Macro Research

1 June 2018

Thinking Macro
‘Trade War’ in perspective
Tomasz Wieladek
• The return of US protectionism after a century has led to fears of a ‘trade war’.
+44 (0) 20 3555 2336
We use a VAR framework to study the effect of US tariffs on global growth and CPI.
tomasz.wieladek@barclays.com
• US tariffs act as a negative supply shock to the world economy. Our estimates for Barclays, UK
the first year after the tariff are large, but uncertain. We illustrate the effect of a 1% US
tariff as a share of imports with second-year estimates. As a result, global growth falls www.barclays.com
by 0.3pp, inflation rises by 0.4pp. With tit-for-tat retaliation, the effects double.

• US steel tariffs represent only 0.33% of US imports, even excluding exemptions.

The Intellectual Property tariff of 25% on $50bn of Chinese goods is 0.5% of US
imports. The threat of a 25% tariff on $100bn of Chinese goods would be double.

• Based on these numbers, the steel tariff could reduce global growth by 0.1pp and
raise inflation by 0.1pp in the second year. If $50bn of IP tariffs are added and
retaliated tit-for-tat, growth would fall by 0.6pp and inflation rise by 0.7pp. With
$100bn and tit-for-tat retaliation, growth falls 0.9pp and inflation rises 1.1pp.

• The impact on EM is larger than on DM. A 1% tariff leads EM GDP to fall by 1.1pp
and inflation to rise 1.1pp. DM growth falls only by 0.5pp, while inflation rises 0.2pp.

• Our approach may underestimate the impact of a rapid tariff rise. We likely omit
second-round price and confidence effects. Global value chains could be captured
by our model, as they emerged when tariffs were reduced in the 1990s. Even if the
‘true’ impact is twice our estimate, large effects still require large tariffs.

• There are several mitigating factors. While the US used tariffs in the 19th century,
large retaliations were rare. A 70% success rate in WTO trade disputes will likely
encourage the US and EU to keep this resolution mechanism. Services trade could
rise with US business deregulation. But likely not enough to offset the EM impact.

FIGURE 1
Econometric analysis shows that large tariffs produce large effects
1.5 % average annual impact in second year after tariff

1.0

0.5

0.0

-0.5

-1.0
Global real GDP growth Global CPI inflation
-1.5
Steel tariff without Steel tariff and Steel and $50bn Steel, $50bn IP Steel, $100bn IP
exemption retaliataion IP tariff and retaliation and retaliation

Note: The first-year estimates from our model are large, yet highly uncertain. We use second-year estimates for
the illustration above, as our model provides greater statistical confidence for these. Source: Barclays Research

PLEASE SEE ANALYST CERTIFICATIONS AND IMPORTANT DISCLOSURES STARTING AFTER PAGE 17
Barclays | Thinking Macro
President Trump announced
several protectionist policies
Negotiations making progress,
but escalation risks remain
US tariff policy acts like a
negative global supply shock
The steel tariff’s small import
share means a limited impact
Large effects require big tariffs
Emerging markets would likely
be more affected than DM
Some factors mitigate the
impact, but not enough for EM
1 June 2018
US trade protectionism: Where do we stand?
This year, the US has implemented a number of protectionist trade actions. In March,
President Trump announced a 25% tariff on steel (10% on aluminium) imports. The US
Trade Representative (USTR) then proposed a 25% intellectual property (IP) related tariff on
1,333 Chinese goods1. President Trump then asked the USTR if it was possible to impose
tariffs on a further $100bn of Chinese goods. Import tariffs in the automotive sector are also
being considered. Some progress has been made in trade negotiations with China (see
China: Tariffs on hold, long negotiations continue, 12 May 2018). But escalation risks
remain, since the steel tariff exemption will expire on 1 June and the White House said it will
impose a $50bn IP tariff on Chinese products, with the list published by 15 June 20182.
We use a VAR model to quantify the potential impact of US tariffs on global growth and CPI
inflation. The first year estimates are subject to high model and parameter uncertainty. We
thus use second year estimates. These show that a 1% unilateral rise in US tariffs as share of
US imports may reduce global growth by 0.3pp and increase inflation by 0.4pp. That said,
the impact of the steel tariff, even without exemptions, would only lead to a 0.1pp decline in
global growth and a rise of 0.1pp in CPI inflation, as steel is only 0.33% of US imports.
Large economic effects larger require big tariffs. If the proposed 25% tariff of $100bn of
Chinese goods is added to the steel tariff, together with tit-for-tat retaliation, our model
shows that such a scenario would raise CPI inflation by 1.1pp and cut growth by 0.9pp.
Our model suggests that the adverse effects of US trade tariffs on emerging markets are
likely to be much larger and more persistent. This is intuitive, as these economies have been
the largest beneficiaries of the most recent globalisation wave. According to our model, a
1% rise in US tariffs leads to a 1.1pp reduction in EM growth in the first year, versus 0.5pp
for DM. For CPI inflation, the numbers are +1.1pp for EM versus +0.2pp for DM.
However, there are a number of mitigating factors. In the first age of globalisation, US tariff
policy was very active, but large retaliations were rare. Similarly, their 70% success rate
incentivises the US and EU to keep the WTO for resolving trade disputes. In addition,
President Trump’s drive for deregulation, by removing entry barriers, could encourage more
services trade, which may mitigate the negative effect of higher tariffs on goods. That said,
any rise in services trade flows is unlikely to fully offset the impact of higher tariffs on EM
countries, given the size and persistence of the effects estimated in this paper.
FIGURE 2
Media attention on protectionism dwarfs previous episodes
600 # of Bloomberg stories containing 'US Trade Protectionism'
500 President Trump changes
US trade policy
400
300
G20 commit to abstain from proctectionism
in response to the 'Great Recession'
200
100
0
Jan-91 Jan-94 Jan-97 Jan-00 Jan-03 Jan-06 Jan-09 Jan-12 Jan-15 Jan-18
Source: Bloomberg, Barclays Research
1 https://ustr.gov/sites/default/files/files/Press/Releases/301FRN.pdf
2 https://www.whitehouse.gov/briefings-statements/president-donald-j-trump-confronting-chinas-unfair-trade-
policies/
2
Barclays | Thinking Macro

A first attempt at modelling the impact of a ‘trade war’

In this section, we discuss the rationales underlying our VAR modelling strategy.

Few studies have looked at the
effects of tariffs on short-term
growth and inflation
The effect of tariffs on the business cycle is not well understood. In a comprehensive survey
paper, Goldberg and Pavcnik (2016) note that most papers on this subject have focused on
the long-term impact of trade policy on trade volumes, industry output, labour markets and
growth. But almost no previous work has studied the short-term impact on quarterly
growth and CPI inflation, which tend to be the focus points for central bankers and
investors. This could be because quarterly tariff data are normally not available, with reserve
causality between trade policy and real activity as an added econometric complication.

BCG (2018) is an exception
One recent paper that attempts to tackle this challenging question is Barattieri, Cacciatore and
Ghironi (2018). They used a VAR model that allowed them to address the reverse-causality
issue, on annual data. Because they base their model on annual panel data, their estimates of
effects cannot be attributed to any particular country. Their main finding is that trade tariffs
act as a negative supply shock, raising inflation and reducing growth.

We estimate a VAR to assess We examine the impact of a potential increase in US tariffs on quarterly global real GDP
the potential effect of US tariffs growth and CPI inflation using a VAR model. The quarterly frequency of our data means we
on global growth and inflation rely on time-series data only, hence, this allows us to better attribute the estimated effects
to US policy. To examine if US tariffs have a global impact, we use the following VAR model:

𝑌𝑡 = 𝐵1 𝑌𝑡−1 + 𝐵1 𝑌𝑡−2 + 𝑒𝑡 𝑒𝑡 ~𝑁(0, 𝛴)

Where 𝑌𝑡 is a vector of the following variables: US tariffs/imports, Global real GDP growth,
Global CPI inflation, US Net Exports/GDP and the natural log of US import prices. The Vector
Autoregression (VAR) approach is one of the least structured empirical methods available
to macroeconomists. As such, it is ideal to understand the impact of economic shocks when
the underlying mechanism is not well understood. This advantage is also a drawback:
beyond quantifying the impact, it is difficult to know why an effect is small or large. It is,
however, important to include those variables in the model that are likely responsible for
driving the transmission mechanism. Therefore, we also add US net exports and US import
prices to our baseline model. To attribute any quantifiable impact to a shock to US tariffs,
additional assumptions need to be made: Following BCG (2018), we order US tariffs in our
model first, meaning that tariffs only respond to all other variables in the model with a lag.

FIGURE 3
US tariffs can have a significant impact on global real GDP growth and CPI inflation
Global Real GDP growth Global CPI inflation

Year 1 Year 2 Year 5 Year 1 Year 2 Year 5

Baseline -1.47 -0.67 -0.54 0.74 0.75 0.52

Simple average tariff series -2.95 -0.80 -0.71 0.35 0.51 0.48
Add US GDP -1.06 -0.58 -0.49 0.98 0.96 0.54
Add US GDP and CPI inflation -1.06 -0.47 -0.51 0.89 1.11 0.56
Add TFP -1.41 -0.71 -0.55 0.69 0.84 0.55
Add TFP and oil prices -1.25 -0.78 -0.55 0.94 0.80 0.52

Average Impact across models -1.53 -0.67 -0.56 0.76 0.83 0.53
Standard deviation across models 0.71 0.13 0.08 0.24 0.20 0.03
Standard deviation across Baseline parameters 0.98 0.53 0.37 0.58 0.28 0.20
Source: Barclays Research. Note: indicates estimates with greatest parameter and model uncertainty.

1 June 2018 3
Barclays | Thinking Macro

To enable analysis at a quarterly frequency, we needed to construct some variables from raw
data. The World Bank’s WDI dataset provides a US Tariff series, which is weighted by import
share at a product level at annual frequency. We then constructed a proxy for quarterly US
tariffs by dividing customs revenue by imports (from the US international trade commission
website). We then interpolate the annual WDI tariff data with this proxy using the Chow-Lin
approach (1971). The resulting US tariff series, shown in Figure 4, is the tariff measure we
focus on in this report. To construct global real GDP growth and CPI inflation, we use
weighted averages, based on IMF PPP weights, of the 33 major countries included in the
GVAR database. We exclude the US from this weighted average, since the domestic effect of
US tariffs is likely to be different from the international one. See appendix A for more detail.

Estimated effects on global GDP and CPI are sizable...

We assume that tariffs react to
real activity only with a lag
The model is estimated from
Q1 90 to Q4 16
As discussed above, identification assumptions are required to attribute the impact of a shock
to US tariff policy. We order our US tariff series first in the VAR, meaning that it reacts to the
other variables only with a lag. As trade policy is slow moving, this assumption is arguably
even more suitable with quarterly than annual data. While no VAR identification is ideal, the
advantage of this ordering, as BCG (2018) demonstrated, is that the impulse responses are not
too dissimilar from a DSGE model. Furthermore, the ordering of the tariff series does not
matter for our results. We estimate the model using data from Q1 90 to Q1 16.

Model results indicate that US
tariffs act as a very persistent
and negative supply shock…
Our VAR model shows that US tariff policy tends to act as a negative supply shock to the
world economy. Impulse response analysis (see appendix B) shows that following an
unexpected rise in US tariffs, global real GDP declines and CPI inflation rises. These effects
are statistically significant even up to 10 years after the initial shock. This suggests that a
rise in US tariffs is a negative and very persistent supply shock to the world economy.

… with a 1% US tariff raising
global inflation 0.75pp, while
lowering global growth 0.67pp
In Figure 3, we show the average impact of an unexpected 1% rise in US tariffs over time
horizons of one, five and 10 years. We average the quarterly median impulse response and
annualise the estimate. For example, in the baseline model, an unexpected 1% rise in US
tariffs would lead to a 0.67pp contraction in global real GDP growth and a rise of 0.75pp in
global CPI inflation in the second year after the policy was announced. We then explore
several modifications to our baseline VAR. The penultimate rows provide the average
impact and the standard deviation across all models. The last row provides the standard
deviation around the base-line estimate. Overall, the first-year estimates are most uncertain
and hence least reliable, which is why we focus on second-year estimates in our discussion.

FIGURE 4 FIGURE 5
Current US steel tariffs are a small fraction of imports… …which is why the global effects are small

4.5 % of all imports 1.0 % average annual impact

4.0 0.8

3.5 0.6

3.0 0.4 Impact without exemption Impact with retaliation

2.5 0.2
2.0 0.0
1.5 -0.2
1.0 -0.4
Effect of President Trump's steel tariff
0.5 -0.6
without exemptions
0.0 -0.8 Real GDP CPI inflation
Q1 1989 Q2 1996 Q3 2003 Q4 2010 Q1 2018 -1.0
WTO implementation US Tariffs as % of imports Year 1 Year 2 Year 5 Year 1 Year 2 Year 5
Prop. IP Tariff - $50bn Prop. IP Tariff - $100bn
Source: US International Trade Commission, Barclays Research Source: Barclays Research. Note: Year 1 estimates are highly uncertain.

1 June 2018 4
Barclays | Thinking Macro
Our weighted tariff series can
change due to trade policy or
consumer preferences…
…but the result are broadly
robust to using an unweighted
series as well
Estimates are broadly robust
across specifications
Our model likely overestimates
the effect of a unilateral US
tariff policy action…
…as tariffs were reduced
across countries after the WTO
was formed
Our best guess is that the
effect of unilateral US policy is
half of our VAR estimates
FIGURE 6
The effect is larger if $50bn IP tariffs are added…
1.0 % average annual impact
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
Year 1
Source: Barclays Research. Note: Year 1 estimates are highly uncertain.
1 June 2018
...and broadly robust across specifications
One potential issue is that our proposed US tariff series could reflect changes in US trade
policy as well as changes in consumer preferences. Weighting tariffs by the import share of
the corresponding product could understate the impact of actual tariff policy, as US
consumers may seek substitutes for products with high tariffs. An alternative approach is to
use the unweighted average of tariffs. But this series could be exaggerated by high tariffs on
products that are not purchased and countries that are not traded with. These results are
shown in the second row (Simple average tariff series) in Figure 3. For growth, the
magnitude of the impact doubles in the first year, but is fairly similar in the second and fifth
years. The impact on inflation is half as large upon introduction, but similar in the fifth year.
Figure 4 shows that most of the global reduction in tariffs occurred after the WTO was
founded in 1995. This coincided with a rise in global total factor productivity (TFP) and low oil
prices. We use John Fernald’s US TFP series as a proxy for global TFP. Adding either TFP or oil
prices to our model does not have a major impact on our results, especially beyond the first
year. We also test if our model is robust to US real GDP growth and CPI inflation. The CPI
inflation results remain broadly robust to all these modifications. The real GDP effect turns out
to be more variable, but only in the first year, with estimates converging over longer time
horizons. Overall, the results are broadly robust to specification. Since there is no ‘right’ model
among these specifications, we use the average impact across models for our calculations.
The estimates are likely exaggerated by the common tariff reduction after the 1994 Uruguay
round, which gave developed countries until 2000 to comply. As this was synchronous
across countries, the constructed US time series very likely picks up the effect of a common
reduction in tariffs across all developed countries, rather than only unilateral US tariff policy.
Ideally, we could separate unilateral from common tariff policy actions. But world tariff data
are available only at annual frequency. All we can do is to recognise this bias and adjust our
estimates. Developing countries had until 2004 to comply, meaning that effects from their
tariff reductions are less likely to affect our model. A very crude measure of the effect of
unilateral US tariff policy action is to consider US market size relative to other developed
economies. This was about 40% during the implementation of tariff reductions mandated
by the WTO. However, between then and today, world trade as a share of world GDP has
expanded by by 30%. Our best-guess, therefore, is that unilateral US tariff policy action has
an effect that is roughly half as large as the estimates shown in Figure 3.
FIGURE 7
…especially if trading partners retaliate
1.0 % average annual impact
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8 Real GDP CPI inflation
Real GDP CPI inflation
-1.0
Year 2 Year 5 Year 1 Year 2 Year 5
Source: Barclays Research. Note: Year 1 estimates are highly uncertain.
5
Barclays | Thinking Macro

The small steel import share leads to a modest impact

President Trump’s steel tariffs
as % of imports are modest…
To quantify the effect of President Trump’s tariff policies, it is important to understand how
they affect our constructed US tariff time series (Figure 4). Without the exemptions the steel
and aluminium tariffs cover $41.4bn or 0.33% of 2017 US imports. Recently, the White
House announced that it will apply a 25% IP-related tariff on $50bn worth of Chinese
products. This corresponds to 0.5% of US imports. Finally, President Trump has instructed
the US Trade representative to look into a 25% tariff on $100bn worth of Chinese goods
(1% of all US imports). If all of these measures were implemented, then US tariffs would rise
to slightly below the level prior to the conclusion of the Uruguay round.

FIGURE 8
For emerging markets, the effects on growth and inflation are larger…
EM GDP growth EM CPI inflation

Year 1 Year 2 Year 5 Year 1 Year 2 Year 5

Baseline -1.91 -0.63 -0.73 2.18 1.84 1.07
World Bank tariff series -2.82 -0.67 -0.73 1.85 1.59 1.08
Add US GDP -1.89 -0.84 -0.80 2.14 1.94 1.11
Add US GDP and CPI inflation -2.01 -0.78 -0.82 2.16 2.06 1.12
Add TFP -2.21 -0.79 -0.74 2.09 1.83 1.11
Add TFP and oil prices -2.27 -1.05 -0.87 2.34 1.77 1.09

Average Impact across models -2.18 -0.79 -0.78 2.13 1.84 1.10
Standard deviation across models 0.35 0.15 0.06 0.16 0.16 0.02
Standard deviation across Baseline parameters 1.31 0.80 0.55 0.95 0.59 0.56
Source: Barclays Research. Note: indicates estimates with greatest parameter and model uncertainty.

…which is why their economic
impact is only modest as well
Figure 5 shows that the steel tariffs are likely to have a small effect. If the proposed 25% IP-
related tariffs are implemented, the effects would be larger (Figure 6). Our model shows
that global growth in the second year could decline by 0.3pp, albeit this is still small relative
to the current rate of 4pp. CPI inflation would rise 0.35pp, which is more sizable compared
with the current rate of 2.4pp. Nevertheless, both of these effects look modest at best.

But if the $50bn IP tariff is
added to steel and there is tit-
for-tat retaliation...
...global growth could fall
0.6pp and inflation rise 0.7pp,
which is sizable
Our model’s flaw, that the estimates are likely picking up synchronous tariff actions, becomes
an advantage when taking into account potential retaliation by trading partners. Retaliation is
a synchronous rise in the tariffs of trading partners of roughly the same magnitude, essentially
similar to the tariff reduction during WTO implementation, but in the other direction. This
means our unadjusted estimate should provide a rough estimate of the effect. The effects of
US trading partners potentially retaliating against the current US steel and proposed IP tariffs
tit-for-tat are shown in Figure 7. Now the effects on growth at -0.6pp and CPI inflation at
0.7pp become sizable. If, in addition to steel, the 25% IP tariff is imposed on $100bn, rather
than $50bn, of Chinese imports and retaliated against, we estimate that the effects would be
large, with inflation rising 1.1pp and global growth falling 0.9pp.

The impact is larger for emerging, than developed, markets

EM economies benefitted most
from free trade and are more
susceptible to protectionism
Emerging markets have likely benefitted the most from the trade hyper-globalisation of the
1990s. The abundant and competitively priced labour supply in these countries, together with
free trade, led to large FDI inflows, allowing these countries to export their way up the
development ladder. Intuitively, this suggests that these countries should also be more
vulnerable to a rise in protectionism. In this section, we split our global real GDP growth and
inflation variables into corresponding variables for emerging and developed markets, to
econometrically examine if EM economies react differently to DM economies.

1 June 2018 6
Barclays | Thinking Macro

FIGURE 9
…than the corresponding impacts for developed economies

DM GDP growth DM CPI inflation

Year 1 Year 2 Year 5 Year 1 Year 2 Year 5

Baseline -0.91 0.00 0.00 0.43 0.28 0.22
World Bank tariff series -1.40 0.00 0.00 0.28 0.20 0.24
Add US GDP -0.84 0.00 0.00 0.41 0.31 0.25
Add US GDP and CPI inflation -0.98 0.00 0.00 0.36 0.37 0.25
Add TFP -0.90 0.00 0.00 0.32 0.30 0.24
Add TFP and oil prices -0.86 0.00 0.00 0.56 0.25 0.26

Average Impact across models -0.98 0.00 0.00 0.39 0.29 0.24
Standard deviation across models 0.21 0.00 0.00 0.10 0.06 0.01
Standard deviation across Baseline parameters 1.01 0.00 0.00 0.53 0.28 0.17

Source: Barclays Research. Note: indicates estimates with greatest parameter and model uncertainty.

Relative to DM, the EM impact
is much larger and persistent
Figure 8 and Figure 9 shows the results for EM and DM economies, respectively. This
breakdown produces several interesting results. First, EM GDP growth is likely to shrink by
roughly twice as much as DM. Second, the DM GDP effect is short-lived and not statistically
significant after one year, but is much more persistent in EM. Finally, the impact on EM CPI
inflation is approximately five times as large as in DM. This could be due to higher USD
denomination of financing flows and trade transactions, as well as different monetary policy
reactions to external shocks in EM than DM.

Even if only the steel tariff was
implemented and retaliated…
…EM growth would fall 0.7pp,
with CPI inflation rising 0.7pp
Not surprisingly, the effect of the current US steel tariffs is much larger for EM economies
(Figure 10.) than DM economies (Figure 11). We compare first year estimates, because of a
lack of statistical significance for the DM GDP response after the first year. With the steel
tariff alone, our model suggests that EM (DM) GDP growth could fall by 0.3pp (0.1pp) and
inflation rise by 0.3pp (0.1pp). With steel tariff retaliation, EM growth could fall by -0.7pp,
with inflation rising by 0.7pp, which are sizable effects. If the US unilaterally implements IP-
related tariffs on $50bn of goods from China, then EM (DM) growth could fall by 0.9pp
(0.4pp) and inflation rise by 0.8pp (0.15pp). In the case of tit-for-tat retaliation, EM (DM)
real GDP growth falls by 1.7pp (0.7pp) and inflation rises by 1.7pp (0.3pp). Overall, DM
would only really feel any effects from tariffs in this very last scenario, while the impact for
EMs is already sizable if the current steel tariffs are retaliated against.

FIGURE 10 FIGURE 11
The impact in emerging market economies… …is much larger than in developed economies.
2.0 % Annual impact in the first year
2.0 % Annual impact in the first year
1.5 1.5
1.0 1.0
0.5 0.5
0.0
0.0
-0.5
-1.0 -0.5
-1.5 -1.0
-2.0 -1.5
-2.5 -2.0
Steel tariff Steel tariff Steel and Steel, $50bn Steel tariff Steel tariff and Steel and Steel, $50bn IP
without and $50bn IP tariff IP with without retaliataion $50bn IP tariff and retaliation
exemption retaliataion retaliation exemption
EM Real GDP Growth EM CPI inflation DM Real GDP Growth DM CPI inflation
Source: Barclays Research Source: Barclays Research

1 June 2018 7
Barclays | Thinking Macro

But confidence effects may amplify the outcome...

Governments have also used
other instruments to aid
domestic industries
Lack of data makes modelling
these policies difficult
Tariffs are not the only instrument that governments have available to aid specific sectors. As
shown in Figure 12, in net terms, measures have become more trade discriminatory in recent
years. For example, the Obama administration implemented several policies aimed at
protecting the steel sector (Figure 13). The US Transportation Investment Generating
Economic Recovery Act, which applies to major public infrastructure projects, has a ‘Buy
American’ clause for Iron and Steel, requiring a preference for US made steel; 94% of all US
steel imports are covered by this public procurement rule. Anti-dumping, preferential lending,
private-sector incentives to rely on US steel producers and tax reliefs have all been deployed to
support this sector since 2009. Modelling these policies is beyond our current framework.

Our estimates are based on
gradual tariff reductions
Rapid tariff rises can generate
spare capacity and important
second-round price effects…
Our calculations likely underestimate the effect of rapidly rising tariffs because they are
based on gradual tariff reductions. But if tariffs rise rapidly and leave crucial intermediate
inputs stranded elsewhere, prices for final product can rise significantly. Aluminium price
dynamics following US sanctions on the Russian firm Rusal support this view. On the other
hand, rapid rises in tariffs on final goods may generate significant global spare capacity in
the affected sector almost immediately. The price of the affected good should therefore
decline. While these price effects are clearly very important for second-round propagation of
large tariffs, modelling them requires product level data on tariffs, prices and quantities,
which makes understanding these effects beyond the scope of our analysis.

…and adverse confidence,
demand and supply effects
If tariffs rise suddenly, economic uncertainty may increase and consumer confidence fall.
Prolonged uncertainty about the future of the multi-lateral international trade framework that
has been in place since World War II will likely tighten financial conditions. However, even
without the pressure from tightening financial conditions, firms and households may choose
to raise pre-cautionary savings in anticipation of an economic slowdown in the near future.
Finally, prolonged uncertainty about future trade restrictions may lead firms in the real
economy to wait before entering foreign markets and producing abroad. The first two effects
would likely reduce demand and, hence, global activity and inflation. However, the last effect
may lead to a reduction in global supply, thereby raising inflation and lowering output, albeit
likely over a longer time horizon.

Given that our estimates are only based on a gradual reduction in tariffs, the above
confidence and second-round price effects are likely not fully picked up by our model.

FIGURE 12 FIGURE 13
Governments use non-tariff discriminatory trade policies… …as the example of the US steel sector demonstrates

12 Measures Recorded by Global Trade Alert (Number of Share of US Steel imports covered by protection measure
1.0
9 products; thousands)
0.8
6
2009 2016 2018 - Pre Trump Tariff
3 0.6

0 0.4
-3
0.2
-6
-9 0.0

-12
2009 2010 2011 2012 2013 2014 2015
Discriminatory Liberalising Net measures

Source: Global Trade Alert, Barclays Research Source: Global Trade Alert, Barclays Research

1 June 2018 8
Barclays | Thinking Macro

...and the GVC impact could be already partly captured

As we discuss below, the emergence of global value chains (GVCs) in recent years makes
understanding the impact of tariffs significantly more difficult. However, as the rise of GVC’s
coincided with tariff reductions, the impact could already be partly captured by our model.

Intermediate goods value
chain disruption has a greater
effect on output than prices…
Clearly, if tariffs are introduced on intermediate goods, which cannot be obtained elsewhere,
such as rare earths, this could lead to a significant disruption in global activity and likely
exacerbate the initial effect of the policy. Of course, if diversification in trade allows the
intermediate product to be sourced from a different location, the effect could be smaller
than initially expected. However, modern production facilities are highly specialised and in
many cases rerouting the production of intermediate goods on a large scale will likely be
very disruptive to the production process. Overall, this suggests that the first-order
adjustment will likely be greater through output than prices.

…relative to final goods value
chain disruption
Tariffs on final goods in the value chain will likely lead to a greater adjustment in prices.
Once a goods item is assembled, it can be sold anywhere. As a result, firms will likely try to
raise sales in markets that do not impose tariffs. This rise in supply to and competition in the
rest of the world will likely lower prices in these markets. Tariffs on final goods will therefore
likely lead to a greater adjustment in prices relative to real activity.

GVCs coincided with tariff
reduction, which means our
estimates may capture these
effects partially
As the emergence of GVCs coincided with US tariff reductions, the GVC effect may already be
partially included in our estimates (Figure 14). A recent OECD study examined the disinflation
impact of GVCs on individual country PPI disinflation with product level data (Figure 15).
When aggregated for the countries in the sample, in the same manner as our global inflation
variable, the average annual reduction in inflation from GVCs is -0.21%. This is not too
dissimilar to our second and fifth year DM CPI inflation estimates (Figure 15). While our
estimates refer to CPI, rather than PPI, inflation, we think this similarity nevertheless suggests
that part of the GVC effect may already be implicitly included in our VAR model.

Even if the true effect is twice
our estimate, only large tariffs,
as % of US imports, will have
large global effects
However, even if, as a result of these omitted channels, the true effect is twice as large as
our estimates, the overall effect of the steel tariff alone will be moderate. The lesson is: small
tariffs, as % of US imports, have small effects; large ones have larger effects. A scenario
where the US imposes tariffs on $100bn worth of Chinese imports and the Chinese retaliate
is plausible. Whether historical evidence and WTO trade institutions suggest that this can be
avoided is where our discussion turns next.

FIGURE 14 FIGURE 15
Lower US tariffs coincided with emergence of GVCs. While …a recent OECD study shows that the impact on DM PPI
correlation does not mean causation,… inflation is close to our estimate for DM CPI inflation

30 % of US imports 3.5 estimated contribution of GVC to average annual

prodicer price inflation, percentage points
25 3.0 0.1
2.5
20 0
2.0
15 -0.1
1.5
10 -0.2
1.0
5 -0.3 Global weighted average = -0.21
0.5

0 0.0 -0.4
1995 1998 2001 2004 2007 2010 2013
GVC Backward Participation indicator (LHS) -0.5
LUX

JPN
AUT

BEL

PRT

ITA
DNK

CHE
NOR

FRA

GRC

LVA
FIN
NLD

CZE

USA
SWE

DEU
SVK

US Tariffs as Share of imports

Source: Various OECD databases Source: Andrews, Gal and Witheridge (2018), Barclays Research

1 June 2018 9
Barclays | Thinking Macro

The US applied protectionist policy during the 19th century…

The US has a long history of The US imposed high tariffs as a policy tool in the first age of globalisation (Figure 16).
active trade policy

FIGURE 16
Compared to 19th century US trade policy, President Trump’s tariffs seem modest…

70 Tariff of
Abominations Smoot-Hawley Tariff (1930)
60 (1828) Underwood Law (1913)

U.S. Average Tariff Rate (%)

Trade Agreement
50 Act (1934)

40 GATT (1947)

30
Trump's Steel and IP Tariffs
20

10 Morrill Tariff (1861)

0
1820 1838 1856 1874 1892 1910 1928 1946 1964 1982 2000 2018
Total Imports Dutiable Imports
Source: US ITC historical database, Barclays Research

Northern states preferred to
protect industry…
…while Southern states
favoured free trade, as their
supply chain was heavily
import dependent
The historian Paul Bairoch argued that the US was “the homeland and bastion of modern
protectionism’ in the 19th century. One aim of this policy was to protect and subsidise
domestic industries. While congressmen representing Northern states voted for tariffs,
Southern states opposed them because they relied more heavily on imports as part of their
supply chain. The large swings in tariffs therefore likely reflect the continuous political
dispute on trade policy between Northern and Southern states during this time. This debate
and movements in tariffs even persisted after the Civil War. However, this changed in 1945,
when the US, as the strongest post-war industrial power, adopted a policy of free trade.

Large scale acts of retaliation during this time period were rare. Many US tariffs were
But retaliation occurred only
designed to protect US workers and businesses from more efficient and lower wage
during the Great Depression
European and British competition. Despite this intention, retaliation by either France or the
UK was infrequent (Figure 18) and happened only during the Great Depression.

FIGURE 17 FIGURE 18
Tariffs were an important source of finance before the While tariffs aimed to protect the US from UK competition,
introduction of income tax in WW1 retaliations only followed during the ‘Great Depression’

25
% Revenue as share of GDP
60
20
Average Tariff Rate (%)

50

15 40

30
10
20
5
10

0 0
1800 1827 1854 1881 1908 1935 1962 1989 2016 1830 1860 1890 1920 1950 1980 2010
Import Tariffs Social Insurance Income Tax France UK US
Source: NBER Macrohistory dataset Source: Various historical datasets

1 June 2018 10
Barclays | Thinking Macro

…but today the WTO can resolve trade conflicts

The WTO helped to resolve the
dispute about President Bush’s
30% steel tariff
The resolution of previous trade disputes through the WTO framework suggests that this
might still be a viable route out of the current situation.
Trump is not the first US president to use tariffs to protect the US steel industry this century.
In March 2002, President George W Bush’s administration imposed a 30% tariff on a range
of imported steel products, in response to the United States’s falling share in global steel
production. Several countries, including Canada, Mexico, Argentina, Thailand and Turkey,
received an exemption. But the EU and eight other countries complained to the WTO. The
disputes settlement panel and the appeals committee ruled that US steel tariffs were illegal
under WTO rules. The EU alone was entitled to $2.2bn in retaliation measures, or $3.1bn in
2018 USD terms, which is greater than the $2.6bn the EU is entitled to in response to
President Trump’s tariffs. President Bush dropped the tariffs in December 2003 in response.

The WTO even helped the US
to win important IP disputes
with China and Argentina
A second area of recent trade disputes is intellectual property rights. In 2007, the US
complained to the WTO that China violated the Agreement on Trade-Related Aspects of
Intellectual Property Rights (TRIPS). The WTO panel recommended that China conform to
the Copyright Law and Customs measures in the TRIPS agreement and these changes were
approved at the 11th National People’s Congress in March 2010. The US also complained to
the WTO that Argentina’s legal regime governing patents, with respect to pharmaceuticals
and agricultural chemicals, was inconsistent with the TRIPS agreement. The dispute
settlement panel found in favour of the US and an agreement was found soon after. These
examples suggest that even in tough intellectual property disputes, which require changes
to sovereign legal frameworks, the WTO’s resolution mechanism has been effective.

US and EU have an incentive to
keep the WTO in place
The data suggest that the US and the EU have strong incentives to maintain the WTO
resolution framework. They have both used this framework extensively to settle disputes,
significantly more than other countries (Figure 19), and have won 70% of the cases they
have raised (Figure 20). This is a lower bound because it does not include any cases settled
outside of the dispute system. But the EU and US lose 41% of cases in which they are the
responding party. The conclusion of these disputes is then applied to all member countries.
In our view, these results provide strong incentives to maintain the current WTO framework.

Trump’s bilateral approach to
trade negotiations could lead
to WTO reform or irrelevance
There is a risk that President Trump’s bilateral trade negotiations approach will make the
WTO less relevant. At present, he has chosen to resolve trade conflicts bilaterally, rather
than through the WTO, perhaps because the latter process is perceived as unfair and slow.
In a best case scenario, this will lead to WTO reform towards greater transparency and
efficiency. But the risk that the WTO becomes less relevant in trade disputes remains.

FIGURE 19 FIGURE 20
US and EU used WTO dispute resolution mechanism most… …winning 70% of their cases
80 % of cases won
140 # of total WTO disputes
70
120
100 60

80 50

60 40

40 30

20 20
0 10
Canada

Japan
India
US

EU

Brazil

China

Russia

0
BRIC Dev Econ EU & USA Other OECD
WTO dispute complaints by member states Complainant Respondent

Source: WTO Source: Johannesson and Mavroidis (2016)

1 June 2018 11
Barclays | Thinking Macro
Services exports have recently
supported global trade…
…and are subject to regulatory
barriers, rather than tariffs
Falling communication costs
remove another barrier to
services trade
The rising share of services will
make world trade more robust
to goods tariffs
If the US lowers barriers to
entry, services exports to the
US could grow further despite
higher goods tariffs
FIGURE 21
Global services trade rose as share of the total…
26
% of total world exports
24
22
20
18
16
14
12
10
1980 1989
World Services Exports
Source: World Bank WDI, Barclays Research
1 June 2018
Deregulation could support services trade
Tariffs apply only to goods, not services, which are becoming increasingly more important
in global trade, rising from a long-run average of 19.5% of total world exports to a high of
24% in 2016 (Figure 21). Indeed, real global services export growth has persistently
outpaced real global goods export growth, including in 2017, when global trade started to
accelerate again (Figure 22). The share of developing countries in services exports has
increased from 15% in 1970 to 39% today.
While services are not subject to tariffs, non-tariff entry barriers can often inhibit services
trade. Indeed, cross-sectional evidence in OECD countries suggests that countries with lower
domestic barriers to entrepreneurship have a significantly higher share of services in total
imports (Figure 21). Anecdotal evidence confirms this. For example, for a long time, only
airlines headquartered in a given European country were allowed to service purely domestic
routes. Once this changed, trade in travel services rose substantially. Similarly, phone
providers were for a long time restricted from providing services in other countries. Domestic
licensing and regulatory requirements still form significant barriers to services trade.
The other barrier to services trade was likely the elevated cost of providing the same quality
service from a place far away. However, falling communications costs, together with a rise
in the availability of skilled workers in developing countries, has changed this significantly.
For example, providing tutoring lessons to pupils in London from India is increasingly
becoming common practice, which was impractical just a few years ago.
Given these falling barriers to services trade, it is perhaps not surprising that services trade
has outpaced goods trade in the past five years (Figure 22). Overall, we believe that falling
communications costs and deregulation will continue to support services trade. This will
likely make world trade more robust to potential increase in tariffs and, if supported by the
removal of domestic barriers to entry, could at least partially, if not completely, offset any
trade reduction resulting from higher US tariffs in the medium term.
One of President Trump’s campaign promises was to reduce business regulations. If this is
implemented and licensing and regulatory requirements are loosened to allow easier entry
into certain professions for US citizens, then the barriers to entry for services trade should
fall to a degree as well. Legislation aimed at domestic deregulation could therefore have the
unintended consequence of encouraging more services trade. Indeed, according to the US
government accountability office, the number of major rules issued under the Trump
administration is significantly lower than under previous presidents (Figure 24).
FIGURE 22
…as services exports outpaced goods export growth
10 5-year moving average of y/y %
9
8
7
6
5
4
3
2
1
1998 2007 2016 0
1984 1988 1992 1996 2000 2004 2008 2012 2016
World Services exports World Goods exports
Source: World Bank WDI, WTO, Barclays Research
12
Barclays | Thinking Macro

FIGURE 23 FIGURE 24
In the cross-section, lower domestic barriers to entry are …meaning a smaller amount of major rules issued in 2017
associated with a higher share of services in imports… could help boost services trade further, as it has in the past

45 % Share of Services in imports # of major rules by presidential year % y/y

y = -10.855x + 38.954 130 15
40 120
R² = 0.3644
35 110
100 10
30 90
25 80 5
70
20 60
50 0
15
40
10 30 -5
20
5
10
0 0 -10
1.00 1.50 2.00 2.50 3.00 1996 1999 2002 2005 2008 2011 2014 2017

Domestic barriers to entrepreneurship Total Major Rules Real US service trade growth (RHS)

Source: Various OECD databases, Barclays Research. Note: Real US services trade growth is the sum of US services imports and
exports, deflated by the US imports and exports price index, respectively.
Source: US Government Accountability Office, Barclays Research

For EMs, it is less likely that
services trade can mitigate the
impact of US tariffs
Mitigating goods tariffs through greater services trade requires the relevant specialisation
and a skilled labour force. Through their more advanced stage in economic development,
developed economies are more likely to specialise in tradable services. On the other hand,
emerging market economies benefit from globalisation by integrating their large supply of
unskilled labour into the global supply chain. Changing the skill set of large sections of the
population to trade services instead is likely not feasible in the short run. Given the much
larger effects on EM real GDP growth and CPI inflation, it is therefore unlikely that services
trade can mitigate the disruption from greater US tariffs in EM countries.

The return of US protectionism can be disruptive

In this section, we review the main lessons from our econometric exploration of US tariffs.

Our results are subject to
caveats, but still informative
Modelling the impact of US tariffs on short-term global growth and inflation is challenging.
Academic work has focused on the long-term effect, and uncertainty about the impact in the
first year after the tariff announcement is large. Our estimates are based on a gradual tariff
reduction, while the current situation is a rapid tariff rise. The estimates presented in this paper
should therefore be interpreted accordingly. However, they nevertheless provide a first
econometric view on how President Trump’s tariffs might affect the world economy.

Tariffs are a negative supply
shock and big effects require
large tariffs
Our results suggest that US tariffs act like a negative supply shock to the world economy,
lowering global growth and raising inflation. However, only large tariffs produce large
effects: the current US Steel tariff is only 0.33% of US imports and would reduce global
growth only by 0.1pp, while raising global inflation by 0.1pp. It is only when $50bn of IP-
tariffs are added and retaliated tit-for-tat that growth falls 0.6pp, while inflation rises 0.7pp.

EM will be affected to a much
greater extent than DM
The impact on emerging markets is much greater than on developed markets. The EM GDP
growth impact is twice as large as on DM and significantly more persistent. The EM CPI
inflation response is approximately five times as large as in DM. While there are a number of
mitigating factors that are not accounted for, the analysis suggests that EM economies will
be affected to a much greater extent than developed markets.

Overall, US protectionism could be disruptive, especially if tariffs are large and are
retaliated. Emerging markets will likely be more affected than developed markets.

1 June 2018 13
Barclays | Thinking Macro

Appendix A: Data
FIGURE 25
Countries used to construct global GDP and CPI inflation

Asia and Pacific North America Europe

Australia Canada Austria

China Mexico Belgium

India South America Finland

Indonesia Argentina France
Japan Brazil Germany
Korea Chile Italy
Malaysia Peru Netherlands

New Zealand Middle East and Africa Norway

Philippines Saudi Arabia Spain
Singapore South Africa Sweden
Thailand Switzerland
Turkey
United Kingdom
Source: Mohaddes and Raissi (2018).

We construct global real GDP growth and CPI inflation, based on IMF PPP weights, from the
countries in Figure 25. To ensure that high EM inflation episodes do not bias our results, we
exclude periods when inflation exceeds 20% q/q on an annualised basis for longer than a
year. The rest of data in this paper are taken from Haver Analytics.

1 June 2018 14
Barclays | Thinking Macro

Appendix B: Impulse responses

The typical output of a VAR model is an impulse response. In the main text, we present the
annualised average of the median quarterly impulse response for years one, two and five. In
this section, we briefly discuss the impulse response output of our model.

Figures 27-29 show all of the impulse responses output for the baseline model. As can be
seen in Figure 27, the confidence bands for the first year after the shock for global GDP
growth are particularly uncertain, which is why we place less weight on these results
throughout the text. Figures 26 and 27 show that the impact of US tariffs on global growth
and CPI inflation is statistically different from zero for the whole time horizon. The US trade
balance also rises, which is what would be expected after a rise in tariffs.

Figures 30-33 show impulse responses when we split the global series into EM and DM
variables. As discussed above, this shows that the DM GDP effect is only short-lived, but quite
persistent for EM. Overlapping confidence interval tests, not shown here, suggest that EM and
DM CPI inflation responses are significantly different at the 68% level from each other.

FIGURE 26 FIGURE 27
Global real GDP growth impulse response to 1% US tariff rise Global CPI inflation impulse response to 1% US tariff rise
0.2 % 0.8
%
0.0
0.6
-0.2
0.4

-0.4 0.2

-0.6 0.0

-0.8 -0.2

-1.0 -0.4
1 4 7 10 13 16 19 22 25 28 31 34 37 40 1 4 7 10 13 16 19 22 25 28 31 34 37 40
16% confidence band Median 16% confidence band Median
86% confidence band 86% confidence band
Source: Barclays Research Source: Barclays Research

FIGURE 28 FIGURE 29
US trade balance impulse response to 1% US tariff rise US import price index impulse response to 1% US tariff rise

1.2 % 2
%
1.0 1

0.8 0
0.6
-1
0.4
-2
0.2
-3
0.0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 -4
1 4 7 10 13 16 19 22 25 28 31 34 37 40
16% confidence band Median
16% confidence band Median
86% confidence band 86% confidence band
Source: Barclays Research Source: Barclays Research

1 June 2018 15
Barclays | Thinking Macro

FIGURE 30 FIGURE 31
EM real GDP growth impulse response to 1% US tariff rise EM real CPI inflation impulse response to 1% US tariff rise

0.9 % 1.6 %
1.4
0.4 1.2
1.0
-0.1 0.8
0.6
-0.6 0.4
0.2
-1.1 0.0
-0.2
-1.6 -0.4
1 4 7 10 13 16 19 22 25 28 31 34 37 40 1 4 7 10 13 16 19 22 25 28 31 34 37 40
16% confidence band Median 16% confidence band Median
86% confidence band 86% confidence band
Source: Barclays Research Source: Barclays Research

FIGURE 32 FIGURE 33
DM real GDP growth impulse response to 1% US tariff rise EM real CPI inflation impulse response to 1% US tariff rise

0.6 % 0.6 %
0.4 0.5
0.2 0.4
0.0 0.3
-0.2
0.2
-0.4
0.1
-0.6
0
-0.8
-0.1
-1.0
-0.2
-1.2
1 4 7 10 13 16 19 22 25 28 31 34 37 40
1 4 7 10 13 16 19 22 25 28 31 34 37 40
16% confidence band Median
16% confidence band Median
86% confidence band 86% confidence band
Source: Barclays Research Source: Barclays Research

1 June 2018 16
Barclays | Thinking Macro

References
Andrews, D, Gal, P and Witheridge, W (2018), “A Genie in a Bottle? Globalisation,
Competition and Inflation”, OECD Economics Department Working Papers, No. 1462

Barattieri, A, Cacciatore, M and Ghironi, F (2018), “Protectionism and the Business Cycle”
NBER Working paper No. 24353

Chow, G. C. and Lin, A.I. (1971). Best Linear Unbiased Interpolation, Distribution, and
Extrapolation of Time Series by Related Series. The Review of Economics and Statistics 53
(4), 372–375.

Fernald, J (2014), “A Quarterly, Utilization-Adjusted Series on Total Factor Productivity”,

Federal Reserve Bank of San Francisco Working Paper 2012-19,

Johannesson, L and Mavroidis, P (2016), “The WTO Dispute Settlement System 1995‒2015:
A Data Set and its Descriptive Statistics”, Research Institute of Industrial Economics, IFN
Working Paper No. 1148, 2017,

Mohaddes, K. and Raissi, M (2018), “Compilation, Revision and Updating of the Global VAR
(GVAR) Database, 1979Q2-2016Q4”, University of Cambridge: Faculty of Economics

Bairoch, P (1993), “Economics and World History: Myths and Paradoxes”, University of
Chicago Press.

Goldberg, P and Pavcnik, N (2016), “The Effects of Trade Policy”, NBER Working paper No.
21957,

1 June 2018 17
Analyst Certification
I, Tomasz Wieladek, hereby certify (1) that the views expressed in this research report accurately reflect my personal views about any or all of the subject
securities or issuers referred to in this research report and (2) no part of my compensation was, is or will be directly or indirectly related to the specific
recommendations or views expressed in this research report.

Important Disclosures:
Barclays Research is produced by the Investment Bank of Barclays Bank PLC and its affiliates (collectively and each individually, "Barclays").
All authors contributing to this research report are Research Analysts unless otherwise indicated. The publication date at the top of the report reflects the
local time where the report was produced and may differ from the release date provided in GMT.

Availability of Disclosures:
For current important disclosures regarding any issuers which are the subject of this research report please refer to https://publicresearch.barclays.com
or alternatively send a written request to: Barclays Research Compliance, 745 Seventh Avenue, 13th Floor, New York, NY 10019 or call +1-212-526-1072.

Barclays Capital Inc. and/or one of its affiliates does and seeks to do business with companies covered in its research reports. As a result, investors
should be aware that Barclays may have a conflict of interest that could affect the objectivity of this report. Barclays Capital Inc. and/or one of its affiliates
regularly trades, generally deals as principal and generally provides liquidity (as market maker or otherwise) in the debt securities that are the subject of
this research report (and related derivatives thereof). Barclays trading desks may have either a long and / or short position in such securities, other
financial instruments and / or derivatives, which may pose a conflict with the interests of investing customers. Where permitted and subject to
appropriate information barrier restrictions, Barclays fixed income research analysts regularly interact with its trading desk personnel regarding current
market conditions and prices. Barclays fixed income research analysts receive compensation based on various factors including, but not limited to, the
quality of their work, the overall performance of the firm (including the profitability of the Investment Banking Department), the profitability and revenues
of the Markets business and the potential interest of the firm's investing clients in research with respect to the asset class covered by the analyst. To the
extent that any historical pricing information was obtained from Barclays trading desks, the firm makes no representation that it is accurate or complete.
All levels, prices and spreads are historical and do not necessarily represent current market levels, prices or spreads, some or all of which may have
changed since the publication of this document. Barclays Research Department produces various types of research including, but not limited to,
fundamental analysis, equity-linked analysis, quantitative analysis, and trade ideas. Recommendations and trade ideas contained in one type of Barclays
Research may differ from those contained in other types of Barclays Research, whether as a result of differing time horizons, methodologies, or otherwise.
The Barclays Research Department operates independently from the Absa Research Department. Absa Research is produced by Absa Bank Limited acting
through its Corporate and Investment Bank division, which is a part of Barclays Africa Group Limited and affiliated with the Investment Bank of Barclays
Bank PLC. Eligible clients may receive research reports from both research departments, which may reach different conclusions and may contain different
and conflicting forecasts, recommendations, or trade ideas. In order to access Barclays Statement regarding Research Dissemination Policies and
Procedures, please refer to https://publicresearch.barcap.com/S/RD.htm. In order to access Barclays Research Conflict Management Policy Statement,
please refer to: https://publicresearch.barcap.com/S/CM.htm.

All pricing information is indicative only. Prices are sourced from Thomson Reuters as of the last available closing price at the time of production of the
research report, unless another time and source is indicated.

Explanation of other types of investment recommendations produced by Barclays FICC Research:

Trade ideas contained herein that have been produced by the Credit teams within Barclays Research are valid at current market conditions and may not
be otherwise relied upon.
Trade ideas contained herein that have been produced by other research teams within Barclays FICC Research shall remain open until they are
subsequently amended or closed in a future research report.

Disclosure of previous investment recommendations produced by Barclays FICC Research:

Barclays FICC Research may have published other investment recommendations in respect of the same securities/instruments recommended in this
research report during the preceding 12 months. To view previous investment recommendations published by Barclays FICC Research in the preceding
12 months please refer to https://live.barcap.com/go/research/Recommendations.

Legal entities involved in producing Barclays Research:

Barclays Bank PLC (Barclays, UK)
Barclays Capital Inc. (BCI, US)
Barclays Securities Japan Limited (BSJL, Japan)
Barclays Bank PLC, Hong Kong branch (Barclays Bank, Hong Kong)
Barclays Capital Canada Inc. (BCCI, Canada)
Barclays Bank Mexico, S.A. (BBMX, Mexico)
Barclays Securities (India) Private Limited (BSIPL, India)
Barclays Bank PLC, India branch (Barclays Bank, India)
Barclays Bank PLC, Singapore branch (Barclays Bank, Singapore)

Disclaimer:
This publication has been produced by Barclays Research Department in the Investment Bank of Barclays Bank PLC and/or one or more of its affiliates
(collectively and each individually, "Barclays"). It has been distributed by one or more Barclays affiliated legal entities listed below. It is provided to our
clients for information purposes only, and Barclays makes no express or implied warranties, and expressly disclaims all warranties of merchantability or
fitness for a particular purpose or use with respect to any data included in this publication. To the extent that this publication states on the front page that
it is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared
for retail investors under U.S. FINRA Rule 2242, it is an “institutional debt research report” and distribution to retail investors is strictly prohibited. Barclays
also distributes such institutional debt research reports to various issuers, regulatory and academic organisations for informational purposes and not for
the purpose of making investment decisions regarding any debt securities. Any such recipients that do not want to continue receiving Barclays
institutional debt research reports should contact debtresearch@barclays.com. Barclays will not treat unauthorized recipients of this report as its clients
and accepts no liability for use by them of the contents which may not be suitable for their personal use. Prices shown are indicative and Barclays is not
offering to buy or sell or soliciting offers to buy or sell any financial instrument.
Without limiting any of the foregoing and to the extent permitted by law, in no event shall Barclays, nor any affiliate, nor any of their respective officers,
directors, partners, or employees have any liability for (a) any special, punitive, indirect, or consequential damages; or (b) any lost profits, lost revenue, loss
of anticipated savings or loss of opportunity or other financial loss, even if notified of the possibility of such damages, arising from any use of this
publication or its contents.
Other than disclosures relating to Barclays, the information contained in this publication has been obtained from sources that Barclays Research believes
to be reliable, but Barclays does not represent or warrant that it is accurate or complete. Barclays is not responsible for, and makes no warranties
whatsoever as to, the information or opinions contained in any written, electronic, audio or video presentations of third parties that are accessible via a
direct hyperlink in this publication or via a hyperlink to a third-party web site (‘Third-Party Content’). Any such Third-Party Content has not been adopted
or endorsed by Barclays, does not represent the views or opinions of Barclays, and is not incorporated by reference into this publication. Third-Party
Content is provided for information purposes only and Barclays has not independently verified its accuracy or completeness.
The views in this publication are solely and exclusively those of the authoring analyst(s) and are subject to change, and Barclays Research has no
obligation to update its opinions or the information in this publication. Unless otherwise disclosed herein, the analysts who authored this report have not
received any compensation from the subject companies in the past 12 months. If this publication contains recommendations, they are general
recommendations that were prepared independently of any other interests, including those of Barclays and/or its affiliates, and/or the subject companies.
This publication does not contain personal investment recommendations or investment advice or take into account the individual financial circumstances
or investment objectives of the clients who receive it. The securities and other investments discussed herein may not be suitable for all investors. Barclays
is not a fiduciary to any recipient of this publication. Investors must independently evaluate the merits and risks of the investments discussed herein,
consult any independent advisors they believe necessary, and exercise independent judgment with regard to any investment decision. The value of and
income from any investment may fluctuate from day to day as a result of changes in relevant economic markets (including changes in market liquidity).
The information herein is not intended to predict actual results, which may differ substantially from those reflected. Past performance is not necessarily
indicative of future results.
This document is being distributed (1) only by or with the approval of an authorised person (Barclays Bank PLC) or (2) to, and is directed at (a) persons in
the United Kingdom having professional experience in matters relating to investments and who fall within the definition of "investment professionals" in
Article 19(5) of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005 (the "Order"); or (b) high net worth companies,
unincorporated associations and partnerships and trustees of high value trusts as described in Article 49(2) of the Order; or (c) other persons to whom it
may otherwise lawfully be communicated (all such persons being "Relevant Persons"). Any investment or investment activity to which this
communication relates is only available to and will only be engaged in with Relevant Persons. Any other persons who receive this communication should
not rely on or act upon it. Barclays Bank PLC is authorised by the Prudential Regulation Authority and regulated by the Financial Conduct Authority and
the Prudential Regulation Authority and is a member of the London Stock Exchange.
The Investment Bank of Barclays Bank PLC undertakes U.S. securities business in the name of its wholly owned subsidiary Barclays Capital Inc., a FINRA
and SIPC member. Barclays Capital Inc., a U.S. registered broker/dealer, is distributing this material in the United States and, in connection therewith
accepts responsibility for its contents. Any U.S. person wishing to effect a transaction in any security discussed herein should do so only by contacting a
representative of Barclays Capital Inc. in the U.S. at 745 Seventh Avenue, New York, New York 10019.
Non-U.S. persons should contact and execute transactions through a Barclays Bank PLC branch or affiliate in their home jurisdiction unless local
regulations permit otherwise.
Barclays Bank PLC, Paris Branch (registered in France under Paris RCS number 381 066 281) is regulated by the Autorité des marchés financiers and the
Autorité de contrôle prudentiel. Registered office 34/36 Avenue de Friedland 75008 Paris.
This material is distributed in Canada by Barclays Capital Canada Inc., a registered investment dealer, a Dealer Member of IIROC (www.iiroc.ca), and a
Member of the Canadian Investor Protection Fund (CIPF).
Subject to the conditions of this publication as set out above, Absa Bank Limited, acting through its Corporate and Investment Bank division, which is a
part of Barclays Africa Group Limited and affiliated with the Investment Bank of Barclays Bank PLC, an authorised financial services provider (Registration
No.: 1986/004794/06. Registered Credit Provider Reg No NCRCP7), is distributing this material in South Africa. Absa Bank Limited is regulated by the
South African Reserve Bank. This publication is not, nor is it intended to be, advice as defined and/or contemplated in the (South African) Financial
Advisory and Intermediary Services Act, 37 of 2002, or any other financial, investment, trading, tax, legal, accounting, retirement, actuarial or other
professional advice or service whatsoever. Any South African person or entity wishing to effect a transaction in any security discussed herein should do so
only by contacting a representative of Absa Bank Limited acting through its Corporate and Investment Bank division in South Africa, 7th Floor, Barclays
Towers West, 15 Troye Street, Johannesburg. Absa Bank Limited is an affiliate of the Barclays group.
All Barclays research reports are distributed to institutional investors in Japan by Barclays Securities Japan Limited. Barclays Securities Japan Limited is a
joint-stock company incorporated in Japan with registered office of 6-10-1 Roppongi, Minato-ku, Tokyo 106-6131, Japan. It is a subsidiary of Barclays
Bank PLC and a registered financial instruments firm regulated by the Financial Services Agency of Japan. Registered Number: Kanto Zaimukyokucho
(kinsho) No. 143.
Barclays Bank PLC, Hong Kong Branch is distributing this material in Hong Kong as an authorised institution regulated by the Hong Kong Monetary
Authority. Registered Office: 41/F, Cheung Kong Center, 2 Queen's Road Central, Hong Kong.
All Indian securities-related research and other equity research produced by Barclays’ Investment Bank are distributed in India by Barclays Securities
(India) Private Limited (BSIPL). BSIPL is a company incorporated under the Companies Act, 1956 having CIN U67120MH2006PTC161063. BSIPL is
registered and regulated by the Securities and Exchange Board of India (SEBI) as a Research Analyst: INH000001519; Portfolio Manager INP000002585;
Stock Broker/Trading and Clearing Member: National Stock Exchange of India Limited (NSE) Capital Market INB231292732, NSE Futures & Options
INF231292732, NSE Currency derivatives INE231450334, Bombay Stock Exchange Limited (BSE) Capital Market INB011292738, BSE Futures & Options
INF011292738; Depository Participant (DP) with the National Securities & Depositories Limited (NSDL): DP ID: IN-DP-NSDL-299-2008; Investment
Adviser: INA000000391. The registered office of BSIPL is at 208, Ceejay House, Shivsagar Estate, Dr. A. Besant Road, Worli, Mumbai – 400 018, India.
Telephone No: +91 2267196000. Fax number: +91 22 67196100. Any other reports produced by Barclays’ Investment Bank are distributed in India by
Barclays Bank PLC, India Branch, an associate of BSIPL in India that is registered with Reserve Bank of India (RBI) as a Banking Company under the
provisions of The Banking Regulation Act, 1949 (Regn No BOM43) and registered with SEBI as Merchant Banker (Regn No INM000002129) and also as
Banker to the Issue (Regn No INBI00000950). Barclays Investments and Loans (India) Limited, registered with RBI as Non Banking Financial Company
(Regn No RBI CoR-07-00258), and Barclays Wealth Trustees (India) Private Limited, registered with Registrar of Companies (CIN
U93000MH2008PTC188438), are associates of BSIPL in India that are not authorised to distribute any reports produced by Barclays’ Investment Bank.
Barclays Bank PLC Frankfurt Branch distributes this material in Germany under the supervision of Bundesanstalt für Finanzdienstleistungsaufsicht (BaFin).
This material is distributed in Mexico by Barclays Bank Mexico, S.A.
Nothing herein should be considered investment advice as defined in the Israeli Regulation of Investment Advisory, Investment Marketing and Portfolio
Management Law, 1995 (“Advisory Law”). This document is being made to eligible clients (as defined under the Advisory Law) only. Barclays Israeli
branch previously held an investment marketing license with the Israel Securities Authority but it cancelled such license on 30/11/2014 as it solely
provides its services to eligible clients pursuant to available exemptions under the Advisory Law, therefore a license with the Israel Securities Authority is
not required. Accordingly, Barclays does not maintain an insurance coverage pursuant to the Advisory Law.
Barclays Bank PLC in the Dubai International Financial Centre (Registered No. 0060) is regulated by the Dubai Financial Services Authority (DFSA).
Principal place of business in the Dubai International Financial Centre: The Gate Village, Building 4, Level 4, PO Box 506504, Dubai, United Arab Emirates.
Barclays Bank PLC-DIFC Branch, may only undertake the financial services activities that fall within the scope of its existing DFSA licence. Related financial
products or services are only available to Professional Clients, as defined by the Dubai Financial Services Authority.
Barclays Bank PLC in the UAE is regulated by the Central Bank of the UAE and is licensed to conduct business activities as a branch of a commercial bank
incorporated outside the UAE in Dubai (Licence No.: 13/1844/2008, Registered Office: Building No. 6, Burj Dubai Business Hub, Sheikh Zayed Road, Dubai
City) and Abu Dhabi (Licence No.: 13/952/2008, Registered Office: Al Jazira Towers, Hamdan Street, PO Box 2734, Abu Dhabi).
Barclays Bank PLC in the Qatar Financial Centre (Registered No. 00018) is authorised by the Qatar Financial Centre Regulatory Authority (QFCRA).
Barclays Bank PLC-QFC Branch may only undertake the regulated activities that fall within the scope of its existing QFCRA licence. Principal place of
business in Qatar: Qatar Financial Centre, Office 1002, 10th Floor, QFC Tower, Diplomatic Area, West Bay, PO Box 15891, Doha, Qatar. Related financial
products or services are only available to Business Customers as defined by the Qatar Financial Centre Regulatory Authority.
This material is distributed in the UAE (including the Dubai International Financial Centre) and Qatar by Barclays Bank PLC.
This material is not intended for investors who are not Qualified Investors according to the laws of the Russian Federation as it might contain information
about or description of the features of financial instruments not admitted for public offering and/or circulation in the Russian Federation and thus not
eligible for non-Qualified Investors. If you are not a Qualified Investor according to the laws of the Russian Federation, please dispose of any copy of this
material in your possession.
This material is distributed in Singapore by the Singapore branch of Barclays Bank PLC, a bank licensed in Singapore by the Monetary Authority of
Singapore. For matters in connection with this report, recipients in Singapore may contact the Singapore branch of Barclays Bank PLC, whose registered
address is 10 Marina Boulevard, #23-01 Marina Bay Financial Centre Tower 2, Singapore 018983.
This material is distributed to persons in Australia by either Barclays Bank PLC, Barclays Capital Inc., Barclays Capital Securities Limited or Barclays Capital
Asia Limited. None of Barclays Bank PLC, nor any of the other referenced Barclays group entities, hold an Australian financial services licence and instead
they each rely on an exemption from the requirement to hold such a licence. This material is intended to only be distributed to “wholesale clients” as
defined by the Australian Corporations Act 2001.
IRS Circular 230 Prepared Materials Disclaimer: Barclays does not provide tax advice and nothing contained herein should be construed to be tax advice.
Please be advised that any discussion of U.S. tax matters contained herein (including any attachments) (i) is not intended or written to be used, and
cannot be used, by you for the purpose of avoiding U.S. tax-related penalties; and (ii) was written to support the promotion or marketing of the
transactions or other matters addressed herein. Accordingly, you should seek advice based on your particular circumstances from an independent tax
advisor.
© Copyright Barclays Bank PLC (2018). All rights reserved. No part of this publication may be reproduced or redistributed in any manner without the prior
written permission of Barclays. Barclays Bank PLC is registered in England No. 1026167. Registered office 1 Churchill Place, London, E14 5HP. Additional
information regarding this publication will be furnished upon request.
EU30683 BARCRES-63852b76