Three months ago, in the aftermath of the February VIXplosion and when the shape of Trump's protectionist agenda was only just starting to take shape, JPM's head quant, Marko Kolanovic made a simple prediction: despite all his posturing, Trump would never dare to pursue an aggressive trade war that destabilizes markets for one simple reason - if stocks crash, it would make a Democratic victory more likely in the midterms, which in turn raises the threat of Trump getting impeached.

The JPM quant also laid out the various possible outcomes between trade war and Fed policy error in terms of a game theoretical "prisoners' dilemma" matrix, as follows:

A significant trade war started by this administration would destabilize global equity markets. Should this happen ahead of the November election, it would impair the administration’s ‘market scorecard’ and likely lead to an election loss. Lost elections open a path to impeachment, and other complications. The game is also non-zero sum, as one can both use tough rhetoric and at the same time do little disruptive action (e.g., players as we defined them can ‘have their cake and eat it’). Setting up a diagram (similar to the well-known ‘prisoners’ dilemma’) points clearly that there will be strong rhetoric, but weak or no action that would destabilize equities. One could argue that a similar analysis can be applied to the risk of the Fed committing policy errors (see figure below, asymmetry in the case of the Fed would be causing a recession, and the public pinning the blame to specific individual central bankers responsible for the decision).

Well, fast forward to this week, when Kolanovic was no longer quite so sure.

In an analysis, in which the quant calculated the adverse hit to the S&P500 from the escalating global trade war at $1.25 trillion...

By attributing the trade-related news flow (positive or negative) to the performance of the US market, we estimated the impact on US equities to be negative 4.5% (with a margin of error of +/-1%, Figure 1). Taking the current market capitalization, this translates into $1.25T of value destruction for US companies.

... Kolanovic admitted that he may have underestimated Trump's determination to overturn the current global regime, after being convinced that Trump would never engage in the kind of bi- and multilateral trade feuds as he has, the JPM strategist now concedes that Trump's trade wars could in fact have an adverse impact on markets, especially if the trade wars persist.

The value destroyed by a trade war might be reversible if policies are reversed, while the positive impact of fiscal  measures is likely to remain. This would likely catalyze a ~4% market rally. However, if this uncertainty hangs over the market for a more extended period of time, the damage becomes more permanent and the probability of a disruptive tail event increases.

Today, with the entire world eagerly looking forward to the imminent fireworks out of the Toronto G7 summit, which UBS previewed as follows:

Because the G6 plus 1 is about spin not substance, the risks for markets come from the visual from this meeting. The US is looking isolated. This is not the isolationism and protectionism of the 1930s. In the 1930s everyone put up trade barriers against everyone else. This time the US is putting up trade barriers against everyone else, and everyone else is carrying on just as before.

... another bank join JPMorgan in looking at a hypothetical "worst case scenario" as a result of the growing risk of an out-of-control global trade war.

As Bank of America economist team led by Ethan Harris writes today, despite the clear escalation of trade tensions, many commentators remain sanguine. The reason, according to Harris, is that "they contend that the recent actions by both sides are merely negotiation tactics and that the final outcome is likely to be benign: that a trade war will rationally be avoided because everyone stands to lose in a trade war."

Bank of America, like JPMorgan however, is "more concerned" and here's why.

First, BofA looks at the "underlying intuition" of the emerging trade conflict.

In a trade negotiation, if each country knows the other's incentives, there is no need for escalation because there is mutual understanding of how much the country in the weaker position is willing to give up. But this is not how most negotiations work in reality, particularly today's reality. In the real world, each country faces considerable uncertainty about the other country's willingness to escalate tensions. Hence, escalation serves the purpose of resolving uncertainty about the other country's willingness to retaliate.

And, according to BofA, "today there are many reasons to fear miscalculation" such as the following four:

  • The Trump administration enters the negotiation with a "mercantilist" perspective. Trade deficits are prima facie evidence of unfairness and reducing the deficit contributes one-for-one to US GDP. Most other countries view trade through the lens of efficiency and supply chains. Trade is not unfair if it reflects natural advantages rather than government support.
  • The Trump administration believes trade wars are good and easy to win. Other countries view them as bad for everyone.
  • The Trump administration believes that it has a clear negotiating advantage because other countries depend more on exporting to the US. Other countries can point to the fact that the US is most at risk because it is battling along so many fronts simultaneously.
  • The two sides may also have differing views of how political and market pressures will constrain each other's actions. Is the Trump Administration willing to ignore pressure from the markets and the Republican Party? Can Europe present a united face? How will Mexico behave as it changes governments?

Before progressing further, Harris explains that this feeling-out process "could take time and multiple rounds of testing."

And this is where the potential tendency for complacency emerges: Since the initial actions are small from a macroeconomic perspective, economists have not changed their forecasts and the markets have shrugged off the shock. Even if the scope of tariffs grows it will take time to see the impact on the economy and the effects can be clouded by other drivers of the economy. Political pressure takes time to build as well.

Which brings us to how BofA lays out the "early rounds" of a trade war, in which we also have a foretaste of what is to come. Here, the initial logic is simple: if  a country is clearly in a weak position, it responds to escalation by making concessions.

The best example of this is South Korea's response to the US' announcement of tariffs on imports of steel and aluminum on 1 March. While other trading partners sought temporary exemptions, threatened retaliation and cried foul about the application of Section 232 (the "national security" clause), South Korea opted to make some concessions to the US on terms of trade in exchange for a permanent exemption from these tariffs.

Countries in stronger negotiating positions, which would be most today, have however responded to protectionist measures with retaliations. "This action is a posturing / signalling device: it teaches the country taking the original action (the US in this case) that the other side will fight back."

As BofA then notes, the first country (the US) must then decide whether to escalate further or de-escalate, having revised up its estimation of the other country's willingness to push back. This decision in turn helps teach the other country about US resolve, and so on.

Rounds of mutually detrimental escalation continue until one country is sufficiently sure about the other's willingness to retaliate that it chooses to de-escalate the situation.

BofA's analytical framework provides several key takeaways that are pertinent to the present situation.

Accidents happen, or what happens when "uncertainty" crosses a key threshold.

First, uncertainty can lead to costly trade disruptions that could otherwise have been avoided. For example, if the US administration knew exactly how much ground (if any) each country was willing to cede on trade, it could calibrate its demands in a manner that resulted in minimal disruptions to international trade. Instead we are seeing a steady rise in trade tensions with the US and its trading allies reported to have "underestimated" one another. This is common parlance for the uncertainty in our framework.

If thing escalated from there, then as BofA puts it, "Really bad accidents happen."

In this case, multiple rounds of back-and-forth tariffs (resulting in a "trade war") are more likely when both countries in a negotiation have substantial capacity to retaliate. But as BofA notes, this is precisely the situation in which the status quo would have been preserved in the absence of any uncertainty: "neither country would escalate trade tensions because both countries would rationally foresee repeated and costly retaliation."

We believe this scenario might depict the US-China impasse. So far, China has retaliated proportionately to each round of measures taken by the US. In response to the steel and aluminum tariffs, China has imposed tariffs on $3bn worth of goods imported from the US. In response to the US threat of another $50bn in sanctions, it has prepared a response of similar size.

Moreover, the proposed Chinese tariffs would target agricultural products such as soybeans and sorghum. Therefore they would hurt the Republican voter base, likely increasing the political costs of escalation faced by the US administration. By targeting products that are heavily consumed by the Chinese population (soybeans) and by livestock (sorghum), the Chinese administration would also be sending a strong signal about its determination to not relent to US demands.

According to the bank, "there is a risk that the US and China will escalate protectionist measures, imposing substantial costs on the two largest economies in the world."

In this case, the rest of the global economy will also probably suffer.

There is one potential failsafe trigger preventing this: both sides eventually seeing what is becoming increasingly evident, that they each have considerable capacity to retaliate, and so both will lose in a full-blown trade war. Unless, of course, one of the actors is irrational.

The visible hand

Finally, what are the mechanisms that could break a spiral of mutually detrimental protectionist measures?  Here, BofA would not count on economic discipline to prevent a trade war. As discussed above, the impact of trade measures is likely to trickle through the macroeconomic data over time and will be hard to isolate from changes in the business cycle and other policy developments.

There is one key trigger that would stop all out "nuclear" trade war: a market crash, i.e, the same gating factor that both Kolanovic, and last weekend, Goldman suggested is the only potential catalyst that both tells Trump he is winning, and at the same time, prevents further escalation. Here is BofA's explanation:

Market discipline could break the spiral. The markets provide immediate and publically observable feedback from investors on the expected effects of policies. So far, US equities have only had a small negative reaction to news on trade. This has likely emboldened the administration to take a more aggressive stance. But if back-and-forth tariffs with China were to cause a large selloff in US equities, China would probably push harder, knowing that continued escalation would be costly for the US.

And then, closing the circle on what Kolanovic noted in March, BofA escalated the impact of a market crash on the political realm, and writes that, in a similar manner, "elections clarify the costs and benefits of protectionism. If the Republican party retains control of the House of Representatives and the Senate in November's midterm elections, free-trade advocates in the Republican party may remain reluctant to push back against the President. The opposite might be true if the Democrats win."

In summary, BofA sees risks to the view that "cooler heads will prevail" on trade, and its concern is not so much irrational action, but that the US and its trading partners could rationally engage in tit-for-tat protectionism, with growing economic costs, in order to test each other's resolve, while markets remain dormant and allow even greater escalation. Still, the Bank of America economists ultimately remain hopeful "because market and electoral discipline could break up this dynamic."

In other words, BofA is hoping that to prevent the worst case outcome, either the market will crash or Republicans will lose the midterms and be kicked out of the House, potentially leading to a Trump impeachment, thus halting the global trade war from escalating too aggressively beyond a level that could potentially crush the global economy.