Global

Recommended Allocation A Series Of Unfortunate Events Markets have taken a series of hits in recent months - sharp drops in emerging market currencies, a political crisis in Italy, and the ongoing trade war between the U.S. and China - not to mention a slowdown in cyclical growth. But risk assets have been remarkably resilient: the U.S. stock market is in the middle of its year-to-date range, and U.S. small cap stocks (more attuned to domestic conditions) are at record highs (Chart 1). The uncertainty is set to continue for a while. But, with global growth likely to settle at an above-trend pace, fiscal and monetary policy still accommodative, and earnings continuing to grow strongly, the recent resilience says to us that risk assets are likely to grind higher and to outperform bonds over the next 12 months. A major underlying cause of the recent volatility has been the growing disparity between growth in the U.S. and in the rest of the world (Chart 2). This is partly due to the strength of the euro and yen last year, which is now dampening activity in these regions, but the slowdown in Chinese industrial growth and a higher oil price may also be having a disproportionate effect on growth outside the U.S. This growth disparity has widened interest rate differentials, which have again become the major driver of currencies, pushing up the U.S. dollar (Chart 3). Chart 1Small Cap Stocks At A Record High Chart 2Disparity Between The U.S. And The Rest... Chart 3...Means Dollar Has Further To Rise In combination with rising U.S. interest rates (the 10-year Treasury yield rose above 3% last month, before correcting a little), dollar appreciation is a threat for emerging markets. EM assets have long shown a consistently strong inverse correlation with the dollar (Chart 4). We expect the EM sell-off to continue. Further Fed hikes and rising inflation expectations in the U.S. (relative to the euro area and Japan) will increase interest-rate differentials and push the dollar up further: we forecast 1.12 for euro/dollar. Investors are still far from capitulating on EM assets after several years of large purchases (Chart 5). Many EM central banks are being forced to raise rates to defend their currencies, which will dent growth. Some may even be forced into reintroducing capital controls. Several emerging economies besides Argentina and Turkey remain vulnerable, having worryingly high amounts of foreign currency debt (Chart 6). Chart 4Strong Dollar Is Bad For Em Assets Chart 5Em Is Still A Consensus Favorite Chart 6Worrying Levels Of FX Debt Chart 7Not Surprising That Italians Are Fed Up Geopolitics is likely to remain a drag on markets for a while, too. Italy remains the biggest threat. The discontent of the Italian population is unsurprising given the country's stagnation since it joined the euro (Chart 7). The probable coalition government of the Lega and Five Star Movement would introduce aggressive fiscal stimulus, putting it in confrontation with the EU's budgetary rules. But BCA's geopolitical strategists see little risk of Italy exiting the euro in the next two years (though 10 years might be a different story).1 Political gyrations may continue for some months, particularly if the new government persists with its plan to blow the fiscal deficit out to 7% of GDP, but the sell-off in short-term Italian bonds looks to be overdone. Developments in trade tariffs, Iran and North Korea could also weigh on markets in coming months. But ultimately economic fundamentals almost always outweigh geopolitical risk. Global growth is slowing, but to an above-trend pace. Fiscal policy is particularly stimulative this year, with 17 of the 33 OECD countries undertaking large fiscal easing, and a further 11 some easing. The overall cyclically-adjusted primary budget balance in OECD countries is forecast to ease by 0.5% of GDP this year and 0.4% next (Chart 8). Monetary policy remains accommodative almost everywhere. The FOMC, in its May statement, by adding the word "symmetric" to describe its 2% inflation objective, was clearly emphasizing that it sees no need to accelerate the pace of rate hikes, despite the recent pickup in core PCE inflation. We expect the Fed to continue to raise rates once a quarter, meaning that monetary policy will not become restrictive until around Q1 next year. With inflation expectations not yet fully normalized (Chart 9), the Fed could still exercise its "put option" by holding for a quarter or two if global risk were to rise significantly. Italy's problems also make it more likely that the ECB will stay easier for longer, and the probability is rising of its deciding to extend asset purchases into next year. Chart 8Fiscal Stimulus (Almost) Everywhere Chart 9Inflation Expectations Have Further To Rise With the consensus already forecasting global GDP to grow 3.4% this year, and U.S. earnings by 22%, there is no obvious catalyst for risk assets to rebound sharply (Chart 10). However, we find it inconceivable that equity markets will not be higher in 12 months' time - and will not have outperformed bonds over that time - if the macro environment plays out as we expect. We, therefore, continue to recommend an overweight on equities and underweight on fixed income, but might start to turn more defensive around the end of the year if the signs are in place that the recession we expect in 2020 is still on the cards. Equities: For the reasons described above, we remain cautious on EM equities. Within EM, our preference would be for markets such as China, Korea and India, which are likely to be less affected by investors' concerns about current account deficits and foreign-currency denominated debt. In DM, our preference remains for late-cyclical sectors, especially energy, financials and industrials. We mainly view regional and country selection as a derivative of the sector call: this supports our preference for euro zone and Japanese stocks over those in the U.S. and U.K. Fixed Income: A combination of quarterly Fed rate hikes, a further normalization of inflation expectations, and moderate rises in the real rate and term premium are likely to push the 10-year U.S. Treasury yield up to 3.5% by year-end (Chart 11). We, therefore, remain underweight duration and prefer TIPs to nominal bonds. We keep our overweights on spread product within the fixed-income bucket, since it should continue to outperform for another couple of quarters. U.S. high-yield spreads are likely to remain steady, giving an attractive carry even after accounting for defaults; investment grade spreads might start to recover, given that the sell-off of quality bonds by companies repatriating short-term investments held offshore ($35 Bn from the 20 largest U.S. companies in Q1) is now mostly over (Chart 12). Chart 10Can Growth Beat These Expectations? Chart 11Treasury Yield To Rise To 3.5% Chart 12Selective Spread Product Remains Attractive Currencies: Interest-rate differentials, as described above, are likely to push the dollar up further, especially against the euro. This should continue until the effect of a strong dollar/weak euro starts to rebalance growth surprises back to the euro area, perhaps around the end of the year. We see less chance of dollar appreciation against the yen (which is still undervalued against its PPP value of 98, and may benefit from its safe-haven status) and against the Canadian dollar (given the Bank of Canada's hawkish stance). Commodities: Industrial commodities are likely to continue to struggle against headwinds from the appreciating dollar, and the continuing moderate slowdown in China (Chart 13). The oil price has become a tougher call recently, with talk that OPEC may agree later this month to bring back as much as 1 million barrels/day in production, but Venezuelan and Iranian supply likely to exit the market. BCA's energy strategists now forecast WTI and Brent to average $70 and $78 in 2H18, and $67/$72 in 2019, but expect higher volatility in the price over coming months (Chart 14).2 Chart 13Continuing Signs Of China Slowdown Chart 14Forecasting Oil Is Getting Harder Garry Evans, Senior Vice President Global Asset Allocation garry@bcaresearch.com 1 Please see BCA Geopolitical Strategy Client Note, "Italy, Spain, Trade Wars... Oh My!," dated 30 May 2018, available at gps.bcaresearch.com 2 Please see BCA Commodity & Energy Strategy Weekly Report, "OPEC 2.0 Guiding To Higher Output: Volatility Set To Rise ... Again," dated 31 May 2018, available at ces.bcaresearch.com GAA Asset Allocation

Highlights The global trade slowdown will intensify, even if U.S. domestic demand remains robust. The large emerging Asian bourses will recouple to the downside with their EM peers. Market-neutral EM equity portfolios should consider going long consumer staples while shorting banks. In Chile, receive 3-year swap rates. Continue to overweight stocks relative to the EM benchmark. Short the Colombian peso versus the Russia ruble. Stay neutral on Colombian equities and local bonds but overweight sovereign credit within their respective EM universes. Feature Performance of large equity markets in north Asia - Korean, Taiwanese and Chinese investable stocks -- has been relatively resilient compared with other EM bourses. Specifically, the EM ex-China, Korea and Taiwan equity index has already dropped 16% in U.S. dollar terms, while the market cap-weighted index of investable Chinese, Korean and Taiwanese stocks is down only 8% from its peak in late January.1 These three markets account for 60% of the MSCI EM stock index. A pertinent question is whether these North Asian markets will de-couple from or re-couple with the rest of EM. Our bias is that they will re-couple to the downside. Global equity portfolios should continue to underweight Asian stocks versus the DM bourses in general, and the S&P 500 in particular. That said, dedicated EM equity portfolios should overweight Korea and Taiwan and maintain a neutral stance on China and Hong Kong relative to the EM and Asian equity benchmarks. The Global Trade Slowdown Will Intensify Emerging Asian stock markets are very sensitive to global trade cycles. Slowing global trade is typically negative for them. There is growing evidence that the global trade deceleration will intensify: The German IFO index for business expectations in German manufacturing - a good leading indicator for global trade - is pointing to a further slowdown in global exports (Chart I-1). Chart I-1Global Trade Slowdown Will Persist Export volume growth has already slowed across manufacturing Asia (Chart I-2). The most recent data points for these series are as of April. Asia's booming tech/semiconductor industry is also slowing. Both Taiwan's export orders growth and Singapore's technology PMI new orders-to-inventory ratio have relapsed (Chart I-3). Chart I-2Asian Exports Growth: Heading Southward Chart I-3Asian Tech: Feeling The Pinch One of the causes of weakness in the global semiconductor cycle could be stagnating global auto sales (Chart I-4). The latter are being weighed down by weakness in auto sales in China and the U.S. Cars require a significant amount of semiconductors, and lack of improvement in global auto sales will suppress semiconductor demand. So far, China has not been at the epicenter of investors' concerns, but this will soon change as its growth slowdown intensifies. Credit conditions continue to tighten in China, which entails downside risks to mainland capital spending and consequently imports. China's imports are set to slump considerably, reinforcing the global trade downturn.2 First, China's bank loan approvals have dropped considerably in the past 18 months, suggesting a meaningful slowdown in bank financing and in turn the country's investment expenditures (Chart I-5). Chart I-4Global Auto And Semiconductor Sales Chart I-5China: Bank Loan Approval And Capex Second, not only are bank loan standards tightening but costs of financing are also rising. The share of loans extended above the prime lending rate has risen to a 15-year high (Chart I-6, top panel). This represents marginal tightening. Finally, onshore corporate bond yields as well as offshore U.S. dollar-denominated corporate bond yields have broken to new highs in this cycle (Chart I-6, bottom panels). Mounting borrowing costs and tighter credit standards in China point to further deceleration in credit-sensitive spending such as investment expenditures and property purchases. On the whole, rising interest rates and material currency depreciation in EM ex-China and credit tightening in China will prompt a considerable slump in imports, depressing world trade. EM including Chinese imports account for 30% of global imports, while the U.S. and EU together make up 24% of global imports values. Hence, global trade will disappoint if and as EM and Chinese imports stumble. A final word on the history of de-coupling among EM regions is in order. There have been a few episodes when emerging Asian and Latin American stocks de-coupled: In 1997-'98, the home-grown Asian crisis devastated regional markets, but Latin American stocks continued to rally until mid-1998 - when they plummeted (Chart I-7, top panel). Chart I-6China: Rising Borrowing Costs Chart I-7De-coupling Between Asia And Latin America In 2007-'08, emerging Asian equities tumbled along with the S&P 500, but Latin American bourses fared well until the middle of 2008 due to surging commodities/oil prices (Chart I-7, middle panel). Finally, the bottom panel of Chart I-7 illustrates that in early 2015, Asian stocks performed well, supported by the inflating Chinese equity bubble. Meanwhile, Latin American stocks plunged. In all of these episodes, the de-coupling between Asia and Latin America proved to be unsustainable, and the markets that showed initial resilience eventually re-coupled to the downside. Bottom Line: Global trade is set to head southward, even if U.S. demand remains robust. China's growth slump will be instrumental to this global trade slowdown. Consequently, Chinese, Korean and Taiwanese equities will be vulnerable. Heeding To Market Signals Financial markets often move ahead of economic data, and simply tracking data is not always helpful in gauging turning points in business cycles. By the time economic data change course, financial markets would typically have already partially adjusted. Besides, past economic and financial market performance is not a guarantee of future performance. This is why we rely on thematic fundamental analysis and monitor intermediate- and long-term trends in financial markets to navigate through markets. There are presently several important market signals that investors should be heeding to: EM corporate bond yields are surging, which typically foreshadows falling EM share prices (Chart I-8). Meanwhile, there is no robust correlation between EM equities and U.S. bond yields. Chart I-8EM Share Prices Always Decline When EM Corporate Bond Yields Rise The basis: So long as the rise in U.S. bond yields is offset by compressing EM credit spreads, EM corporate bond yields decline and EM share prices rally. But when EM corporate (or sovereign) yields rise, irrespective of whether this is due to rising U.S. Treasury yields or widening EM credit spreads, EM equity prices come under considerable selling pressure. Lately, both EM credit spreads have been widening, offsetting the drop in U.S. bond yields. Hence, a drop in U.S. bond yields is not in and of itself sufficient to halt a decline in EM share prices. So long as EM corporate and sovereign credit spreads are widening by more than the decline in U.S. Treasury yields, EM corporate and sovereign bond yields will rise, heralding lower EM share prices. The ratio of total return (including carry) of six commodities currencies relative to safe-haven currencies3 is breaking below its 200-day moving average after having bounced from this technical support line several times in the past 12 months (Chart I-9). This could be confirming that the bull market in EM risk assets is over, and a bear market is underway. Chinese property stocks listed onshore have broken down, and those trading in Hong Kong seem to be forming a head-and-shoulder pattern (Chart I-10). In the latter case, such a technical formation will likely be followed by a considerable down-leg. Chart I-9An Important Breakdown Chart I-10Chinese Property Stocks Look Very Vulnerable Further, China's onshore A-share index has already dropped by 15% from its cyclical peak in late January. Finally, both emerging Asia's relative equity performance against developed markets, as well as the emerging Asian currency index versus the U.S. dollar (ADXY) seem to be rolling over at their long-term moving averages (Chart I-11). The same technical pattern is presenting itself for global energy and mining stocks in absolute terms, and also in the overall Brazilian equity index (Chart I-12). Chart I-11Asian Equities And Currencies Are ##br##At Critical Juncture Chart I-12Commodity Equities And Brazil ##br##Are Facing Technical Resistance The failure of these markets to break above their long-term technical resistance levels may be signalling that their advance since early 2016 has been a cyclical - not structural - bull market, and is likely over. These technical chart profiles so far confirm our fundamental analysis that the EM and commodities rallies since early 2016 did not represent a multi-year secular bull market. If correct, the downside risks to EM including Asian markets are substantial, and selling/shorting them now is not too late. Bottom Line: EM including Asian stocks, currencies and credit markets are at risk of gapping down. Absolute-return investors should trade these markets on the short side. Asset allocators should underweight EM markets relative to DM in general and the U.S. in particular. A complete list of our currency, fixed-income and equity recommendations is available on pages 20-21. An EM Equity Sector Trade: Long Consumer Staples / Short Banks EM consumer staples have massively underperformed banks as well as the overall EM index since January 2016 (Chart I-13). The odds are that their relative performance is about to reverse. Equity investors should consider implementing the following equity pair trade: long consumer staples / short banks: Consumer staples are a low-beta sector because their revenues are less cyclical. As EM growth downshifts, share prices of companies with more stable revenue streams will likely outperform. Bank stocks are vulnerable as local interest rates in many EMs rise in response to the selloff in their respective currencies (Chart I-14). Consumer staples usually outperform banks when local borrowing costs are rising. Chart I-13Go Long EM Consumer Staples / ##br##Short EM Banks Chart I-14EM Banks Stocks Are Inversely Correlated With##br## EM Local Bond Yields We expect more currency depreciation in EM, which will exert further upward pressure on local rates, including interbank rates. Further, growth weakness in EM economies typically leads to rising non-performing loan (NPL) provisions. Chart I-15A and Chart I-15B demonstrates that weakening nominal GDP growth (shown inverted on the charts) leads to higher provisioning. Hence, a renewed EM growth slowdown will hurt bank profits. Chart I-15AWeaker Nominal GDP Growth Entails ##br##Higher NPL Provisions Chart I-15BWeaker Nominal GDP Growth Entails ##br##Higher NPL Provisions Our assessment is that banks in many EM countries have provisioned less than what is probably necessary following years of a credit boom. Indeed, in the last 12-18 months or so, many banks have even been reducing their NPL provisions to boost profits. Hence, a reversal of these dynamics will undermine banks' earnings. Bottom Line: Market-neutral EM equity portfolios should consider going long consumer staples while shorting banks. This is in addition to our long-term strategy of shorting EM banks versus U.S. banks as well as shorting banks in absolute terms in individual markets such as Brazil, Turkey, Malaysia and small-cap banks in China. Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com 1 These calculations are done using MSCI investible stock indexes in U.S. dollars terms. 2 Please see Emerging Markets Strategy Weekly Report, "The Dollar Rally And China's Imports", dated May 24, 2018, available at ems.bcaresearch.com. 3 Average of cad, aud, nzd, brl, clp & zar total returns (including carry) relative to average of jpy & chf total returns (including carry). Chile: Stay Overweight Equities, Receive Rates 31 May 2018 Chart II-1Chilean Equities Relative Performance And Copper Prices It is often assumed that Chilean financial markets are a play on copper. While this largely holds true for the Chilean peso, it is not always correct regarding its stock market's relative performance to its EM peers. Chile has outperformed in the past amid declining copper prices (Chart II-1). Despite our negative view on copper prices, we are reiterating our overweight allocation to this bourse within an EM equity portfolio. There are convincing signs that growth in the Chilean economy is moving along fine for now (Chart II-2). While weakness in global trade will weigh on the economy, the critical variable that makes Chile stand out from other commodities producers in the EM universe is its ability to cut interest rates amid currency depreciation. Chart II-3 illustrates that interest rates in Chile can and do fall when the peso depreciates. This stands in stark contrast with many others economies in the EM universe. There are a number of factors that suggest inflationary pressures will remain dormant for some time. This will allow the Central Bank of Chile (CBC) to cut rates as and when required. Chart II-2Chile: Economic Conditions Chart II-3Interest Rates In Chile Can Fall When Peso Depreciates First, the output gap is negative and has been widening, which has historically led to falling core inflation (Chart II-4). Second, a wide range of consumer inflation measures - services and trimmed-mean inflation rates - are very low and remain in a downtrend (Chart II-5). Chart II-4Chile: Output Gap And Inflation Chart II-5Chile: Inflation Is Very Low And Falling Finally, there are no signs of wage inflation, which is the key driver of genuine inflation. In fact, wage growth is decelerating sharply (Chart II-6). Odds are that this disinflationary rout will go on for longer, given Chile's demographic and labor market dynamics. The country's labor force growth has accelerated and the economy does not seem able to absorb this excess labor supply (Chart II-7). Consistently, our labor surplus proxy - calculated as the number of unemployed looking for a job divided by the number of job vacancies - has surged to all-time highs (Chart II-8). Chart II-6Chile: Wage Growth Is Very Weak Chart II-7Chile: Rising Labor Force Chart II-8Chile: Excessive Labor Supply... Interestingly, this is not happening because of weak employment. Chart II-9 shows that the employment-to-working population ratio is at a record high, while employment growth is robust. This upholds that decent job growth is not sufficient to absorb the expanding supply of labor. All in all, a structural excess supply of labor as well as a cyclical slowdown in global trade and lower copper prices altogether will likely warrant a decline in interest rates in Chile. Consequently, we recommend a new fixed income trade: Receive 3-year swap rates. The recent rise provides a good entry point (Chart II-10). Chart II-9...Despite Robust Employment Growth Chart II-10Chile: Receive 3-Year Swap Rates The ability to cut interest rates will mitigate the effect of weaker exports on the economy. We recommend dedicated EM investors maintain an overweight allocation in Chile in their equity, local currency bond and corporate credit portfolios. For absolute return investors, the risk-reward profiles for Chilean stocks and the currency are not attractive. The peso will depreciate considerably, and shorting it versus the U.S. dollar will prove profitable. Consistent with our negative view on copper prices, we have been recommending a short position in copper with a long leg in the Chilean peso. This allows traders to earn some carry while waiting for copper prices to break down. Stephan Gabillard, Senior Analyst stephang@bcaresearch.com Colombia: The Currency Will Be A Release Valve The structural long-term outlook for Colombia is positive, as a combination of pro-market orthodox policies and reform initiatives amid positive tailwinds from demographic should ensure a reasonably high potential GDP growth rate. In the first round of presidential elections held last weekend, the gap between right wing candidate Ivan Duque and left-wing candidate Gustav Petro came out large enough to make a Duque victory highly likely in the second round to be held on June 17. His election would entail a positive backdrop for the reform agenda and business investment over the coming years. Yet despite the positive structural backdrop, Colombia is still facing a major imbalance - excessive reliance on oil in sustaining stable balance of payments (BoP) dynamics. The trade balance deficit - including oil - is $8 billion, while excluding oil it stands at $20 billion, or 7.5% of GDP (Chart III-1). Hence, if oil prices drop materially in the second half of this year - as we expect - Colombia's balance of payments will be strained. Consequently, the currency will come under depreciation pressure. The peso is presently fairly valued as the real effective exchange rate based on unit labor costs is at its historical mean (Chart III-2). Chart III-1Colombia's Achilles' Hill: Trade Balance Excluding Oil Chart III-2The Colombian Peso Is Fairly Valued The central bank has adopted a "hands-off" approach toward the exchange rate, and is likely to allow the peso to depreciate if the BoP deteriorates. Weak economic conditions will likely prevent it from hiking interest rates to bolster the peso: Even though the central bank has reduced its policy rate by 350 basis points since the end of 2016, lending rates remain restrictive when compared with the nominal GDP growth rate (Chart III-3, top panel). Fiscal policy has been tight, with government expenditures subdued and the primary deficit narrowing (Chart III-3, bottom panel). This is unlikely to change for now if conservative candidate, Ivan Duque, wins the election. Consumer and business demand has failed to pick up, and shows little sign of recovery (Chart III-4). Non-performing loans (NPL) continue to rise, forcing banks to raise their NPL provisioning (Chart III-5). Weak nominal GDP growth suggests provisions may rise further. Chart III-3Colombia: Little Sign Of Recovery Chart III-4Colombia: Little Sign Of Recovery Chart III-5Colombian Banks: NPL And NPL Provision Continue Rising Overall, banks' balance sheets remain impaired, hampering their ability to extend loans. Investment Recommendations Despite a favorable structural outlook, Colombia's cyclical growth and financial market outlooks remain poor. Chances are that the peso will come under selling pressure as the external environment deteriorates - i.e., the currency will act as a release valve. We recommend staying neutral on Colombian stocks and local bonds relative to their EM peers, and to overweight Colombian sovereign credit within an EM credit portfolio. The basis is that sound and tight fiscal policies and a continuation of supply side reforms will benefit this credit market. To capitalize on potential currency depreciation while hedging for the uncertainty of oil price decline, we recommend shorting the peso against the Russian ruble. Although Colombia's structural outlook is more promising than Russia's, the latter's BoP dynamics is healthier and its cyclical growth outlook is better than Colombia's. Andrija Vesic, Research Analyst AndrijaV@bcaresearch.com Equity Recommendations Fixed-Income, Credit And Currency Recommendations

Highlights Chart of the WeekBCA's Ensemble Forecast Vs. Base Case With OPEC 2.0 signaling it will consider raising production in 2H18 to cover unexpected losses from Venezuela, and rising odds that state's output will cease, we've adopted an ensemble approach to forecast benchmark crude oil prices. This ensemble includes: i) our existing base case - steady demand and a loss of 500k b/d from Iran; ii) OPEC 2.0 restoring production cuts in 2H18; and, iii) explicit odds Venezuela's ~ 1mm b/d of exports collapse (Chart of the Week).1 We expect definitive output guidance following OPEC 2.0's June 22 meeting. For now, our base case dominates our 2H18 forecast, given our expectation any increase in production will be slowly restored to the market. Next year we see a higher probability most of OPEC 2.0's cuts will be restored. The odds that Venezuela's exports collapse goes from 20% in 2H18 to 30% in 2019. This ensemble forecast takes our 2H18 Brent forecast to $76/bbl from an average $78/bbl, and our WTI forecast to $70/bbl from $72/bbl. For next year, our Brent forecast goes to $73/bbl from $80/bbl, and our WTI expectation goes to $67/bbl from $72/bbl. We expect higher volatility, as well. Highlights Energy: Overweight. Spot Brent and WTI prices fell ~ 6% in the past week, as OPEC 2.0 signaled member states were considering restoring production. We remain long call spreads and the energy-heavy S&P GSCI, believing markets over-reacted to the news. Base Metals: Neutral. India's Tamil Nadu state government ordered the country's largest copper smelter shut, following rioting over alleged pollution from the plant, according to Bloomberg. This removes 400k MT of capacity from the market.2 Precious Metals: Neutral. Rising geopolitical risks in Italy are supporting gold prices, despite a stronger USD. Ags/Softs: Underweight. The re-emergence of U.S.-Sino trade tensions weighed on corn and soybean futures this week. This comes despite an ongoing truckers' strike in Brazil, which has been supporting soybean prices.3 Feature Just when it looked like OPEC 2.0 would keep its production cuts in place for the rest of the year, the coalition's leadership is signaling it will consider reversing production cuts during 2H18. Needless to say, this makes the task of forecasting prices more difficult. Guidance coming from the St. Petersburg Economic Forum at the end of last week was not definitive - it resembled more of a trial balloon. Press reports suggest as much as 1mm b/d of product cuts could gradually be restored to the market over 2H18, which would loosen global balances relative to our previous expectation (Chart 2). Still, Russia's energy minister Alexander Novak declined to confirm these cuts would be made.4 By our reckoning, some 1.2mm b/d of production actually has been cut by OPEC 2.0 since January 2017, mostly from KSA and Russia, which together account for close to 1mm b/d of the total. The big surprise on the production side has been the collapse of Venezuela, which went from just under 2.1mm b/d of crude output in Nov/16 - the month against which production targets were set under the OPEC 2.0 Agreement - to ~ 1.4mm b/d at present. We have Venezuela's production falling to 1.2mm b/d by the end of this year, and 1.0mm b/d by the end of 2019. We expect Iranian exports to fall ~ 200k b/d at the end of 2018, and another 300k b/d by the end of 1H19 in our base case model, as a result of the re-imposition of U.S. sanctions against it. This takes total Iranian export losses to 500k b/d by 2H19 in our base case. The only substantial growth on the production side is coming from U.S. shales in our base case, with production expected to be up 1.28mm b/d this year to 6.52mm, and 7.98mm b/d in 2019. Even this growth, however, could be constricted/delayed due to pipeline bottlenecks in the Permian. With demand expected to remain strong - growing at 1.7mm b/d this year and next in our models - market balances were tightening, and OECD inventories were falling appreciably (Chart 3). Chart 2Restoring OPEC 2.0 Production Cuts##BR##Would Loosen Global Balances Chart 3Inventories Would Draw Less If##BR##OPEC 2.0 Production Is Restored In 2018 The collapse of Venezuela's output did appreciably accelerate the tightening of the market, and lifted prices beyond the level that would have prevailed had this production not been lost to the market. This contraction, combined with the threatened re-imposition of sanctions on Iran, prompted leaders in important consumer markets to warn growth could be at risk with the oil-price rise potentially fueling inflation and inflation expectations - leading central banks, particularly the Fed, to continue tightening monetary policy. As gasoline, jet fuel and diesel prices rise, a greater share of household budgets goes toward purchasing hydrocarbons, which, all else equal, stifles growth if rising incomes cannot absorb the higher prices.5 Consumer Protests Registered With OPEC 2.0 Leaders in large oil-consuming states - particularly India, China and the U.S. - registered their dissatisfaction with high energy prices over the past month with OPEC 2.0, most notably when U.S. President Donald Trump tweeted his displeasure in April. OPEC Secretary General Mohammad Barkindo recalled the tweet at the St. Petersburg Economic Forum last week, saying, "I think I was prodded by his excellency Khalid Al-Falih that probably there was a need for us to respond. We in OPEC always pride ourselves as friends of the United States."6 Consumers in many states no longer are shielded from high oil prices, as governments around the world used the collapse in prices beginning in 2014 to remove/reduce fuel subsidies.7 This changes the dynamics of EM oil demand considerably, even if governments feel compelled to step into markets and order suppliers to not pass through the entire price increase. KSA and Russia appear largely united in their view of what is required to keep oil markets balanced over the long haul, so as not to disincentivize consumers from purchasing motor fuels. But over the short term, their goals differ. KSA is looking to IPO Saudi Aramco - next year, according to the latest reports - and this sale would most definitely benefit from higher prices. Indeed, KSA's oil minister Khalid al-Falih appeared to be comfortable with prices pushing toward $80/bbl recently. Russia's Novak has said in the past he favors an oil price somewhere between $50 and $60/bbl.8 Moving To Ensemble Forecasts Reconciling OPEC 2.0's short- and long-term goals, particularly the coalition's apparent new-found desire to be responsive to consumer interests; rising geopolitical tensions involving significant exporting states; and rising odds Venezuela implodes, and its exports are lost to the market, complicates the price-forecasting process considerably. In order to give full account to the different paths these uncertain influences will have on prices, we've adopted an ensemble model, in which we forecast three separate price paths: A base case, using our existing fundamental inputs and econometric modeling, which we published last week; A production-restoration case, where 870k b/d of production is restored to markets by OPEC 2.0 over 2H18 to compensate for the unexpected loss of Venezuela's output; The complete collapse of Venezuela's oil exports - amounting to ~ 1mm b/d - which we also published last week.9 In our base case, we use our standard fundamental model inputs - global production, consumption and OECD inventories - to forecast prices for this year and next (Table 1). The production-restoration and the Venezuela-export collapse models are boundary cases for our ensemble forecast, which is particularly important in 2019. The production restoration case leads to 870k b/d of OPEC 2.0 production coming back on line over the course of 2H18, with Venezuelan production deteriorating slowly, which is bearish for prices. The Venezuela-export collapse case results in a significant loss in production - 1mm b/d of Venezuela exports beginning in Jun/18 - which is bullish for prices, even with 1.2mm b/d of output being restored by OPEC 2.0 over the course of 2H18. Table 1BCA Global Oil Supply - Demand Balances (mm b/d) To generate the ensemble forecast, we weight the three cases above, with our base case dominating the model in 2H18, and falling off in 2019, while the production-restoration case dominates our outlook in 2019 (Chart 4). We also increase the probability of Venezuela's 1mm b/d collapsing over this interval - going from a 20% chance in Jun/18 to 30% in Dec/19. We will be continually updating these estimated probabilities (Table 2). Table 2BCA Ensemble Forecast Components As we approach OPEC 2.0's June 22 meeting in Vienna, we expect more definitive guidance from KSA and Russia, which will allow us to refine these probabilities. In addition, we expect volatility to increase, as changes in forward guidance and uncertainty in physical markets increases the rate at which speculators react to the arrival of new information (Chart 5).10 Chart 4Ensemble Forecast Accounts For##BR##Collapse In Venezuela's Exports Chart 5Spec Positioning Will##BR##Push Volatility Higher Bottom Line: OPEC 2.0 injected a new element of uncertainty into the markets this past week by signaling it would consider restoring oil-production cuts over 2H18, which could be as high as 1mm b/d, in response to consumer complaints at the highest levels. The guidance from the coalition's leadership in these early days does not allow us to definitely adjust our oil supply estimates, so we're simulating what we consider to be a highly likely schedule of production restoration. In addition, we are assigning explicit odds to the collapse of Venezuela's exports, which would remove ~ 1mm b/d of exports from the market. We combine these separate assessments with our existing forecasting model to create an ensemble forecast for prices in 2H18 and 2019. In this approach, our existing base-case model, which assumes OPEC 2.0's production cuts will be maintained this year and slowly restored over 1H19 is maintained; a production-restoration case is introduced, which assumes 870k b/d of production is brought back on line over the course of 2H18. Lastly, we assume Venezuela's production is lost to the market in Jun/18, and that OPEC 2.0 restores the 1.2mm b/d of actual production cuts it made beginning in Jan/17 over 2H18. We weight these different cases to produce our ensemble forecast. Using this approach, we are revising our 2H18 Brent forecast to $76/bbl from an average $78/bbl, and our WTI forecast to $70/bbl from $72/bbl. For next year, we are lowering our Brent forecast to $73/bbl from $80/bbl, and our WTI expectation to $67/bbl from $72/bbl. We expect higher volatility, as well. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com Matt Conlan, Senior Vice President Energy Sector Strategy mattconlan@bcaresearchny.com Hugo Bélanger, Senior Analyst Commodity & Energy Strategy HugoB@bcaresearch.com 1 OPEC 2.0 is the name we coined for the producer coalition led by the Kingdom of Saudi Arabia (KSA) and Russia, which agreed to cut 1.8mm b/d of production. By our reckoning, some 1.2mm b/d have been cut voluntarily - mostly by KSA and Russia. Alexander Novak, Russia's oil minister, stated actual cuts are closer to 2.7mm b/d, mostly because of the freefall in Venezuela's production. Non-Gulf states also have seen significant production losses. 2 See "Copper Supply Shock Hits India As Top Plant Ordered To Close," published by Bloomberg.com, May 29, 2018. 3 See "GRAINS-Corn, Soybeans Sag On Renewed U.S.-China Trade Jitters," published by Reuters.com, May 29, 2018. 4 Please see "OPEC, Russia Prepared To Raise Oil Output Amid U.S. Pressure," published by uk.reuters.com on May 25, 2018. 5 The OECD makes this point explicitly in its just-released report "OECD sees stronger world economy, but risks loom large," published May 30, 2018. 6 Please see fn. 3 above. 7 Please see "With the Benefit of Hindsight: The Impact of the 2014 - 16 Oil Price Collapse," published by the World Bank in January 2018. See fn. 11 for a list of EM countries that reformed their oil subsidies, which includes oil exporters in OPEC like KSA, Kuwait and Nigeria. 8 We discuss this at length in "OPEC 2.0 Getting Comfortable With Higher Prices," published February 22, 2018, by BCA Research's Commodity & Energy Strategy. It is available at ces.bacresearch.com. 9 We presented the Venezuela-production collapse simulation in last week's Commodity & Energy Strategy. Please see "Brent, WTI Average $80, $72 Next Year; Upside Risk Dominates, $100/bbl Possible In 2019." It is available at ces.bcaresearch.com. 10 We explore the relationship between price volatility and spec positioning in "Feedback Loop: Spec Positioning & Oil Price Volatility," published May 10, 2018, by BCA Research's Commodity & Energy Strategy. It is available at ces.bcaresearch.com. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed in 2018 Summary of Trades Closed in 2017

Highlights In this Weekly Report, we review all of the individual trades in our Tactical Overlay portfolio. These are positions that are intended to complement our strategic Model Bond Portfolio, typically with shorter holding periods, and sometimes in smaller or less liquid markets that are outside our usual core bond coverage (like Swedish government bonds or euro area CPI swaps). This report includes a summary of the rationale for each position, as well as a decision on whether to retain the position, close it or switch it into a new trade that has more profit potential for the same theme underlying the original trade (Table 1). Table 1Global Fixed Income Strategy Tactical Overlay Trades Feature U.S. Long 5-year U.S. Treasury bullet vs. 2-year/10-year duration-matched barbell (CLOSE AND SWITCH TO NEW TRADE) Long U.S. TIPS vs. nominal U.S. Treasuries (HOLD) Short 10-year U.S. Treasuries vs. 10-year German Bunds (HOLD) Chart 1UST Curve Trading More Off The Funds##BR##Rate Than Inflation Expectations We have three U.S.-focused tactical trades that are all expressions of our core views on U.S. inflation expectations and future Fed monetary policy moves. We first recommended a U.S. butterfly trade, going long the 5-year U.S. Treasury bullet and short a duration-matched 2-year/10-year Treasury barbell (Chart 1), back on December 20th, 2016. We have kept the recommendation during periodic reviews of our tactical trades since then. This is a position that was expected to benefit from a bearish steepening of the U.S. Treasury curve as the market priced in higher longer-term inflation expectations. The trade has not performed according to our expectations, however, generating a loss of -0.40% since inception.1 There was a positive correlation between the slope of the Treasury curve, the butterfly spread and TIPS breakevens shortly after trade inception. However, the Treasury curve flattened through 2017 as the Fed continued to hike rates, even as realized inflation fell (2nd panel), pushing the real fed funds towards neutral levels as measured by estimates like r* (3rd panel). This has left the 2/5/10 Treasury butterfly cheap on our valuation model (bottom panel), Looking ahead, the case for a renewed bear-steepening of the U.S. Treasury curve, and widening of the 2/5/10 butterfly spread, rests on the Fed accommodating the current rise in U.S. inflation by being cautious with future rate hikes. Recent comments from Fed officials suggest that policymakers are in no hurry to rapidly raise rates in order to cool off an "overheating" U.S. economy. Yet at the same time, U.S. inflation continues to rise and the economy is in good shape, so the Fed can't take a pause on rate hikes. This will likely leave the Treasury curve range bound, with the potential for some periods of bear-steepening as inflation expectations rise. Our conviction on this Treasury butterfly spread trade has fallen of late. Yet with our model suggesting that the belly of the curve is somewhat cheap to the wings, and given our view that U.S. inflation expectations have not reached a cyclical peak, we are reluctant to completely exit this position. Instead, we are opting to switch out of the 2/5/10 U.S. Treasury butterfly into another butterfly that our colleagues at BCA U.S. Bond Strategy have identified as cheap within their newly-expanded curve modeling framework - the 1/7/20 butterfly (long the 7-year bullet vs. short a duration-matched 1/20 barbell).2 That butterfly offers better carry than the 2/5/10 butterfly (Chart 2), and is nearly one standard deviation cheap to estimated fair value. Another of our U.S.-focused tactical trades has been to directly play for rising U.S. inflation expectations by going long TIPS versus nominal U.S. Treasuries. This is a long-held trade (initiated on August 23rd, 2016) which has performed very well, delivering a return of 4.13%.3 We continue to see the potential for TIPS breakevens to widen back to levels consistent with the market believing that inflation can sustainably return to the Fed's 2% target on the PCE deflator, which is equivalent to 2.4-2.5% on CPI-based 10-year TIPS inflation expectations. Given the persistent strong correlation between oil prices and breakevens, and with the BCA Commodity & Energy Strategy team continuing to forecast Brent oil prices jumping above $80/bbl over the next year (Chart 3), there is still solid underlying support for wider breakevens. This is especially true given the uptrend in overall global inflation (middle panel), and the likelihood that core U.S. inflation can also continue to rise alongside an expanding U.S. economy (bottom panel). We are sticking with our long TIPS position vs. nominal Treasuries. Chart 2Switch The UST Butterfly##BR##Trade From 2/5/10 to 1/7/20 Chart 3Stay Long U.S. TIPS##BR##Vs. Nominal Treasuries Our final U.S.-focused tactical trade is actually a cross-market trade where we are short 10-year U.S. Treasuries versus 10-year German Bunds. We initiated that trade on August 8th, 2017 when the Treasury-Bund spread was at 179bps. With the spread now at 252bps, the trade has delivered a solid total return of 4.23%. This was driven primarily by the rapid move higher in Treasury yields in response to faster U.S. growth (Chart 4), more rapid U.S. inflation and Fed rate hikes versus a stand-pat European Central Bank (ECB).4 From a medium-term perspective, those three fundamental drivers of the Treasury-Bund spread continue to point to U.S. bond underperformance (Chart 5). From this perspective, the peak in the spread will not be reached until U.S. economic growth and inflation peak and the Fed signals an end to its current tightening cycle. None of those outcomes is on the horizon, and we continue to target an eventual cyclical top in the 10-year Treasury yield in the 3.25-3.5% range as inflation expectations move higher. Yet the Treasury-Bund spread has reached an overvalued extreme according to our "fair value" model (Chart 6). In other words, the markets have moved to more than fully discount the cyclical differences between the U.S. and euro area - a trend that surely reflects the huge short positioning in the U.S. Treasury market. Yet it is also important to note that the fair value spread continues to steadily climb higher. In our model, the spread is primarily a function of differences in central bank policy rates between the Fed and ECB, relative unemployment rates and relative headline inflation rates. All three of those factors continue to move in a direction favorable to a wider Treasury-Bund spread, and the gap is only growing wider with both growth and inflation in the euro zone losing momentum. Chart 4Stay Long 10yr UST##BR##Vs. 10yr German Bund Chart 5UST-Bund Spread Widening##BR##Due To Relative Fundamentals... Chart 6...But The Spread##BR##Has Overshot A Bit The spread is currently being pushed to even wider extremes by the current turmoil in Italy, which is pushing money out of Italian BTPs into safer assets like Bunds. The situation remains fluid and new elections are likely in Italy later this year, thus it is unlikely that any more to restore investor confidence in Italy is on the immediate horizon. This will keep Bund yields depressed versus Treasuries, even as the ECB continues to signal that it will fully taper its asset purchases by year-end (rate hikes remain a long way off in Europe, however). We continue to recommend staying short Treasuries versus Bunds, and would view any tightening of the spread back towards our model estimate of fair value as an opportunity to enter the position or add to an existing position. Euro Area Long 10-year euro area CPI swaps (HOLD, BUT ADD A STOP AT 1.5%) Short 5-year Italy government bonds vs. 5-year Spain government bonds (HOLD) Chart 7Stay Long 10-Year Euro Area CPI Swaps We have two tactical trades that are purely within the euro area: positioning for higher inflation expectations through a long position in 10-year euro CPI swaps, and playing relative credit quality within the Peripheral countries by shorting 5-year Italian bonds versus a long position in 5-year Spanish debt. The long 10-year CPI swaps trade, which was initiated on December 20th, 2016, has generated a total return of +0.45% over the life of the trade so far (Chart 7).5 The rationale for the recommendation, and our conviction behind it, has evolved over that time. We first recommended the trade when the ECB was aggressively easing monetary policy and there was clear positive momentum in euro area economic growth that was driving down unemployment. At a time when oil prices were steadily climbing and the euro was very weak, the case for seeing some improvement in inflation expectations in the euro area was a strong one. Inflation expectations stayed resilient in 2017, however, despite the unexpected strength of the euro. Continued gains in oil prices and above-trend economic growth that rapidly absorbed spare capacity in the euro area more than offset any downward pressure on inflation from a stronger currency. Looking ahead, the combination of renewed weakness in the euro and firm oil prices should allow headline inflation in the euro area to drift higher from current levels in the next 3-6 months (2nd panel). However, the euro area economy has lost the positive momentum seen last year with steady declines in cyclical data like manufacturing PMIs, industrial production and exports (3rd panel). Admittedly, that deceleration has come from a high level and leading indicators are not yet pointing to a prolonged period of below-potential growth that could raise unemployment and reduce domestic inflation pressures. Yet with core inflation still struggling to climb beyond the 1% level (bottom panel), any worsening of euro area economic momentum could lead to inflation expectations stalling out well before getting close to the ECB's 2% target level. Thus, we continue to recommend this long 10-year CPI swaps position, but we are adding a new stop-out level at 1.5% to protect against downside risks if the euro area growth outlook darkens. On our other euro area tactical trade, we have been recommending shorting Italian government bonds versus Spanish equivalents. We initiated that trade on December 16th, 2016 and it has produced a total return of +0.57% over the life of the trade. The original logic for the trade was based on an assessment that Italy's medium-term growth potential, sovereign debt fundamentals and political stability were all much worse than that of Spain (Chart 8), yet Italian bond yields were still trading at too low a spread to Spanish debt. The cyclical improvement in the Italian economy in 2017 helped pushed Italian yields even closer to Spanish yields, yet we stuck with the trade given the looming political risk from the Italian parliamentary elections. The recent political turmoil in Italy has justified our persistence with this trade, with the 5-year Italy-Spain spread widening out by 46 bps over just the past two weeks. With the situation remaining highly fluid as the Italian coalition partners (the 5-Star Movement and the League) struggle to form a new government, Italian assets will continue to trade with a substantial risk premium to Spain and other European bond markets. Yet with the Italian economy now also showing signs of losing cyclical momentum, the case for continued Italian bond underperformance is a strong one, and we moved to a strategic underweight stance on Italian debt last week.6 Looking ahead, we see the potential for additional spread widening between Italy and Spain in the coming months. Spain is enjoying better economic growth, the deficit outlook is worsening for Italy with the new coalition government proposing a stimulus that could widen the budget deficit by as much as 6% of GDP, and Spanish support for the euro currency is far higher than it is in Italy. All those factors justify a wider risk premium for Italian debt over Spanish bonds (Chart 9). Chart 8Spain Trumps Italy On All Fronts Chart 9Stay Short 5-Year Italy Versus 5-Year Spain Our view on Italian debt, both from a tactical and strategic viewpoint, is bearish. We are maintaining our tactical trade, and we also advise selling into any rallies in Italy rather than buying the dips. U.K. Long 5-year Gilt bullet vs. duration-matched 2-year/10-year Gilt barbell (HOLD) We entered into a U.K. Gilt butterfly trade, long the 5-year bullet versus the duration-matched 2-year/10-year barbell, back on March 27th, 2018.7 The logic of the trade was a simple one. We simply did not believe that the Bank of England (BoE) would follow through on its hawkish commentary by hiking rates as much as was discounted in the Gilt curve. Our view came to fruition as the BoE held rates steady at the May monetary policy meeting, which resulted in a bullish steepening at the front end of the Gilt curve. Our butterfly trade has returned +0.25% since inception, and we see more to come in the coming months.8 The U.K. economy has lost considerable momentum, with no growth shown in Q1 (real GDP only expanded +0.1%). The OECD leading economic indicator for the U.K. is at the weakest level in five years, and now consumer confidence is rolling over as rising oil costs are offsetting the pickup in wages (Chart 10). Overall headline inflation has peaked, however, after the big currency-fueled surge in 2016 and 2017 (bottom panel). With both growth and inflation slowing, and with the lingering uncertainty of the Brexit negotiations weighing on business confidence and investment, the BoE will have a tough time hiking rates even one more time this year. There are still 34bps of rate hikes priced into the U.K. Overnight Index Swap (OIS) curve, which leaves room for 2-year Gilts to decline as the BoE stays on hold for longer (Chart 11). This will cause the front-end of the Gilt curve to steepen. Meanwhile, longer-term Gilt yields will have a difficult time falling given the deceleration of global central bank asset purchase programs that is slowly raising depressed term premia on government bonds (3rd panel). Another factor that will help keep the Gilt curve steeper, all else equal, is the path of the inflation expectations curve. Shorter-dated expectations are likely to fall faster as growth slows and headline inflation continues to drift lower (bottom panel). Chart 10Fading Momentum For##BR##U.K. Growth & Inflation Chart 11Stay Long The 5yr U.K. Gilt Bullet##BR##Vs. The 2/10 Gilt Barbell Although some narrowing of the butterfly spread is already priced in the forwards (top panel), we see that outperformance of the 5-year happening faster, and by a greater amount, than the forwards. Stay long the belly of the Gilt curve versus the wings. Canada Long 10-year Canada inflation-linked government bonds vs. nominal Canada government bonds (HOLD) We recommended entering a long Canada 10-year breakeven inflation trade on January 9th, 2018.9 Since then, the 10-year breakeven inflation rate rose by 6bps along with the rise in oil prices denominated in Canadian dollars (Chart 12). This has helped our tactical trade deliver a return of +0.64% since inception.10 More fundamentally, the breakeven has risen as strong Canadian growth has helped close the output gap and push realized Canadian inflation back to the middle of the Bank of Canada (BoC)'s 1-3% target band. The rapid rate of real GDP growth has decelerated a bit after approaching 4% last year, and the OECD leading economic indicator for Canada may be peaking at a high level (Chart 13). Growth in consumer spending is also look a bit toppy, with bigger downside risks evident in the sharp declines in the growth of retail sales and house prices (3rd panel). Both were affected by a harsher-than-usual Canadian winter, but the cooling of the overheated Canadian housing market (especially in Toronto) is a welcome development for financial stability. Chart 12Stay Long Canadian##BR##Inflation Breakevens Chart 13Canadian Inflation At BoC Target,##BR##But Has Growth Peaked? On balance, however, the current state of Canadian economic data shows an economy that is slowing a bit from a very overheated pace, but is still likely to grow above potential with no spare capacity available. Both headline and core inflation will remain under upward pressure against this backdrop, at a time when the BoC's policy rate is still well below neutral. We continue to recommend staying long Canadian inflation-linked government bonds over nominal equivalents with a near-term target of 2% on the 10-year breakeven inflation rate. We will re-evaluate the position with regards to Canadian growth and inflation trends once that target is reached. Australia Long December 2018 Australian Bank Bill futures (SELL AND SWITCH TO NEW TRADE). We entered into a long December 2018 Australian Bank Bill futures trade on October 17, 2017 as a focused way to express the view that the Reserve Bank of Australia (RBA) would stay on hold for longer than markets expect. The trade has worked out nicely, generating a profit of +0.25%. The potential for further upside is fairly low at these levels so we are now closing the trade. However, our view remains that the RBA will not be able to hike as early as markets are pricing. As such, we are opening a new position - long October 2019 Australia Bank Bill futures. Markets expect the first rate hike will occur in nine months' time. The October 2019 Australia Bank Bill futures are currently pricing in a massive 180bps of rate hikes over the next sixteen months. That will not happen. The RBA will not be able to hike this much given the lack of inflation pressures and a wide output gap. Our Australia Central Bank Monitor, which measures cyclical growth and inflation pressures, has pulled back to the zero line, confirming that there is no current need to tighten policy (Chart 14). Real GDP growth slowed to 2.4% in Q4 2017, from 2.9% the previous quarter. Weakness in the OECD leading economic indicator and Citigroup economic surprise index for Australia suggest that the Q1 reading will also disappoint. Consumer spending will be dampened by weak wage growth, softening consumer sentiment and the recent decline in house prices in multiple major cities. As a result of easing house prices, the growth rate of household net wealth was considerably lower in 2017 relative to the previous four years. Additionally, credit growth has been slowing, even before the recent news of the bank scandals that will force banks to be more stringent with lending practices. Most importantly, however, inflation remains below the RBA's target and there is a lack of inflationary pressures. The inflation component of our Central Bank Monitor has collapsed and is now well below the zero line. Both headline and core inflation readings are stable but remain persistently below 2%. Tradeable goods prices have declined for nine consecutive months despite the currency weakness seen in the Australian dollar over the past twelve months. The IMF is not projecting Australia to have a closed output gap until 2020, and that is with the optimistic expectation that Australia achieves 3% growth. Labor markets have plenty of slack as evidenced by rising unemployment rate, nonexistent wage growth and elevated level of underemployment. The RBA estimates that the current unemployment rate is still approximately 0.5% above full employment. Against this backdrop, it is unlikely that inflation will sustainably rise enough to force the RBA's hand, leaving scope for interest rate expectations to decline (Chart 15). Chart 14The RBA Will##BR##Stay Dovish Chart 15Switch Long Australia Bank Bill Futures##BR##Trade From Dec/18 Contract To Oct/19 Contract New Zealand Long 5-year New Zealand government bonds vs. 5-year U.S. Treasuries, currency-hedged into U.S. dollars (HOLD) Long 5-year New Zealand government bonds vs. 5-year German government bonds, with no currency hedge (HOLD) One of our more successful tactical trades has been in New Zealand (NZ) government bonds. We entered long positions in 5-year NZ debt versus 5-year U.S. Treasuries and 5-year German Bunds on May 30th, 2017, but we reviewed, and decided to maintain, those positions in a recent Weekly Report.11 The NZ-US spread trade has returned 4.67% since inception, hedged into U.S. dollars (Chart 16).12 The NZ-Germany trade, however, was a very rare instance where we recommended a cross-country spread trade on a currency UN-hedged basis, based on the negative view on the euro that we had last year. With the euro rising sharply against the New Zealand dollar, the unhedged return on that trade has been -2.87% (a return that, if hedged back into the euro denomination of the German bonds, would have generated a return of +3.56%). Looking ahead, we see continued scope for NZ bond outperformance, although the return potential is far less than it was when we first put on the trade. NZ economic growth is in the process of peaking, with export growth already rolling over (Chart 17, top panel). Net immigration inflows, which have been a major support for the NZ housing market and overall consumer spending over the past five years, have already begun to slow with the Reserve Bank of New Zealand (RBNZ) projecting bigger declines in the next couple of years (2nd panel). Both headline and core CPI inflation took a surprising downward turn in Q1 of this year, and both are well below the midpoint of the RBNZ target band (3rd panel). Chart 16Stay Long NZ 5yr Bonds##BR##Vs. The U.S. & Germany... Chart 17...With NZ Growth &##BR##Inflation Losing Momentum With both growth and inflation slowing, the RBNZ can remain dovish on monetary policy. An additional factor is the NZ government has recently changed the mandate of the RBNZ to include both inflation targeting and "maximizing employment" in a similar fashion to the Federal Reserve. With inflation posing no threat, the RBNZ can focus on its employment mandate by maintaining highly accommodative policy settings. With the NZ OIS curve still discounting one full 25bp RBNZ hike over the next year (bottom panel), there is scope for NZ bonds to outperform as that hike will not happen. This will allow NZ bond spreads to tighten, or at least outperform versus the forwards where some modest widening is currently priced. We are sticking with both spread trades, but we are choosing to leave the NZ-Germany trade currency unhedged given the renewed weakness in the euro (the unhedged return has already improved by over two full percentage points since the euro peaked earlier this year). We will monitor levels of the NZD/EUR currency cross rate to determine when to potentially hedge the currency exposure of our trade back into euros. Sweden Long Sweden 10-year government bond vs. 2-year government bond Short 2-year Sweden government bond vs. 2-year German government bond We recently entered two Sweden tactical bond trades on May 8, 2018, going long the Swedish 10-year vs. the 2-year and shorting the Swedish 2-year vs. the German 2-year (Chart 18).13 We expect that strong growth momentum, rising inflation and a tight labor market will force the Riksbank to raise rates earlier, and by more, than markets expect. Since inception for these "young" trades, each has returned -1bp.14 Sweden's economy made a solid recovery in 2017, with year-over-year real GDP growth reaching 3.3% in Q4. Going forward, export growth will remain supported by strong global activity, low unit labor costs, and a weak krona. Our own Swedish export growth model is already signaling a pickup over the rest of 2018. Consumption has been resilient and should continue to be supported by steadily recovering wages. Capital spending has been robust and industrial confidence remains in an uptrend. Additionally, leading indicators are still signaling positive growth momentum. The Riksbank's preferred measure of inflation, CPIF, slowed to 1.9% in April after briefly touching the central bank's target last month (Chart 19). In our view, this is a minor pullback rather than the start of a sustained reversal. Our core inflation model projects a gradual increase in the coming months, driven by above-trend growth that has soaked up all spare capacity. Labor markets have tightened considerably, and the unemployment rate is now more than one percentage point below the OECD's estimate of the full-employment NAIRU. During the last period when unemployment was this far below NAIRU, wage growth surged to over 4%. Chart 18Stay In A Sweden 2/10 Curve Flattener##BR##& Short 2yr Swedish Bonds Vs Germany Chart 19The Riksbank Will Not Ignore##BR##The Coming Inflation Overshoot For the curve flattener trade, our expectation is that the Riksbank will shift to a more hawkish tone in the coming months, leading markets to reprice the shape of the Swedish yield curve, as too few rate hikes are discounted in the short-end. With their mandates met, the Riksbank will be forced to act more aggressively. Importantly, there is no flattening currently priced into the Swedish bond forward curve, thus there is no negative carry associated with putting on a flattener now. In the relative value trade, we shorted the Swedish 2-year relative to the German 2-year. Growth in Sweden is likely to outpace that of the euro area once again in 2018. Swedish inflation is almost at the Riksbank target while euro area inflation continues to undershoot the ECB benchmark. The ECB is signaling that it is in no hurry to begin raising interest rates, therefore policy rate differentials will drive the 2-year Sweden-Germany spread wider over the next 12-18 months, with no spread move currently priced into the forwards. South Korea Short Korea 10-Year Government Bonds Vs. Long 2-Year Korea Government Bonds (CLOSE) We first introduced this trade on May 30th, 2017, after the election of Moon Jae-In as the South Korean president.15 The new government made major campaign promises to greatly expand fiscal spending on social welfare, public sector job creation, and increased aid to North Korea. With the central government's budget balance set to worsen significantly, we expected longer-term Korean bond yields to begin to price in faster growth and rising future debt levels, resulting in a bearish steepening of the yield curve (Chart 20). Since the new president was elected, however, the Korean economy worsened - even as much of the global economy was enjoying a cyclical upturn - with the trend likely to continue (Chart 21). The OECD leading economic indicator for Korea is weakening, while the annual growth in industrial production now sits at -4.2% - the worst level since the 2009 recession. Capital spending and exports are also slowing rapidly. Chart 20Close The 2yr/10y Korean##BR##Government Bond Curve Steepener Chart 21Korean Curve Stable,##BR##Despite Slower Growth & Fiscal Stimulus Due to the slowdown in the economy, Korean firms' capacity utilization is now at the worst level since the middle of 2009. Although businesses were already suffering from downward pressure on revenues, the Moon administration dramatically increased the minimum wage last year, directly leading to a rise in bankruptcies for small and medium size firms (the bankruptcy rate rose from 1.9% in the first half of 2017 to 2.5% in the latter half). Looking ahead, the Moon government will continue to increase spending on welfare and financial aid for North Korea, especially if the domestic economy continues to struggle. We still believe that the rise in deficits and debt will eventually lead the market to price in some increase in the fiscal risk premium and a steeper Korean yield curve. Yet with the Bank of Korea (BoK) having already surprised the markets last November with a rate hike, and with Korean inflation now ticking higher alongside a stable won, we fear that any renewed steepening of the Korean curve awaits a shift to a more dovish BoK that is not yet on the horizon. For now, given the competing forces on the Korean yield curve, we are choosing to close our 2/10 Korea curve steepener at a loss of -0.63%.16 We will continue to monitor the Korean situation to look for opportunities to re-enter the trade at a later date. Robert Robis, Senior Vice President Global Fixed Income Strategy rrobis@bcaresearch.com Patrick Trinh, Associate Editor Patrick@bcaresearch.com Ray Park, Research Analyst ray@bcaresearch.com 1 Returns are calculated using Bloomberg pricing of the total return of a 2/5/10 butterfly. 2 Please see BCA U.S. Bond Strategy Special Report, "More Bullets, Barbells And Butterflies", dated May 15th 2018, available at usbs.bcaresearch.com. 3 Return is taken directly from Bloomberg Barclays index data on the duration-adjusted excess return of the entire TIPS index versus the entire Treasury index. 4 This return is calculated using Bloomberg data on actual U.S. and German bonds, and is shown on a currency-hedged basis into U.S. dollars - the currency denomination of the bond we are short in this spread trade. 5 Returns are calculated using Bloomberg Barclays inflation swap index data for a euro area CPI swap with a rolling 10-year maturity. 6 Please see BCA Global Fixed Income Strategy Weekly Report, "Is It Partly Sunny Or Mostly Cloudy?", dated May 22nd 2018, available at gfis.bcaresearch.com. 7 Please see BCA Global Fixed Income Strategy Weekly Report, "Nervous Complacency", dated March 27th, 2018, available at gfis.bcaresearch.com. 8 Returns are calculated using Bloomberg data on actual Gilts, rather than bond index data. 9 Please see BCA Global Fixed Income Strategy Weekly Report, "Let The Good Times Roll", dated January 9th 2018, available at gfis.bcaresearch.com. 10 This return is measured as the total return of the Canadian inflation-linked bond index less that of the nominal Canadian government bond index from the Bloomberg Barclays family of bond indices. 11 Please see BCA Global Fixed Income Strategy Weekly Report, "Serenity Now", dated May 15th 2018, available at gfis.bcaresearch.com. 12 Returns are calculated using Bloomberg data on actual New Zealand government bonds, with our own adjustments for the impact on returns from currency hedging. 13 Please see BCA Global Fixed Income Strategy Special Report, "Sweden: The Riksbank Cannot Kick The Can Down The Road Anymore", dated May 8th 2018, available at gfis.bcaresearch.com. 14 Returns are calculated using Bloomberg data for actual individual Swedish government bonds, rather than bond index data. Both legs of the trade are duration-matched. 15 Please see BCA Global Fixed Income Strategy Weekly Report, "Distant Early Warning", dated May 30th 2017, available at gfis.bcaresearch.com. 16 Returns are calculated using Bloomberg data for actual individual Korean government bonds, rather than bond index data. Both legs of the trade are duration-matched and funding costs are included. Recommendations The GFIS Recommended Portfolio Vs. The Custom Benchmark Index Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Highlights Last year's broad-based global growth recovery has given way to slower growth and increasing differentiation in growth rates across economies. The U.S. has gone from laggard to leader in the global growth horse race, helping to drive the dollar to a five-month high. The biggest risk to our cautious view on emerging markets is that China stimulates the economy proactively as an insurance policy against a possible trade war. So far, there is little evidence that this is happening, but we are watching the data closely. The turmoil in Italy's bond markets is a timely reminder that if the European periphery wants more stimulus, this has to happen through a weaker euro rather than through larger budget deficits. Stay short EUR/USD. We expect to take profits at around the 1.15 level. Feature From Convergence To Divergence 2017 was the year of synchronized global growth. For the first time since 2007, all 46 countries tracked by the OECD experienced positive GDP growth. The euro area economy surprised on the upside, recording real GDP growth of 2.3%. This was slightly above U.S. levels, despite the fact that trend growth is about half a percentage point lower in the euro area. Growth in Japan nearly doubled to 1.7% from the prior year. Emerging markets, which succumbed to a broad-based slowdown starting in 2015, came roaring back. The U.S. dollar tends to perform poorly when global growth is accelerating and the composition of that growth is shifting away from the United States. This was precisely the setting that the global economy found itself in last year, which is why the greenback came under pressure. Things are looking sharply different this year. Global growth has cooled, as evidenced by both the PMIs and economic surprise indices (Chart 1). Euro area growth was sliced in half in the first quarter; U.K. growth decelerated further; and Japanese growth fell into negative territory for the first time since 2015. In contrast, the U.S. has held up relatively well. While growth did dip to 2.3% in Q1, the latest tracking estimates suggest a rebound in the second quarter. Retail sales accelerated in April. The Philly Fed PMI also surprised on the upside, with the new orders component reaching the highest level since 1973. The New York's Fed model is pointing to growth of 3.2% in Q2, while the Atlanta Fed's Nowcast is signaling growth of 4.1%. The divergence in growth rates between the U.S. and most major economies has been mirrored in recent inflation prints. U.S. core inflation has moved higher, but has stumbled elsewhere (Chart 2). Chart 1Global Growth Has Cooled With The U.S.##br## Faring Best Chart 2Inflation Is Accelerating In The U.S., ##br##Decelerating Elsewhere The relatively strong pace of U.S. growth has led to a widening in interest-rate differentials between the United States and its peers. The 10-year U.S. Treasury yield has risen by 95 basis points since its September lows, compared to 20 points for German bunds, 47 points for U.K. gilts, and 4 points for JGBs. With the exception of the U.K., the increase in spreads has been dominated by the real rate component (Chart 3). Chart 3Widening Interest Rate Differentials Between The U.S. And Its Peers ##br##Have Been Driven By The Real Component King Dollar Reigns Supreme Conceptually, it is real, rather than nominal, interest rate differentials that ought to move currencies. We noted earlier this year that the dollar's failure to strengthen on the back of rising Treasury yields was an anomaly that was unlikely to persist. Sure enough, the dollar has now begun to recouple with real interest rate differentials (Chart 4). Our sense is that this year's trends can last a while longer. Leading Economic Indicators have continued to move in favor of the U.S., suggesting that U.S. outperformance is not likely to end anytime soon (Chart 5). Fiscal policy should also help prop up U.S. aggregate demand. The U.S. structural budget deficit is set to widen much more than elsewhere over the next few years (Chart 6). Chart 4Dollar Is Recoupling With Rate Differentials Chart 5U.S. Is Outshining Its Peers Chart 6U.S. Fiscal Policy Is More Stimulative The U.S. economy is now back to full employment. For the first time in the 17-year history of the Bureau of Labor Statistics' Job Openings and Labor Turnover Survey (JOLTS), the number of job openings exceeds the number of unemployed workers (Chart 7). Our composite labor survey indicator has continued to move higher (Chart 8). Core PCE inflation has already accelerated to 2.3% on an annualized 6-month basis and 2.6% on a 3-month basis. The New York Fed's Inflation Gauge, which leads inflation by about 18 months, is pointing to higher inflation over the coming quarters (Chart 9). This means that the bar for further gradual rate hikes is quite low. Chart 7There Are Now More Vacancies Than Jobseekers Chart 8U.S. Wage Growth Is Set To Grind Higher Chart 9U.S. Inflation: Upside Risks Recent revelations by Kevin Warsh - who was once the favorite to lead the Federal Reserve - that Trump was dismissive of the Fed's historic independence during their interview, is only likely to strengthen Jay Powell's resolve to avoid being seen as a Trump flunky.1 China: Shifting Into The Slow Lane? Of course, the outlook for the dollar and bond spreads will also hinge on what happens in the rest of the world. We are watching two economies especially closely: China and Italy. The latest data suggest that China has lost some growth momentum. Retail sales and fixed asset investment decelerated in April. Property sales also declined from an elevated level. Sales tend to lead prices. Home prices were flat in most tier 1 cities over the prior year, reflecting elevated inventory levels, tighter lending standards, and stricter administrative controls (Chart 10). Further price weakness is likely, which could dampen construction activity in the months ahead. Industrial production beat expectations in April, but the overall trend in industrial activity remains to the downside. Electricity production, freight traffic, and excavator sales have all been decelerating (Chart 11). Import growth has also come down, which is one reason why GDP growth in the rest of the world has moderated (Chart 12). Chart 10China: Housing Has Cooled Chart 11China: Industrial Activity Is Slowing Chart 12China: Import Growth Has Decelerated Trade War Fears: Will China Overcompensate? In addition to the regular cyclical growth risks, concerns about a trade war loom in the background. The Trump Administration's decision last weekend to defer imposing tariffs on China caused investors to breathe a sigh of relief, but much remains unresolved, including ongoing allegations that China is stealing intellectual property from the U.S. and other countries. Trump's decision to pull out of June's summit with North Korea will only strain America's relationship with China. Considering the damage to China that a full-out trade war would cause, it would be sensible for the government to take out some insurance against a possible downturn. Thus far, any evidence that the authorities are trying to stimulate the economy through either fiscal or monetary means is sketchy (Chart 13). Reserve requirements were cut by 100 basis points in April, but corporate borrowing costs remain elevated. Fiscal outlays are growing at broadly the same pace as last year. The trade-weighted RMB has continued to strengthen. Still, it is hard to believe that the government has not put together a contingency plan that it could roll out if circumstances warrant it. The biggest risk to our fairly cautious view on emerging markets is that China launches a stimulus package in response to a trade war that quickly ends in détente. Similar to what occurred in 2008/09, this would leave China with more stimulus than it actually needed. Italy: From Fiscal Austerity To Bunga Bunga Unlike in China, Italy's incoming coalition government - forged through an uneasy alliance between the populist Five Star Movement (M5S) and the right-leaning League - has made no secret about its desire to ease fiscal policy. The M5S wants more social spending while the League has lobbied for a flat tax. These measures, along with a host of others, would add €100 billion, or 6% of GDP, to the budget deficit. Given that the Italian unemployment rate stands at 11% - 5.3 percentage points above its 2007 low - one could make a compelling case that Italy would benefit from temporary fiscal stimulus. However, the proposed policies are being marketed as permanent in nature. Moreover, several policies, such as the proposal to roll back the planned increase in the retirement age, would actually reduce potential GDP by shrinking the size of the labor force. It is no wonder that bond markets are worried (Chart 14). Chart 13China: No Clear Evidence Of Stimulus ... Yet Chart 14Mamma Mia! Propping Up Demand In Italy Much has been written about what Italy should be doing, but the fact is that there are no simple solutions. Italy suffers from a shrinking working-age population and anemic productivity growth, both of which reduce the incentive for firms to expand capacity. Like many other European countries, Italy also suffers from a debt overhang. This is obviously true for government debt but it is also true, to some extent, for private debt. While the ratio of private debt-to-GDP is below the euro area average, it stills stands at 113%, up from 65% in the mid-1990s (Chart 15). The desire to save more in order to pay back debt, coupled with a reluctance to invest in new capacity, has left Italy with what economists call a private-sector financial surplus (Chart 16). Chart 15Italian Private Sector Has Been Taking ##br## On Less Debt Since The Crisis Chart 16Italy: The Private Sector Wants To Save If the private sector earns more than it spends, the excess savings have to be absorbed either by the government through its own dissaving or by the rest of the world through a current account surplus. Both options are problematic for Italy. Running large budget deficits for a prolonged period of time would take the level of government debt-to-GDP to stratospheric levels. Japan has been able to get away with this strategy because it issues debt in its own currency. This is a luxury that is not at Italy's disposal. Despite Mario Draghi's pledge to do "whatever it takes" to preserve the euro area, it is far from clear that the ECB would keep buying Italian debt if the country began to openly skirt the EU's deficit rules. Absent an effective lender of last resort, the Italian bond market could fall victim to a speculative attack - a process in which higher yields lead to even higher yields, and eventually a default (Chart 17). Chart 17When A Lender Of Last Resort Is Absent, Multiple Equilibria Are Possible This just leaves the option of trying to bolster aggregate demand by exporting excess production abroad via a current account surplus. To its credit, Italy has been able to shift its current account balance from a deficit of 1.4% of GDP in 2007 to a projected surplus of 2.6% of GDP this year. However, some of that surplus simply reflects the fact that a weak economy has suppressed imports. Progress in reducing unit labor costs relative to its euro area peers has been painfully slow (Chart 18). Chart 18Italy: More Work To Be Done To Improve Competitiveness If Italy had a flexible exchange rate, it could simply devalue its currency to gain competitiveness. Since it does not have one, it has to improve competitiveness by restraining wage growth and implementing productivity-enhancing structural reforms. The former requires the presence of labor market slack, while the latter, even in a best-case scenario, will take substantial time to achieve. And neither option is politically popular. Given the difficulty of raising Italy's competitiveness relative to the rest of the euro area, the only realistic short-term solution is to boost it relative to the rest of the world. That requires a weak euro which, in turn, requires a dovish ECB. Investment Conclusions In our Second Quarter Strategy Outlook, published on March 30th, we predicted that the dollar was poised to experience a violent rally as short sellers rushed to cover their positions. This view has played out in spades. As we go to press, the nominal broad-trade weighted dollar has gained 4% since early April. It is up 30% since bottoming in July 2011 and is only 6% below its December 2016 peak (Chart 19). The dollar rally has brought our views closer in line with the market. Notably, EUR/USD is now less than two percent above our target of $1.15. The dollar is an ultra-high momentum currency. Chart 20 shows that a simple strategy of buying the DXY when it was above its moving average and selling it when it was below its moving average would have delivered a sizable profit over the past two decades (the exact moving average does not matter much, but the 50-day seems to work best). As such, while we intend to turn neutral on the dollar if it gains another few percent or so, an overshoot is quite probable. Chart 19The Dollar Has Bounced Back Chart 20The Dollar Trades On Momentum About 80% of EM foreign-currency debt is denominated in dollars. In many cases, dollar borrowers have non-dollar revenue streams. Thus, a stronger dollar automatically hurts their businesses. In the past, this has often ignited a feedback loop where a stronger dollar triggers capital outflows from emerging markets, leading to an even stronger dollar. Our EM strategists strongly feel that such a vicious cycle is fast approaching, especially if China's economy continues to slow. In the late 1990s, brewing EM tensions triggered several brutal equity selloffs. For example, the S&P lost 22% between July 20 and October 8, 1998. However, EM stress also restrained the Fed from tightening too quickly. The resulting dose of liquidity set the stage for a massive blow-off rally between the fall of 1998 and the spring of 2000. A similar dynamic could unfold this time around. We remain overweight global equities for now, but are hedging the risk by being short AUD/JPY, a trade that has gained 5% since we initiated it on February 1st. Peter Berezin, Chief Global Strategist Global Investment Strategy peterb@bcaresearch.com 1 Ben White, "How Trump could break from the Fed's independence," Politico, May 9, 2018. Strategy & Market Trends Tactical Trades Strategic Recommendations Closed Trades

Highlights Stable global demand; steady declines in Venezuela's crude oil output; and the cumulative loss of 500k b/d of Iranian exports to U.S. sanctions by 2H19 will lift average Brent and WTI prices to $80 and $72/bbl in 2019, respectively (Chart of the Week). Brent prices will average $78/bbl in 2H18, while WTI goes to $72/bbl, as these supply-side effects are not material to prices this year. We lowered our estimate of Venezuela output to 1.2mm b/d by end-2018 (vs. 1.3mm b/d previously), and to 1.0mm b/d by end-2019 (vs. 1.2mm b/d). Offsetting these losses and continued deterioration in non-Gulf OPEC supply in 2019, we assume OPEC 2.0 slowly restores 1.2mm b/d in 1H19, and U.S. shale oil grows 1.4mm b/d. Even so, balances tighten significantly (Chart 2).1 Chart of the WeekBrent Will Average $80/bbl In 2019 Chart 2Balances Tighter As Supply Falls If Venezuela collapses, and its ~ 1mm b/d of crude exports are lost, Brent crude oil could go to $100/bbl by end 2019, in the simulation we ran assuming exports collapse in 2H18. Uncertainty over supply and demand responses to higher prices makes this difficult to model. Highlights Energy: Overweight. Our options recommendations - long Brent call spreads spanning Dec/18 to Aug/19 delivery - are up an average 50.5%. Our long S&P GSCI position, recommended Dec 7/17 to take advantage of increasing backwardation, is up 18.9%.2 Base Metals: Neutral. Copper rallied earlier this week on an apparent easing of trade tensions between the U.S. and China. However, a statement by U.S. President Trump suggesting uncertain progress in talks led to a reversal in most of these gains by mid-day Wednesday. Precious Metals: Neutral. Our long gold portfolio hedge and tactical long silver position were relatively flat over the past week, as the broad trade-weighted USD moved higher. Ags/Softs: Underweight. China's Sinograin, the state grain buyer, reportedly was in the market this week showing interest in purchasing U.S. soybeans, according to agriculture.com's Successful Farming website. Feature Barring the immediate collapse of Venezuela's oil industry and the loss of its ~ 1mm b/d of oil exports, which we discuss below beginning on page 7, the global crude market will continue to tighten from the supply side, on the back of ratcheting geopolitical pressures. Chief among these are the continuing loss of Venezuelan crude oil production, which, even without a total collapse that wipes out its ~ 1mm b/d of exports, will see production fall to 1.2mm b/d by the end of this year from ~ 1.44mm b/d at present. This represents a decline in our previous estimate of 100k b/d. By the end of 2019, we expect Venezuela production to fall to 1.0mm b/d, 200k b/d below our previous estimate. One year ago, Venezuela was producing just under 2.0mm b/d of crude. The other supply source affected by geopolitics is Iran, where we expect export volumes to fall later this year, due to the re-imposition of U.S. nuclear-related sanctions (Chart 3). We are modeling a loss of 200k b/d by year-end 2018, and a cumulative loss of 500k b/d by the end of 1H19.3 Lastly, we have raised the probability OPEC 2.0 keeps its production cuts in place in 2H18 to 100% from 80%. This added $2/bbl to our 2018 Brent forecast. We expect a wider Brent - WTI differential this year, and left our 2018 WTI forecast at $70/bbl. Chart 3Iran Exports Down 500k b/d By 2H19, In BCA Model The steady decline in Venezuelan production and the loss of Iranian exports, coupled with an extension of OPEC 2.0's production cuts to end-2018, will take total OPEC crude oil production to 32.0mm b/d this year (down 300k b/d y/y), and 31.7mm b/d next year. Non-Gulf OPEC production also falls: coming in at 7.5mm b/d this year, these producers account for a 300k b/d y/y loss, and, at 7.0mm b/d next year, a 500k b/d y/y loss in 2019. Once again this leaves non-OPEC production as the leading source of new supply: We have total non-OPEC liquids (crude, condensates and other liquids) up 2.12mm b/d to 60.7mm b/d this year, and up 2.11mm b/d next year. This is led - no surprise - by U.S. shales, which we expect to increase by 1.3mm b/d this year to 6.52mm b/d, and 1.5mm b/d next year to 7.98mm b/d, respectively (Chart 4). Net, we expect global crude and liquids supply to average 99.73mm b/d this year, and 101.76mm b/d in 2019. On the demand side, our growth estimates are unchanged in our latest balances model. We continue to expect global demand growth of 1.7mm b/d this year and next - the prospects of which strengthened with an apparent dialing back of U.S. - China trade animosities over the past week (Chart 5). This will move the level of global consumption up to 100.3mm b/d this year and 102mm b/d next year, as can be seen in Table 1. Chart 4Steady Decline In Venezuela Exports,##BR##Iran Sanctions Tighten Markets Chart 5Global Demand Remains Strong In##BR##Our Updated Balances Models The effect of the supply-side adjustments to our model - holding our demand assumptions pretty much constant - can be seen in the new path of OECD inventories vis-à-vis the 2010 - 2014 five-year average level of stocks (Chart 6). OPEC 2.0's strong compliance with its production-management agreement, along with losses of Venezuelan and Iranian exports and above-average demand growth caused estimated OECD commercial inventories to fall ~ 303mm bbls versus Jan/17 levels. Table 1BCA Global Oil Supply - Demand Balances (mm b/d) Chart 6Tighter Markets, Lower Inventories,##BR##Keep Forward Curves Backwardated Keeping OECD inventories below their 2010 - 2014 average levels means Brent and WTI forward curves will remain backwardated at least to the end of 2019, which, we believe, is OPEC 2.0's ultimate goal. This will ensure the coalition's member states receive the highest price along these forward curves, while the coalition's U.S. shale-oil rivals are forced to hedge at a lower price a year or two forward. Backwardation also works to the advantage of commodity index investors, particularly when the investable index is heavily weighted to oil and refined products like the S&P GSCI.4 This recommendation is up 18.9% since it was recommended Dec 7/17. Net, we expect Brent prices to average $78/bbl in 2H18, while WTI goes to $72/bbl. For next year, we expect Brent to average $80/bbl and WTI to average $72/bbl. Simulation Of A Venezuela Supply Shock To Oil Markets The likelihood Venezuela manages to maintain exports of ~ 1mm b/d this year and next falls daily.5 Were markets to lose these export volumes, they initially would scramble to replace them, leading to a short-term price spike, in our view. We simulated the loss of Venezuela's ~ 1mm b/d of exports, assuming these volumes fall off in June, and starting, in Jul/18, OPEC 2.0 gradually restores the 1.2mm b/d it actually cut from production over 2H18. By Jan/19 OPEC 2.0's 1.2mm b/d cuts are fully restored, in our simulation. However, the loss of Venezuela exports is only fully realized in 2H19, assuming oil consumption stays strong. Brent prices end 2019 ~ $100/bbl (Chart 7). OECD inventories fall to ~ 2.65 billion bbls by end 2018, and to ~ 2.32 billion bbls by end-2019 (Chart 8). This is not unreasonable, given the inelasticity of demand to price over the short term, but we would expect that in 1H20, demand would fall in response to higher prices. Chart 7Oil Prices Move Higher In Our Simulation,##BR##If Venezuela's Exports Collapse... Chart 8... OECD Inventories Drop Sharply,##BR##As Well Of course, by that time, the supply side likely would have adjusted as well. We will be exploring this further and developing additional simulations to understand the evolution of prices beyond 2020. How this plays out is unknowable at present. But, as a starting point for understanding the implications of losing Venezuela's exports, this is a reasonable set of assumptions, given the challenges in not only returning OPEC 2.0 volumes removed from the market, but getting them to refining centers in 2H18. What is unclear at present is how governments will use their strategic petroleum reserves (SPRs), and whether OPEC will fire up spare capacity to handle the loss of Venezuela's exports, should this occur. Much will depend on how OPEC 2.0 and consumer governments' SPRs interact if exports collapse. Production Cuts, Inventories, SPRs And Spare Capacity In the simulation above, we reckon OPEC 2.0 flowing production can be brought back to market in fairly short order, and that still-ample inventories and spare capacity would be available to cover the sudden loss of Venezuela's exports, to say nothing of strategic petroleum reserves held in the U.S., China, Japan, and the EU. The key, though, is how long it would take to get this supply to market, and how governments holding SPRs react. We estimate it will take anywhere from one to three months to begin to restore the volumes OPEC 2.0 took off the market if Venezuela goes offline. It will take a few months for the restored crude production to start flowing into pipelines and on to ships, followed by 50- to 60-day journeys from the Gulf to be delivered to refining centers. Chart 9OPEC Spare Capacity ~ 2% Of Global Supply,##BR##Lower Than 2003 - 2008 Price Run-Up In the meantime, refiners would continue to draw crude inventory to supply product markets, along with product inventories, a critical consideration going into the northern hemisphere's summer driving season. In a short-term pinch, governments could draw their strategic petroleum reserves to fill the gaps while OPEC 2.0 production is being restored, and markets get back to the status-quo ante prevailing prior to the loss of Venezuela's exports.6 OPEC's ~ 1.9mm b/d of spare capacity - most of which is located in KSA - could be called upon in an emergency; however, this requires 30 days to be brought on line, per U.S. EIA, and can only be sustained for at least 90 days (Chart 9). The EIA is forecasting OPEC spare capacity will fall from current levels of 1.9mm bbls to ~ 1.3mm bbls by end-2019.7 Given these uncertainties, we continue to recommend investors remain long Brent crude oil option call spreads, which we recommended over the course of the past few months.8 We expect prices and volatility to move higher, both of which are positive for option positions. Bottom Line: Venezuela's crude oil production is in free-fall. We estimate it will drop to 1.2mm b/d by the end of this year, and to 1.0mm b/d by the end of next year. Iran's exports could fall 500k b/d by the end of 1H19, as a result of the re-imposition of nuclear sanctions by the U.S. These geopolitically induced supply losses tighten markets in 2019, raising our prices forecasts for Brent and WTI to $80 and $72/bbl, respectively. We are raising our Brent forecast for 2018 by $2/bbl, expecting prices to average $76 and $70/bbl, respectively, since these risks likely do not kick in until late in 2018. A collapse in Venezuelan production could spike prices to $100/bbl by the end of 2019, even as OPEC 2.0 restores the 1.2mm b/d of production it removed from markets beginning in 2H18. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com Hugo Bélanger, Senior Analyst Commodity & Energy Strategy HugoB@bcaresearch.com 1 OPEC 2.0 is the name we coined for the producer coalition led by the Kingdom of Saudi Arabia (KSA) and Russia. Its production cuts of ~ 1.2mm b/d and natural declines have removed ~ 1.8mm b/d from the market. 2 Backwardation is a term of art used in commodity markets to describe an inverted forward price curve - i.e., prompt-delivery commodities trade higher than the same commodity delivered in the future. The opposite of backwardation is contango. 3 There is an extremely high degree of uncertainty around this estimate, which is why we are treating it as our Bayesian prior, and will be revising it as additional information becomes available. We do not believe all of the production restored by Iran post-sanctions - 1mm b/d - will be lost to export markets, but starting with a prior of ~ half of it being lost due to less-than-full re-imposition of sanctions is reasonable. 4 Commodity-index total returns are the sum of price appreciation registered by being long the index; "roll yield," which comes from buying deferred futures in backwardated markets, letting them roll up the forward curve as they approach delivery, selling them, then replacing them with cheaper deferred contracts in the same commodity; and collateral yield, which accrues to margin deposits on the futures comprising the index. Roll yield can be illustrated by way of a simplistic example: Assume the oil exposure in an index is established in a backwardated market - say, spot is trading at $62/bbl and the 3rd nearby WTI future trades at $60/bbl. Assuming nothing changes, an investor can hold the 3rd nearby contract until it becomes spot, then roll it (i.e., sell it in the spot month and replace it with another 3rd nearby contract at $60/bbl) for a $2/bbl gain. This process can be repeated as long as the forward curve remains backwardated. 5 Matters have only gotten worse since the Council on Foreign Relations published its so-called Contingency Planning Memorandum No. 33 February 13, 2018, titled "A Venezuelan Refugee Crisis," which opened with the following: Venezuela is in an economic free fall. As a result of government-led mismanagement and corruption, the currency value is plummeting, prices are hyperinflated, and gross domestic product (GDP) has fallen by over a third in the last five years. In an economy that produces little except oil, the government has cut imports by over 75 percent, choosing to use its hard currency to service the roughly $140 billion in debt and other obligations. These economic choices have led to a humanitarian crisis. Basic food and medicines for Venezuela's approximately thirty million citizens are increasingly scarce, and the devastation of the health-care system has spurred outbreaks of treatable diseases and rising death rates. The CFR's memo is available at https://www.cfr.org/report/venezuelan-refugee-crisis 6 There is no way to model exactly how this will play out, absent a detailed plan put forward by the IEA and China, where the largest SPRs reside. IEA members have bound themselves to hold reserves equal to 90 days of net petroleum imports. Among the largest SPRs, U.S. holds just over 660mm barrels of oil in its SPR; China held ~ 290mm barrels at the end of last year, based on IEA estimates. Germany and Japan together hold close to 550mm bbls, according to the Joint Organizations Data Initiatives (JODI). KSA's crude oil inventories - not exactly SPRs - stood at ~ 235mm barrels in March, according to JODI. We are highly confident disposition of these reserves in the event of a shock to Venezuela's exports is being discussed in Washington, Paris, Riyadh and Beijing. Please see p. 2 of the U.S. Government Accountability Office's Testimony Before the subcommittee on Energy, Committee on Energy and Commerce, House of Representatives, "Strategic Petroleum Reserve, Preliminary Observations on the Emergency Oil Stockpile," released for publication Nov. 2, 2017. 7 This actually is a fairly low level of spare capacity, amounting to ~ 2% of global supply. During, the price run-up of 2003 - 2008, OPEC's total spare capacity was near or below 3% of supply and that was considered tight at the time. 8 Please see p. 11 for a summary of these trades' performance. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed in 2018 Summary of Trades Closed in 2017

Special Report

"Amongst all unimportant subjects, football is by far the most important." - Pope John Paul II From June 14 to July 15, billions of people will tune into the 2018 FIFA World Cup, with untold loss of productivity, working hours, and quality family time as a consequence. At BCA Research, we decided to stop pretending that we are indifferent to the quadrennial sporting pilgrimage - or to our staff hogging bandwidth by live-streaming games during business hours - and instead harness the best young minds of the company to deliver this most eminently unimportant forecast. To our clients who may worry that their hard-earned dollars are subsidizing fleeting research pursuits, we want to assure you that the research herein was produced "off the clock." In our firm's 70-year history, the principal question driving all analysis has been "so what?" Each and every piece of analysis produced at BCA Research is intended to conclude with an investment angle that makes sense of the noise produced by the cacophony of data. We do not intend to evolve out of that genetic material. At the same time, we are passionate about research. Even though this report does not have obvious investment implications, it is an excellent example of our research fundamentals.1 We develop a macro framework through qualitative analysis, enrich it with data manipulated in novel ways, and articulate it through empirical testing. Just as we produced this report "off the clock," we hope that our clients consume it in their down time. The following pages are a respite from President Trump's tweets, elevated equity valuations, prospects of tepid returns, renewed confrontation in the Middle East, and trade wars. It is also the first in a series of in-depth, multi-disciplinary reports that we intend to produce. Stay tuned. This BCA Special Report presents a unique two-step empirical approach to predict the outcome of each World Cup match. Our approach includes micro (player level) and macro (team level) factors to forecast game matches. We present this model in Section I. Section II then introduces several interesting qualitative narratives, focusing on specific teams attending this year's competition. Section III builds on the existing academic literature and our own unique framework to establish whether the World Cup has any market and economic implications. I. Two-Step Model: Forecasting The 2018 FIFA World Cup In this section we introduce a two-step simulation of the upcoming World Cup. Building on academic literature that has shown that teams respond differently to the various stages of the tournament, we develop two separate models for the group and knockout stages of the competition (Box 1).2 Box 1: The World Cup: A Quick Overview The quadrennial World Cup finals will take place in Russia over a period of one month. The tournament is referred to as the "finals" because the actual competition to select the 32 teams began in 2015. The qualification tournament whittled down the 209 FIFA member states (minus the already qualified host, Russia) over the course of three years via continental qualification tournaments. The 32 teams competing in the finals are separated into eight groups. Each team will play one game against each of their group opponents. The top two performers advance to the knockout stage of the competition. In this stage, the World Cup resembles the NCAA March Madness tournament in that teams face immediate elimination. The 2018 FIFA World Cup is the 21st edition of the competition that goes back to 1930. In the history of the competition, only eight nations have won the cup. Brazil has won five times, Germany and Italy four, Argentina and Uruguay twice, and France, England, and Spain once each. We are not the first to attempt to predict the World Cup. In a recent innovative approach, Clemente et al. (2015) use parameters from network analysis to analyze the tournament's matches. More traditional models rely on the FIFA World Ranking, which ranks teams based on past performance.3 Other conventional models draw on economic fundamentals to explain the success and failure of national teams. A literature overview from 2004 of both quantitative and qualitative models found that a simulator running a commercially produced computer game offered the most successful prediction of the World Cup.4 This should not come as a surprise. The computer gaming industry has overtaken Hollywood in terms of total revenue, while production costs of some computer games are running higher than those of blockbuster movies.5 Given that realism is an important feature of sport simulation, it should follow that computer games have a better forecasting track record than humans. In this report, we combine macro and micro variables to develop a unique two-step model. Our micro factors are based on an extensive database consisting of individual player statistics. Inspired by the 2004 study above, we rely on the computer gaming industry for player ranking, modifying it with our own collective soccertise. The list of potential explanatory variables that we considered including in our model is available in Table 1. Table 1Variables Considered And Used For The Models By relying on individual player attributes we believe that we have avoided the pitfalls of more "macro"-oriented models, such as those relying on overall team rankings. Overall team rankings are established on the basis of team performance over a multi-year period of time. We find that this approach overstates team performance over team quality in present time. These models also fail to incorporate tactical analysis into the model. We introduce these macro and qualitative factors - such as reputation and long-term historical performance - in the qualitative part of our evaluation, while letting objective, player- and team-specific data, dominate our model. Data To determine which variables from Table 1 had the highest predictive power we relied on the 192 matches that occurred during the 2006, 2010 and 2014 FIFA World Cups. There are two reasons we focused only on the last three tournaments. The first is data availability. The second is that football has undergone dramatic evolution in strategy in the twenty-first century. Among the 192 matches, 48 took place in the knockout stage of the competition, while the remaining 144 games occurred during the group stages. We rely on the database of player statistics used in Electronic Arts (EA) Sports FIFA computer simulation. The database includes player statistics from August 2004 to present. At the time of the publication of this report, not all 32 teams had made official their final list of 23 players (the lists will be published on June 4th by the FIFA). Therefore, we relied on the most recent World Cup qualifiers, as well as all the friendly matches played year-to-date to come up with the most likely line-ups for these teams. Step One: The Group Stage Model To simulate the 2018 group stage matches, we developed an Ordered Probit (OP) model estimated on past World Cups group stage games. Ordered Probit models are powerful when modeling an ordinal outcome (i.e. the response value has a strictly increasing ordering known prior to the estimation). Football matches can end in either a loss, a draw, or a win, with a well-defined order from loss to win. The Ordered Probit model therefore allows us to use this information, increasing the predictive ability of the model.6 The Ordered Probit model selected is represented using a continuous latent variable yi* that is linearly determined by a set of explanatory variables χι: Where φ is the cumulative normal distribution function and ϒ are arbitrary thresholds selected via log-likelihood maximization.7 Based on our sample of group-stage matches from the past three World Cups, we found that the best explanatory variables are: Team Average Player Rating Average Age - Forwards Average Number of Caps - Defenders Speed Positions Average Rating Group Stage Explanatory Variables Team Average Rating "Talent wins games, but teamwork and intelligence wins championships." Thus said six-time NBA champion Michael Jordan. Our model confirms that the insight from the basketball legend applies to football as well, as average team talent - based on the mean of individual player overall attributes taken from the EA Sports database - is the most significant predictor of game success in both stages of the competition. However, this variable becomes less important in the knockout stages, as the ability to play as a team ("teamwork"), as well as particular tactical matchups ("intelligence"), become paramount in winning the tournament. This is because, in the late stages of the competition, the talent differential between teams narrows substantively. Average Age - Forwards Teams with younger forwards tend to perform better than teams with older forwards due to the highly physical demands of the position. Forwards perform the highest amount of high-intensity efforts (sprinting and sudden changes of direction) and experience the most amount of physical contact among all positions in football.8 Considering that these abilities tend to peak in the early to mid-20's for most professional athletes, it is no surprise that our model gives a premium to youth in this position.9 Moreover, the annals of World Cup history are chock-full of tales of youthful forwards making a name for themselves when it matters.10 Figure 1Position Definitions Average Number Of Caps - Defenders Teams with more experienced defenders will tend to do better than teams with less experienced defenders. The physical demands on defenders tend to be less strenuous relative to other positions, given that they dish out the pain. This means that defenders tend to peak later than any other field position in football and are able to maintain peak performance sometimes well into their 30's. What defenders lack in athleticism they must make up in game IQ, as anticipation, composure, and tactical awareness are all acquired skills that improve with experience. Speed Positions Average Rating Teams with superior talent in their speed positions (full backs and wingers) will perform better in group stages (Figure 1). Academic research has shown that counterattacking football is the most effective tactic against unbalanced defenses.11 Wingers and full backs are crucial for both developing and defending against a counterattack as they cover the least congested part of the pitch, where there is generally the most space to run. Interestingly, this variable becomes less important in the knockout stages. This is most likely due to the fact that teams with skilled and fast wingers can easily exploit the space conceded by the tactically disorganized defenses of easier opponents that populate the early stages of the competition. In the later stages, teams are generally more tactically disciplined. Modeling Group Stage Games For each game, our model uses the spread between Team 1 and Team 2 of each explanatory variable to determine the probability of Team 1 winning the game. Table 2 lists all 32 teams and their descriptive statistics on the four explanatory variables. Table 3 presents how the 32 teams are ranked based on their probability of passing to the next stage inferred by these explanatory variables. Table 2Descriptive Statistics: Group Stage Table 3Group Stage Ranking Our predictive variables were selected based on the results of the 2006, 2010 and 2014 World Cups. The estimated coefficients are then used to produce 1,000 simulations of each game in order to obtain a distribution of the outcome. Table 4Marginal Effect Of Selected##BR##Variables: Group Stage The team average ratings are the core variables in our group stage model. However, due to the high level of multicollinearity in the disaggregated player rating variables, we could not capture the entire set of information from these individual variables. As a result, our final probabilities are derived from the weighted average of two separate estimations, Model 1 and Model 2. This allows our model to capture the importance of the speed positions average rating in predicting the outcome of the group stage matches, which displays the highest marginal impact on the winning probability (Table 4). Therefore, our final model for the probability of winning a game is: Where: Model 1 (M1)=ƒ (Team Average Player Rating, Forward Average Age, Defenders Average Caps) and, Model 2 (M2)= ƒ(Speed Positions Average Rating, Forward Average Age, Defenders Average Caps), and α and (1- α) are the weights given to each models.12 The ultimate product of our modeling is a coefficient we have termed E(points). As a reminder, teams are awarded three points for a win, one point for a draw, and zero points for a loss. The maximum amount of points a team can gain is nine, given that there are three matches (Box 2). BOX 2: E(points) Calculation E(points) allows us to determine which teams have the highest probability of passing to the knockout stages. We caution readers from reading too much into E(points) in terms of which teams move on to the knockout rounds as the competition in each group greatly determines the value. For example, Denmark has a higher E(points) coefficient than Serbia, but that is a function of its relatively easy group (it faces weaker competition). Group Stage: Results Group A Group A was one of the most challenging to forecast (Table 5). First, Uruguay is probably one of the weakest of the group favorites in the competition.13 Second, host Russia and Egypt are separated by few points in terms of overall quality. Table 5Group A Summary Results To make our lives easier, we decided to attribute a significant home advantage bonus to Russia.14 History tells us that hosting the World Cup provides a clear advantage to almost any team (Table 6). One out of every three hosts has won the competition, going back to the first tournament in 1930. A whopping 65% of all hosts have made it all the way to the semi-finals. It pains us to see the "Egyptian King," Liverpool F.C. superstar Mohamed Salah, exit at the group stage. However, history is arrayed against his squad. Russia would have to be only the second host team ever - aside from South Africa in 2010 - to fail to capitalize on its home court in the group stage. And, as history teaches us, you just don't beat Russians on Russian soil. Group B Group B was, by far, the easiest to forecast (Table 7). The two Iberian giants will make it through, a high-conviction view. Portugal will try to become the fourth team to win the European Championship and the World Cup consecutively, joining such great teams as West Germany (Euro 1972, World Cup 1974), France (World Cup 1998, Euro 2000), and Spain (Euro 2008, World Cup 2010, Euro 2012). Our knockout stage model likes Portugal, giving it the fifth-best probability of winning the entire competition. Its performance in the group stage will go far in validating our confidence in the team. Table 6Home Advantage Is Real Table 7Group B Summary Results Group C A pre-tournament favorite, France will look to avenge its stunning loss to Portugal sans Cristiano Ronaldo in the finals of the 2016 Euro. France starts its competition off in one of the weakest groups (Table 8). Group C lacks real competition, with Denmark having only a 25% probability of beating France. This is the lowest probability of success for any second-ranked team in the group stage of the competition. Anything less than 9 points for France would be a concern for the fans of Les Bleus. Table 8Group C Summary Results Group D Argentina and Croatia are the favorites to go through in Group D, but our model likes Nigeria's and Iceland's chances (Table 9). Iceland keeps turning forecasters into fools. As we discuss in Section II, its success in international football is empirical evidence of the divine. Its run in the 2016 Euro Cup, and win against England, may be the greatest upset in any major sporting competition. However, the element of surprise is gone. Table 9Group D Summary Results Nigeria, on the other hand, could be a dark horse. Our model likes their chances, with 31% probability that they advance to the next stage. They have also won only one of their last 12 World Cup matches, which suggests that they may be overlooked coming into the tournament. Group E Brazil is likely to dominate Group E, but the fight for second place, and thus qualification into the knockout round, will be vicious. Switzerland and Serbia represent a real challenge to our model. Their values in the four explanatory factors are tantalizingly close (Chart 1). Serbia gets the narrowest of pushes on three out of the four, but our model assigns identical probabilities of winning (39%) to each team when they face each other head-to-head (Table 10). The model gives the slightest of chances to Switzerland, mainly due to its higher probability of stealing points in its game against Brazil (48% compared to 46% for Serbia). It is a stunning revelation backed by history. Serbia has no luck against non-European competition in the tournament. Other than its impressive win against the eventual finalist Germany in 2010, it has lost to Argentina (by 6-0!), Ivory Coast, Ghana, and Australia. Furthermore, momentum is not on the side of the Orlovi, given that they backed their way into the World Cup with a 3-2 loss to Austria and a nail-biter against Georgia. These results prompted the Serbian federation to replace the head coach months ahead of the World Cup, not a reassuring sign. Chart 1Serbia Vs. Switzerland Table 10Group E Summary Results Group F Germany, the 2014 World Cup winner, will find very little opposition in Group F and should easily avoid the fate of the past two reigning champions who were eliminated early (Table 11). Die Mannschaft was the only team in Europe that won all its qualification games, ending the qualifying tournament with an otherworldly +39 goal difference. Our model likes Mexico over Sweden. This makes sense given that El Tri have progressed past the group stage in the previous six World Cups (only to be promptly eliminated in the first game of the knockout stage every time). Table 11Group F Summary Results Group G Our model finds Group G exceedingly easy to predict. Belgium and England will go through (Table 12). Both teams have a lot in common, starting with quality squads that have failed to make a mark in recent international competitions. Belgium comes to the 2018 World Cup with top quality players in many positions on the field (Table 13). The two squads will find no real competition in this group, but the pressure will be immense on both young squads. How they handle lowly Tunisia and Panama will determine their mental readiness for the knockout stage of the competition. Table 12Group G Summary Results Table 13Top Players In Every Position On The Pitch Group H Group H is by far the most difficult to forecast (Table 14), with all four teams clustered around a similar E(points) value (Chart 2). Poland has the best player - Bayern Münich superstar Robert Lewandowski - and Colombia the best team. However, both Senegal and Japan could surprise. The forecasting error in Group H should have little bearing on further results given the relatively low ranking of all four teams (for example, both Switzerland and Serbia from Group E are ranked as superior teams). However, the top two teams from this group will "cross" against Belgium and England, the two chronic underachievers. As such, winning the group carries considerable upside as the winner would face the young, untested, and skittish England. This means that Colombia, Poland, or perhaps even Senegal and Japan, could end up stunning the world and finishing at least in the top eight. Table 14Group H Summary Results Chart 2Probability Of Proceeding To The Knockout Stage Group Stage Results Table 15 illustrates group stage results. We assign points based on the highest probability outcomes (win, draw, loss) of each game. Table 15Group Stage Results How confident are we of the results? Other than groups E and H, we are highly confident. Box 3 goes over our three most likely "Dark Horses" to go through to the second round. BOX 3: BCA's "Dark Horses" Serbia The Serbian national football team has faced many challenges over the past two decades. Despite these hurdles, Serbia has produced top-class footballers that are integral parts of several European super teams. An experienced defense and aggressive midfield - headlined by one of the world's premier defensive midfielder Nemanja Matic of Manchester United - makes Serbia a hard team to play against. There is little of the Balkan flair on this team that defined Yugoslav football in the 1980s. But perhaps that is a good thing, given that World Cup games are tight, nervous, and often devolve into a defensive slog. Senegal It has been 16 years since Les Lions de la Teranga last appeared in a World Cup - and what a memorable appearance that was. In 2002, Senegal reached the quarter finals, the second African team in history to make it that far in the competition. The 2018 squad shows similar potential. The Lions' forwards pose the kind of threat to opponents' defenses that many of their World Cup rivals would envy. Keep in mind some of these names: Sadio Mané (Liverpool F.C.), Ismaila Sarr (Rennes), and Keita Balde (A.S. Monaco). Peru Peru is our choice for a long shot at this year's tournament. The Peruvian national football team is unbeaten since November 2016, winning eight matches and drawing four, with many coming against elite opposition. Although we do not take overall team rankings into consideration when making forecasts, it is notable that Peru is ranked 11th in the world, according to the ELO index. Peru's path to Russia took it through the South American qualifying tournament, known for being the most competitive in the world. Teams have to play 18 games, many at high altitude or in unforgiving climates, against some of the world's most talented squads. While it is true that they were the last team to qualify, having to beat New Zealand in the intercontinental playoffs, Peru came ahead of Copa America champions Chile and just one and two points behind Colombia and Argentina, respectively. While our model does not believe in the relatively young and unproven Peruvian team, placing it last in the group, the team's excellent coaching and fearless play make it a potential candidate to come out of Group C along with France, especially given that our model may be overstating Denmark's potential. Step Two: The Knockout Stage Model The knockout stage is somewhat easier to model given that the set of possible outcomes is reduced to only {loss; win}. This difference with the group stage is not only relevant for the math behind our model. It is also relevant for the strategy teams employ during the games. Therefore, we simulated this part of our analysis using a probit model estimated on a sample of only knockout stage games from the 2006, 2010 and 2014 World Cups. The binary choice probability of observing a specific outcome becomes: In this stage of the competition, we found the following factors to be the most important: Team Average Player Rating Club Level Synergy Player GINI Coefficient Average Rating - Midfielders Knockout Stage Explanatory Variables Team Average Rating As with the group stage, the overall rating of the team - based on the average of individual rankings from the EA Sports database - is the most powerful explanatory variable. Despite the higher marginal effect of the rating variables in the knockout stage sample, the standard deviation and average of these variables are significantly smaller than in the group stage.15 In other words, the gap in player quality between teams in the group stage is often vast. However, the knockout stage culls the minnows, narrowing the gap in overall player quality between teams. At this stage of the competition, our model has to be supplemented with variables that test for teamwork and synergy. Club Level Synergy Teams with more players playing in the same club tend to perform better in the knockout stages. This is evident from all the World Cup winners in our sample.16 Given the limited practice time that national teams have ahead of the tournament, the year-round experience of playing with teammates in club competition can provide a huge advantage. Especially for football teams from countries with major leagues - such as Germany, Spain and Italy. Their players are more likely to cluster on the major clubs in those leagues, whereas players from smaller footballing nations have to ply their trade in dispersed leagues and teams across the globe. Great Man Theory (Player GINI coefficient) Teams win games, but heroes win World Cups. To test whether superstars are relevant to winning games, we designed a player quality GINI measure. We find that teams with a higher GINI coefficient outperform those with a lower measure. It seems that having a superstar, or two or three, surrounded with role players is a superior strategy to having balanced talent across all positions. This variable only becomes significant in the knockout stages, where the overall talent level between teams narrows. While more skilled teams tend to be more balanced (Chart 3), once we normalize for skill, a higher GINI becomes a predictor of success. Chart 3The Great Man Theory Intuitively we agree with this finding. When the stress and tension of knockout stages peak and the weight of one's entire nation begins to crush a lesser player's shoulders, the Great Man shines through. Exceptional players have the skill, stamina, and otherworldly confidence to unlock the highly disciplined and tactically sound defensive schemes found in the latter stages of the competition. This should be good news for Belgium, Portugal, and Argentina, but really bad news for England. Average Rating - Midfielders Once we decompose the different positions in the field, we find that the midfield is more important to success than other positions in the knockout matches. Research has shown that midfielders, particularly those forming the "spine" of the team, are the most involved in a team's passing play, regardless of the tactics or strategy used.17 Precise and creative passing is key in knockout matches where tactically disciplined defenses are difficult to unlock. Defensive prowess in the midfield is also paramount to prevent the opposition from developing their attack.18 As with the group stage model, the final probabilities for the knockout stage games were derived from the average of two models in order to maximize the information contained in the average midfield player rating variable (Table 16). Table 16Marginal Effect Of Selected Variables: Knockout Stage Therefore, our final probability is: Where: Model 1 (M1) = ƒ(Team Average Player Rating, Club Level Synergy, Player GINI Coefficient) and, Model 2 (M2) = ƒ(Midfielders Average Rating, Club Level Synergy, Player GINI Coefficient), and a and (1 - α) are the weights given to each models.19 Table 17 summarizes the descriptive statistics of each team according to the four variables used to model their performance. Table 17Descriptive Statistics: Knockout Stage Knockout Stage: Results The Round of 16 There were no surprises in the round of 16 (Table 18). The two matches worth paying attention to are England vs. Colombia and Portugal vs. Uruguay. Our model gives the two European teams a 60% chance of winning their respective games. We see the "Great Man Theory" carrying Portugal forward to the quarter-finals. We are essentially willing to bet that Ronaldo is at least worth a quarter-finals berth at the World Cup. Table 18Round Of 16 Summary Results The England-Colombia matchup is essentially a coin toss. The young and untested Three Lions will face a tough challenge in Colombia. However, it is safer to carry England over to the next round as their "conditional probability" of progressing to the quarter-finals is significantly higher than that of Colombia (53% to 17%) (Chart 4). Why the difference? Conditional probability takes into account the original probability of passing the group stage. Given England's easy group, and Colombia's challenging competition, it is mathematically safer to bet on England. Chart 4Probability Of Advancing To The Quarter-Finals Quarter-Finals Making it into the quarter-finals of the World Cup is an extraordinary success reserved for only eight footballing nations. At this point, teams have played four intense and decisive games over three weeks. Fatigue sets in, especially given that the superstars are playing at the end of a grueling club season in what is normally their off-season. Our model bets strongly on the previous two World Cup winners (Table 19), while Brazil and France struggle against their opponents. Belgium and Portugal are left to wonder what might have been, although for Belgium the pain will be greater. Not only will they yet again fail to meet expectations, but also they will waste the highest conditional probability of advancing to the next stage of the four teams that do not advance (Chart 5). Table 19Quarter-Finals Summary Results Chart 5Probability Of Advancing To The Semis France Vs. Portugal: Setting The Record Straight The loss to Portugal in the final of the 2016 Euro, on home soil no less, still stings for France. The loss was particularly painful given that Portugal's superstar, Real Madrid's Cristiano Ronaldo, had to leave the game due to an injury and spent the majority of the game hopping on one leg, yelling instructions to his teammates from the sidelines. Our model gives France a 66% probability of winning the game. The French team is superior in every facet of the game, other than in the speed positions (Diagram 1). France also sports two of the game's best defensive midfielders, Chelsea's N'Golo Kanté and Juventus's Blaise Matuidi, to whom it will fall on to neutralize Portugal's Great Man. French coach Didier Deschamps has also benefited from the overall improvement in the French Ligue 1, calling upon players that play together in the local league. Diagram 1Les Bleus Vs. A Selecao das Quinas Diagram 2A Selecao Vs. The Red Devils We expect a tight match, with intense man-on-man coverage of Ronaldo. France will dominate the ball, slowly building chances against Portugal's defense. In 2016, a young and inexperienced French team wasted a plethora of chances against Portugal's version of the Maginot Line. We do not see history repeating itself. Brazil Vs. Belgium: Red Devils Don't Dance Samba The second tight matchup predicted by our model pits Brazil against Belgium. As with the France-Portugal matchup, the underdog has a solid one-in-three chance of winning the game. What makes Belgium so dangerous is their overall midfielder rating, which we identified as one of the four most relevant factors for winning the game. The problem for Belgium is that it trails Brazil by a lot in other facets of the game (Diagram 2). Its defense is particularly suspect. Belgium has no players ranked in the top five of their defensive position, whereas Brazil sports a fifth of all the best defensive players in the entire tournament. Spain Vs. Argentina: Don't Cry For Me Lionel Our model has no time for La Albiceleste, giving Argentina a paltry 10% chance of defeating Spain. As a reminder, our model is completely ignorant of head-to-head matchups between teams, but it has essentially predicted the 6-1 drubbing that La Roja gave Argentina in a friendly in late March. Granted, Argentina's superstar, and F.C. Barcelona icon, Lionel Messi watched the spanking from the bench. But unless Messi can play defense at the same high level at which he finishes attacks, Argentina is in trouble. Argentina's 2018 squad is essentially the stereotype of every team we have watched from the nation in the last half-century. Its forwards are world class, perhaps the best in the tournament (Diagram 3). Four Argentines are amongst the 15 best forwards in our sample, an extraordinary number (see Table 13). Nevertheless, the rest of the team is subpar relative to other favorites, and particularly against Spain. Germany Vs. England: Hard Brexit Defending champions rarely meet expectations at the World Cup. The tournament is simply too grueling, the pressure too great, and the chasm of time between tournaments too wide to bridge within a single generation. However, Germany faces a young, untested English team in our bracket. Getting to the quarter finals would be seen as a success for England, one upon which to build a winning generation for the 2022 tournament in Qatar. Germany tops England in all facets of the game and across the pitch (Diagram 4). Our model gives England little chance against Die Mannschaft. England has greater chances of extracting a soft Brexit from Berlin than of penetrating Germany's clinical press. Diagram 3La Furia Roja Vs. A Albiceleste Diagram 4Die Mannschaft Vs. The Three Lions Semi-Finals Our model gives Belgium and Portugal decent odds for advancing to the semis, leaving a door open for surprises in the quarter-final round. In the semi-finals, however, the model tightens the odds, snuffing out the chances for an aging Germany and an inexperienced Brazil (Table 20). Brazil's and Germany's chances fall to just 15% and 11%, respectively (Chart 6). Table 20Semi-Finals Summary Results Chart 6Conditional Probability Of##BR##Advancing To The Finals Spain's conditional probability of reaching the finals hovers at an extraordinarily high 54%. France remains at a respectable, and yet uncertain, 41%. France Vs. Brazil: Painful Memories We are not surprised that our model assigns Brazil just a 27% probability against France. France practically coasted to the finals of the Euro 2016, defeating Germany. In terms of quality, the two teams are even on forwards. Brazil takes a big advantage in speed positions and defenders, but France wins where it matters: midfield (Diagram 5). For many years now, Brazil has been known to produce some of the best forwards and defenders, but apart from Ballon d'Or winner Kaka, who retired a couple of years ago, it is difficult to name an outstanding Brazilian midfielder in recent years. Diagram 5Les Bleus Vs. A Selecao The French squad also displays better results on the club synergy variable, the second-best reading for the teams in the knockout stage behind Spain. This is the result of improvements to its domestic league. Meanwhile, Brazil's top players continue to be dispersed across a number of different top clubs. Brazil will avenge its disastrous result from 2014 with a solid showing in Russia. Its pride will be reestablished and memories of the 7-1 drubbing softened. However, it will fail yet again against a European power. Spain Vs. Germany: The Battle Of Champions The other semi-final game will feature the 2010 champion (La Furia Roja) against the 2014 champion (Die Mannschaft). You have to go back all the way to the 1990 World Cup to see the two previous World Cup winners face each other in the semi-finals. Our model is indifferent to Germany's more recent success, giving Spain an overwhelming 85% probability of winning. In 2018, Spain has superior quality across the pitch (Diagram 6). Diagram 6 La Furia Roja Vs. Die Mannschaft In 2010, the year of Spain's last title, the team relied heavily on what made F.C. Barcelona unbeatable in Europe: heavy possession of the ball and gameplay built on crisp passes. This still holds true today, as tiki-taka has become part of Spain's footballing DNA. However, after underperforming in the 2014 World Cup and at the 2016 Euro, new coach Julen Lopetegui has slowly improved overall play. This has particularly involved raising the quality of the Spanish attack. As Spain discovered in the last two major competitions, it is not enough to have possession for 80% of the game if one cannot do anything with the ball. The Finals: Spain Vs. France BCA's Two-Step World Cup model predicts that, on July 15, 2018, the world will see Spain dispatch France in the finals of the world's sporting pilgrimage (Diagram 7). The two teams have the highest conditional probability of traveling down the grueling camino a la gloria eterna, from the group stages to the final (Chart 7). Intriguingly, of the teams knocked out in the quarter-finals, Portugal shows up with a slight conditional probability of winning the tournament. The England-Colombia matchup is essentially a coin toss. Diagram 7Road Map Of The World's Sporting Pilgrimage Chart 7Conditional Probability Of Entering Elysium Our model gives Spain a 59% probability of beating France (Chart 8), large enough to give us confidence, but not an overwhelming figure. It is undeniable that Spain has superior quality across the pitch (Diagram 8). Chart 8Final Matchup Probabilities Diagram 8La Furia Roja Vs. Les Bleus Moreover, Spain clearly excels over France in its defense - it conceded just three goals in its ten qualifier matches - and in its club synergy value, where it scores the highest mark. Five players sampled in our data play for F.C. Barcelona and five play for Real Madrid. What of the game for third place? Surprisingly, it may be the game of the tournament.20 Sure, third place means little. However, this game will mean a lot. If our model is correct, Brazil will face nemesis Germany in a revenge game. The young Brazilian team, completely revamped after the 7-1 Blitzkrieg on home soil, will be playing for the future, for revenge, and for national pride. Our model gives Brazil only a slight edge over Germany (Table 21). This is unsurprising, given that Brazil is, by far, superior in defense, forward, and speed positions, even though it is considerably inferior where it matters: the midfield (Diagram 9). Table 21Third Place Match Summary Results How does our forecast compare with current betting odds? Chart 9 shows that our model gives Spain an extraordinarily high probability of winning the tournament. Spain's 32% odds are double what the bookies' favorites, Brazil and Germany, command. Overall, we think that the betting market is underestimating the odds of a Mediterranean champion, while overstating the odds of both Germany and Brazil. Diagram 9A Selecao Vs. Die Mannschaft Chart 9We Do Not Condone Betting!##BR##(But If You Were Wondering...) II. Teams And Narratives To Watch In Russia: A Qualitative Assessment The World Cup is not just about winning. It is the most watched event - sporting or non-sporting - in the world because it weaves so many narratives and themes together. In this section, we explore some of these narratives and themes. Some are investment relevant, some are geopolitically relevant. Our qualitative assessments are ordered from the lowest-ranked teams to the highest-ranked. South Korea - A New Era Dawns, With The Same Football Results Coming from a country that has imprisoned every former president, including the one it just impeached, it should come as no surprise that South Korea's national team lacks stable leadership. The Koreans had a remarkable coach - the German sweeper Uli Stielike - who had dedicated three years to preparing them for this tournament. They fired him last year, however, due to a handful of losses. He has been replaced with a native son - a known advantage in World Cup football - Shin Tae-yong. Shin Tae-yong threw a special winter training camp for the team in order to establish himself as coach and get to know the squad given such a short time remaining before the tournament. Activities consisted of drinking tiger's blood and hiking up to the Unicorn Lair of Kiringul where, rumor has it, former North Korean dictator Kim Jong Il was conceived. South Korea made it to fourth place when it co-hosted the World Cup in 2002. While it had a lot to do with the Flying Dutchman coach, Guus Hiddink, he was not playing on the field. Rather, it was the insatiable Korean thirst to perform better than their co-host and erstwhile colonial overlord, Japan.21 In 2018, South Korea has a good defense: they seldom have let the ball in the net in recent tournaments. The only problem is that they do not know how to score goals. Young star player Son Heung-min, of Tottenham Hotspurs fame, has shown that he can score multiple goals late in the game - but only when playing against Uzbekistan. Of course, South Korea does have a tremendously compelling national story this year: the war with North Korea is finally over! President Moon Jae-in and his North Korean counterpart, dictator Kim Jong Un, held the third Inter-Korean summit on April 27, and declared for the first time since 1952 that they would stop trying to kill each other, at least formally. Peace is the word of the day. Can the optimism of an era free of war, with potential Korean reunification on the horizon, translate to a morale booster that helps Korean athletes? Probably not. Our model assigns Korea only a 3% chance of making it out of the group stage. The 2018 Winter Olympics were held in South Korea - they even entered the opening ceremony jointly with their communist neighbors. But while the South Koreans did well, winning several golden medals, they did not set a new record. Unlike the South Korean women's hockey team, the South Korean men's football team won't receive an infusion of North Koreans. Nor will they host a North Korean art troupe. Although we doubt that either would help their odds in Russia. Egypt, Morocco, Tunisia, And Saudi Arabia - Renaissance Or Regression? It is now eight long and turbulent years since the Arab Spring ignited with protests in Tunisia. Since then, stagnant regimes have been toppled, hundreds of thousands of people killed in civil wars, and an Islamic State has been created and crushed. Tunisia and Morocco stand out as clear outliers in terms of geopolitics. Since the 2014 parliamentary elections, Tunisia has begun paving the way towards becoming the first country in the Arab world with a functional democracy. Morocco is far from a democracy, but has remained stable over the past decade in stark contrast to the region. King Mohammed VI has enacted a number of reforms that have forestalled, for now, expression of grievances from the public. Saudi leadership has decided to follow the Moroccan example and preempt social unrest by enacting wide ranging reforms from the top. The 33-year old Crown Prince Mohammad Bin Salman has not only pledged to reform the economy, but also to fundamentally change conservative Saudi society. Women have been given the right to drive, the religious police has been regulated into irrelevance, and there is even talk of eliminating public gender segregation. BCA believes that these reforms are genuine and that they will be executed with vigor.22 Saudi leadership is reacting to the reality that the majority of the population is under 30 years of age (Chart 10). Mohammad Bin Salman therefore did not "come out of nowhere," as many commentators claim. He is the regime's response to the socio-economic realities of Saudi Arabia. Chart 10Saudis Are Catering To The Youth In contrast with these three countries, Egypt has stagnated in terms of governance and political reforms. In many ways, the country has regressed back to the military-backed rule that preceded the Arab Spring protests. However, President Abdel Fattah el-Sisi has launched and followed through on some private-sector and macro reforms which should result in marginal improvements for the economy.23 If football is a reflection of socio-economic conditions, then the participation of these four states at the World Cup is a harbinger of a brighter future. In Russia, Arab countries will have the highest number of representatives ever in the World Cup. Given that three of the four countries qualified in the highly competitive African continental qualification tournament, the result is impressive. Our quantitative model gives Morocco, Saudi Arabia, and Tunisia little chance to reach the knockout rounds. Egypt, on the other hand, is only eliminated because we have decided to give Russia a host premium. Since 1930, as a host, only South Africa failed to qualify for the round of 16 in 2010. Egypt has a real chance to be one of the dark horses in this year's tournament. It has the requisite footballing tradition to be able to withstand the pressure of a major tournament. Egypt has won seven African Cup of Nations, the most among its continental peers, and nearly half of its likely squad play in major international leagues. The most important among these is Mohamed Salah, "The Egyptian King," of Liverpool F.C. Salah may be the best forward in the world at the moment and will certainly wreak havoc in the group stage against Russia and Saudi Arabia. These four Arab states have already accomplished a lot by qualifying. But if one of them progresses to the knockout stage, it would be notable. BCA Emerging Markets Strategy research has shown that genuine structural reforms that improve the quality of governance are required for long-term productivity growth, and thus, asset performance.24 But a change in attitude comes first. Confidence is an important part of that change, and seeing Mo Salah and Egypt take on the world's best could light the spark for a Middle East Renaissance. Iceland - A Footballing Black Swan Iceland's football team success is a great reminder to investors that the probability of unexpected events is often greater than the consensus expectation. Iceland has a population of 330,000, just slightly larger than Swansea, Wales' second-largest city and home of "the Swans," the perennial bottom-feeders of the English Premiership League. If one eliminates Iceland's female population, the infirm, and the too old and too young, the pool of available humans for the country's football team is about 40,000. Iceland's qualification for the World Cup is extraordinary. Its success at the Euro 2016, where the country advanced to the last eight teams after defeating England, is nothing short of evidence of the divine. It may be the greatest upset in sports. Ever. Consider that in order to even participate at the Euro 2016, Iceland had to survive a qualification group that included the Netherlands (which failed to qualify!), Czech Republic, Turkey, Kazakhstan, and Latvia.25 Iceland finished second in the group and thus qualified directly for the event. Then, to get to the final eight at the Euro, Iceland had to survive a group made up of Hungary, Portugal, and Austria (where it finished ahead of the eventual Euro champion Portugal!).26 It defeated England in the round of 16 by a score of 2 to 1, only to finally be vanquished by the host nation France.27 What makes Iceland's success so astonishing is statistics, specifically the concept of conditional probability. Because qualification for a major event, such as the Euro, requires successive results, the conditional probability of Iceland getting through to the quarter-finals eventually is close to zero.28 Especially when one considers that there is no element of surprise for a country like Iceland once it shocks the world by qualifying. This is not college basketball, where the knockout nature of the March Madness tournament - and lack of scouting of small universities - creates the potential for upsets. This is international football, played by grown men getting paid millions of dollars to do their job (i.e., not lose to Iceland). What is the secret behind the success of Strákarnir okkar (Our Boys)? Several theories have been advanced: investment in coaching, heated football pitches, "Nordic mentality," "The Breath of Odin," etc.29 To us, all of this smells of the hindsight rationalization that Nassim Taleb identified as the hallmark of Black Swan events.30 Icelandic footballing success is a shock, a rift in the space-time continuum, and unlikely to be replicated. Our model gives Iceland little chance against Argentina, Croatia, and Nigeria. Of course, the whole point of a Black Swan event is that it is impossible to model. So, sit back, relax, and enjoy the Viking War chant!31 Japan - Bad Timing For New Leadership Japan is a nation in ascendancy, having emerged from its "Lost Decades" in recent years to reclaim a leading position among the nations of the world. What stirred the giant from its slumber? A global financial crisis, earthquakes, tsunamis, nuclear meltdowns, debt-deflation, and the revival of its ancient Chinese foe. A new era is dawning. That is fortunate, as it means that Japan's failures in the World Cup will fall under the final year of the outgoing emperor's reign. The problem, once again, is the lack of stable leadership. Just as factions in the ruling Liberal Democratic Party are busy trying to remove the hugely successful Prime Minister Shinzo Abe before a critical party leadership vote, so factions within the Japan Football Association have removed the head coach, Vahid Halilhodzic, just two months ahead of the tournament. Unlike his Korean counterpart, Akira Nishino has not held a winter camp to familiarize himself with the team. Instead, he will rely on the rock-solid supposition that the Japanese would think it ignoble to be led into battle by a foreigner. There is a basis for believing that teams do better under the leadership of a coach who is a national. But there is only so far that ethnic solidarity can go. After all, this is a team that wrestled Haiti and Mali to a draw. To throw its coach under the bus at the last minute is to ensure that Japan's squad remains, in the words of William Durant, "inclined to picturesque suicide". Croatia & Serbia - Strength In Numbers Extrapolating into the future from their vantage point in 1990, leaders of Yugoslav federal republics Slovenia and Croatia made an understandable forecast. First, communism was dead, and it was time for economic and political reforms. Second, "economies of scale" no longer mattered. The future was in global trade and open borders. Therefore, membership in a still-communist federal Yugoslavia dominated by semi-authoritarian Serbia was suboptimal. Looking back at the decision today, it is hard to argue with the results. Slovenia is in the Euro Area with a GDP per capita of $21,304 (2016), while Croatia is close behind, in the EU with a GDP per capita of $12,090 (2016). Meanwhile, their neighbor and aggressive "big brother" Serbia, which fought a war against both to keep them in Yugoslavia, is on the outside looking in. Its GDP per capita ($5,348 in 2016) has only recently recovered to the 1989 Yugoslav levels! There is no argument that Croatia and Slovenia made the right choice, at the time, by choosing secession. But the question is whether they would have still done it knowing everything we know about the present. For one thing, we cannot extrapolate globalization into the future. We are, in fact, at its apex.32 Strength in numbers will matter in a de-globalized, multipolar world of the twenty-first century. And while Croatia and Slovenia are part of a bigger whole - the EU - their size relative to the EU population and economy is laughable relative to their importance in Yugoslavia (Chart 11). Chart 11Small Parts Of A Bigger Whole And then, of course, there is sport. Both Serbia and Croatia are present in Russia. Our model ranks them 14th and 15th respectively. If we combine their rosters into 23 top-ranked players, their ranking jumps to eight, high enough that a semi-final berth becomes a possibility given good luck. This is not the best team that former Yugoslavia could have featured. In 1998, Croatia came third in France with an over-the-hill squad. But in 1994, Serbia and Montenegro (still called Yugoslavia at the time) were banned from competing while that same Croatia was unprepared to compete in the qualifications, which began just after the country's independence. That 1994 team would have featured in-prime Croatian stars33 - who came within two Lilian Thuram goals of the 1998 finals - and a team of Serbian,34 Montenegrin,35 and Macedonian36 players who took Red Star Belgrade to its Champions League title in 1991. Given the relatively weak competition in 1994 (Bulgaria and Sweden made the semi-finals), Yugoslavia could have been crowned World Cup champion in 1994. Of course, if pigs had wings, they would fly. On the other hand, the fact is that Slovenia and Croatia are among the two most Euroskeptic countries in the EU. Clearly, there is some buyers' remorse in both. Perhaps as part of a reformed, democratic, and federal Yugoslavia, both would have enjoyed greater clout in Brussels and thus, greater say in what kind of policies the EU imposed on them. And if that did not work, at least looking at the shiny 1994 FIFA World Cup Trophy would have helped put everything else into perspective. Russia - The World Cup As The Ultimate Lagging Indicator Expectations are low for the Russian national team at the 2018 World Cup. Despite its size, population, relative wealth, and strong tradition of government support for sports, Russia is just not a footballing nation. Its best result at the World Cup was way back in 1966 (fourth place), although it did win a European title in 1960. That was more than half a century ago! As a host, we would expect Russia to get a bump out of its woefully easy group. But from there, the focus will shift away from the Sbornaya to the organization of the tournament. This is the second time Russia is hosting a major international competition, following the 2014 Sochi Winter Olympics, and all eyes will be on the updated infrastructure and tight security measures. The World Cup presents a much greater challenge than the Sochi Olympics because there are 11 venues spread out over a geography the size of Western Europe. Three of the venues - Volgograd, Rostov-on-Don, and Sochi - are also close to the Caucasus region. Russian security forces essentially encircled Sochi in 2014, severely limiting movement into and out of the Black Sea resort town.37 This is unfeasible to do for the country's 11 largest cities. Chart 12Russia - The World Cup As##BR##The Ultimate Lagging Indicator There is an additional problem of limiting hooligan violence. We assume that Russian law enforcement will be much better at limiting the influx of troublemakers than its European neighbors. However, Russia has plenty of domestic hooligans to create violence, which was on clear display at the Euro 2016 clash between English and Russian "fans." All of this brings up the question: Why is Russia organizing the World Cup in the first place? Especially given its semi-pariah status following a spate of controversial geopolitical events thus far in 2018? The answer is that major sporting events are the ultimate lagging indicator of a country's economy, its geopolitics, and market performance. The Sochi selection was made in 2007, amidst an epic commodity bull market and at the peak of the "BRICS are taking over" narrative. The World Cup selection was made in 2010, well before Crimea was annexed, Ukraine was destabilized, and Russia decided to play "peacemaker" in Syria (Chart 12). Just like Brazil in 2014, the World Cup therefore arrives to Russia at an odd time, with the market, economy, and the country's relations with the West hitting rock bottom. Of course, the 2014 World Cup was the bottom for Brazil, with a spate of seismic economic and political changes following. Brazilian equities are up nearly 50% from the closing ceremony of the World Cup to today. Can the same happen in Russia? Mexico - Will Anything Change? Mexican voters will go to the polls on July 1 as its national football team competes in Russia. The election has been dubbed the "biggest election in Mexican history" by the country's National Electoral Institute. We agree. For the first time in Mexico's history, a left-wing, anti-establishment movement has a chance to win (Chart 13). Chart 13The "Biggest Election In Mexican History"? Will Mexico's football team also break with tradition? For the past six World Cups, Mexico has emerged from the group stage only to be promptly eliminated in the knockout stage every time. In 2018, El Tri again have a great chance to emerge from the group. While Germany is a formidable foe, the other European nation in the group, Sweden, barely qualified and is playing without its only top-class footballer, Zlatan Ibrahimovic. As such, Mexico should accomplish the usual feat relatively easily. The question is whether it can do more. Unfortunately, second-place finish in its group will pair it, baring a major surprise, with Brazil. This is where its road will most certainly end. Will politics mirror football? Is Andres Manuel Lopez Obrador (AMLO) just more of the same? We are hopeful that Mexico's left-wing answer to President Donald Trump could actually make some changes for the better. First, the Mexican Congress is unlikely to give AMLO free rein in governing the country. Second, AMLO has moderated his campaign, purposefully hiring centrists and technocrats to signal moderation. Third, AMLO has the charisma and the gravitas to sit down with President Trump and negotiate face-to-face, unlike his predecessor, Enrique Peña Nieto, who never quite got used to Trump's rhetoric and aggressive style. In the end, if AMLO fails to make serious inroads with economic reforms, ongoing drug-related crime, and negotiations with President Trump, at least Mexicans will have one major success to take from 2018. Mexico qualified for the 2018 World Cup, whereas its eternal football rival the U.S. did not! Belgium & Switzerland - The Upside Of Immigration Let us state the facts. Without footballers of immigrant descent, we would not be writing an analysis on the Belgian and Swiss football teams today. Of the 23 call-ups for the international friendlies ahead of the World Cup, 13 were of immigrant background for the Swiss team and 12 for the Belgian. This includes the majority of the stars of both teams, such as Granit Xhaka (Albanian) and Ricardo Rodriguez (Spanish-Chilean) for Switzerland, and Vincent Kompany (Congolese-Belgian), Marouane Fellaini (Moroccan), and Romelu Lukaku (Congolese) for Belgium. In large part, this is a tired topic already well covered by the Belgian and Swiss media. It also draws on the main narrative following the 1998 French World Cup win. Half of the legendary Les Bleus team was made up of players with immigrant backgrounds, including essentially all of its stars. Of the 14 goals scored by the French team, nine came from its multicultural stars Zinedine Zidane, Youri Djorkaeff, Thierry Henry, Lilian Thuram, and David Trezeguet. What separates the 1998 French team from the Belgian and Swiss teams today, however, is the geopolitical context. The 2015 migration crisis in Europe - when the influx of illegal immigrants reached a peak of 220,000 per month in October of that year - still dominates politics on the continent (Chart 14). Although the crisis is now definitively over - the monthly influx figure is below 3,000 - both anti-establishment and establishment parties have swung hard against immigration. Chart 14What Migration Crisis? But wait, are these players really immigrants to begin with? The answer is an obvious and definitive no. Only six Swiss and none of the Belgian players were born outside the country. Even the 1998 French team had only two players who were actually born outside of France (and an additional two were from overseas French territories). This reflects one of the greatest misunderstandings in the political narratives regarding European immigration. There essentially is no longer any non-European immigration to Europe. At least not via legal channels. Players who are visible minorities on these teams are almost universally the children and grandchildren of the post-World War II immigrants from the former colonies (and three of the six foreign-born Swiss players are not visible minorities at all, given that they are from the former Yugoslavia). Several European countries undoubtedly have a problem integrating non-European immigrants. Our colleague Peter Berezin has made a connection between a generous social welfare state and lack of successful integration, with Belgium and Sweden as prominent examples of the connection.38 However, few commentators understand that there is no ongoing large-scale influx of immigrants to Europe, aside from occasional migration crises prompted by wars. The migrations of the 1950s and 1960s, from former colonies, was followed in the 1960s and 1970s by "guest worker" migration. Since then, Europe has progressively tightened its migration policies to all but internal migration from EU member states. What does this mean for Europe's national football teams? Nothing. Football is a sport played by the middle and lower classes almost universally.39 The high proportion of visible minorities on Europe's football rosters is simply a reflection of the fact that many immigrants and their kids grow up in precisely such communities. Eventually, they become French, Belgian, or Swiss. After all, nobody would today doubt Michel Platini's Frenchness. But he is as much a product of immigration as Zinedine Zidane or Zlatan Ibrahimovic. England - A Football Brexit It is well known that the English national team is cursed in World Cup football. If not for the home team advantage, they never would have won their single trophy in 1966. Many commentators in England lament the strength of the English Premier League. Best teams are collections of superstars from other countries who receive the highest wages in football, yet naturally return every four years to play for their mother country. In other words, the English fight wars with mercenaries hired from abroad whose loyalties lie elsewhere. Imagine that! The English national team has chronically underperformed despite having one of the most expensive, and productive, leagues in the world. Like many other U.K. industries, English football leverages the high productivity of extraordinary immigrants. Indeed, the percentage of foreign players in the English Premier League is the highest in the world (Chart 15).40 Chart 15The Most Globalized Labor Market? The immigrant free-for-all, however, almost ended. Just as concerns over immigration prompted Brexit, concerns that English football was being dominated by the non-English almost led to changes to the domestic league Homegrown Player Rule. A proposal to put more stringent rules on foreign players in order to fix the talent pipeline in English football almost took hold under the previous FA chairman, Greg Dyke. Fortunately, the worst effects of the Brexit vote have already passed: The team suffered their worst upset in 66 years at the hands of Iceland in the 2016 Euro Cup, just days after the referendum was held. Remember, there are fewer people in Iceland than there are Sikhs in the United Kingdom. This devastating upset was the first material instance of "Bremorse" in the wake of the referendum. The team is unlikely to suffer such an upset again. Instead, it will perform just like it did in 1066: defeating the likes of the King of Norway and Iceland only to be invaded and enslaved by the King of France. But unfortunately for England, it does not get a rematch with Iceland in group play. Rather, its group of four affords an occasion for an even greater humiliation, perhaps even the crowning humiliation of the entire Brexit saga: a loss to Brussels. Such a loss is likely if only because the signal lesson of British history teaches that Europeans can never successfully invade the island unless it happens to be suffering from deep internal divisions. And with Jeremy Corbyn and a second Brexit referendum on the horizon, such "intestine" divisions are utterly assured. Given all of the above, does it seem likely that this is the year in which England will break a half-century-long curse? No. But our model gives the young team great odds of reaching the knockout stage. Argentina - Promise Vs. Performance Every investor knows the old adage that at the turn of the twentieth century, Argentina was "one of the ten richest countries in the world," ahead of European heavyweights France, Germany, and Italy. And that in 1950, Argentina's GDP-per-capita was six times greater than the likes of South Korea. After half a century of Argentine economic mismanagement and unfulfilled promises, South Korea has today doubled Argentina's national wealth (Chart 16). Chart 16Argentina - Promise Vs. Performance At probably the height of the country's economic and political mismanagement - during the nearly decade-long military rule between 1976 and 1983 - Argentina's football team won two World Cups: 1978 and 1986. Since then, however, it has left behind a legacy of unfulfilled promises and lost opportunities that largely mirror its last thirty years of economic performance. Since the 1990 World Cup in Italy, Argentina has entered every World Cup as one of the favorites. Its fans consider the team one of the world's five greatest footballing nations. To its credit, it has made the finals twice, in 1990 and 2014, and the quarter-finals three other times in that span. But its status as a perennial superpower is under threat. We would argue that Argentina is, in fact, the greatest underperformer in the last three decades of international football. Despite being the fourth highest-ranked team over the past two decades - according to Elo scores, which we interpret to represent the consensus view - the country has failed to lift the World Cup. The greatest threat to Argentina's legacy as a footballing superpower is that, in another twenty years of unfulfilled promises, fans will look back at its past success the way we think of Uruguay's two World Cups, won in 1930 and 1950. One could argue, for example, that its two titles are a product of a single, golden generation, not of eternal status as a superpower. Since Diego Maradona stopped playing, Argentina has only advanced past the quarterfinals once. Coming into the 2018 World Cup, the expectations are once again high. Argentina still has one of the world's top-two players in Lionel Messi. And Elo again ranks it as the fourth-best team, with expectations of at least a semi-final appearance. Our model is more skeptical, dropping it to a sixth-best ranking and predicting yet another quarter-final exit. If Argentina wastes its "Messi generation," the chances that its unfulfilled football promises mirror that of its economy will grow. Portugal - The "Great Man" Theory The "Great Man Theory" posits that a large portion of human history can be explained by the impact of "Great Men" (and women) on global events. Popular from the Greek and Roman classics through the nineteenth century, the theory has been contradicted by post-modern philosophers, sociologists, and historians. BCA's Geopolitical Strategy largely rejects the idea in its methodology. We focus on the constraints to "Great People," rather than their preferences. After all, preferences are optional and subject to constraints, whereas constraints are neither optional nor subject to preferences. Enter Cristiano Ronaldo dos Santos Aveiro. Perhaps the greatest footballer ever, Ronaldo is indeed a Great Man. He has willed his teams (whether at club level or internationally) to extraordinary feats. Even in injury, his defeat on the battlefield inspired his teammates to play incredible football in the Euro 2016 final against a superior France. Ronaldo lacks mortal constraints on the football pitch. He has scored 308 goals in 289 appearances for Real Madrid (since 2009), an incredible per-game ratio (Chart 17). In the 2017-2018 season, at the age of 33, he has scored 41 goals in 38 appearances, maintaining a per-goal average of his entire career at the Spanish club. His game appearance total this season is lower than at any other time since 2009-2010, suggesting that he will come into the World Cup fresh and rested. Chart 17The Ronaldo Effect Football, however, is a game played on a wide pitch by 11 players. How can one man, even one as great as Ronaldo, make a difference? Gravity. Ronaldo is such a dangerous player with the ball that defensive midfielders and defenders have to constantly gravitate towards his presence on the pitch. This opens up lanes for his Portuguese teammates to counterattack. And Ronaldo's teammates are not as bad as the consensus thinks. Our quantitative model ranks Portugal seventh, just a few points below Brazil. They are also fortunate to be in a relatively easy group. While the matchup with Spain will be a tense affair, Portugal will easily dispatch Iran and Morocco. And if they come second in the group, they will face the easiest group winner of the next round, Group A's Uruguay. While we do not subscribe to the Great Man theory as a methodology, we respect it. Over the long term, and over a great number of iterations, focusing on great individuals is folly. But when faced with specific events, there is definitely value in thinking about how extraordinary human effort matters. Portugal's path to the quarter finals is easy. Once there, games will become tighter, the pressure unbearable, the stage infinitely larger. It is in these moments that legends play like Great Men, while lesser footballers play like mere mortals. Brazil - Redemption? As host of the last World Cup in 2014, Brazil didn't so much hit rock bottom as descend to the ninth circle of hell. Its merciless 7-1 obliteration by Germany in the semifinals - the sharpest rebuke to home-team self-confidence in history - would have scarred the national psyche even if the country had not proceeded into the worst economic recession in a hundred years. At that point the country's politics became a Shakespearean tragedy in which a handful of corrupt aristocrats slaughtered each other on stage, for all the world to see. Chart 18No Neymar Home "Brazil is back," say the fans - just as the bulls in the financial markets. The argument goes like this: The crisis has proven that the country has resilient institutions. Inflation (read: hubris) has collapsed to the lowest point in years, while the country has tightened its belt (both financially and figuratively), in a long-overdue act of self-discipline. The only difference is that the football comeback, unlike the economic one, is built on firm foundations. The Brazilian squad has made an impeccable run since Dunga stepped down as coach in 2016. It was the first team to qualify for this year's World Cup, going undefeated for quite a stretch of time. Today, the team is coached by Tite, a former player. The least we can say is that he reinvented the way Brazil plays. Brazil's lamentable disappointment under coach Luiz Felipe Scolari ultimately comes down to the need for a new football identity and a new crop of players. That's what they have today. They are a very good team, and freshly humbled. We expect them to go far, though not all the way. It is not certain that all of Brazil's demons have been purged. By way of illustration, the leading presidential candidate is an advocate of military dictatorship, with no pro-market candidate in sight (Chart 18). France - Marchons, Marchons! In 2016, France lost to Portugal in the final of the Euro Cup on home court. This was not the first gut-wrenching loss in the final of a major tournament. In 2006, its aging 1998 generation lost a bitterly fought final against Italy. In Russia, France comes as one of the favorites. Its team is celebrating the 20-year anniversary of its epic 1998 run. Meanwhile, at home, France is marching in lockstep to President Emmanuel Macron's reforms. The economy is booming, cantankerous unions are either defeated or capitulating, and even equity market performance is leaving traditional competitors in the dust (Chart 19). Chart 19France - Marchons, Marchons! Our model sees France reaching the final of the World Cup. However, it also predicts that France will lose this final, a repeat of the 2016 and 2006 performances. Les Bleus will come close, but not light the victory cigar. Football in France is embedded in the cultural fabric; its team is portrayed by the media as a reflection of the contemporary socio-economic narratives. When France won in 1998, for example, the team's multicultural heritage was touted as a model of integration and as a symbol of open-minded, post-colonial France. The 2010 debacle in South Africa, on the other hand, where French players literally went on strike following an altercation with their coach, was presented as a symbol of declining French relevance and growing national impotence. Since 2014, however, positive momentum has been building both on the football pitch and in real life. A young team managed to qualify for the 2014 World Cup, losing to the eventual champion Germany in a closely contested quarter-final game. Two years later, that same team defeated World Champion Germany in a semi-final of the Euro tournament. On the political front, 39-year old Emmanuel Macron swept aside populists and left-wing firebrands, stunning the world by campaigning from a staunchly centrist perspective. The question for France in 2018 is whether it will fall short of victory, yet again. We ask the question both in terms of the performance on the football pitch and in politics. The youthful, energetic, president mirrors the youthful, energetic, squad. But a lot is at stake and second place may not be good enough for either. We remain optimistic that genuine reforms are coming to France.41 And despite our model's preference for Spain in the finals of the World Cup, France has a great chance of repeating the glory from twenty years ago. Spain - Divided It Stands, And Wins On October 1, 2017, 92% of Catalan voters decided that Catalonia should secede from Spain in a referendum deemed illegal by Madrid.42 Soon after, images of old ladies and students being beat up by riot police flooded the internet, as Mariano Rajoy's government signaled its refusal to compromise on the question of independence. Chart 20Catalan Separatism: Overstated Risk Intriguingly, Spain's most independent-minded regions have often produced some of its best footballers. Cesc Fàbregas, Gerard Piqué, Carles Puyol, Xavi Hernandez, and Sergio Busquets are all of Catalan descent and have dutifully defended the Spanish colors in several successful Euro and World Cup campaigns. Meanwhile, the bitter Barcelona-Real Madrid rivalry has been at the heart of Spanish football and deeply embedded in the country's political and socioeconomic history for generations. In other countries, such divisions and rivalries would let politics get in the way. In Yugoslavia, the civil war is famously said to have begun with fan rioting at a 1991 Dinamo Zagreb vs. Crvena Zvezda match. But Spain, somehow, channels the chaos into beautiful, and effective, football. Pro-independence sentiment has begun to waver in Catalonia, as BCA's Geopolitical Strategy predicted (Chart 20).43 This is bullish for the Spanish economy and assets, but also for its football team. Spain conquered the world of football at the height of its economic crisis, in 2010. Will it do so again, on the heels of its political crisis? Our model predicts yes. And we agree. It would seem that Spain's greatest attribute as a nation is to produce diamonds under pressure. III. Investment Implications: Does The World Cup Matter? The FIFA World Cup is the most widely watched sporting event in the world by far (Chart 21). Does all that passion and emotion translate into any economic or market implications? Chart 21World Cup Is The Most Important Unimportant Event In The World Academic literature on the topic is, at best, inconclusive. International football, as an industry, is largely inconsequential. For instance, the wealthiest football club in the world - Manchester United - generates less than 8% of the revenue of the 500th company in the Fortune 500 index. To assess whether the World Cup matters, we take two approaches. First, we explore the implications of hosting the event on the economy of the host nation. Second, we examine the implications of the World Cup on equity markets. Hosting The World Cup Despite the size and the reach of the World Cup, we find little evidence that hosting the event is beneficial to the host's economy. There appears to be little to no effect on either a short or long-term horizon. We found no "World Cup effect" on any of the economic variables one would assume may be impacted. Nonetheless, some patterns are worth highlighting (Table 22). Table 22Is It Worth It? We chose to look at the behavior of each economic variable one year before and after the World Cup, with the exception of core consumer price inflation and the unemployment rate, which we observed on a three-year horizon around the event. The rationale behind this is that the labor market tends to react slowly to changes in underlying economic activity, due to factors such as the rigidity of employment contracts and stickiness of wages. The study starts with the 1990 World Cup hosted by Italy. Interestingly, some patterns are apparent. Out of all the variables we analyzed, inward FDI tends to consistently increase in the host country following the World Cup. Our suspicion is that hosting the event demonstrates, on the margin, that the country meets suitable conditions (governance, return on capital, etc.) to attract foreign capital. On the other hand, we may be capturing a trend already underway, which itself was revealed by the fact that the country was selected to host the World Cup in the first place. As Table 23 illustrates, the selection process lags the actual event, as with any major sporting event (such as the Olympics). Intriguingly, the time-lag is, on average, six years, almost precisely the length of an economic cycle. As such, the decision to award the hosting of the event to a particular country may already take into account the bullish macroeconomic cycle. By the time the event is hosted, the cycle is in full swing and may actually be peaking. Table 23Soccernomic Cycles? Consistently, capital expenditures also tend to increase following the Word Cup. As more foreign capital flows in the country, more investment takes place. Currencies, however, on a one-year horizon, do not react to these FDI flows since the short-term gyrations are dictated by more volatile short-term portfolio flows. As FDI flows in, the unemployment rate falls and tends to continue falling after a World Cup. However, some caution is advised in interpreting these conclusions. Macroeconomic business cycles, commodity prices, and other structural factors are still dominating factors in determining the trend and health of economies. As we indicated in our analysis of Russia in the qualitative section above, the World Cup itself can be considered a lagging indicator of the economy. This certainly appears to be the case with the three members of the BRICS who have hosted the event: South Africa, Brazil, and Russia. As such, hosting the World Cup may signal the top of the bullish cycle, as it certainly did for Brazil. This warrants a few key observations: Brazil did not see a rise in capital expenditure after hosting the World Cup in 2014. In fact, it suffered one of the deepest recessions in the last century. Consistently, unemployment did not fall as it did for other countries as commodities entered a bear market in 2015 and weighed heavily on the economy. Although South Africa did see its unemployment rate fall before the World Cup, the trend did not continue following the event. In the few years leading up to the global financial crisis, the world economy was in the midst of a period of stellar growth. Therefore, it makes sense that Germany was experiencing rising FDI, capital expenditure, and inflationary pressures prior to hosting the event in 2006. In addition, the implementation of the Hartz IV reforms in the early part of the decade put an end to the long-term structural uptrend in the unemployment rate, which has since declined by more than 7%. The year 2002 was an important one for Japan, as the turn of the millennium marked the end of the "lost decade". Subsequently, Japan experienced a falling unemployment rate and rising GDP growth. However, the end of the "lost decade" also saw the beginning of the self-feeding deflationary expectations which have anchored expected inflation levels near zero. As such, consumer prices fell before and after the event. France went through a period of intense reforms in the late 1990s. Prior and subsequent to 1998, growth rates were falling simply because 1998 recorded a higher growth rate of 3.3%, compared to 1.6% in 1997 and 3.1% in 1999. The lowering of the VAT rate, income, and corporate tax by the Chirac government led to real capital expenditure and real household consumption increases during the late 1990s. Furthermore, substantial labor market reforms helped to decrease the unemployment rate. The 1990s were the longest period of economic expansion in U.S. history. In fact, 1994 was the year where the number of jobs created was one of the largest on record, at 3.85 million. As such, real GDP growth and business investment expenditures were high, and the unemployment rate was falling. The prospects of high growth also brought in increased foreign direct investment in the same period in which the World Cup coincided. However, emerging markets were seeing an equally stellar period of growth, which was supported by high capital inflows and appreciating EM currencies, thus leading to a lower U.S. dollar. Although the Federal Reserve began hiking interest rates in 1994, the effect on the dollar was not registered until the beginning of 1995, which also marked the beginning of a dollar bull market. In the 1990s, the Italian government was heavily liberalizing the economy, leading to its eventual inclusion in the Euro Area at the end of the decade. Italian growth during the late '80s and early '90s was high enough for it to surpass Britain and France in terms of GDP growth, consumption and investment growth in 1990 were high. However, the 1990s oil price shock depressed consumer and business activity. Bottom Line: In line with the academic literature, the results of our assessment are at best inconclusive. However, the data does suggest that hosting of the World Cup coincides with some positive macroeconomic developments, particularly in terms of FDI inflows. This, however, may reflect the fact that the decision to host the event is usually made at the beginning of a bullish upcycle, which means that the actual event marks the peak, or top, of the cycle. Equity Market Implications The World Cup may be most relevant, not as a catalyst for market action, but rather as a distraction. The European Central Bank, for example, concluded in a 2012 study that trading volumes dropped considerably during match times.44 The World Cup in Russia could be particularly distracting. Because of Russia's time zone, most matches will be played between 8:00 am and 2:00 pm in New York, and 1:00 pm and 7:00 pm in London. As opposed to the economic impact, there seems to be a general consensus in the academic literature regarding the equity market implications of World Cup matches. For instance, a cross-sectional analysis of 39 countries looking at World Cup and relevant qualifying matches between 1973 and 2004 (that represents 1,162 matches!) concludes that losses have an economically and statistically significant negative effect on the losing country's stock market.45 Furthermore, the study finds that this effect is stronger in small caps, which are disproportionately held by local investors and therefore are more strongly affected by sentiment following a football match. Another study found that being short NYSE and long Treasuries during World Cups would produce a higher return from 1950 to 2007.46 This is due to the decisions taken by foreign investors in U.S. bond and equity markets. During the World Cup, these investors would withdraw capital to invest in their local markets, which would put downward pressures on NYSE prices. Table 24 presents the daily stock return deviation from what one would expect outside of World Cups for all the matches in the past seven World Cups, starting with the quarter-finals (for the period from 1990 to 2014). We report the local equity market movement of the teams playing up to three days after a match took place. Table 24Do World Cup Matches Impact The Local Stock Returns Of The Teams Playing? Several observations can be made: We observe positive "abnormal" returns, i.e., deviations from trend, in the day following the matches, for any of the teams playing, and larger "abnormal" returns when it comes to the winning team. While we observe positive abnormal returns for the winners at t+1, it appears that more than 70% of the daily returns used to compute this average turn out to be negative - implying that, on average, the winners' local equity markets experience small negative returns and, in a few instances, large abnormal positive returns. In line with the literature, we observe negative abnormal returns on t+1 for the losers, regardless of how we define our benchmark. A high number of abnormal daily returns, up to three days after the match, appear to be negative (this is also the case regardless of whether we compare it to the past average, median, or non-World Cup daily returns). Bottom Line: We do not recommend that investors "play" the World Cup. Generally speaking, equity markets react, on average, positively to wins and negatively to losses. However, this is not always the case. 1 There are, however, some less obvious investment implications ... but we do not encourage gambling on sporting events! 2 C. Cotta, A. M. Mora, and J. J. Merelo, "A Network Analysis of the 2010 FIFA World Cup Champion Team Play," Journal of Systems Science 26:1 (2013), 21-42. 3 For instance, please see Gregory T. Papanikos, "Economic, population and political determinants of the 2014 World Cup match results," Soccer & Society 18:4 (2017), 516-532; and Fabian Wunderlich and Daniel Memmert, "Analysis of the predictive qualities of betting odds and FIFA World Ranking: evidence from the 2006, 2010 and 2014 Football World Cups," Journal of Sports Sciences 34:24 (2016), 20176-20184. 4 Please see O'Donoghue et al, "An evaluation of quantitative and qualitative methods of predicting the 2002 FIFA World Cup. Part II: Game activity and analysis," Journal of Sports Sciences 22:6 (2004), 513-514 5 Please see, The Economist, "Why video games are so expensive to develop," September 25, 2014, available at www.economist.com. 6 The OP model makes two critical assumptions: (1) it assumes that the (χ'ι, β, γ) function has the form of a continuous probability distribution function which is the standard normal distribution function of a linear combination of our explanatory variables; (2) it assumes proportional odds between each category in the dependent variable. Assumption (2) was confirmed using the proportional odds Brant test, confirming that the ordered probit is the best suited model for our purpose. Please see Alexander, Carol, Market Risk Analysis: Practical Financial Econometrics (John Wiley & Sons) 2008, 426 pages. 7 Please see EViews 8.1 User's Guide II, pp.259-284. 8 Please see J. Bloomfield, R. Polman, and P. O'Donoghue, "Physical Demands of Different Positions in FA Premier League Soccer," Journal of Sports Science & Medicine 6:1 (2007), 63-70. 9 Please see Seife Dendir, "When do soccer players peak? A note," Journal of Sports Analytics 2 (2016), 89-105. 10 Think 22-year old Mario Götze's stunning game-winner four years ago. 11 Please see H. Sarmento et al, "Match analysis in football: a systematic review," Journal of Sports Sciences 32:20 (2014), 1831-1843. 12 In order to favor our core model (M1), which uses the team average player rating variable, we assigned a weight α = 0.66. 13 We are sure that this is a pure coincidence. Being paired with the host Russia had nothing to do with this. Nothing. 14 A "Novichok bonus" perhaps? 15 Remember that the marginal effect represents the impact of a 1-unit change in the rating variable on the probability of winning. This is why we need to be careful when comparing both stages' marginal effect. A 1-unit change in the rating variable is less frequent, but extremely important in the knockout stage, hence the higher coefficient. 16 In 2006, Italy had 10 players from either Juventus or AC Milan; in 2010, Spain had 12 players from either F.C. Barcelona or Real Madrid; and in 2014, Germany had 11 players from either Bayern Munich or Borussia Dortmund. 17 Please see M. Clemente et al, "Midfielder as the prominent participant in the building attack: A network analysis of national teams in FIFA World Cup 2014," International Journal of Performance Analysis in Sport 15:2 (2015), 704-722. 18 For instance, please see F. M. Clemente et al, "Using Network Metrics in Soccer: A Macro-Analysis," Journal of Human Kinetics 45 (2015), 123-134. 19 In order to favor our core model (M1), which uses the team average player rating variable, we assigned a weight α = 0.66. 20 This would not be the first time that the third place game steals the spotlight. Going back to 1978, the third place game has outscored the final by a considerable margin. Teams usually enter the game with no pressure and therefore commit to a flowing, attacking style of play. Some of the most exciting games in World Cup history were played for third place: think Germany's 3-2 win over Uruguay in 2010, or Turkey's thrilling 3-2 win against South Korea. 21 As well as ... yes, atrocious refereeing. 22 Please see BCA Geopolitical Strategy Special Report, "The Middle East: Separating The Signal From The Noise," dated November 15, 2017, available from gps.bcaresearch.com. 23 Please see BCA Frontier Market Strategy Special Report, "Egypt: Giving Benefit Of Doubt," dated February 6, 2018, available at fms.bcaresearch.com. 24 Please see BCA Emerging Markets Strategy and Geopolitical Strategy Special Report, "Ranking EM Countries Based On Structural Variables," dated August 2, 2017, available at ems.bcaresearch.com. 25 The Netherlands has a population 50 times greater than Iceland and a GDP 35 times greater. 26 Portugal has a population 30 times greater than Iceland and a GDP 10 times greater. 27 England has a population 200 times greater than Iceland and a GDP 110 times greater. 28 Just in case any of our Icelandic clients take offense to our premise, we want to point out that the country has competed in a number of Games of the Small States of Europe. Participants at this biannual event includes such sporting powerhouses as Andorra, Cyprus, Liechtenstein, Luxembourg, Malta, Monaco, Montenegro, and San Marino. Iceland actually trails Cyprus in the total tally of medals collected since 1985. 29 Please see Frode Telseth and Vidar Halldorsson, "The success of Nordic football: the cases of the national men's teams of Norway in the 1990s and Iceland in the 2010s," Sport In Society, November 1, 2017. 30 Please see Taleb, Nassim, Fooled By Randomness (New York: Random House), 316 pages. 31 Unless you are a U.S. soccer fan. The U.S. has a population 1,000 times greater than Iceland and a GDP 800 times greater. Please light yourself on fire responsibly. 32 Please see BCA Geopolitical Strategy Special Report, "The Apex Of Globalization - All Downhill From Here," dated November 14, 2014, available at gps.bcaresearch.com. 33 Robert Prosinecki would have been 25-years old, and at the time one of the highest paid players in the world. In 1998, due to a number of injuries, he was playing for Croatia and largely an afterthought in the team. Davor Šuker and Zvonimir Boban would have also been 25 in 1994 and both at the peak of their careers. 34 Vladimir Jugovic and Sinisa Mihajlovic, were both in their prime in 1994, and both were stars in Italy. 35 The eventual Real Madrid legend Predrag Mijatovic would have been in-prime 25 years of age in 1994 (and unlikely to make the starting 11 of the combined Yugoslav team!), while, by then, the three-time Champions League winner Dejan Savicevic would have already been A.C. Milan's best player. 36 Shout out to Darko Pancev, the 1991 European Golden Boot award winner! 37 For example, for the duration of the 2014 Winter Olympics, only vehicles registered in Sochi, or those with special permission, were allowed through security checkpoints set a 100 kilometers outside the city. 38 Please see BCA Global Investment Strategy Weekly Report, "The End Of Europe's Welfare State," dated June 26, 2015, available at gis.bcaresearch.com. 39 Except in the U.S., which is maybe why it continues to underperform in global competitions. 40 Contrary to most leagues, the Premier League does not have a cap on the number of foreign players a team can have. It just requires eight players to be home grown (three years with an English team before the age of 21). 41 Please see BCA Geopolitical Strategy Special Report, "The French Revolution," dated February 3, 2017, available at gps.bcaresearch.com. 42 The turnout of the referendum was a woeful 43%, however. 43 Please see BCA Geopolitical Strategy Weekly Report, "Why So Serious?" dated October 11, 2017, available at gps.bcaresearch.com. 44 Ehrmann, M. and Jansen D., "The Pitch Rather Than The Pit: Investor Inattention During FIFA World Cup Matches," European Central Bank, Working Paper 1424, February 2012. 45 Please see Edmans, A., Garcia, D., & Norli, O. "Sports Sentiment and Stock Returns," The Journal of Finance 62:4 (2007), 1967-1998. 46 G. Kaplanski, and H. Levy, "Exploitable Predictable Irrationality: the FIFA World Cup effect on the U.S. Stock Market," Journal of Financial and Quantitative Analysis 45:2 (2010), 535-553. References Alexander, Carol, Market Risk Analysis: Practical Financial Econometrics (John Wiley & Sons) 2008, 426 pages. J. K. Ashton, B. Gerrard, and R. Hudson (2011), "Do national soccer results really impact on the stock market?" Applied Economics, 43:26, 3709-3717. M. H. Berument, and N. B. Ceylan (2012), "Effects of soccer on stock markets: the return-volatility relationship," The Social Science Journal 49, 368-374. J. Bloomfield, R. Polman, and P. O'Donoghue (2007), "Physical Demands of Different Positions in FA Premier League Soccer," Journal of Sports Science & Medicine, 6(1): 63-70. F. M. Clemente, M. S. Couceiro, F. Martins, and R. Mendes (2015), "Using Network Metrics in Soccer: A Macro-Analysis," Journal of Human Kinetics 45/2015, 123-134. M. Clemente et al. (2015), "Midfielder as the prominent participant in the building attack: A network analysis of national teams in FIFA World Cup 2014," International Journal of Performance Analysis in Sport 15(2): 704-722. C. Cotta, A. M. Mora, and J. J. Merelo (2013) "A Network Analysis of the 2010 FIFA World Cup Champion Team Play," Journal of Systems Science 26(1): 21-42. Seife Dendir (2016), "When do soccer players peak? A note," Journal of Sports Analytics, 2: 89-105. A. Edmans, D. Garcia, and O. Norli (2007), "Sports Sentiment and Stock Returns," The Journal of Finance, 62 (4), 1967-1998. J. Geyer-Klingeberg, M. Hang, M. Walter, and A. Rathgeber (2018), "Do stock markets react to soccer games? A meta-regression analysis," Applied Economics, 50:19, 2171-2189. G. Kaplanski, and H. Levy (2010), "Exploitable Predictable Irrationality: the FIFA World Cup effect on the U.S. Stock Market," Journal of Financial and Quantitative Analysis, 45(2): 535-553. Jorge Knijnik (2014), "Playing for freedom: Sócrates, futebol-arte and democratic struggle in Brazil," Soccer & Society, 15:5, 635-654. Christian Koller (2017), "Le football suisse: Des pionniers aux professionnels, " Soccer & Society, 18:4, 597-598 Lindsay Sarah Krasnoff (2017), "Devolution of Les Bleus as a symbol of a multicultural French future," Soccer & Society, 18:2-3, 311-319 S. Kuper, and S. Szymanski (2009), "Soccernomics: Why England Loses, Why Germany and Brazil Win, and Why the U.S., Japan, Australia, Turkey - and Even Iraq - Are Destined To Become the Kings of the World's Most Popular Sport". H. Liu, M.A. Gomez, C. Lago-Peñas, and J. Sampaio (2015), "Match statistics related to winning in the group stage of 2014 Brazil FIFA World Cup," Journal of Sports Sciences, 33:12, 1205-1213. R. Mackenzie, and C. Cushion (2013), "Performance Analysis in Football: A Critical Review and Implications for Future Research", Journal of Sports Sciences 31(6). J.A. Mangan, Hyun-Duck Kim, A. Cruz, and Gi-Heun Kang (2013), "Rivalries: China, Japan and South Korea - Memory, Modernity, Politics, Geopolitics - and Sport," The International Journal of the History of Sport, 30:10, 1130-1152. M. Martiniello, and G. W Boucher (2017), "The colours of Belgium: red devils and the representation of diversity," Visual Studies, 32:3, 224-235. Brank Milanovic (2010), "The World at Play: Soccer Takes on Globalization," YaleGlobal Online. Richard Mills (2009), "'It All Ended in an Unsporting Way': Serbian Football and the Disintegration of Yugoslavia, 1989-2006," The International Journal of the History of Sport, 26:9, 1187-1217. Shakya Mitra (2014), "Spanish football: from underachievers to world beaters," Soccer & Society, 15:5, 709-719. O'Donoghue et al. (2004), "An evaluation of quantitative and qualitative methods of predicting the 2002 FIFA World Cup. Part II: Game activity and analysis," Journal of Sports Sciences 22(6):513-514. Gregory T. Papanikos (2017), "Economic, population and political determinants of the 2014 World Cup match results," Soccer & Society, 18:4, 516-532. Guy Podoler (2008), "Nation, State and Football: The Korean Case," The International Journal of the History of Sport, 25:1, 1-17. Alejandro Quiroga (2017), "Spanish football and social change: sociological investigations," Soccer & Society, 18:1, 160-162. H. Sarmento et al. (2014), "Match analysis in football: a systematic review," Journal of Sports Sciences, 32:20, 1831-1843. B. Scholtens, and W. Peenstra (2009), "Scoring on the stock exchange? The effect of football matches on stock market returns: an event study," Applied Economics, 41:25, 3231-3237. Stefan Szymanski (2010), "The Economic Impact Of The World Cup," Football Economics and Policy 226-235. Taleb, Nassim, Fooled By Randomness (New York: Random House), 316 pages. F. Telseth, and V. Halldorsson (2017), "The success culture of Nordic football: the cases of the national men's teams of Norway in the 1990s and Iceland in the 2010s," Sport in Society. Dag Tuastad (2014), "From football riot to revolution. The political role of football in the Arab world," Soccer & Society, 15:3, 376-388. Scott Waalkes (2017), "Does soccer explain the world or does the world explain soccer?" Soccer and globalization, Soccer & Society, 18:2-3, 166-180. D. Wong, and S. Chadwick (2017), "Risk and (in)security of FIFA football World Cups - outlook for Russia 2018," Sport in Society, 20:5-6, 583-598. F. Wunderlich, and D. Memmert (2016), "Analysis of the predictive qualities of betting odds and FIFA World Ranking: evidence from the 2006, 2010 and 2014 Football World Cups," Journal of Sports Sciences, 34:24, 20176-20184. D. Zec, and M. Paunovic (2015), "Football's positive influence on integration in diverse societies: the case study of Yugoslavia," Soccer & Society, 16:2-3, 232-244.

Highlights Global Yields: Relative growth and inflation trends continue to favor the U.S., with divergences widening as non-U.S. is downshifting. This means that the cyclical peak in spreads between U.S. Treasuries and other developed market government bonds has not been reached yet, and the latest bout of U.S. dollar strength can continue. Stay underweight U.S. Treasuries in global government bond portfolios. Italy: Concerns over the future policies of the new Five-Star/League populist coalition government in Italy have triggered a selloff in Italian financial markets. While investors are right to be worried about the potential for greater fiscal stimulus and move vocal euroskepticism from those in charge in Italy, slowing economic growth is an even bigger immediate problem for debt sustainability concerns. Downgrade Italy to underweight (2 of 5) in global government bond portfolios. Feature After knocking on the door of the 3% threshold several times this year, the 10-year U.S. Treasury yield finally blew through that level last week. The ease with which this move occurred was a bit surprising, given that bond investor sentiment has stayed consistently bearish and Treasury market positioning remains extremely short. This raises the odds of a potential pullback in yields if the U.S. economy or inflation were to lose upside momentum. The only problem for the Treasury market is that neither of those trends is occurring at the moment. Chart of the WeekTreasuries Are Losing##BR##For The Right Reasons U.S. real GDP expanded at a 2.3% annualized rate in the first quarter of 2018, and the latest real-time GDP estimates for the second quarter from the Atlanta Fed (+4.1%) and New York Fed (+3.0%) are calling for an acceleration. The leading economic indicators produced by both the OECD and the Conference Board continue to climb higher, in stark contrast to the lost momentum in hard data and lead indicators in other major regions like Europe and Japan (Chart of the Week). Similar divergences are occurring in the inflation data, where core CPI inflation is accelerating in the U.S. and languishing elsewhere. The ability of U.S. Treasury yields to ignore the negative international headlines coming from typical trouble spots like Turkey, Argentina, Italy, Iran and North Korea is impressive. Clearly, none of these developments are big enough (yet!) to have any negative impact on U.S. growth expectations and, in turn, Fed rate hike expectations. At the same time, Fed officials continue to signal that another two or three rate increases are still likely over the remainder of the year. Add in the steady climb in inflation expectations, supported by oil prices reaching multi-year highs, and it is no surprise that those aggressive Treasury short positions have been on the right side of the market. If we were to apply a weather analogy to the global economy, conditions appear "partly sunny" if looking at the U.S, but "mostly cloudy" when looking elsewhere. This has major implications for the future path of U.S. Treasury yields versus other government bond markets, and for the U.S. dollar as well. Expect U.S. Bond Relative Underperformance To Continue From a more global perspective, the ability of non-U.S. bond yields to rise has become more limited. The overall OECD leading economic indicator - which is correlated to real global bond yields - looks to be rolling over, and our diffusion index of individual country indicators shows that this trend is broad-based (Chart 2). Within the major developed economies, only the U.S. stands out as having a rising leading economic indicator (although the Canadian index is holding up at a high level). The most depressed readings come from the three markets we are overweight in our model bond portfolio - the U.K., Japan and Australia (Chart 3). These growth divergences are not only visible in "soft" economic data like leading indicators and purchasing manager indices. U.S. retail sales showed a surprising burst of strength in April, and the release of that data last week was the trigger for pushing the 10-year Treasury yield above 3%. Meanwhile, readings on real GDP growth in the first quarter for the euro area and Japan were quite weak compared to the acceleration seen throughout 2017. In the case of Japan, GDP actually contracted at a 0.6% annualized rate in Q1, ending a run of eight consecutive quarters of positive growth which was the longest such streak in 28 years (Chart 4). Chart 2A Stagflationary Tug-Of-War##BR##On Global Yields Chart 3U.S. Growth##BR##Stands Out Chart 4Is China To Blame For##BR##Slowing Non-U.S. Growth? At the same time, China's domestic economy has seen some slowing of growth, as well, as evidenced by the rapid deceleration of import growth (bottom panel). For the economies in Europe and Japan where growth is still heavily geared towards exports, and where domestic demand still struggles to gain sustainable upward momentum in the absence of an export/production cycle, a slowing China poses a big problem - one that is less of an issue for the more domestically-focused U.S. economy. The divergence of growth and inflation accelerating in the U.S. but potentially peaking out elsewhere, can be seen in the widening of government bond yield spreads between the U.S. and its developed market peers. In Table 1, we show the change in the bond yield spread between 10-year U.S. Treasuries and similar maturity government debt from the U.K., Germany, Japan, Canada and Australia since the last major trough in global yields in September 2017. The spread changes are broken down into movements in inflation expectations and real yields to see which was more influential. For example, of the 75bps widening in the 10-year U.S. Treasury-German Bund spread, 55bps has been due to widening real yield differentials and only 20bps has come from higher inflation expectations in the U.S. Table 1Cross-Country Yield Spread Changes (in bps) Since The September 2017 Low In U.S. Treasury Yields These changes show that the underperformance of U.S. Treasuries (i.e. spread widening) has come mostly though higher real yields in the U.S. Inflation expectations are widening in the U.S., but are also moving higher in all other countries except the U.K. So the relative change in inflation expectations between the U.S. and the other countries has been more modest than the absolute change in U.S. TIPS breakevens (Chart 5). The fact that the real yield differentials are moving increasingly in favor of the U.S. has implications for the U.S. dollar. The greenback has finally begun to appreciate after the weakness seen in 2017, with potentially a lot more room to run judging by the levels implied by those wide real yield gaps. This is most evident for the euro, yen and British pound (Chart 6). Chart 5Higher Inflation Expectations##BR##& Yields In The U.S. Chart 6USD Finally Responding To Wide##BR##Real Yield Differentials The path of the U.S. dollar is the key to how this U.S./non-U.S. growth divergence story will end. If the dollar continues to strengthen as the Fed lifts rates in the coming months, then monetary conditions in the U.S. run the risk of moving into restrictive territory. This could spur a bout of renewed U.S. market turbulence not unlike that seen in 2015 and 2016 when the Fed was trapped in what we described at the time as a "policy loop", where a higher dollar and rising market volatility (especially in the emerging markets) prompted the Fed to delay planned rate hikes. The circumstances are different now compared to three years ago. The dollar is only mildly appreciating from the depressed levels of 2017, U.S. core inflation is approaching the Fed's 2% target, and the U.S. economy is at full employment with fiscal stimulus on the way. In other words, the hurdle for the Fed to alter its current rate hike plans is much higher than it was in 2015/16 when the U.S. economy and inflation were in more fragile states. For now, we continue to see relative growth and inflation trends pushing in a direction for continued U.S. government bond underperformance over the balance of 2018. One-sided bearish positioning may create a backdrop where Treasury yields could fall for a brief period, but the true cyclical peak in yields - somewhere in the 3.25-3.5% range - and in U.S./non-U.S. yield spreads has not been reached yet. Bottom Line: Relative growth and inflation trends continue to favor the U.S., with divergences widening as non-U.S. is downshifting. This means that the cyclical peak in spreads between U.S. Treasuries and other developed market government bonds has not been reached yet, and the latest bout of U.S. dollar strength can continue. Stay underweight U.S. Treasuries in global government bond portfolios. Italy: Worry More About Slowing Growth Than Politics Italian political risk returned to European financial markets last week after details of the policy program for the new Five-Star Movement/League coalition government were leaked to the press. Some of the more alarming proposals included: Having the European Central Bank (ECB) "freeze" or "cancel" the €250bn in Italian government debt it holds via its asset purchase program. Revising the rules of the European Union (EU) Growth and Stability Pact, specifically its fiscal rules on debt and deficits, while also asking for Europe to, more generally, return to a "pre-Maastricht" (pre-euro?) position. These headlines were interpreted as a sign that the populists taking over Italy were looking for a way to loosen fiscal policy in excess of EU rules, if not abandon the euro currency entirely. This would be a realization of the outcome from the March election that investors feared the most. Markets responded as expected, with Italian government bond yields soaring across the entire yield curve and Italian equities and the euro selling off (Chart 7). We last discussed Italy back in February in a Special Report co-written with our colleagues at BCA Geopolitical Strategy.1 We concluded that, even though euroskepticism would continue to have appeal in Italy because support for the common currency is much weaker than in the rest of the euro area (Chart 8), none of the likely coalition partners in a new government would make noise about potentially bringing back the lira with the economy in a cyclical expansion. All of the likely winning coalitions would seek to ease Italian fiscal policy, however, which would bring back investor worries about Italian debt sustainability. Chart 7The Return Of##BR##The Italy Risk Premium Chart 8The Euro Is Still Less Popular##BR##In Italy Than Elsewhere The first part of our conclusion went in a fashion that we did not expect, with the anti-establishment Five-Star party joining forces with the far-right League in a populist coalition that could embrace euroskepticism more emphatically. The second part of that conclusion does appear to be panning out, with the new government already looking to cut taxes and ramp up fiscal spending. These outcomes would be enough for investors to begin pricing in a higher fiscal risk premium in Italian assets, thus justifying the market moves seen last week. Yet there was one other conclusion from our report that is more relevant now for fixed income investors. Italian government bonds would not begin to underperform until there were signs that Italy's economy was slowing - which is what appears to be happening now. Like the rest of the euro area, Italy saw a deceleration of economic growth in the first quarter of the year. The most cyclical components of the Italian economy, manufacturing and exports, have both shown a considerable deceleration. Exports to non-EU countries, in particular, have noticeably slowed (Chart 9), which is likely yet another sign of how slowing Chinese growth is spilling over into much of the global economy through trade channels. Domestic demand has seen some cyclical strength on the back of the surge in exports, production and employment seen in 2016/17. However, the risk now is that slowing exports feed back into slowing production and weaker hiring activity. Any sign of a slowdown would only embolden the new coalition government to aim for easier fiscal policy. That would be a logical response by any government, particularly with current budget forecasts calling for tightening fiscal policy over the next few years. The latest set of debt and deficit projections from the IMF show that Italy is expected to have a balanced budget by 2021 (Chart 10). This would imply that the primary budget balance (i.e. net of interest payments) would rise to as high as 3.6% of GDP - an enormously restrictive policy stance that no advanced economy currently runs. Chart 9Italian Cyclical Momentum##BR##Has Peaked Chart 10This Rosy Trajectory For##BR##Italian Debt Will Not Happen That degree of fiscal tightening also makes the debt dynamics of Italy look much more sustainable, with debt/GDP projected to fall by ten percentage points by 2021 according to the IMF (bottom panel). Given the leanings of the new government, and with the economy starting to lose some momentum, there is zero chance that the IMF deficit and debt projections will come to fruition. In fact, the opposite is likely to happen under the new government, with the fiscal deficit likely to widen and debt/GDP likely to increase. While a return to the "bad old" economic policies of Italy might harken back to the days of the 2011 European debt crisis, there are two major differences between then and now: Italy's borrowing costs are far lower, thanks to the hyper-easy monetary policies of the ECB (both zero/negative interest rates and outright bond purchases). The average debt on newly-issued Italian government debt has plunged from the 6-7% levels around the time of the debt crisis to less than 1% over the past three years, according to the Bank of Italy (Chart 11). This has helped substantially reduce the amount of net interest payments made by the Italian government - by one full percentage point of GDP, according to the IMF. Less Italian debt is owned by non-Italian residents than during the crisis. According to data from the Bruegel think tank in Brussels, the percentage of Italian sovereign debt held by non-Italian residents is now 36%, compared to 50% during the years before the crisis (Chart 12). As that crisis unfolded, those investors rapidly dumped their Italian bonds, cutting their ownership share by ten percentage points in less than one year. Domestic Italian banks were forced to pick up the slack, which increased the already significant fiscal exposure of the Italian banking system. Now, the ownership mix is much more balanced, including the 20% of Italian bonds owned by the ECB. This means that, today, 64% of Italy's debt is owned by those with a vested interest in Italian stability, rather than fickle foreign investors who would be much more willing to dump their bonds when the Italian news turns less favorable. Chart 11The Big Difference Between 2011 & Today Chart 12A Smaller Share Of Italy's Debt Is Held By Fickly Foreigners Now Vs 2011 This is not to say that another Italian debt crisis could not happen, especially if the Five-Star/League coalition were to more seriously discuss a potential exit from the euro. The only difference now is that Italy's debt sustainability issues are not as acute as in 2011 because of the low borrowing costs and more diverse ownership of Italian debt. Chart 13Downgrade Italian Debt To Underweight From a bond strategy perspective, however, we are more focused on the growth dynamics in Italy than the current political noise. As we also concluded in our February Special Report, the time to downgrade Italian debt was when the economy was clearly about to slow, as heralded by a decline in the OECD's leading economic indicator for Italy. That series has been highly correlated to the relative performance of Italian government debt (Chart 13) and, therefore, is a useful indicator to follow to determine Italian bond strategy. With the leading indicator now falling for four consecutive months, and with hard Italian data also starting to slow, a period of Italian bond underperformance has likely just begun - an outcome that can only be made worse by the new euroskeptic and free spending Italian government. Thus, we are downgrading Italy in our country rankings this week to underweight (2 out of 5), and cutting our recommended allocations to Italian debt in our model bond portfolio to ½ index weight. We place the proceeds of that reduction into German bonds across the yield curve. Bottom Line: Concerns over the future policies of the new Five-Star/League populist coalition government in Italy have triggered a selloff in Italian financial markets. While investors are right to be worried about the potential for greater fiscal stimulus and move vocal euroskepticism from those in charge in Italy, slowing economic growth is an even bigger immediate problem for debt sustainability concerns. Downgrade Italy to underweight (2 of 5) in global government bond portfolios. Robert Robis, Senior Vice President Global Fixed Income Strategy rrobis@bcaresearch.com 1 Please see BCA Global Fixed Income Strategy/Geopolitical Strategy Special Report, "Italy: Growth Cures All Ills ... For Now", dated February 21st 2018, available at gfis.bcaresearch.com and gps.bcaresearch.com. Recommendations The GFIS Recommended Portfolio Vs. The Custom Benchmark Index Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Highlights The Swan Diagram depicts four different "zones of economic unhappiness," each one corresponding to a case where unemployment and inflation is either too high or too low, and the current account position is either too large or too small. The global economy has made significant progress in moving towards both internal and external balance over the past few years, but shortfalls remain. A number of large economies, including Japan, China, and Italy, continue to need stimulative fiscal policy to prop up domestic demand. In Italy's case, investor unease about the country's fiscal outlook is likely to raise borrowing costs for the government, curb capital inflows into the euro area, and push the ECB in a more dovish direction. All this will weigh on the euro. The U.S. should be tightening fiscal policy at this stage in the cycle. Instead, President Trump has pushed through significant fiscal easing. This is the main reason the 10-year Treasury yield hit a seven-year high this week. An overheated U.S. economy will pave the way for further Fed hikes, which will likely result in a stronger dollar. Rising U.S. rates and a strengthening dollar will hurt emerging markets. Turkey, South Africa, Brazil, and Indonesia are among the most vulnerable. Feature The Dismal Science, Illustrated Last week's report discussed the market consequences of the tug-of-war that policymakers often face in trying to achieve a variety of economic objectives with a limited set of policy instruments.1 In passing, we mentioned that some of these trade-offs can be depicted using the so-called Swan Diagram, named after Australian economist Trevor Swan. This week's report delves further into this topic by estimating where various economies find themselves inside the Swan Diagram, and what this may mean for their currency, equity, and bond markets. True to the reputation of economics as the dismal science, the Swan Diagram depicts four "zones of economic unhappiness" (Chart 1). Each zone represents a different way in which an economy can deviate from "internal balance" (low and stable unemployment) and "external balance" (an optimal current account position). This amounts to saying that an economy can suffer from one of the following: 1) high unemployment and an excessively large current account deficit; 2) high inflation and an excessively large current account surplus; 3) high unemployment and an excessively large current account surplus; and 4) high inflation and an excessively large current account deficit. Box 1 describes the logic behind the diagram. Chart 1Four Zones Of Unhappiness BOX 1 The Logic Behind The Swan Diagram As noted in the main text, the Swan Diagram depicts four different "zones of economic unhappiness," each one corresponding to a case where unemployment and inflation are either too high or too low, and the current account balance is either too large or too small. A rightward movement along the horizontal axis can be construed as an easing of fiscal policy, whereas an upward movement along the vertical axis can be thought of as an easing in monetary policy. All things equal, easier monetary policy is assumed to result in a weaker currency. The internal balance schedule, which corresponds to the ideal state where the economy is at full employment and inflation is stable, is downward sloping because an easing in fiscal policy must be offset by a tightening in monetary policy in order to keep the economy from overheating. The external balance schedule is upward sloping because easier fiscal policy raises aggregate demand, which results in higher imports, and hence a deterioration in the trade balance. A depreciation of the currency via an easing in monetary policy is necessary to bring imports back down. Any point to the right of the internal balance schedule represents too much inflation; any point to the left represents too much unemployment. Likewise, any point to the right of the external balance schedule represents a larger-than-acceptable current account deficit, whereas any point to the left represents an excessively large current account surplus. Note that according to the Swan Diagram, an economy that suffers from high unemployment may still need a weaker currency even if it already has a current account surplus. Intuitively, this is because a depressed economy suppresses imports, leading to a "stronger" current account balance than would otherwise be the case. We use two variables to estimate the degree to which an economy has diverged from internal balance: core inflation and the output gap (Chart 2). If the output gap is negative, the economy is producing less output than it is capable of. If the output gap is positive, the economy is operating beyond full capacity. All things equal, high core inflation and a large and positive output gap is symptomatic of an economy that is showing signs of overheating. Chart 2The Two Dimensions Of Internal Balance When it comes to estimating the extent to which an economy is deviating from external balance, we include both the current account position and the net international investment position (NIIP) in our calculations (Chart 3). The NIIP is the difference between an economy's external assets and its liabilities. If one were to sum all current account balances into the distant past and adjust for valuation effects, one would end up with the net international investment position. If a country has a positive NIIP, it can run a current account deficit over time by running down its accumulated foreign wealth.2 Chart 3The Two Dimensions Of External Balance Policy And Market Outcomes Within The Swan Diagram Chart 4 shows our estimates of where the main developed and emerging markets fall into the Swan Diagram. The top right quadrant depicts economies that need to tighten both monetary and fiscal policy. The bottom left quadrant depicts economies that need to ease both monetary and fiscal policy. The other two quadrants denote cases where either tighter fiscal/looser monetary policy or looser fiscal/tighter monetary policy are appropriate. In order to gauge progress over time, we attach an arrow to each data point. The base of the arrow shows where the economy was five years ago and the tip shows where it is today. Chart 4Policy Prescription Arising From The Swan Diagram From a market perspective, an economy's currency is likely to weaken if it finds itself in one of the two quadrants requiring easier monetary policy. Among developed economies, the best combination for equities in local-currency terms is usually an easier monetary policy and a looser fiscal policy. That is also the configuration that results in the sharpest steepening of the yield curve. Conversely, the worst outcome for developed market stocks in local-currency terms is tighter monetary policy coupled with fiscal austerity. That is also the policy package that is most likely to result in a flatter yield curve. In dollar terms, a stronger local currency will typically boost returns. This is particularly the case in emerging markets, where stock markets are likely to suffer in situations where the home currency is under pressure. A few observations come to mind: The global economy has made significant progress in restoring internal balance over the past five years. That said, negative output gaps remain in nearly half of the countries in our sample. And even in several cases where output gaps have disappeared, a shortfall in inflation suggests the presence of latent slack that official estimates of excess capacity may be missing. External imbalances have also declined over time. Since earth does not trade with Mars, the global current account balance and net international investment position must always be equal to zero. Nevertheless, the absolute value of current account balances, expressed as a share of global GDP, has fallen by half since 2006 (Chart 5). Chart 5Shrinking Global Imbalances The decline in China's current account balance has played a key role in facilitating the rebalancing of demand across the global economy. The current account showed a deficit in Q1 for the first time in 17 years. While several technical factors exacerbated the decline, the current account will probably register a surplus of only 1% of GDP this year, down from a peak of nearly 10% of GDP in 2007. The Chinese economy also appears to be close to internal balance. However, maintaining full employment has come at the cost of rapid credit growth and a massive quasi-public sector deficit, which the IMF estimates currently stands at over 12% of GDP (Chart 6). Thus, one could argue that a somewhat weaker currency and less credit expansion would be in China's best interest. Similar to China, Japan has been able to reach internal balance only through lax fiscal policy (Chart 7). The lesson here is that economies such as China and Japan which have a surfeit of savings - partly reflecting a very low neutral real rate of interest - would probably be better off with cheaper currencies rather than having to rely on artificial means of propping up demand. Chart 6China's 'Secret' Budget Deficit Chart 7The Cost Of Propping Up Demand Germany has overtaken China as the biggest contributor to current account surpluses in the world. Germany's current account surplus now stands at over 8% of GDP, up from a small deficit in 1999, when the euro came into inception. In contrast to China and Japan, Germany is running a fiscal surplus. Solely from its perspective, Germany would benefit from more fiscal stimulus and a stronger euro. The problem, of course, is that a stronger euro would not be in the best interest of most other euro area economies. While external imbalances within the euro area have decreased markedly over the past decade, they have not gone away (Chart 8). Investors also remain wary of fiscal easing in Southern Europe. This week's spike in Italian bond yields - fueled by speculation that a Five-Star/League government will abandon plans for fiscal consolidation - is a timely reminder that the bond vigilantes are far from dead (Chart 9). The Italian government's borrowing costs are likely to rise over the coming months, which will curb capital inflows into the euro area and push the ECB in a more dovish direction. All this will weigh on the common currency. Chart 8The Euro Club: Imbalances Have Been Decreasing Chart 9Uh Oh Spaghettio! The U.S. is the opposite of Germany. Unlike Germany, it has a large fiscal deficit and a current account deficit. The Swan Diagram says that the U.S. would benefit from tighter fiscal policy and a weaker dollar. President Trump and the Republicans in Congress have other plans, however. They have pushed through large tax cuts and significant spending increases (Chart 10). This will likely prompt the Fed to raise rates more aggressively than the market is currently discounting, leading to a stronger dollar. Chart 10The U.S. Budget Deficit Is Set To Widen Even If The Unemployment Rate Continues To Decline Rising U.S. rates and a strengthening dollar will hurt emerging markets, particularly those with current account deficits and negative net international investment positions. High levels of external debt could exacerbate any problems (Chart 11). On that basis, Turkey, South Africa, Brazil, and Indonesia are among the most vulnerable. Chart 11External Debt And Debt Servicing Across EM Investment Conclusions Chart 12The U.S. Economy Is Doing ##br##Better Than Its Peers The global economy is approaching internal balance, but this may produce some unpleasant side effects. Productivity growth is anaemic and the retirement of baby boomers from the workforce will reduce the pace of labor force growth. In such a setting, potential GDP growth in many countries is likely to remain subpar. If demand growth continues to outstrip supply growth, inflation will rise. Heightened stock market volatility this year has partly been driven by the realization among investors that the Goldilocks environment of above-trend growth and low inflation may not last as long as they had hoped. The U.S. economy has now moved beyond full employment, and bountiful fiscal stimulus could lead to further overheating. This is the main reason the 10-year Treasury yield reached a seven-year high this week. Continued above-trend growth is likely to prompt the Fed to raise rates more than the market expects, which should result in a stronger dollar. The fact that the U.S. economy is outperforming the rest of the world based on economic surprise indices and our leading economic indicators could give the dollar a further lift (Chart 12). A resurgent dollar will help boost competitiveness in developed economies such as Japan and Europe. Emerging markets will also benefit in the long run from cheaper currencies, but if the adjustment happens rapidly, as is often the case, this could exact a short-term toll. For the time being, investors should overweight developed over emerging markets in equity portfolios. Peter Berezin, Chief Global Strategist Global Investment Strategy peterb@bcaresearch.com 1 Please see Global Investment Strategy Weekly Report, "Tinbergen's Ghost," dated May 11, 2018. 2 To keep things simple, we assume that a country's Net International Investment Position (NIIP) shrinks to zero over 50 years. Thus, if a country has a positive NIIP of 50% of GDP, we assume that it should target a current account deficit of 1% of GDP; whereas if it has a negative NIIP of 50% of GDP, it should target a current account surplus of 1% of GDP. Tactical Global Asset Allocation Recommendations Strategy & Market Trends Tactical Trades Strategic Recommendations Closed Trades