Emerging Markets

Highlights So What? Markets remain complacent about U.S.-China trade. Why? The U.S. has escalated the trade war by threatening sanctions on key Chinese tech firms. Chinese President Xi Jinping is preparing his domestic audience for protracted struggle. U.S. domestic politics do not prohibit, and likely encourage, a tough stance on China. Farmers are not a constraint on Trump — economic growth is. Go long spot gold and JPY-USD. Feature Markets remain complacent. Chart 1 suggests that while the combination of unilateral trade tariffs and spiking U.S. 10-year Treasury yields was enough to sink the S&P 500 in 2018, the former alone cannot do so today. Chart 1Tariffs Alone Not Enough To Sink Equities? Wrong. Specifically, the increase in the Section 301 tariff rate from 10% to 25% on $200 billion worth of Chinese imports and the threat of a new 25% tariff on the remaining $300 billion worth of Chinese imports in just a month’s time has only led to a 3% pullback in equities since May 3. That was the last trading day prior to President Donald Trump’s infamous tweet about hiking the tariff. Unlike the trade war escalation in October through November of last year, the Federal Reserve is no longer hiking rates, China’s economic indicators have bottomed, and U.S. equity investors have now fully imbibed the “Art of the Deal.” The consensus holds that the escalation of trade tensions with China is contained within the context of Trump’s well-known routine of inflicting pain and then compromising. We would wager that the bond market is right and equities are wrong. Equities will converge to the downside, unless the market receives a concrete positive catalyst that improves the near-term outlook for U.S.-China relations and hence global trade. The problem is that for equities such a catalyst could happen at any time in the form of additional Chinese stimulus. Therefore, higher volatility is the only guaranteed outcome. The sudden onslaught of U.S. pressure makes it harder for Chinese President Xi Jinping to offer structural concessions to his American counterpart without looking weak. It was easier to do so when the threat of tariffs was under wraps, as was the case between December 1 and May 5. This new obstacle informed our decision to close out our long China equities and long copper trades and downgrade our end-June trade deal probability from 50% to 40%. But the escalation of tensions makes stimulus more likely to surprise to the upside, which will at least partially offset the negative hit to global sentiment and the trade outlook. Waiting For A Positive Political Intervention Three negative geopolitical catalysts loom in plain sight, while investors are still waiting on a positive catalyst. The negatives: China has not yet announced retaliation to the U.S. Commerce Department’s blacklisting of Huawei and a handful of other Chinese tech firms; the U.S. could implement the blacklist within three months, increasing the risk of a broader “tech blockade” against China; and the U.S. authorities are prepared to extend tariffs to all Chinese goods in one month. Meanwhile there are no high-level talks currently scheduled between the principal Chinese and American negotiators as we go to press. This could change quickly. But if negotiating teams do not hold substantive meetings with positive reports afterwards, then investors cannot be sure that Presidents Donald Trump and Xi Jinping will speak to each other, let alone finalize a substantive trade deal, at the G20 in Japan on June 28-29. The macro backdrop is hardly encouraging: global export volumes are contracting and the dollar’s fall may be arrested amid a huge spike in global policy uncertainty. Any rebound in the greenback will pile additional pressure onto trade flows, at least until the market sees a substantial increase in Chinese stimulus (Chart 2). Furthermore, it is concerning that President Trump, a businessman president and champion of American manufacturing, is raising tariffs at a time when lending and factory activity are already slowing in the politically vital Midwestern states (Chart 3). The implication is that he is unfazed by economic risks and therefore less predictable. He is pursuing long-term national foreign policy objectives at the expense of everything else. This may be patriotic but it will be painful for global equity investors. Chart 2Trump Unfazed By Deteriorating Global Economy Chart 3Economic Activity Is Already Slowing Chart 4Markets Blasé About Looming Risks It is not only the S&P 500 that is failing to register the dangerous combination of weak global trade and escalating U.S.-China strategic conflict. Our colleague Anastasios Avgeriou of the BCA U.S. Equity Strategy points out that the “Ted spread,” the premium charged on interbank lending over the risk-free rate, is as docile as the safe-haven Japanese yen (Chart 4). President Xi Jinping, however, is not so blasé. He took a trip to Jiangxi province on May 20 to declare that China is embarking on a “new Long March.” This is a reference to the legendary strategic withdrawal executed by the early Chinese Communist Party in its civil war against the nationalists in 1934-35. It was an 8,000-mile slog across the rugged terrain of western and central China, peppered with battles against warlords and nationalists, in which nearly nine-tenths of the communist troops never made it. It is a historical event of immense propagandistic power used to celebrate the CPC’s resilience and ultimate triumph over corrupt and capitalist forces backed by imperialist Western powers. Most importantly, the Long March culminated in Mao Zedong’s consolidation of power over the party and ultimately the nation. In short, President Xi just told President Trump to “bring it on,” as he apparently believes that a conflict with the U.S. will strengthen his rule. The S&P 500 and the “Ted spread” are failing to register the dangerous combination of weak global trade and escalating U.S.-China strategic conflict. Trump, meanwhile, operates on a much shorter time horizon. He is coming closer to impeachment, as House Speaker Nancy Pelosi sharpens her rhetoric and negotiations over a bipartisan infrastructure bill collapse. Impeachment will fail and in the process will most likely help Trump’s reelection chances. But gridlock at home means that one of our top five “Black Swan” risks for 2019 is now being activated: Trump is at risk of becoming a lame duck and is therefore looking for conflicts abroad as a way of stirring up support at home. Bottom Line: The bad news in the trade war is all-too-apparent while good news is elusive. Yet key “risk off” indicators have hardly responded. We recommend going long JPY-USD on a cyclical basis on the expectation that the market will continue to have indigestion until a positive catalyst emerges in the trade talks. Trump’s Trade War Calculus The trade war is focused on China more so than other states – and Trump likely has the public backing for such a conflict. President Trump delayed any Section 232 tariffs on auto and auto parts imports this month as the China trade war escalated (Chart 5). This confirms our reasoning that the nearly 50/50 risk of tariffs on car imports from Europe and Japan (recently upgraded from 35%) is contingent on first wrapping up a China deal. Another signal that Trump is conscientious not to saddle the equity market with too many trade wars is the decision finally to exempt Canada and Mexico from Section 232 aluminum and steel tariffs (Chart 6). It is now possible for Canada to ratify the deal before parliament dissolves in late June and for the U.S. and Mexico to follow. American ratification will involve twists and turns as the Democrats raise challenges but their obstructionism is ultimately fruitless as it will not hurt Trump’s approval ratings and labor unions largely support the new deal. Meanwhile a major hurdle relating to Mexican labor standards has already been met. These are positive developments for these markets and yet they call attention to a critical point about the Trump administration’s trade strategy: Trump has not shown much willingness to compromise his trade demands with allies in order to secure their cooperation in pressuring China. The threat of car tariffs is still looming over Europe (and even Japan and South Korea). In fact, a united front among these players would have made it much harder for China to resist structural changes (Chart 7). Chart 6Canada And Mexico Are Off The Hook Chart 7A 'Coalition Of The Willing' Would Be More Effective Nevertheless, we have long held that China, not NAFTA or Europe, would be the focus of Trump’s ire because there is much greater consensus within the U.S. political establishment on the need for a more muscular approach to China grievances, and hence fewer constraints on Trump. This view has now come full circle, at least for the time being. Bear in mind that while Republicans and even Democrats have a favorable view of international trade, in keeping with an improving economy (Chart 8), the U.S. as a whole is more skeptical of free trade than most other countries (Chart 9). The economy is insulated and globalization has operated unchecked for several decades, generating resentment. This is especially relevant with China. Americans have an unfavorable view of China’s trade practices and China in general (Charts 10 and 11). This perception is getting worse as the great power competition heats up. Even a majority or near-majority of Democrats view China’s cyber-attacks, ownership of U.S. debt, environmental policies, and economic competition as causes of real concern (Chart 12). This means Trump is closer to the median voter when he is tough on China. The result is a lower chance of a “weak deal,” i.e. a short-term deal to reduce the trade deficit primarily through Chinese purchases of commodities, since this will be a political liability for Trump. He may be forced into such a deal if the market revolts (say 35% odds). But otherwise he will hold out for something better, which Xi Jinping may be unwilling to give. China, not NAFTA or Europe, is the focus of Trump’s ire. This is why we rank “no deal” at 50%, more likely than any kind of deal (40%), though there is some chance of an extension of talks beyond the June G20 (10%). Bottom Line: The delay of auto tariffs and progress in replacing NAFTA suggest that the Trump administration is cognizant of the negative market impact of its trade wars and the need to focus on China. However, the risks to Europe and Japan are not yet removed. And any Chinese concessions will be weaker than might otherwise have been possible had Trump created a “coalition of the willing” to prosecute China’s violations of global trading norms. A weak deal makes it more likely that strategic conflict is the result. Trump Beats Bernie Beats Biden? Or Vice Versa? U.S. domestic politics are also pushing Trump in the direction of conflict with China. The American voter’s distrust of China explains why former Vice President Joe Biden, and leading contender for the Democratic Party nomination in 2020, recently caught flak from both sides of the aisle for being soft on China. At a campaign stop in Iowa on May 1, Biden said, “China is going to eat our lunch? Come on, man … They’re not competition for us.” He has made similarly dovish comments in the recent past. It makes sense, then, that Trump is trying to link “Sleepy Joe” (as he calls Biden) with weakness on China and trade. Biden, who is still enjoying a very sizable bump to his polling a month after formally announcing his candidacy (Chart 13), is a direct threat to Trump’s electoral strategy of maximizing white blue-collar turnout and support, particularly in the Midwestern swing states. Biden was on the ticket when President Barack Obama won these states in 2008 and 2012. He is a native son of Pennsylvania. And he appeals to the same voters as a plain-talking everyman. Both Biden and Democratic Socialist Bernie Sanders of Vermont are beating Trump in the very early head-to-head polling for the 2020 presidential race. In fact, Sanders has a bigger lead over Trump than Biden in many of these polls (Chart 14). Yet Sanders has a narrower path to victory in the general election – he is heavily dependent on the Rustbelt, where he could either win based on repeating the 2016 results in a new demographic context (the “Status Quo” scenario in Chart 15), or by winning back the blue-collar voters who abandoned the Democrats for Trump in 2016 (the “Blue Collar Democrats” scenario). Sanders performed well in these states in the Democratic primary in 2016, whereas he struggled in the South. Chart 16Democrats Swung Too Far Left For Many Independents Biden, on the other hand, is capable of winning not only in these two scenarios, but also by rebuilding the Obama coalition. He has a better bid to win over the black community due to his close association with Obama and his command of Democratic Party machinery, plus potentially his choice of running mate (the “Obama vs. Trump” scenario). By this means Biden, unlike Sanders, can compete against Trump in the Sun Belt and South in addition to the Midwest. Therefore, it is all the more imperative for Trump to try to corner Biden and frame the debate about Biden early. Trump may also be betting that despite the head-to-head polling, Sanders is too far left for the median voter. While the Democratic Party swings sharply to the left, the median voter remains more centrist, judging by the fact that independent voters (who make up half the electorate now) only slightly favor Democrats over Republicans, a trend that is only slightly rising (Chart 16). Biden’s polling is strong enough that he holds out the prospect of winning the Democratic nomination relatively smoothly, without deepening the ideological split in the party too much. Whereas Trump would benefit in the general election if Democrats suffered an internal split over a bloody primary season in which Bernie Sanders clawed his way to the nomination. The hit to American farmers is probably not a significant political constraint on President Trump waging his trade war. The upshot is that Trump is vulnerable in U.S. politics and will attempt to take action to strengthen his position. Meanwhile if Biden’s position on trade changes then we will know that he reads the Midwestern voter the same way Trump does – as a protectionist. Bottom Line: Trump’s eagerness to attack Biden reveals the specific threat that Biden poses to Trump’s electoral strategy as well as Trump’s calculus that a belligerent position on China is a vote-getter in the key Midwestern swing states. We expect Biden to become more hawkish on China, which will emphasize the long-term nature of the U.S.-China struggle and confirm the median voter’s appetite for hawkish policy. American Farmers Unlikely To Alter The 2020 Playing Field Yet can Trump’s political base withstand the trade war? And can he possibly win the swing states if the trade war is escalating and damaging pocketbooks? There are many stories about farmers in the Midwest and other purple states who are deeply alarmed at Trump’s trade policies, prompting questions about whether he could be unseated there. American farmers have been among the hardest hit in the trade war. China was a major market for U.S. agricultural exports prior to the conflict (Chart 17). Since then U.S. agriculture has struggled, as exports to China have declined by more than 50% y/y in 2018 (Chart 18). Agricultural commodity prices are down ~10% since a year ago, with soybeans – the poster child of the conflict – trading at 10 year lows. Net farm incomes – a broad measure of profits – were on a downward trend prior to the trade war (Chart 19). While the USDA estimates that overall U.S. farm income will increase by 8.1% y/y this year, this follows a nearly 18% y/y decline in 2018 to reach the lowest level since 2002 (Chart 20). The recent escalation of the trade war will weigh on these incomes. A common narrative in the financial media is that this hit to American farmers is a significant political constraint on President Trump in waging his trade war. He could be forced to accept a watered-down deal with China to preserve this voting bloc’s support ahead of November 2020, the thinking goes. Possibly, but probably not because of farmers abandoning the Republican Party en masse. First of all, rural counties and small towns continued supporting the Republican Party in the 2018 midterms, at a time when the initial negative impact of the trade war was front-page news (Chart 21). Second, some of the key farm states are unlikely to be key swing states in the election. Take soybeans, for example. Prior to the trade war, nearly 60% of U.S. soybean exports, and more than a third of U.S. soybeans, ended up in China. Illinois is the top producer, followed by Iowa and Minnesota. Last year soybean production in these three states accounted for 15%, 13%, and 8% of total U.S. production, respectively. As such, agriculture and livestock products exports to China in 1Q2019 are down 76% y/y in Illinois and 97% y/y in Minnesota. However, Trump won Iowa by nearly 150 thousand votes, a 9.4% margin, and there are not enough farmers in the state to overturn that margin. The negative impact on soybeans could prevent Trump from picking up Minnesota, where he lost by only 1.5% of the vote. But Minnesota is unlikely to cost him the White House in 2020. The picture is different in the key swing states of Michigan, Pennsylvania, and Wisconsin. Farming accounts for only ~1% of jobs in Michigan, Ohio, and Pennsylvania – and 2.3% of jobs in Wisconsin – and thus farmers represent a small share of the voting bloc in these states (Chart 22). But Trump won Michigan by a mere 0.23% of the vote, Pennsylvania by 0.72%, and Wisconsin by 0.77%. If one-fifth of farmers in these states switched their vote, Trump’s 2016 margin of victory would vanish. Of course, manufacturers are a much larger voting bloc (Chart 23). And rural voters are unlikely to shift to the Democrats on such a large scale. Moreover, ag exports from these states have generally held up (Chart 24), the majority of their exports are destined for North America rather than China. The benefit from the recent thaw in North American trade relations will outweigh the loss of China as a market (Chart 25). The Trump administration is also producing an aid package worth at least $15 billion to shield farmers at least partially from the trade war impact.1 This compares to an estimated $12 billion loss in net farm income in 2018. Ultimately, Trump is much more threatened by other voting groups in these states. Young voters, women, minorities, suburbanites, and college-educated white voters all pose a threat to his thin margins if they turn out to vote and/or increase their support for the Democratic Party in 2020. A surge in Millennials, for instance, played the chief role in unseating Republican Governor Scott Walker in Wisconsin in 2018 (Chart 26). While midterm elections differ fundamentally from presidential elections, the Republicans lost 10 out of 12 significant elections in the Midwest during the midterms (Table 1). Table 1Republicans Lost Almost All Significant Midwest Elections In The Midterm It is true that the winning Democratic candidates in the six major statewide races in Michigan, Pennsylvania, and Wisconsin all had voters who believed Trump’s trade policies were more likely to “hurt” the local economy than help it, according to exit polls (Chart 27). At the same time, a majority of voters believed that the trade policies either “helped” the local economy or “had no impact,” as opposed to hurting it. And Democrats are somewhat divided on this issue. Health care, not the economy, was the primary concern of voters. Moreover, health care, not the economy, was the primary concern of voters, especially Democratic voters (Chart 28). Republicans cared more about the economy and tended to support Trump’s trade policies. In sum, unless the trade war causes a general economic slowdown that changes voter priorities, Trump’s chief threat in 2020 comes from urban and suburban voters angry over his attempt to dismantle the Affordable Care Act, rather than from farmers suffering from the trade war. The large bloc of manufacturing workers in the Midwestern battleground states helps to explain why Trump is willing to wage a trade war at such a critical time: loyal rural counties bear the brunt of the economic pain yet a tough-on-China policy could bring out swing voters from the manufacturing sector in suburbs and cities. Bottom Line: Trump could very well lose agriculture-heavy swing states in 2020, but it would not be because of losing his base among rural voters. Rather, it would be a result of a broader economic slowdown – or a superior showing of key demographic groups in favor of Democrats for other reasons like health care. The large bloc of manufacturing voters relative to Trump’s margins of victory helps to explain his aggressive posture on the trade war. Investment Conclusions Go long JPY-USD on a cyclical, 12-month horizon in the context of escalating trade war, complacent markets, and yet the prospect of additional Chinese stimulus improving global growth. This trade should be reinforced by the specific hurdles facing Japan over the next three to 18 months. While we would not be surprised if a trade agreement with the U.S. is concluded quickly, even ahead of any U.S.-China deal, nevertheless Japan faces upper house elections, a potential consumption tax hike, and preparations for a contentious constitutional revision and popular referendum on the cyclical horizon. On the expectation of greater Chinese stimulus, we are maintaining our long China Play Index call, which is up 2.2%. As a hedge against both geopolitical risk and the impact of Chinese stimulus over the cyclical horizon, go long spot gold.   Matt Gertken, Vice President Geopolitical Strategy mattg@bcaresearch.com Roukaya Ibrahim, Editor/Strategist Geopolitical Strategy RoukayaI@bcaresearch.com Footnotes 1 While the plan is yet to be finalized, payments of ~$2/bushel to soybean farmers, $0.63/bushel to wheat farmers, and $0.04/bushel to corn farmers are under consideration. Unlike last year when the payments were distributed according to farmers’ current production, a potential modification to this year’s plan is that the payments will be distributed based on this years’ planted acreage and past yields.

Special Report

Highlights This report explores the structural potential for the A-share market by reviewing the performance of several MSCI factor indexes, as well as a number of our own factor portfolios. The persistent outperformance of several fundamental factors in China’s domestic equity market may suggest that A-shares are a less efficient market than other emerging or developed markets, but the evidence does not support the idea that A-shares are a “casino” market that is purely driven by speculation. An A-share portfolio formed of industry groups with above-median return on equity and below-median ex-post beta has significantly outperformed over the past decade, and we are comfortable recommending it relative to the MSCI China A Onshore index to global investors who are looking to increase their secular exposure to domestic Chinese stocks. Feature Chart 1Interest In A-Shares Rose Sharply In Q1, In Lockstep With Prices Structural interest in domestic Chinese stocks is growing among global investors, which has been heightened somewhat over the past year by MSCI’s decision to progressively include A-shares in the MSCI Emerging Markets Index as well as the ongoing global search for yield/value. BCA’s China Investment Strategy team received many inquiries about the domestic market in Q1 of this year; this was mostly a response to the eye-popping 33% return for the market in the first quarter (Chart 1), but it also reflects a growing recognition of the domestic market and a desire among investors to increase their awareness. The performance of domestic stocks over the past decade has caused some investors to refer to A-shares as a “casino” market; i.e., a highly-volatile and speculative market divorced from fundamental performance, that is primarily driven by momentum-oriented retail investors. But, we have pushed back against the “strong-form” of this view in previous reports,1 arguing that while the A-share market is certainly volatile, it does not appear to be particularly irrational. In this Special Report we explore the structural potential for the A-share market by reviewing the performance of several MSCI factor indexes, both to identify profitable investment opportunities and to further investigate the claim that playing in the A-share market is akin to betting at a casino. We also build several of our own factor portfolios based on similar attributes at the industry group level, and compare the performance of these portfolios to the MSCI factor indexes. Despite obvious near-term risks facing the Chinese economy and equity market, we conclude by recommending a long position in a specific high-performance factor strategy relative to the A-share benchmark, and highlight how the factor performance of Chinese and global stocks challenges the view that the domestic market is driven largely by sentiment rather than fundamentals. The Performance Of MSCI’s Factor Indexes: Onshore China Vs. Global Table 1 presents the six MSCI factor indexes available for both the All Country World and China A Onshore parent indexes, along with a brief description of MSCI’s construction methodology. The methodology for the high dividend yield, size, and value portfolios are relatively straightforward, whereas the construction of the quality, momentum, and low volatility indexes is more involved and definition-dependent. Table 1Description Of MSCI Single Factor Indexes MSCI defines quality stocks as those having a high return on equity, a low debt/equity ratio, and low earnings variability. Momentum portfolios are build using an average of 6- and 12-month return measures (adjusted for the most recent month’s return to eliminate any short-term reversal effects). Finally, MSCI’s low volatility methodology employs a constrained portfolio optimization approach, where the weight of each constituent stock is set in order to minimize overall portfolio variance. Charts 2 & 3 present the historical price performance of the six factor portfolios for both China A and global, relative to the main benchmark indexes for both regions. We note our observations below: Chart 2The Relative Performance MSCI’s Factor Indexes (Part I) Chart 3The Relative Performance MSCI's Factor Indexes (Part II) Until 2015, the relative price (under)performance of high dividend yielding (“HDY”) stocks had been similar for domestic Chinese and global stocks this cycle.2 Following 2015, high yield stocks continued to underperform at the global level, but massively outperformed in the domestic market to the tune of 11% per year in excess price return. This appears to reflect a shift in domestic Chinese investor sentiment following the 2015 collapse in the A-share market, and has caused a meaningful compression in the average yield of the onshore HDY index relative to the domestic benchmark. It remains unclear whether continued relative outperformance in the domestic market is likely from this factor given how aggressively the strategy has been pursued over the past four years. A low volatility strategy has paid off in the domestic market over the past several years, in contrast to the global market where there has been, at best, a slight uptrend in relative performance over time. Importantly, while the inception date of the MSCI China A Minimum Volatility index is fairly recent, the relative outperformance of the index appears to be much more consistent than that of the HDY index. The apparent success of the low volatility factor within a high volatility equity market is interesting, and serves as the basis of some important investment strategy conclusions to be highlighted later in the report. Similar to the HDY and minimum volatility factors, quality also has outperformed within China’s domestic market. However, this is also true for the global version, in a very consistent fashion over the past decade. The coincident outperformance of quality in both the domestic Chinese and global equity markets is interesting, and also links to some of our later conclusions. In one of the most surprising results of our report, the relative performance of the MSCI momentum index is completely different for the domestic Chinese market than for global stocks. Chart 3 shows that a cyclical momentum strategy has underperformed the MSCI China A onshore index, whereas it has steadily and fairly consistently outperformed at the global level. To us, this finding seems at odds with the common characterization of the A-share market as being driven heavily by momentum-oriented retail investors. In addition, it raises some questions about the efficiency of the global equity market over the past decade. By far, size has exhibited the most impressive potential for outperformance in the A-share market, with domestic small caps having risen three fold relative to large- and mid-cap stocks. However, this massive outperformance has stunningly reversed, with small caps losing 20% per year in relative terms since late-2016. The timing of the collapse appears to be strongly linked with the crackdown on shadow banking by the Chinese government over the past three years (Chart 4), and is consistent with the fact that small firms were disproportionately reliant on alternative financing relative to large firms. This also suggests that at least a portion of the prior massive run-up in small cap performance was boosted by easy access to credit, making it difficult to assess the true underlying impact of size on domestic equity performance. At the global level, small cap stocks have also outperformed the global benchmark, although nearly 80% of the uptrend in relative performance occurred within the first three years of the global expansion. Since mid-2011, global small caps have outperformed the All Country World index by approximately 30 bps per year. Based on MSCI’s indexes, quality and minimum volatility appear to be the most consistently beneficial factors in the A-share market. Finally, the relative performance of value stocks is similar in trend to that of the HDY indexes for both China and global, which is not especially surprising given that the dividend yield is one of three valuation metrics used by MSCI (and many other financial market participants) to define value. However, A-share value has achieved roughly 2/3rds of the annualized outperformance as the HDY factor, in a more consistent fashion that appears to be less influenced by the sentimental shift that occurred in 2015. At the global level, the underperformance of value and HDY stocks likely reflects sector effects, particularly the decade-long underperformance of financials in the aftermath of the global financial crisis (Chart 5). Chart 4Shadow Banking Crackdown = Major Small Cap Underperformance Chart 5Globally, Long-Term Underperformance Of HDY And Value Reflects Financial Sector Decline   Building And Examining New Factor Portfolios CHART 6An Equally-Weighted Industry Group Portfolio Has Outperformed In China And Globally Given the methodological complexity of some of the factor indexes provided by MSCI, we have created 10 of our own factor portfolios for both the China A onshore and All Country World benchmarks using a simpler approach. Several of these portfolios are conceptually similar to the six MSCI factor indexes, but in practice there are several differences: We identify attributes at the industry group level (GICS level 2), rather than at the individual security level. We use all 24 industry groups at the global level, but owing to certain data limitations, we use a mix of sectors and industry groups for domestic China (for a total of 20 groups). Our approach equally-weights stock groups based on whether the attribute in question is above or below the median of all of the groups. This means that attributes are compared across groups based on their actual current values, not relative to their own history (which is, in practice, what occurs when MSCI calculates a Z-score for a particular attribute). Our equally-weighted group approach means that a portfolio equally-weighted across all groups is the appropriate benchmark to test the performance of our portfolios, rather than the MSCI China A onshore index itself (which is weighted by market capitalization). This is an important difference, because Chart 6 shows that an equally-weighted industry group portfolio has outperformed the main equity benchmark both for global stocks and for A-shares. Thus, outperformance of our factor portfolios versus an equally-weighted industry group portfolio would imply even more outperformance vs the benchmark, value-weighted index. Table 2 describes our 10 factor portfolios, as well as the methodology used to construct them. Besides the general differences between our methodology and MSCI’s noted above, there are also a few specific differences even within factors: Table 2Description of BCA-Calculated Industry-Group Factor Portfolios For Onshore China And Global Stocks (US$) Our version of a low volatility portfolio allocates to industry groups with below-median beta (rolling 12-month) versus the MSCI China A onshore index. Our momentum portfolio allocates to industry groups based on a simple rolling 12-month return without adjustment, and we include a short-term momentum portfolio for comparison purposes. We specifically test above-median ROE as a factor, rather than building a composite quality index. Our size factor measures whether smaller mid- and large-cap industry groups outperform over time, rather than the impact of size on individual stocks. We use a univariate measure of value, namely a low 12-month trailing P/E ratio. Finally, we include two new factors not explicitly measured by the MSCI indexes: rolling 12-month ex-post alpha (based on a simple regression against benchmark returns), and high 12-month trailing EPS growth. We present two versions of the alpha factor, one portfolio formed on alpha magnitude, and the other on alpha significance. Charts 7-9 present the relative performance of these factor portfolios versus an equally-weighted industry group benchmark, and we list our key findings below: Chart 7The High-Dividend Factor Has Indeed Outperformed Since 2015 Chart 8Momentum Works Better Globally Than In The A-Share Market   Chart 9ROE And Volatility Are Clearly Persistent Factors In China   Similar to the relative performance of the MSCI factor indexes, our factor portfolios performed better for China than they did for global stocks. With the exception of high (cyclical) momentum, high ROE, and to a lesser extent smaller industry group performance, the factors did not deliver consistently positive excess returns at the global level. Ex-post alpha worked well for global industry groups until 2015, but has since retraced a significant portion of its cumulative excess relative return. Among the domestic China factor portfolios that we created, 7 out of 10 delivered a positive cumulative excess return over the period in question. Size was the worst performing factor, followed closely by short-term momentum and by HDY. Industry groups with a high dividend yield have indeed outperformed substantially since 2015 (consistent with the MSCI China A High Dividend Yield index), but when measured against an equally-weighted industry group portfolio, this outperformance merely made up for a substantial downtrend in relative performance from 2009 to 2015. Among the successful factor portfolios, two stand out as having delivered exceptional relative performance: high ROE, and low beta. Chart 10 shows the performance of a portfolio that averages the industry group weights implied by these two factors, and Table 3 presents whole-period regression results for the portfolio versus both our equally-weighted industry group benchmark and the MSCI China A index. Taken together, Chart 10 and Table 3 make it clear that the combination of high ROE and low beta factors has not only significantly outperformed over time, but has done so with lower volatility. Chart 10High ROE And Low Beta Has Been, And Probably Will Remain, A Winning Combination Table 3Risk And Excess Return Analysis Of Chinese Onshore High ROE / Low Beta Factor Portfolio* Investment Conclusions What conclusions can investors draw from the above analysis? In our view, there are three key takeaways. Our factor indexes confirm that ROE and low volatility are persistent factors in the A-share market. First, the persistent outperformance of several fundamental factors in China’s domestic equity market may suggest that A-shares are a less efficient market than other emerging or developed markets. However, to us, the outperformance of legitimate fundamental factors, and the underperformance (or mediocre performance) of momentum is not consistent with the idea that A-shares are a “casino” market that is purely driven by speculation. China’s domestic market is certainly more volatile than others, but we draw a sharp distinction between a stock market that frequently moves in spite of fundamentals and one that moves in an outsized fashion but in a direction that is consistent with fundamental developments. Chart 11Different “Equally-Weighted” Approaches Can Have A Huge Impact On Performance Second, an accidental conclusion of our report is that an equally-weighted industry group approach appears to have generated persistently positive excess returns over the past decade, for both domestic China and the global equity benchmark. Chart 11 shows the extreme contrast between the official MSCI equal-weight index for All Country World (which equally weights each component stock), and our equally-weighted industry-group portfolio. Stunningly, buying the latter over the former would have generated 3% extra return per year over the past 10 years. For investors seeking increased secular exposure to the A-share market, these results simply suggest that outsized sector or industry-group weightings should be viewed with caution and avoided if possible. Third, industry groups with high return on equity have persistently outperformed their peers over the past decade, both in China and around the world. However, we think that the joint outperformance of high ROE and low beta stocks in the A-share market carries special significance, one that is linked to China’s enormous increase in corporate and household debt over the past decade. Investors who are familiar with the DuPont approach to decomposing return on equity will recall that ROE (Net Income / Equity) can be broken down into the product of profit margins (Net Income / Sales), asset turnover (Sales / Assets), and financial leverage (Asset / Equity). Since higher financial leverage tends to increase the operating risk of a firm (and thus, presumably, its stock price), and the product of profit margins and asset turnover equals return on assets (ROA), the persistent outperformance of high ROE and low beta industry groups suggest that domestic investors have been focused on buying firms with a high ROA and relatively low financial leverage. Table 4Current High ROE / Low Beta Factor Industry Group Portfolio Weights* Chart 12A High ROE & Low Beta Portfolio Isn’t Cheap, But Is No More Expensive Today Than In The Past This is a sensible approach, and we are comfortable recommending it to global investors who are looking to increase their secular exposure to A-shares. We are also opening a long relative position in this high ROE / low beta portfolio in our trade book today to track the call, and to provide investors with timely updates on the weights in the portfolio (currently shown in Table 4). While it is true that valuation of this portfolio today is not as attractive as it is for the value-weighted benchmark, Chart 12 highlights that this has been true for the entire sample period, and it has not prevented the substantial outperformance that we have documented. This implies that, while not necessarily widespread, some “cheap” onshore industry groups represent a value trap, rather than true value.   Jonathan LaBerge, CFA Vice President Special Reports jonathanl@bcaresearch.com   Footnotes 1      Please see China Investment Strategy Weekly Report, “A-Shares: Stay Neutral, For Now,” dated March 14, 2018 and “A Shaky Ladder,” dated June 13, 2018, available at cis.bcaresearch.com. 2      While Charts 2 & 3 illustrate relative price instead of total return indexes, our conclusions are generally the same even when showing the latter. Cyclical Investment Stance Equity Sector Recommendations

Analysis on central Europe and Pakistan is published below. Highlights There are several reasons why Chinese authorities will likely allow the yuan to depreciate 6-8% from current levels in the coming months. RMB depreciation will weigh not only on emerging Asian but also on other EM currencies via several channels. We continue to recommend shorting a basket of the following currencies versus the U.S. dollar: ZAR, CLP, COP, IDR, MYR, PHP and KRW. Feature Chinese authorities will likely allow the yuan to fall 6-8% vis-à-vis the U.S. dollar from current levels in the coming months. The value of the RMB holds the key to the broader trend in EM currencies. In turn, dynamics in EM exchange rates typically define the trajectory for EM asset classes: stocks, credit spreads and local currency bonds. Odds are that the RMB along with other emerging Asian currencies will continue to depreciate (Chart I-1). There are several reasons why Chinese authorities will likely allow the yuan to fall 6-8% vis-à-vis the U.S. dollar from current levels in the coming months. Chart I-1Emerging Asian Currency Index Chart I-2Deflating Export Prices Herald Currency Depreciation First, currency depreciation will help mitigate the impact of U.S. import tariffs. With global trade volumes shrinking and U.S. import prices from China deflating (Chart I-2, top panel), China will benefit from a cheaper currency. Second, RMB devaluation goes against the Trump’s administration’s preference. The U.S.-China trade talks have flopped, and both sides seem to be jockeying to better position themselves ahead of another round of discussions. From a negotiation strategy perspective, it makes sense for China to devalue the yuan before a new round of negotiations starts again. Third, China needs lower interest rates to reduce the strain on numerous debtors. However, narrowing interest rate differential with the U.S. has often coincided with periods of RMB depreciation over the past nine years (Chart I-3). Chart I-3CNY/USD And Interest Rate Differential One reason why policymakers in China were previously reluctant to explore RMB/USD depreciation beyond the 7 mark was due to the risk of rampant capital exodus and a potential spike in financial market volatility. In other words, authorities were mindful that even mild and controlled depreciation could spiral out of control. However, with Chinese nationalistic rhetoric on the rise and the nation rallying around the flag, authorities now have more room to maneuver. They will not have a problem restricting capital outflows by residents, and there will be little general public dissatisfaction with a devaluation. Finally, at around $3 trillion, the central bank’s foreign exchange reserves are equivalent to only 14% of all yuan deposits, and 11% of broad money (M2) supply. In turn, the overhang of local currency money supply will exert structural downward pressure on the renminbi’s exchange rate in the coming years. This may be a convenient time to release some proverbial air out of the balloon – namely, the lingering money bubble in China – by devaluing the yuan. Bottom Line: The path of least resistance for the RMB is down. EM Currencies Are In Danger In recent months, we have been highlighting that the Korean won has been at a critical technical juncture and has broken down (Chart I-4, top panel). The Taiwanese and Singaporean dollars seem to be the next shoes to drop (Chart I-4, middle and bottom panels). Chart I-4 Chart I-5No Recovery In Asian Exports So Far U.S. import prices from various Asian countries are deflating, as shown in the bottom panel of Chart 2 on page 2. This typically warrants currency depreciation to mitigate the impact of export price deflation on national producers. Furthermore, emerging Asian exports are still shrinking, as evidenced by the latest trade numbers. Korea’s total exports for the first 20 days of May and Taiwan’s exports of electronics parts as of April are still contracting at a rapid pace (Chart I-5). The latter leads cyclical turning points in global trade by a couple of months. Finally, the RMB is the anchor currency in emerging Asia, and its depreciation will filter through the exchange rates of other regional, export-dependent economies. Regarding other EM currencies, they are also at risk of Chinese yuan depreciation. Apart from manufacturing sector competitiveness (discussed above), China’s exchange rate affects other economies in two distinct ways: Less Chinese imports: An RMB devaluation reduces the amount of China’s U.S. dollar inflows/payments to its trade partners (Chart I-6). Many EM and some DM currencies will be negatively affected since China is a major source of demand for these economies. Less capital outflows from China: RMB depreciation will likely be accompanied with heightened controls over capital outflows from China. In fact, various proxies for capital flight out of the mainland suggest the authorities have already substantially clamped down on outflows (Chart I-7). Economies that have profited from capital flight from China over the years are already feeling pain. For example, relapsing Australian property prices can be attributed to reduced capital flows from China. Chart I-6Chinese Imports In RMB And USD Chart I-7China: Reduced Capital Flight Likewise, there will be a period of painful adjustment in many emerging economies in Asia and elsewhere that have profited from Chinese capital flows – both via official and non-official channels. Bottom Line: RMB depreciation will affect various currencies via diverse channels: (1) deteriorating export competitiveness for manufacturing-based economies; (2) diminished mainland imports from China’s trade partners; and (3) reduced capital flows from China to economies that have typically relied on Chinese capital flows. The U.S. Dollar: A Review Of The Indicators We believe that the cyclical and structural backdrops remain favorable for the dollar, and that it will likely overshoot before a major top sets in. The U.S. dollar bull market is extended, but that does not mean it is over. Odds are that the greenback will overshoot before making a major top. In our last weekly report, we revisited currency valuations and found the dollar to be only moderately (one standard deviation) expensive, according to the real effective exchange rate based on unit labor costs. The latter is our favored currency valuation metric. The greenback has been in a major structural bull market since 2011. Secular bull/bear markets do not typically end before valuations reach 1.5-2 standard deviations. We believe that the cyclical and structural backdrops remain favorable for the dollar, and that it will likely overshoot before a major top sets in. Chart I-8U.S. Equity And Economic Outperformance Warrants Dollar Appreciation U.S. stocks are outperforming the rest of the world in local currency terms, not only based on market-cap equity benchmarks but also when measured using equal-weighted equity indexes (Chart I-8). This signals that return on capital is higher in the U.S. relative to the rest of the world. The latter has historically been positively related with the primary trend in the trade-weighted dollar (Chart I-8). The U.S. dollar currently offers an attractive yield relative to many of its peers. Chart I-9 illustrates the interest rate (3-month swap rate) differentials between the dollar and various EM and DM currencies. For each individual exchange rate, the bar denotes the U.S. interest rate spread over other markets, and the dot is the mean of this spread over the past 20 years. Not only is the current interest rate differential in favor of the greenback in the case of many currencies, but the spread is well above its 20-year mean for virtually all of the currencies included in Chart I-9. The sole exception is the Mexican peso – the latter’s current interest rate differential versus the U.S. is wider than its 20-year mean. In fact, the peso is among our most preferable EM currencies. The U.S. dollar currently offers an attractive yield relative to many of its peers. Bottom Line: Odds are in favor of a U.S. dollar overshoot, especially versus cyclical currencies such as EM and commodities-based ones. We continue to recommend shorting a basket of the following currencies versus the U.S. dollar: ZAR, CLP, COP, IDR, MYR, PHP and KRW. We are also structurally short the RMB versus the dollar. For investors who are looking for currencies with the least downside versus the U.S. dollar, our picks are MXN, RUB, THB, TWD, SGD and central European currencies. EM Credit Markets And Domestic Bonds: It’s All About Exchange Rates From a macro perspective, EM dollar-denominated and local currency bonds are primarily driven by exchange rates. EM sovereign spreads are very sensitive to both EM exchange rates and industrial metals prices (Chart I-10). The latter two are primarily driven by global trade cycles in general and China’s growth in particular. EM corporate spreads have been less sensitive to EM exchange rates. Yet they are unlikely to defy a major down-leg in EM currencies. The basis is as follows: when currencies depreciate, foreign-currency debt becomes more expensive to service warranting a period of wider credit spreads. Exchange rate fluctuations account for the bulk of domestic bonds’ total returns for foreign investors. We discussed this topic in our report titled Asset Allocation For EM Assets. Chart 11 shows the total return indexes in dollars and euros for the EM GBI local currency government bond index. Euro-based investors have fared much better than dollar-based ones. The euro’s depreciation versus the dollar explains this gap. However, from a technical point of view, total return in euros is facing a major resistance level (Chart I-11, bottom panel). European investors should take note. Chart I-10EM Sovereign Spreads Correlate With EM Currencies And Commodities Chart I-11Total Returns on EM Local Bonds In USD And Euros   Arthur Budaghyan Chief Emerging Markets Strategist arthurb@bcaresearch.com   Continue Favoring Central Europe Within EM Equities: An overweight position in Central European (CE) equities within an EM equity portfolio is still warranted due to the following reasons: First, CE economies are the least exposed to a Chinese and global trade slowdown - the main causes of the EM selloff. Even though these economies are leveraged to German and euro area manufacturing - both of which are currently weak - they have little direct exposure to China and commodities. Second, currency trends are critical for the relative performance of equities. We expect EM currencies will depreciate versus the euro and against CE currencies. This argues in favor of CE stocks within an EM portfolio. Third, CE domestic demand remains strong and private credit growth robust (Chart II-1). Additionally, the authorities are maintaining a loose fiscal policy stance. As to European equity portfolios, we recommend underweighting CE bourses versus the core European markets. Chart II-2 illustrates that when EM equities underperform DM ones, CE share prices lag behind euro area stocks. Chart II-1Private Credit Growth Is Robust Chart II-2CE Underperforms Core Europe When EM Underperforms DM Currencies and Fixed-Income Markets: CE growth outperformance relative to EM suggests that CE exchange rates will outperform the majority of EM currencies. Critically, odds are that the euro has made a major bottom versus most EM currencies. This will facilitate CE exchange rate appreciation versus many other EM currencies. The latter warrants overweighting CE fixed-income markets against respective EM benchmarks. Currency Trades: Today we recommend closing our long CZK / short euro position. This trade has generated a 4.4% gain since September 28, 2016 with extremely low volatility. The basis for closing this position is there are signs that Czech growth and labor market tightness are peaking, warranting an end to rate hikes. Specifically, both economic activity and wage growth are slowing. This will lead the central bank to halt its rate hikes. Instead, we are opening a new trade: Go long CZK versus an equal-weighted basket of PLN and HUF. For the first time, Czech short rates have risen above those in Poland and Hungary (Chart II-3). This will be a major driver for Czech koruna appreciation against the other two currencies. The PLN and HUF will underperform the CZK because their monetary and fiscal policies are much easier than is currently warranted. Chart II-3Czech Interest Rate Differentials Versus Hungary And Poland Are Positive Chart II-4Trade Balances Favor CZK vs HUF & PLN The Hungarian central bank will launch its corporate QE program in July 2019 with a total of HUF 300 billion in corporate bond purchases. This will likely weigh on the HUF as the central bank monetizes some of the country’s outstanding corporate debt. Additionally, the Polish government has announced large fiscal stimulus ahead of this year’s elections. The fiscal deficit is projected to widen from 1% currently to 2% of GDP by 2020. Finally, trade balances in Poland and Hungary are deteriorating while the Czech Republic is running a large trade surplus (Chart II-4). Bottom Line: Continue overweighting CE within both EM equity and local currency bond portfolios. We are taking profits on our long CZK / short the euro trade and initiating a new position: Long CZK / short an equal-weighted basket of HUF and PLN.   Andrija Vesic, Research Analyst andrijav@bcaresearch.com   Pakistan: No Pain, No Gain Pakistan’s economy and stock market are currently going through painful but necessary adjustments. The country has been suffering from a severe balance-of-payment crisis. Its exchange rate has already depreciated by 30% versus the U.S. dollar since December 2017. Its stock market in U.S. dollar terms has plunged 55% from its May 2017 peak. A bottom in the stock market is likely to occur when the currency stabilizes. Odds are that the Pakistani rupee is in its late phase of adjustment (Chart III-1). First, a US$ 6 billion worth IMF bailout fund is on its way. The country reached a staff-level agreement with the IMF on May 12. The IMF will release the funds in phases over a period of 39 months. Meanwhile, Pakistan will likely also receive US$ 2-3 billion from the World Bank and the Asian Development Bank (ADB) in the next three years. Altogether, multilateral financing will amount to about US$3 billion per year over the next three years. The country will also likely continue its bi-lateral borrowings from China, Saudi Arabia and the UAE. Last year, about US$10 billion of external borrowing and a nearly US$7 billion reduction in the central bank’s foreign reserves helped fund the US$18 billion current account deficit. Over the next 12 months, we expect the financing needs to be considerably smaller due to shrinking twin deficits (Chart III-2). Chart III-1Pakistan's Rupee: Close To A Bottom? Chart III-2Twin Deficits Are Likely To Shrink Both trade and current account balances have started showing improvement in U.S. dollar terms due to a steep contraction in imports. Going forward, we expect export growth to turn positive on the back of currency devaluation but import contraction will deepen. Lastly, the IMF agreement might allow Pakistan to issue some Eurobonds while higher local rates might attract some foreign portfolio capital. Second, Pakistan’s top leadership has cooperated with the IMF. Just earlier this month IMF economist Reza Baqir was appointed the new central bank governor. In addition, the Finance Minister and the Federal Bureau of Revenue chairman have been replaced. These new appointments increase the odds that the IMF program will be enforced in Pakistan. Indeed, after only two weeks on the job the new central bank governor raised the policy rate this Monday by 150 basis points to 12.25%. Meanwhile, significant fiscal consolidation is on the way, as the new policymakers will be committed to the IMF program. The budget for the next fiscal year (June 2019 – May 2020), which will be presented in Parliament on May 24, will likely show a considerable reduction in non-interest expenditures. Finally, the IMF is also pushing for increased central bank independence. In the last 17 months, the central bank purchased massive amounts of government securities – a de facto monetization of public debt. This has exacerbated domestic inflation and currency depreciation. So long as the country is under the IMF program, it is reasonable to expect no public debt monetization. In summary, the ongoing substantial monetary and fiscal tightening and accompanying reduction in the twin deficits, coupled with the increased availability of foreign funding are positive for the exchange rate. It is possible that Pakistan will follow the 2016-2017 Egyptian roadmap. Egypt experienced a severe balance-of-payment crisis and agreed to a similar IMF bailout program. In the case of Egypt, a 55% depreciation in its currency in late 2016 was followed by a 77% rally in share prices in U.S. dollar terms over the subsequent 18 months (Chart III-3). We are putting Pakistani stocks on our upgrade watch list. We are reluctant to upgrade it now because currency weakness might persist for a couple of months. Further, monetary and fiscal tightening will amplify the economic downturn weighing on corporate earnings. Banks’ NPL ratios and provisions will likely rise considerably. Chart III-3The 2016-2017 Egyptian Roadmap Chart III-4Pakistani Equities: A Long-Term Profile Bottom Line: We are putting Pakistani equities on an upgrade watch list. This bourse’s technicals are becoming interesting – it might bottom at its previous highs (Chart III-4). In addition, both absolute and relative valuations of Pakistani stocks appear attractive (Charts III-5 & Chart III-6). Chart III-5Equity Valuations Look Attractive Chart III-6Relative Equity Valuations Also Look Attractive We are waiting for share prices and the currency to stabilize before recommending an overweight position in Pakistani equities.   Ellen JingYuan He, Associate Vice President ellenj@bcaresearch.com Footnotes Equity Recommendations Fixed-Income, Credit And Currency Recommendations

Highlights The risk premium in crude oil prices is rising again, as policy risk – and the potential for large policy-driven errors – increases (Chart of the Week).1 This is not being fully reflected in options markets, where implied volatilities are trading close to their long-term average levels (Chart 2). In the past month, risks to oil flows – military and otherwise – and supply have risen, which is keeping a bid under prices. The Sino – U.S. trade war has worsened, and threatens to put global supply chains at risk, along with EM demand growth in the medium term. Meanwhile, amid global monetary easing, the USD has strengthened, producing a more immediate headwind for EM commodity demand. Against this backdrop of opposing forces, oil prices remain elevated and relatively stable in the low $70/bbl range for Brent. Our balances estimates and price forecasts have not changed materially this month. However, the balance of risks has widened in both tails of the price distribution. We expect implied volatilities in the crude oil options markets – particularly Brent – to move higher, as a result. As for prices, we continue to expect Brent to average $75/bbl this year and $80/bbl next year, with WTI trading $7/bbl and $5/bbl below those levels in 2019 and 2020, respectively. Energy: Overweight. The U.S. EIA moved closer to our fundamental assessment and Brent forecast in its most recent market update, lifting its Brent spot-price expectation for this year to an average of $70/bbl, ~ $5/bbl above its April forecast. The EIA’s revision reflects “tighter expected global oil market balances in mid-2019 and increasing supply disruption risks globally.” Base Metals: Neutral. In the wake of Vale’s January supply disaster at its Córrego do Feijão mine, iron ore shipments from Brazil were down 60% in April y/y. Cyclones disrupted supply in Western Australia, pushing 62% Fe iron ore prices to a 5-year high above $100/MT last week. Chinese steelmakers registered a 12.7% y/y gain in crude steel output last month, which, along with dockside iron ore inventory draws of ~ 20 MT ytd, is supporting prices generally. Precious Metals: Neutral. A stronger USD is weighing on gold. Global geopolitical tensions – chiefly in the Persian Gulf and in Sino – U.S. trade relations – are keeping prices above $1,270/oz. We remain long gold as a portfolio hedge. Ags/Softs: Underweight. Severe weather conditions in the Midwest continues to delay corn planting, and is contributing to a rally this week in corn prices to $3.94/bushel on Tuesday, up $3.48/bushel from last week’s level. Feature The risk of a military confrontation between the U.S. and Iran is higher than it was a month ago and rising. Should it erupt, such a confrontation would threaten oil exports from the Persian Gulf through the Strait of Hormuz, where ~ 20% of global supply transits daily.2 Bellicose rhetoric from the U.S. – some of it directed at materially reducing Iran’s influence in Iraq – alternately is ramped up and walked back, while attacks on soft targets in the Kingdom of Saudi Arabia (KSA) – e.g., oil shipping and west-bound oil pipelines – draw attention to the exposure of this critical infrastructure, upon which global oil markets rely.3 Iran, meanwhile, uses the media to prepare its population for further economic deprivation, and to lob its own vituperative rhetoric at the U.S. Venezuela’s collapse as an oil producer and exporter continues unabated, keeping markets for the heavier sour crude favored by U.S. refiners tight. Civil war threatens to cut into Libyan production, which we are carrying at just over 1mm b/d, while whiffs of another Arab Spring can be detected in Algeria, where popular discontent with ruling elites grows.4 On the demand side, the summer driving season is about to kick off in the Northern Hemisphere, heralding increased gasoline demand. Countering that, the Sino – U.S. trade war shows signs of devolving into a Cold War, which could force a re-ordering of supply chains globally, lifting costs and consumer-level inflation in the process. Longer-term, this could work against central-bank easing globally, and retard growth in EM consumer demand. The risk of a military confrontation between the U.S. and Iran is higher than it was a month ago and rising. Should it erupt, such a confrontation would threaten oil exports from the Persian Gulf through the Strait of Hormuz. For the present, we continue to expect EM demand growth to hold up, expanding by 1.5mm b/d this year and 1.6mm b/d next year. This will be supported by continued monetary easing globally, and additional fiscal stimulus from China if its trade war with the U.S. worsens. There is a chance weakness in DM demand will persist, but we think the odds of a normal seasonal pick-up in 2H19 will continue to support demand overall (Chart 3). That said, given the threats to demand growth – an expanded Sino – U.S. trade war and stronger USD, in particular – we will continue to monitor the health of EM demand closely. Chart 2Brent Implied Volatility Will Move Higher Chart 3DM Oil Demand Growth Wobbles, EM Steady   OPEC 2.0 Maintains Production Discipline Chart 4OPEC 2.0's Production Discipline, Strong Demand Drained Inventories The goal of OPEC 2.0 from its inception at the end of 2016 has been to drain OECD inventories, which swelled to 3.1 billion barrels in July 2016, on the back of a market-share war launched by the old OPEC under the leadership of KSA, and a surge in U.S. shale-oil production. KSA continues to stress the need to restrain crude oil production so as to draw down global oil inventories, and has done much of the heavy lifting this year to make that happen (Chart 4). The other putative leader of OPEC 2.0, Russia, continues to express misgivings with such a strategy, arguing instead the producer coalition should make more oil available to the market. We are more aligned with Russia’s view, and continue to believe OPEC 2.0 will need to increase production. In our balances (Table 1), our base case assumes those producers that can lift production – core OPEC and Russia – will do so to keep prices below $85/bbl (Chart 5). We expect OPEC 2.0 will be able to offset the loss of ~ 700kb/d from Iran exports by increasing production gradually from May to September in proportion to its quota agreement. In our base case, we have Iranian exports falling to 600k b/d. We continue to expect OPEC 2.0 to be able to offset the loss of Venezuela’s production throughout the year, which we expect to fall to 500k b/d by December (vs. ~ 735k b/d presently). Going into next month’s Vienna meeting, we do not expect KSA to dramatically increase production, but would not be surprised if it took production from its current 9.8mm b/d level closer to its OPEC 2.0 quota of 10.33mm b/d in 2H19. Table 1BCA Global Oil Supply - Demand Balances (MMb/d, Base Case Balances) Going into next month’s Vienna meeting, we do not expect KSA to dramatically increase production, but would not be surprised if it took production from its current 9.8mm b/d level closer to its OPEC 2.0 quota of 10.33mm b/d in 2H19. We also expect Russia to lift its production closer to 11.6mm b/d from ~ 11.4mm b/d at present. Even with OPEC 2.0 lifting production ~ 900k b/d in 2H19 vs. 1H19, the bulk of global production increases will be concentrated in the U.S., where we expect shale-oil output to grow 1.2mm b/d this year, and 840k b/d next year. This will account for 85% of the overall increase of 2.4mm b/d we expect in the U.S. this year and next. Our estimates of production growth in the U.S. shales is tempered by a growing conviction the large integrated oil majors and stand-alone E&P companies will continue to put the interests of shareholders above their desire to increase production just for the sake of increasing it, as was done in the past. This is driven by a desire to attract and retain capital, which will be critical to the majors and the big E&Ps in the years ahead.5 We continue to see demand growth exceeding supply growth this year. This will produce a physical deficit, which will continue to drain inventories. Even with these production increases, we continue to see demand growth exceeding supply growth this year. This will produce a physical deficit, which will continue to drain inventories (Chart 6). Chart 5Core OPEC 2.0 Will Lift Production Chart 6Balances Continue To Tighten   Spare Capacity Will Be Stretched In addition to Iran and Venezuela, we are closely following what appears to be the early stages of another civil war in Libya, which threatens the ~ 1mm b/d of production flowing from there. In addition, we are seeing signs of growing civil discontent in Algeria not unlike that of 2011, which was sparked by popular dissatisfaction with ruling elites throughout the Middle East in the lead-up to the Arab Spring. We have maintained existing spare capacity can handle the loss of Iranian and Venezuelan production and exports we’ve built into our balances and price-forecast models. However, covering these losses will stretch the capacity of global supply to accommodate unplanned outages, which could leave markets extremely tight in the event of production losses in Libya or Nigeria, or in producing provinces prone to natural disasters (e.g., Canadian wildfires or U.S. Gulf hurricanes). At present, markets appear to be comfortable with OPEC 2.0’s ability to cover losses from Iran and Venezuela, given current spare capacity of ~ 3mm b/d, most of which remains in KSA, and continued growth in non-OPEC output (Chart 7). As inventories continue to draw globally, markets’ attention will turn more toward this spare capacity.   Expect Higher Volatility We remain long Brent call spreads in July and August 2019, which are up an average 101% since they were recommended in February. These positions benefit from higher prices and higher volatility. Chart 8Geopolitics, Increasing Backwardation Support Higher Brent Implied Volatility Our fundamental assessments of supply, demand and inventory levels remain fairly steady. Thus, our price forecasts – $75 and $80/bbl this year and next for Brent, with WTI trading $7 and $5/bbl under that – remain unchanged. With OPEC 2.0 maintaining production discipline and U.S. shale producers maintaining capital discipline, the rate of growth on the supply side will be restrained, and below the rate of growth in global demand. These forces combine to keep inventories drawing this year, which will lead to a steeper backwardation in forward curves, particularly Brent’s (Chart 8). Coupled with true uncertainty re how the U.S. – Iran confrontation in the Persian Gulf is resolved, and how the Sino – U.S. trade war plays out, this steepening backwardation will lead to higher implied volatility in crude oil options markets. Bottom Line: Our expectation of higher prices and steepening backwardation in forward curves is supported by our analysis of fundamentals and the current political economy of global oil markets, which emphasizes policy risk arising from the actions of geopolitically significant states. These factors also will push implied volatility in options markets higher. As a result, we remain long Brent call spreads in July and August 2019, which are up an average 101% since they were recommended in February. These positions benefit from higher prices and higher volatility. We also remain long 2H19 Brent vs. short 2H20 Brent futures in line with our view backwardation will increase; this position is up 155.4% since it was initiated in February, as a result of the steepening of backwardation in the forward curve. Steepening backwardation also will benefit our long S&P GSCI recommendation, which is heavily weighted to energy markets; this position is up 8% since inception. Lastly, we remain long spot WTI, which is up 34.6% since it was recommended in January.   Robert P. Ryan, Chief Commodity & Energy Strategist rryan@bcaresearch.com Hugo Bélanger, Senior Analyst Commodity & Energy Strategy HugoB@bcaresearch.com Footnotes 1 In the price decomposition shown in our Chart of the Week, we account for the contribution that changes in global supply, demand and inventory levels make to the evolution of Brent prices, using a proprietary econometric model. We treat the residual term of the model – what’s left of the price decomposition after these fundamental variables are accounted for – as a measure of the risk premium in prices. An expansion of the risk premium – in the positive or negative direction – is coincident with an expansion of the implied volatility of Brent crude oil options typically expands (sometimes with a lag or two), and vice versa. This is intuitively appealing, since risk premia and volatility expand as uncertainty in the market rises. 2 We considered this topic in depth in a Special Report written with BCA Research’s Geopolitical Strategy entitled “U.S., OPEC Talk Oil Prices Down; Gulf Tensions Could Become Kinetic,” published July 19, 2018, and in “Brinkmanship Fuels Chaos In Oil Markets, And Raises The Odds Of Conflict In The Gulf,” published July 5, 2018. Both reports are available at ces.bcaresearch.com. 3 Iran’s influence in Iraq is an internally divisive issue, and a focal point of the U.S., a view we share. Please see, “Iraq: The Fulcrum Of Middle East Geopolitics And Global Oil Supply,” a Special Report we published with BCA Research’s Geopolitical Strategy September 5, 2018. KSA and Western intelligence agencies allege Iran is behind the attacks on Saudi oil infrastructure. Please see “Saudi Arabia accuses Iran of ordering drone attack on oil pipeline,” published by reuters.com. The westbound pipelines in KSA are critical to maintaining the Kingdom’s export capacity, as we noted in “Risk Premium In Oil Prices Rising; KSA Lifts West Coast Export Capacity,” published by BCA Research’s Commodity & Energy Strategy October 25, 2018. This report is available at ces.bcaresearch.com. 4 Please see “Algeria Has a Legitimacy Problem,” posted on the LSE’s Middle East Centre Blog by Benjamin P. Nickels on May 20, 2019, and “Algeria’s Second Arab Spring?” by Ishac Diwan posted at project-syndicate.org March 28, 2019. 5 We will be exploring this topic in depth in a Special Report next month. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades TRADE RECOMMENDATION PERFORMANCE IN 2019 Q1 Commodity Prices and Plays Reference Table Trades Closed in 2019 Summary of Trades Closed in

Highlights Duration: We see current bond market behavior as very similar to mid-2016, when heightened political uncertainty obscured the economy’s true strength and kept bond yields lower for longer than was justified by the economic fundamentals. The correct strategy at that time was to sell into the bond market’s strength, and we advocate a similar strategy today. China: Any attempt by the Chinese government to retaliate in the trade war by selling U.S. Treasury securities would be either self-defeating or ineffective, depending on the exact strategy employed. In either case, U.S. Treasury yields will be unaffected. Fed: At least part of the Fed’s dovish turn might represent a desire to send the labor share of national income higher. We introduce a new data series for Fed Watchers to track. Feature The Trump Administration fired the latest salvo in the trade war two weeks ago, expanding tariffs to a broader swathe of Chinese imports. Then last week, the escalation of tensions spilled over to the bond market, sending global yields abruptly lower. Chart 1Flight To Safety The 10-year U.S. Treasury yield bounced off 2.35% last Thursday and has since settled at 2.39% (Chart 1). Meanwhile, the overnight index swap curve is now priced for 44 bps of Fed rate cuts over the next 12 months (Chart 1, bottom panel). It is possible, and even likely, that geopolitical tensions will keep yields low during the next month or two. In fact, our Geopolitical Strategy service places the odds of a complete breakdown in trade negotiations by the end of June at 50%.1  But we would encourage investors to sell into rallies, positioning for higher yields on a 6-12 month horizon. To see why, we return to a Weekly Report from early April where we walked through different factors that would be useful in the creation of a macroeconomic model for the 10-year U.S. Treasury yield.2 We consider what has changed during the past six weeks and what those developments mean for bond yields going forward. Back In The Bond Kitchen In early April, we ran through four different factors that should be included in any bond model and suggested macroeconomic indicators that best capture the trends in each. The four factors are: Global Growth: Best proxied by the Global Manufacturing PMI and Bullish Dollar Sentiment Policy Uncertainty: Best proxied by the Global Economic Policy Uncertainty Index Output Gap: Best proxied by Average Hourly Earnings Sentiment: Best proxied by the U.S. Economic Surprise Index We consider each factor in turn. Global Growth Chart 2Monitoring Global Growth The Global Manufacturing PMI, our preferred series for tracking global growth, ticked down during the past month, continuing the free-fall that has been in place since the end of 2017 (Chart 2). At 50.3, it is now only slightly above the 50 boom/bust line and is close to where it was in mid-2016, when the 10-year yield hit its cyclical low. But on a positive note, several leading indicators have hooked up in recent months, suggesting that the Global PMI could soon trough and move higher in the second half of the year. Specifically, the ZEW survey of global economic sentiment is off its lows, as is the BCA Global Leading Economic Indicator (LEI). Meanwhile, the Global LEI Diffusion Index has surged, indicating that 74% of the 23 countries in our sample are seeing improvement in their LEIs. Historically, the Global LEI Diffusion Index leads changes in both the Global LEI and the Global Manufacturing PMI (Chart 2, panel 3). Financial market prices that are highly geared to global growth had been singing a similar tune, but they rolled over as trade tensions flared during the past two weeks. For example, cyclical equity sectors recently started to underperform defensive sectors (Chart 2, bottom panel), and the important CRB Raw Industrials index took a nosedive. We place particular importance on the CRB Raw Industrials index as a timely indicator of global growth, because the ratio between the CRB index and gold correlates nicely with the 10-year Treasury yield (Chart 3).3 Unsurprisingly, the ratio’s recent dip coincides with last week’s drop in the 10-year. Several leading indicators have hooked up in recent months, suggesting that the Global PMI could soon trough and move higher in the second half of the year.  In addition to the Global Manufacturing PMI, we recommend including a survey of bullish sentiment toward the U.S. dollar in any bond model. More bullish dollar sentiment coincides with lower Treasury yields, and vice-versa. Our preferred survey shows that dollar sentiment remains elevated, but hasn’t changed much since April (Chart 4). The dollar itself, however, has begun to appreciate during the past two weeks (Chart 4, bottom panel). Chart 3A Falling CRB/Gold Ratio... Chart 4...And The Greenback Is On The Rise Bottom Line: The coincident global growth indicators that correlate best with bond yields – the Global Manufacturing PMI and Dollar Bullish Sentiment – are sending a similar message as in April. Meanwhile, leading economic indicators continue to suggest that we should expect improvement in the second half of the year. The biggest change from April is that global growth indicators derived from financial market prices – cyclical versus defensive equities, the CRB Raw Industrials index and the trade-weighted dollar – have responded negatively to heightened political risk. If this weakness persists and eventually infects the economic data, then it could prevent a second-half rebound in global growth, keeping Treasury yields low for even longer.   Policy Uncertainty Spikes in the monthly Global Economic Policy Uncertainty Index often cause capital to seek out the safety of U.S. Treasuries, and we recommend including this index in any macroeconomic bond model (Chart 5A). Spikes in the monthly Global Economic Policy Uncertainty Index often cause capital to seek out the safety of U.S. Treasuries. While there have been no updates to the monthly index since the trade war’s recent escalation, one of its components – a daily index that tracks the number of relevant news stories – has surged during the past two weeks (Chart 5B). This clearly illustrates that a sharp increase in political uncertainty has been the catalyst for the bond market rally. Investors are obviously concerned that an ongoing and intensifying trade war might derail the economic recovery, and they are seeking out Treasuries as a hedge. Chart 5AGlobal Uncertainty Set To Spike Chart 5BMarkets Are Concerned In such situations, the traditional playbook is to fade any purely uncertainty-driven rally, on the view that markets tend to overreact to headline risk. This strategy worked well following the mid-2016 Brexit vote. The uncertainty shock from the vote sent the 10-year quickly down to 1.37%, but it then increased in the second half of the year when it became apparent that the economic recovery would continue. While higher tariffs will certainly be a drag on growth going forward, accommodative Fed policy and a probable increase in Chinese economic stimulus will mitigate the impact, keeping the economic recovery intact.4 Output Gap Chart 6Wages Are Headed Higher The output gap is a concept that represents where the economy is operating relative to its peak capacity, and its progress during the past three years is the main reason why bond yields will not re-test 2016 lows. We have found that wage growth is the most reliable way to measure the output gap: higher wage growth signals less spare capacity, and less spare capacity coincides with higher bond yields. We recommend Average Hourly Earnings as the best wage measure to include in any bond model. Since April, average hourly earnings growth has been roughly flat, but leading indicators suggest that further acceleration is highly likely in the coming months (Chart 6). While the Fed is keen to let wage growth accelerate, rising wage growth also makes a rate cut difficult to justify. The combination of rising wage growth and an on-hold Fed should put a rising floor under long-maturity bond yields. Sentiment The final factor that should be included in any bond model is sentiment. In April, we suggested that the U.S. Economic Surprise Index is the best measure of sentiment. When the surprise index has been deeply negative for a long time, it usually means that investors are downbeat on the economy and that the bar for a positive surprise is low. This has actually been the case in recent months, and our simple auto-regressive model suggests that the surprise index is biased higher (Chart 7). Positioning data confirm this message, and in fact show that investors are taking as much duration risk as they were when yields troughed in mid-2016 (Chart 8). Chart 7Low Bar For Positive Surprises Chart 8Similar Positioning As In Mid-2016 The overall message is that bond investors have a very dim view of the economy, and it will not take much positive news to send yields higher. Investment Strategy We see current bond market behavior as very similar to mid-2016, when heightened political uncertainty obscured the economy’s true strength and kept bond yields lower for longer than was justified by the economic fundamentals. The correct strategy at that time was to sell into the bond market’s strength, and we advocate a similar strategy today. Timing when the next move higher in bond yields will occur is difficult, but we take some comfort in the fact that the flatness of the yield curve makes it less costly than usual to carry below-benchmark duration positions. In fact, the average yield on the Bloomberg Barclays Cash index is 7 bps higher than the average yield on the Bloomberg Barclays Treasury Master Index. Bond investors have a very dim view of the economy, and it will not take much positive news to send yields higher. To further mitigate the cost of keeping duration low, we advocate taking duration-neutral positions that are short the belly (5-year & 7-year) part of the yield curve and long the very long and very short ends of the curve. Such trades are also provide a positive yield pick-up, and will earn capital gains when Treasury yields move higher.5 A Quick Note On China’s Treasury Purchases Chart 9Do Not Expect Treasuries To Be Used As A Weapon In This War The trade war’s recent escalation has led some to speculate that China could retaliate against higher tariffs by dumping U.S. Treasury securities onto the open market. The speculation only increased when the TIC data revealed that Chinese net Treasury purchases totaled -$24 billion in March, the most deeply negative figure since October 2016 (Chart 9).   We see low odds that China will employ this tactic in the trade war, and no meaningful impact on Treasury yields in any case. To see why, let’s consider two possible scenarios. In the first scenario, China sells a large amount of U.S. Treasury securities and keeps the proceeds from the sales in its domestic currency. Assuming the amounts in question are sufficiently large, these transactions would cause the RMB to appreciate and lead to a tightening of Chinese monetary conditions. Tighter monetary conditions are exactly what the Chinese government does not want as it seeks to counteract the negative economic impact from tariffs. In fact, China is much more likely to engineer a further easing of monetary conditions, much like in 2015/16 (Chart 9, bottom panel). In the second scenario, China could sell U.S. Treasuries and purchase other foreign bonds (German bunds, for example). This would nullify any impact on Chinese monetary conditions, but it would not have much impact on U.S. Treasury yields. With Chinese money still flowing into global bond markets, the re-balancing would only push other investors out of non-U.S. bond markets and into U.S. Treasuries. Without changing the overall demand for global bonds, it is difficult to envision much of an impact on U.S. yields. Bottom Line: Any attempt by the Chinese government to retaliate in the trade war by selling U.S. Treasury securities would be either self-defeating or ineffective, depending on the exact strategy employed. In either case, U.S. Treasury yields will be unaffected. A New Data Series For Fed Watchers: Rich’s Ratio A number of recent Fed speeches have referred to the time series plotted in Chart 10: The share of national income going to labor, as opposed to corporate profits. Chart 10Introducing Rich's Ratio Vice-Chair Richard Clarida brought this analysis to the Fed, and the data series was actually once dubbed “Rich’s Ratio” by Clarida’s old PIMCO colleague Paul McCulley. The idea behind Rich’s Ratio is that while some late-cycle wage gains are passed through to prices, a portion also eat into corporate profits. Notice in Chart 10 that Rich’s Ratio has a tendency to rise late in the economic recovery. Based on his past writings, we would not be surprised if at least part of the Fed’s recent dovish turn represents a desire to send Rich’s Ratio higher, even if that goal might entail a modest overshoot of the Fed's 2 percent inflation target. We will have more to say about Rich’s Ratio in the coming weeks. For now, we simply want to make Fed Watchers aware that they have a new series to track. Stay tuned. Ryan Swift, U.S. Bond Strategist rswift@bcaresearch.com Footnotes 1 Please see Geopolitical Strategy Weekly Report, “How Trump Became A War President”, dated May 17, 2019, available at gps.bcaresearch.com 2 Please see U.S. Bond Strategy Weekly Report, “Bond Kitchen”, dated April 9, 2019, available at usbs.bcaresearch.com 3 The rationale for why the CRB/Gold ratio tracks the 10-year Treasury yield is found in U.S. Bond Strategy Weekly Report, “The Search For Aaa Spread”, dated March 12, 2019, available at usbs.bcaresearch.com 4 Please see Global Investment Strategy Weekly Report, “Tarrified”, dated May 16, 2019, available at gis.bcaresearch.com 5 Please see U.S. Bond Strategy Weekly Report, “Paid To Wait”, dated February 26, 2019, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification