Just how damaging have recent attacks on free trade been? Back in May, we explained in a detailed CIO Special that "the results of this battle are likely to impact portfolio returns for years to come." Few months later, the time has come to take stock and update our thinking. For this brief update, we introduce two of the four yardsticks, which we are watching. The first concerns what actually happened, and what new trade risks and opportunities have lately appeared on the horizon. The second looks at the potential consequences on gross domestic product (GDP), both of the countries impacted and the world as a whole. So far, this looks limited. But as we explain in our conclusion, this is hardly reassuring.
In January, the United States imposed tariffs on washing machines and solar panels, claiming these imports had caused "serious injuries" to U.S. manufacturers. Those tariffs mainly hit China, Vietnam, Korea and Malaysia. Citing national security, the Trump administration followed up with tariffs on steel and aluminum from a wide range of countries. Those countries include not just China, but also such traditional allies as the European Union (EU).
Mexico and Canada have not been sparred either, even as the United States was trying to renegotiate the North American Free Trade Agreement (NAFTA). The NAFTA talks have lately resumed, but it is unclear how much progress can be made ahead of the U.S. midterm elections on November 6. But other sources of tensions are likely to keep investors on edge throughout the next couple of months.
China initially reacted with only very modest counter measures, targeting items such as fresh and dried fruits and pork products. The EU and Canada retaliated to the extent allowed under the rules of the World Trade Organisation (WTO) and initiated complaint proceedings against the United States at the WTO.
So far, relatively harmless. Unfortunately, things have not stopped there. Alleging that China has been violating the intellectual property rights of U.S. companies for years, the Trump administration has imposed a 25% punitive on 34 billion dollars worth of Chinese imports. It plans to follow up shortly with tariffs on another 16 billion dollars. Any Chinese retaliation is likely to be countered by U.S. tariffs on an additional 200 billion dollars of Chinese imports. In total, the United States has imposed tariffs on imports worth some 89 billion dollars so far. Retaliatory measures already implemented or announced by its trading partners amount to 79 billion dollars of U.S. exports.
And the next targets are already in sight. President Trump has announced that he is prepared to impose tariffs on all 500 billion dollars worth of Chinese imports. Add threatened U.S. tariffs on some 350 billion dollars worth of cars and automotive parts, and you can see why many investors are worried. The EU reacted by dispatching the president of the European Commission, Jean-Claude Juncker to Washington on July 25. The visit went well, suggesting the EU car sector might be spared tariffs – at least for now.
The immediate relief in markets was tangible. However, keep in mind that not all that much was actually agreed. It remains exceedingly unlikely that the United States and the EU will make much progress "toward zero tariffs, zero non-tariff barriers and zero subsidies on non-auto industrial goods", the latest goal stated by President Trump. Given the prior history of transatlantic trade talks and the political realities on both sides of the Atlantic, anything along those lines is likely to take years. Our first test can focus as a corrective to market overreactions. By focusing on protectionist measures firmly announced, implemented or taken back, it avoids reading too much into any particular meeting or non-binding statement by any of the parties.
How bad could things get? That rather depends on how you look at the potential fallout. The impact on global GDP has been the focus of most attention by financial-market commentators. Estimates on the impact of even quite aggressive measures over the next few months or years tend to be fairly low. The next section will explain why this is hardly reassuring.
Estimating the impact of trade tensions on global GDP is far from straightforward. The difficulties in coming up with estimates start with the fact that nobody really knows how many billions of dollars will actually be impacted by any particular measure. That depends on how much buyers of any given good or service react to the increase of its price, due to tariff increases. It turn, this is a function of how tariffs and tariff exemptions will be enforced. More broadly, it depends on how critical and special a good is, and how easy it is to source a similar product elsewhere – be it domestically or from a foreign source exempted from the measures. Moreover, both domestic and foreign producers will react to the new price signals. For example, the EU is rightly worried that in addition to hurting its steel producers directly, U.S. tariffs are also leading steel from other countries to be diverted to European markets.
Such effects are just one of many reasons why truly reliable estimates require the use of what is known as a "computable general equilibrium" (CGE) model. In such a model, everything sort of depends on everything else. What happens in the goods sector in any large country due to a tariff, for example, can potentially spill over to what happens to interest rates, currencies and unemployment both domestically and in the rest of the world. We said "sort of", because by necessity, such a model is a simplified version of the world. It can give you an indication of how a national economy and global GDP might react to changes in trade policy, not necessarily how they will react.
Central banks and international organizations, like the International Monetary Fund (IMF), are avid users of CGE models. They are also well aware of their limitations. One particular problem is that while CGE models use actual economic data, the results can only be as good as the data you put in. In the case of trade wars, there are preciously few historic data points. That's fortunate from the point of view of global welfare in the past, but unfortunate in terms of the reliability of resulting estimates.
What CGE models do really well, however, is create awareness of the mental pitfalls that could easily result from looking at any given variable in isolation. For example, a surprising number of sell-side analysts claim that unless other countries retaliate, U.S. tariffs could actually increase U.S. GDP. The reasoning behind this idea is that tariffs increase import prices, which will give a boost to the domestic production of similar goods and services.
There are many problems with this reasoning. Three big ones stand out. First, it ignores the impact of a new U.S. tariff on the dollar. If exchange rates are flexible, CGE modelling (well founded on economic theory and history) suggests the strengthening of the dollar's real exchange rates. Second, new manufacturing capacity will not magically appear overnight, especially as the U.S. labor market is already tight. As we recently pointed out , the United States is close to, if not a shade beyond full employment. Third, international manufacturing supply chains are a lot more complicated than it is possible to model in even the most sophisticated CGE model. A large chunk of Chinese exports to the United States (500 billion dollars or so) just reflects the attractiveness of China for the final assembly stage of goods designed by U.S. and European multinational firms – think of smartphones, for example.
As a result, any protectionist measures will almost certainly cause immediate efficiency losses. Even if domestic production substitutes for some of the items that were hit with tariffs, that production is likely to lead to a less efficient allocation of domestic resources. After all, goods and services – including intermediate goods and services sourced from profit-maximizing companies abroad – are generally imported for good reasons. It is often more efficient and cheaper to buy or make them elsewhere than to produce them domestically, which is why they are imported in the first place. Some of the efficiency gains that trade enables will likely be lost almost immediately due to tariffs, while in the longer term, productivity growth is also likely to be negatively affected. If competition is reduced, this may be made worse, for example by locking out more innovative foreign suppliers or competitors.
The size of such efficiency losses is hard to measure and estimate, but likely to be a key part of the costs of protectionism. Trade wars also increase uncertainty (how will supply chains be affected, will there be retaliation, etc.), and this can adversely affect activity by causing firms to cut back on capital spending and hiring. It also tends to lead investors to demand higher risk premiums, thereby tightening financial conditions.
So what does all that mean for global GDP? Well, the IMF recently estimated in its world economic outlook that the measures announced so far will only reduce the world's GDP by about 0.5 percentage points in total by 2020. Things could clearly get a lot worse, but the estimate also shows that the world economy can probably handle trade headwinds.
Overall, the results from our second test can be summed up as follows: For now we think that trade remains unlikely to trigger enough of a broader economic slowdown to cause market jitters in and of itself. Unfortunately, the immediate impact on GDP is not the only thing investors are likely to worry about, if trade tensions get out of hand.
As a measure of the economic damage done, GDP has several disadvantages. The first is that it is inherently backward looking. One way around this is to rely on forecasts, which incorporate the latest findings from such sources as company surveys or financial-market indicators. Needless to say, we do that anyway in deriving our growth forecasts for the United States and other major economies. For now, we see some early warning signs, but no reason to panic. As we recently wrote, commenting on our latest growth forecasts : "Make no mistake, in our opinion the impromptu attacks by the Trump administration on free trade are, to put it mildly, fundamentally misguided and frighteningly uninformed. (…) they also risk reducing the growth prospects of the U.S. and the world economy as a whole. To us, it seems obvious that eroding the rules-based international order underpinning global trade and undermining international supply chains will hurt everyone involved, including emerging markets. However, it is also unlikely to cause a global recession anytime soon."
This hints at a second problem. It is quite possible for a trade war to cause damage to portfolios, without triggering a recession. GDP is designed to measure an economy's output in any given year, not how wealthy its citizens are. And when it comes to trade, it is helpful to return to the analogy we introduced in our May CIO Special on the topic. A useful way to think about the effects of trade is to imagine a secret technology, which "allows the United States to convert wheat, lumber and soy beans into cheap high-quality consumer goods, such as toys or washing machines." As with any new production process, some workers and sectors will inevitably lose out, while others will win. Our imaginary secret technology "will make U.S. farmers and farm workers better off. At the same time, those owning now uncompetitive factories making toys or washing machines will be worse off, and so will their workers. For the United States as a whole, though, trade is generally a good thing - in the same way, and for the same reasons that new technologies are. It benefits consumers."
To which we would only add that import tariffs on intermediates tend to have even worse welfare implications for the country imposing them. By targeting such inputs and intermediates as steel and aluminum car parts, they are likely to hurt even some of the industries other measures are trying to protect. Washing machines and carmakers are good examples.
Now, imagine the sole factory where this secret technology is applied being hit by a hurricane. How bad will the impact be? If the damage can quickly be repaired, not all that bad. That, broadly is how financial markets currently think about the ongoing trade conflict – a temporary interruption in business as usual. The owners of the factory might lose some profits, but profits should quickly recover, while much of the rest of the economy might remain largely unaffected. Indeed, it is even possible that after taking a hit from the hurricane, GDP growth could accelerate as the factory is rebuilt.
A lasting trade war, with protectionist measures permanently staying in place, is a different matter altogether. You can think of it as a man-made disaster that results in the factory being destroyed and our secret technology lost forever. That would clearly be very bad news for the owners of the factory, but also for anyone else who had made an investment decision thinking they could sell to the factory – the farmers, who specialized in wheat, lumber and soy beans, in our example. The damage to consumers, from having to pay more for their toys or washing machines, would be just as permanent. Very little of this wealth destruction would be readily visible in GDP growth over the short term.
A third problem with GDP as a measure of trade damage is that while investors rightly worry a lot about the next recession, what they really care about is the long-term prospects. International competition tends to spur innovation, by allowing countries to specialize in the bits of the global value chain they are best at.
So, if neither GDP nor our forecast for the next 12 months can tell you how trade might impact financial markets, what might? We are currently working on two additional measures that we hope will help. Our third yardstick to think about the damage done so far can be thought of as a market-derived trade fear index. It is going to include trade sensitive stock and commodity-markets indicators to give an indication of how long the conflict might last and how much it might hurt. Of course, there is no guarantee that markets will get it right. Our fourth yardstick will attempt to use game theory to predict how things might work out. This basically comes down to the question whether Donald Trump is serious about dismantling global value chains – or merely bluffing in search of a better deal. In the former case, the results could be devastating, even if countries try to maintain the rules-based global trading system. 
Political processes are notoriously difficult to forecast. Indeed, it is somewhat surprising that trade tensions have escalated as much as they already have. It is becoming increasingly clear that the Trump administration faces plenty of domestic opposition. The U.S. economy and U.S. financial markets appear a lot more exposed than GDP forecasts suggest. Estimates vary on how much, but the effects of disruptions to global supply chains could be severe. After all, much of U.S. trade in goods and services appears more like smartphones (designed by U.S., even if assembled in China) than, say, steel. The imports of final goods made elsewhere contain plenty of U.S. value added. Meanwhile, U.S. manufacturers get much of their inputs overseas.
The other week, the U.S. car industry provided an interesting example of how tricky protectionist measures are in an era of cross-border supply chains. At the public hearing on the proposed tariffs on cars and car parts, Europeans were not the only ones begging for mercy. So were representatives of U.S. carmakers. They pointed out that a 25% tariff on imported cars and parts might raise the cost of a typical imported car by 6000 dollars – but at the same time increase the cost of U.S. built ones by 2000 dollars. Shortly thereafter, the largest U.S. carmaker cut its earnings outlook, blaming tariffs.
Such pressures might yet help secure the outcome many market observers were expecting earlier this year. As we previously wrote: "the United States might secure crowd-pleasing "concessions" for the benefit of politically well-connected industries, mainly at the expense of U.S. consumers.(…) the path towards "compromises" should be fairly easy. All it requires is for knowledgeable Chinese, European, Mexican and Canadian trade negotiators to follow the Korea strategy – allowing the United States to secure the occasional "concession" mostly at the expense of U.S. importers, and, ultimately, U.S. consumers."
The Juncker visit could prove another case in point. For now, it appears the EU's mere willingness to continue to talk about liquid natural gas and soy beans was enough to avoid a further escalation.
As we will describe when introducing our last two tests, investors would be well advised not to overestimate the effects of both further escalation and climb-downs by any of the parties. The real question remains whether and how the rules-based multilateral trading framework under the WTO umbrella can adapt and survive. This probably becomes less likely with every new protectionist measure. And every protectionist measure also increases pressures for more interventions into the functioning of free markets – as the recent U.S. aid package for farmers shows.
From a market perspective, it would be very regrettable if the WTO and support for free trade and free markets more broadly continues to be undermined. For similar reasons, most U.S. businesses and their representatives do not appear all that keen on protectionist measures, however well intended. Top business groups such as the U.S. Chamber of Commerce are fighting publicly and hard against U.S. tariff plans. Quietly, so are their political allies among Republicans in Congress, especially in the Senate. Traditional free traders, including many of Republicans, are understandably worried. So are the business interests Donald Trump claims to promote. That old aphorism of Ronald Reagan comes to mind: "The most terrifying words in the English language are: I'm from the government and I'm here to help."
1. The Economist (2018a) has recently provided a very handy breakdown of the various measures (i.e. how many billion per sector) each country has implemented so far: https://www.economist.com/graphic-detail/2018/07/20/donald-trump-is-fighting-trade-wars-on-severalfronts. Its July 19 issue also contains an excellent briefing on the topic: https://www.economist.com/briefing/2018/07/19/the-world-trading-system-is-under-attack
2. For an interesting account of the impact so far, see Crooks, E. and Fei, F. (2018) "Trade war winners and losers grapple with Trump tariff chaos: Companies bewildered by judgments on exemptions as backlog of decisions rises", Financial Times, July 24, https://www.ft.com/content/675e439c-8c25-11e8-bf9e-8771d5404543
3. Davies (2018) provides an excellent introduction to the problem. Also see Obstfeld (2016) for details on the estimates cited in Davies (2018). "Tariffs Do More Harm Than Good at Home", International Monetary Fund, https://blogs.imf.org/2016/09/08/tariffs-do-more-harm-than-good-at-home/
4. This is based on "the increase in US tariffs on imported solar panels, washing machines, steel, aluminum, and a range of Chinese products, and the announced retaliatory measures by trading partners as of July 6", i.e. it does not, as yet consider the threat of tariffs on an additional 200 billion dollars worth of Chinese imports. For details, see https://www.imf.org/en/Publications/WEO/Issues/2018/07/02/world-economic-outlook-update-july-2018#_ftn1
5. Irwin, N. (2018) has come up with quite a useful list of indicators investors can keep an eye on, in assessing the trade impact on the U.S. economy.
6. For an indication and some estimates how bad things might get, see Krugman, P. (2018)