between all used predictive factors. Nine independent risk factors translate into 36 correlations, which are hard to keep track of. The S&P Global® Buyside Risk system controls the compatibility of these correlations by monitoring metrics estimated for the set of independent risk factors based on the previously determined time frame. This allows us to keep the prediction error in check.

Segmentation of the Stress Scenario

The final scenario modelling choice is the segmentation of the scenario. We divide portfolio assets into groups according to asset class and economic region and use different sets of independent risk factors to apply decoupled shocks to these asset groups. In this way we significantly enhance the compatibility in the scenario design.

As a result of the segmentation, the correlation matrix is taken apart, allowing us to remove the less stable correlations from the propagation step. A closer look at the numbers in the matrix (Fig. 5) reveals that while the correlations between interest rates and yields are relatively stable, the relationship between rates and equities is the opposite.

Another benefit of segmentation lies in the capability of shifting to a non-inferred stress test for an individual portfolio segment. In our trade war scenario, we predict an 18% drop in Chinese equity value, whereas the underlying historical correlations were obtained from far less volatile market period. Should we infer this extreme, nearly three-sigma shock, it would propagate to other instruments causing unreasonable price moves and significant prediction noise.

The inferred stress test method used in this scenario required careful and repetitive diagnostic of the results. Using the S&P Global® Buyside Risk system we meticulously selected independent risk factors and decomposed the portfolio into multiple segments to mitigate the potential for high prediction noise or counterintuitive outcomes. This setup, combined with an understanding of historical factor correlations, ensures a sufficient level of compatibility among the shocks in our scenarios.

Simulated impact of Trade Scenarios on Financial Markets

The outcome of our trade scenarios reveals significant effects on market performance. The trade-war scenario indicates a negative impact across all markets, while the trade-deal scenario presents potential benefits for developed markets. Upon performing these scenarios in the S&P Global® Buyside Risk system, we observe substantial impact on asset returns in specific regions and countries.

a) Impact on the Equity Portfolio

From the USD perspective, both US and Chinese equities experience significant declines in both scenarios, whereas other developed equity markets remain relatively stable or experience only minor losses in the trade-war scenario.

b) Impact on the Corporate Bond Portfolio

Further depreciation of the US dollar, coupled with increased costs for credit default protection due to the escalating trade conflict, suggests that corporate fixed income in the U.S. may incur substantial losses. In contrast, other developed bond markets, particularly in Euroland, are expected to not only gain in both scenarios but may also be viewed by investors as safe havens. 

c) Impact on the US Treasury Portfolio

Finally, our scenario predicts considerable losses in the US Treasury bond market, driven by recurring selloffs triggered by Chinese creditors in the event of a trade war. These losses could be exacerbated if China uses proceeds from matured US Treasury bonds to flood global currency markets with the US dollar.

Concluding the considerations about outcomes of the tariff scenarios, we need to point once again to the fact that financial markets around the globe are currently driven by substantial uncertainty. For that reason, risk professionals need to be prepared to revise regularly their predictive stress scenarios and update them by the most recent turn of events. After all, it may be worthwhile to take to heart the best practice approach shared by Donald Trump himself in his book “Art of the deal”: “I try to learn from the past, but I plan for the future by focusing on the present.”