BCA Indicators/Model

Highlights Q3/2018 Performance Breakdown: The Global Fixed Income Strategy (GFIS) recommended model bond portfolio outperformed its custom benchmark in the third quarter of 2018 by +9bps. This raised the overall 2018 year-to-date performance to +6bps. Winners & Losers: The outperformance came mostly from our defensive duration positioning, which benefitted as global bond yields rose during the quarter, but also from successful country selection (overweight Australia & New Zealand, underweight the U.S., Canada & Italy). Our underweight tilts on EM credit were the largest drag on performance after the sharp EM rally in September. Scenario Analysis: The combination of defensive overall duration positioning and underweight allocations to EM and European credit should allow the model bond portfolio to outperform its custom benchmark index over the next year. Feature This week, we present the performance numbers of the BCA Global Fixed Income Strategy (GFIS) model bond portfolio for the 3rd quarter of 2018. We also update our scenario analysis of the future expected performance of the portfolio based on the risk-factor based return forecasting framework we introduced earlier this year. As a reminder to existing readers (and for new clients), the portfolio is a part of our service that is meant to complement the usual macro analysis of global fixed income markets. The model portfolio is how we communicate our opinion on the relative attractiveness between government bond and spread product sectors, by applying actual percentage weightings to each of our recommendations within a fully invested hypothetical bond portfolio. Broadly speaking, the portfolio did slightly outperform its benchmark index over the past three months, driven mostly by defensive duration positioning during a period of rising developed market bond yields. The portfolio would have done considerably better if not for a September rally in emerging market (EM) credit that flew in the face of our maximum underweight position in EM. We still have strong conviction in those two main themes - higher global bond yields and EM underperformance - and we fully expect our model portfolio to generate larger outperformance over the next year. Q3/2018 Model Portfolio Performance Breakdown: Duration Underweights Pay Off The total return of the GFIS model bond portfolio was +0.12% (hedged into U.S. dollars) in the third quarter of the year, which outperformed the custom benchmark index by +9bps (Chart of the Week).1 The main driver of the outperformance was our structural below-benchmark portfolio duration stance, which benefited as the overall Bloomberg Barclays Global Treasury Index yield rose to 1.54% - the highest level since April 2014. The portfolio's excess return got as high as +19bps on September 4th, before seeing some pullback in recent weeks as our main spread product tilt - underweight EM hard currency sovereign and corporate debt - enjoyed a counter-trend rally in September from the bearish spread widening seen since the start of 2018. Chart of the WeekDefensive Duration Stance = Q3 Outperformance Table 1GFIS Model Bond Portfolio Q3/2018 Overall Return Attribution In terms of the specific breakdown between the government bond and spread product allocations in our model portfolio, the former generated +17bps of outperformance versus our custom benchmark index while the latter lagged the benchmark by -8bps (Table 1). The bar charts showing the total and relative returns for each individual government bond market and spread product sector are presented in Charts 2 and 3. Chart 2GFIS Model Bond Portfolio Q3/2018 Government##BR##Bond Performance Attribution By Country Chart 3GFIS Model Bond Portfolio Q3/2018 Spread##BR##Product Performance Attribution By Sector The main individual sectors of the portfolio that drove the excess returns were the following: Biggest outperformers Underweight Japanese government bonds (JGBs) with maturities beyond 10 years (+7bps) Underweight U.S. Treasuries with maturities beyond 7 years (+6bps) Underweight French government bonds with maturities beyond 7 years (+2bps) Underweight Italian government bonds (+2bps) Overweight JGBs with maturities up to 10 years (+1bp) Biggest underperformers Underweight EM USD-denominated sovereign debt (-3bps) Underweight EM USD-denominated corporate debt (-3bps) Underweight euro area investment grade corporate debt (-2bps) Underweight euro area high-yield corporate debt (-1bp) Chart 4 presents the ranked benchmark index returns of the individual countries and spread product sectors in the GFIS model bond portfolio. The returns are hedged into U.S. dollars (we do not take active currency risk in this portfolio) and also adjusted to reflect duration differences between each country/sector and the overall custom benchmark index for the model portfolio. We have also color coded the bars in each chart to reflect our recommended investment stance for each market during the third quarter (red for underweight, blue for overweight, gray for neutral weight). Chart 4Ranking The Winners & Losers From The Model Portfolio In Q3/2018 Spread product sectors dominate the left half of that chart, as credit spreads have tightened across the board since the early September peak. The best performing sector during Q3 in our model portfolio universe was EM hard currency sovereign debt, which has delivered a total return of +2.8% since September 4th (with spreads tightening by 50bps) after losing -0.7% in July and August. Similar performance stories occurred in corporate debt in the U.S. and Europe during the quarter. That credit outperformance comes after the sustained spread widening seen in virtually all global credit markets (excluding U.S. high-yield) since January of this year. The main drivers that prompted that widening - Fed tightening, a stronger U.S. dollar, diminishing asset purchases from the European Central Bank (ECB) and Bank of Japan (BoJ), some cyclical slowing of non-U.S. growth - are still in place. With our geopolitical strategists continuing to highlight the additional risks of U.S.-China and U.S.-Iran tensions intensifying after next month's U.S. Midterm elections, a cautious stance on global spread product - as we have maintained since downgrading our recommended overall credit exposure to neutral in late June - is still warranted.2 Outside of spread product, our model portfolio tilts generally lined up with the sector returns shown in Chart 4. We have overweights on two of the best performing government bond markets (Australia and New Zealand) and underweights on three of the worst performers (U.S., Canada, Italy). Interestingly, despite having overweights on two of the worst performing government bond markets - Japan and the U.K. - the excess return contribution from those countries did not hurt the model bond portfolio return in Q3 (+8bps and 0bps, respectively). This was due to the curve steepening bias embedded within our overweight country tilts (i.e. more duration allocated to shorter-maturity buckets, see the model portfolio details on Page 14), which benefitted as yield curves in those countries bear-steepened. Net-net, we are satisfied with the modest portfolio outperformance seen in Q3, given that the rally in global credit markets went against our more defensive posture on spread product exposure. Bottom Line: The GFIS recommended model bond portfolio outperformed its custom benchmark in the third quarter of 2018 by +9bps. This put the overall 2018 year-to-date performance into positive territory (+6bps). The outperformance came entirely from our defensive duration positioning, which benefitted as global bond yields rose during the quarter, and from successful country selection. Our underweight tilts on EM credit were the largest drag on performance after the sharp EM rally in September. Future Drivers Of Portfolio Returns Looking ahead, the performance of the model bond portfolio will continue to benefit from two primary trends: rising global bond yields and growth divergences that continue to favor the U.S. In terms of the specific weightings in the GFIS model bond portfolio, we still prefer owning U.S. corporate debt versus equivalents in Europe and EM. When we downgraded our recommended allocation to U.S. and investment grade corporates to neutral from overweight back in July, we also cut the portfolio exposure to euro area corporates, as well as to all EM hard currency debt, to underweight. The latter changes were necessary to maintain our desired higher exposure to U.S. corporate debt versus non-U.S. corporates, although it did leave the model portfolio with a small overall underweight stance on global spread product (Chart 5). Importantly, we are maintaining a below-benchmark stance on overall portfolio duration, which is now one full year shorter than our benchmark index duration (Chart 6), even as we have grown more cautious on credit exposure. This is because we still see potential medium-term upward pressure on bond yields coming from tightening monetary policies (Fed rate hikes, ECB tapering of bond purchases) and increasing inflation expectations. The majority of global central bankers are dealing with tight labor markets and slowly rising inflation rates. While global growth has cooled a bit from the rapid pace seen in 2017, it has not been by enough to force policymakers to shift to a more dovish bias. Chart 5Spread Product Allocation:##BR##Neutral U.S., Underweight Non-U.S. Chart 6Maintaining##BR##Below-Benchnmark Duration Our underweights on EM and euro area spread product have left the portfolio in a "negative carry" position where it yields 34bps less than the benchmark index (Chart 7). In a backdrop of stable markets and low volatility, being short carry will be a drag on the model bond portfolio performance as we saw over the past month. Yet we do not see the recent market calm as being sustainable, with all plausible outcomes pointing to more volatile markets, largely driven by U.S.-centric events (more Fed tightening, a stronger dollar, U.S. growth convergence to slower non-U.S. growth, increased trade protectionism, higher oil prices due to U.S.-Iran tensions). We continue to suggest a cautious allocation of investor risk budgets against this backdrop. We have been targeting a tracking error (relative volatility versus the benchmark) for our model bond portfolio in the 40-60bp range, well below our 100bps maximum. Our current allocations give us a tracking error right at the bottom of that range (Chart 8).3 Chart 7The Cost Of Being More Defensive On Credit Chart 8Maintaining A Cautious Allocation Of The Risk Budget Scenario Analysis & Return Forecasts Back in April of this year, we introduced a framework for estimating total returns for all government bond markets and spread product sectors, based on common risk factors.4 For credit, returns are estimated as a function of changes in the U.S. dollar, the Fed funds rate, oil prices and market volatility as proxied by the VIX index (Table 2A). For government bonds, non-U.S. yield changes are estimated using historical betas to changes in U.S. Treasury yields (Table 2B). This framework allows us to conduct scenario analysis based on projected returns for each asset class in the model bond portfolio universe by making assumptions on those individual risk factors. Table 2AFactor Regressions Used To Estimate##BR##Spread Product Yield Changes Table 2BEstimated Government Bond##BR##Yield Betas To U.S. Treasuries With these tools, we than can attempt to forecast returns for each bond sector under different scenarios. We can then use those forecasts to predict the expected return for our model bond portfolio under those same scenarios. In Tables 3A & 3B. we show three differing scenarios, with all the following changes occurring over a one-year horizon. Table 3AScenario Analysis For The GFIS Model Portfolio Table 3BU.S. Treasury Yield Assumptions For The Scenario Analysis Our Base Case: the Fed delivers another 100bps of rate hikes, the U.S. dollar rises +5%, oil prices rise by +10%, the VIX index increases by five points from current levels, and U.S. Treasury yields rise by 40bps across the curve. A Very Hawkish Fed: the Fed delivers 150bps of rate hikes, the U.S. dollar rises by +10%, oil prices rise by +10%, the VIX index increases by ten points from current levels and there is a sharp bear flattening of the U.S. Treasury curve (2yr yield +75bps, 10yr yield +40bps). A Very Dovish Fed: the Fed only hikes rates by 25bps, the U.S. dollar falls by -5%, oil prices fall by -20%, the VIX index increases by fifteen points from current levels and there is a modest bull steepening of the U.S. Treasury curve. In this scenario, the Fed puts the rate hiking cycle on hold in response to a sharp tightening of U.S. financial conditions. Table 3A shows the expected returns for all three scenarios based on our risk-factor framework. The model bond portfolio is expected to outperform the custom benchmark index in all three scenarios we have laid out. This occurs even with the negative carry coming from the credit underweights in EM and Europe, with losses from credit spread widening projected to be larger than the yield give-up from being underweight. The excess returns are modest, however, with only 6bps of outperformance expected in our base case scenario and 13bps expected in the "Very Hawkish Fed" and "Very Dovish Fed" scenarios. This return distribution, with better outcomes occurring in the "tails", is a desirable property to have as it relates to the VIX/volatility forecasts embedded in the scenarios. Both of the non-base case scenarios have a higher VIX (Chart 9), even in the case of the "Very Dovish Fed" outcome where a severe U.S. financial market selloff (coming complete with a higher VIX) would be the necessary trigger for the Fed to reverse course and begin cutting interest rates (Chart 10). Such a backdrop would obviously hurt our below-benchmark duration stance, but would help our underweight EM/Europe spread product recommendations. Chart 9Risk Factors For Scenario Analysis Chart 10UST Yield Moves For Scenario Analysis Of course, our recommendations will not be static at current levels throughout the next twelve months. We increasingly expect that our next major allocation move will be downgrade U.S. spread product exposure and raise U.S. Treasury allocations, especially after the Fed delivers a few more 25bps-per-quarter rate hikes and the U.S. dollar rises further. This will provide a boost to the portfolio's expected returns through renewed spread widening and, potentially, a reduction of our below-benchmark overall duration stance as Treasury yields reach likely cyclical peaks. Bottom Line: The combination of defensive overall duration positioning and underweight allocations to EM and European credit should allow the model bond portfolio to outperform its custom benchmark index over the next year. Robert Robis, CFA, Senior Vice President Global Fixed Income Strategy rrobis@bcaresearch.com 1 The GFIS model bond portfolio custom benchmark index is the Bloomberg Barclays Global Aggregate Index, but with allocations to global high-yield corporate debt replacing very high quality spread product (i.e. AA-rated). We believe this to be more indicative of the typical internal benchmark used by global multi-sector fixed income managers. 2 Please see BCA Global Fixed Income Strategy Weekly Report, "Time To Take Some Chips Off The Table: Downgrade Global Spread Product Exposure To Neutral", dated June 26th 2018, available at gfis.bcaresearch.com. 3 In general, we aim to target a tracking error no greater than 100bps. We think this is reasonable for a portfolio where currency exposure is fully hedged and less than 5% of the portfolio benchmark is in bonds with ratings below investment grade. 4 Please see BCA Global Fixed Income Strategy Weekly Report, "GFIS Model Bond Portfolio Q1/2018 Performance Review: A Rough Start", dated April 10th 2018, available at gfis.bcareseach.com. Recommendations Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Feature GAA DM Equity Country Allocation Model Update The GAA DM Equity Country Allocation model is updated as of September 30, 2018. The quant model has not made significant allocation changes as shown in Table 1. As shown in Table 2 and Charts 1, 2 and 3, the overall model underperformed its benchmark by 53 bps in September, largely driven by Level 2 model which underperformed its benchmark by 156 bps. Japan was the largest underweight in the model, yet Japan was the best performing country in September, which contributed largely to the model's underperformance. Since going live, the overall model has outperformed its benchmarks by only 7 bps, driven by the Level 2 outperformance of 46 bps offset by the 8 bps of Level 1 underperformance. Even though the model underperformed significantly in both August and September, it's still within the back-tested range based on one-year and four-year changes. Table 1Model Allocation Vs. Benchmark Weights Table 2Performance (Total Returns In USD %) Chart 1GAA DM Model Vs. MSCI World Chart 2GAA U.S. Vs. Non U.S. Model (Level 1) Chart 3GAA Non U.S. Model (Level 2) Please see also the website http://gaa.bcaresearch.com/trades/allocation_performance. For more details on the models, please see Special Report, "Global Equity Allocation: Introducing The Developed Markets Country Allocation Model," dated January 29, 2016, available at https://gaa.bcaresearch.com. Please note that the overall country and sector recommendations published in our Monthly Portfolio Update and Quarterly Portfolio Outlook use the results of these quantitative models as one input, but do not stick slavishly to them. We believe that models are a useful check, but structural changes and unquantifiable factors need to be considered too in making overall recommendations. GAA Equity Sector Selection Model Dear Client, As advised last month, we have suspended the GAA Equity Sector Selection Model due to the significant changes in the GICS sector classifications, implemented at the end of September. We will rebuild the model using the newly constituted sectors once full back data is available from MSCI, which we understand will be in December. We thank you for your understanding. Xiaoli Tang, Associate Vice President xiaoliT@bcaresearch.com Aditya Kurian, Senior Analyst adityak@bcaresearch.com

Special Report

Highlights The Global Golden Rule (GGR): The gap between market expectations of global central bank policy rates and realized interest rate outcomes is a reliable predictor of government bond returns. Thus, "getting the policymaker call right" is the key to outperformance for bond investors. Implied Government Bond Yields: Given the strong correlation between policy rate surprises and government bond yield changes, we can use the GGR to forecast yields one year from now based on our own assumptions of how many rate hikes (cuts) will be delivered versus what is discounted in money market yield curves. Total Return Forecasts: We can use implied government bond yield changes from the GGR to generate expected 12-month total returns for government bond indexes of different maturities, taking into account different rate hike assumptions for various central banks. Feature Chart 1Global Monetary Divergences? This month marked the ten-year anniversary of the 2008 Lehman Brothers default, which set off a worldwide financial crisis and a massive easing of global monetary policy. Extraordinary measures - zero (or negative) interest rates, large-scale asset purchases and dovish forward guidance from policymakers - were all successful in suppressing both global bond yields and volatility over time, helping the global economy slowly heal from the crisis. Now, a decade later, such hyper-easy monetary policies are no longer required given low unemployment rates and rising inflation in the major developed economies. That can be seen today with the Federal Reserve shifting to "quantitative tightening" (letting bonds run off its swollen balance sheet) alongside steady rate hikes, the European Central Bank (ECB) set to stop net new buying of euro area bonds at year-end, and the Bank of Japan (BoJ) dramatically slowing its pace of asset purchases. BCA's Central Bank Monitors, which assess the cyclical pressure on policymakers to tighten or ease monetary policy, have collectively been calling for interest rate increases since the start of 2017. Yet our Central Bank Monetary Barometer, which measures the percentage of central banks that have tightened policy over the previous three months, shows that only 1 in 5 banks have actually delivered rate hikes of late (Chart 1). Thus, the risks are tilted towards more countries moving away from highly accommodative monetary conditions given tightening labor markets and rising inflation pressures. This now-global shift towards policy normalization has major implications for global bond investing. The focus is now returning back to more traditional drivers of government bond returns, like changes in central bank policy rates. We recently shared a Special Report published by our colleagues at our sister BCA service, U.S. Bond Strategy, describing a methodology they dubbed "The Golden Rule of Bond Investing".1 That report introduced a numerical framework that translates actual changes in the U.S. fed funds rate relative to market expectations into return forecasts for U.S. Treasuries. The historical results convincingly showed that investors who "get the Fed right" by making correct bets on changes in the funds rate versus expectations were very likely to make the right call on the direction of Treasury yields. In this Special Report, we extend that Golden Rule analysis to government bonds in the other major developed markets (DM). Our conclusion is that utilizing a "Global Golden Rule" (GGR) framework that links bond returns to unexpected changes in policy rates can help bond investors correctly forecast changes in non-U.S. bond yields. The report is set up in two sections. First, we illustrate how the GGR works and how it empirically tends to generally succeed over time for different DM bond markets. In the second section, we make use of the GGR to generate expected return forecasts for non-U.S. government bonds for a variety of interest rate "surprise" scenarios. ECB Policy Rate Surprises Dovish surprises from the ECB do reliably coincide with positive German government bond excess returns versus cash (Chart 2A). Chart 2AECB Policy Rate Surprise & Yields I Chart 2BECB Policy Rate Surprise & Yields II The 12-month ECB policy rate surprise and the 12-month change in the Bloomberg Barclays German Treasury index yield displays a strong positive correlation (Chart 2B). The excess returns during periods of dovish surprises is 14.4% on average and are positive 85% of the time. Hawkish surprises on the other hand, coincide with negative average excess returns of -1.5% (Chart 2C). In terms of total return, the picture is roughly the same except that under hawkish surprises, the average total return you would expect is now positive, given that it factors in coupon income (Chart 2D). Chart 2CGermany: Government Bond Index Excess Return & ECB Policy Rate Surprises (2004 - Present) Chart 2DGermany: Government Bond Index Total Return & ECB Policy Rate Surprises (2004 - Present) Table 1Germany: 12-Month Government Bond Index Returns And Rate Surprises (2004 - Present) Looking ahead, the ECB should not deviate from its current dovish forward guidance of no interest rate hikes until at least the third quarter of 2019. That is somewhat consistent with the reading of the ECB monitor being almost equal to zero. Bank Of England (BoE) Policy Rate Surprises The GGR works well for the U.K. as can be seen in Chart 3A. Chart 3ABoE Policy Rate Surprise & Yields I Chart 3BBoE Policy Rate Surprise & Yields II The 12-month BoE policy rate surprise and the 12-month change in the Bloomberg Barclays U.K. Treasury index yield displays a strong positive correlation except for a major divergence in 1997-1998 (Chart 3B). Dovish surprises coincide with positive excess returns over cash 78% of the time and are on average equal to 6.2% over the full sample (Chart 3C and Chart 3D). As you would expect if the GGR applies, hawkish surprises coincide with negative excess returns. Chart 3CU.K.: Government Bond Index Excess Return & BoE Policy Rate Surprises (1993 - Present) Chart 3DU.K.: Government Bond Index Total Return & BoE Policy Rate Surprises (1993 - Present) Table 2U.K.: 12-Month Government Bond Index Returns And Rate Surprises (1993 - Present) Looking ahead, outcomes will be biased toward dovish surprises over the next six months given the uncertain outcome of the U.K.-E.U. Brexit negotiations. Against that backdrop, the BoE will remain accommodative despite inflationary pressures building up. Bank Of Japan (BoJ) Policy Rate Surprises The GGR does not seem to work when it comes to the Japanese bond market. This reflects the fact that both the markets and the Bank of Japan (BoJ) have understood that chronic low inflation has required no changes in BoJ policy rates (Chart 4A, second panel). Chart 4ABoJ Policy Rate Surprise & Yields I Chart 4BBoJ Policy Rate Surprise & Yields II While the 12-month BoJ policy rate surprise and the 12-month change in the Bloomberg Barclays Japan Treasury index yield displayed a strong positive correlation pre-1998, the correlation has broken down since then (Chart 4B). Negative excess returns over cash both coincide with dovish and hawkish surprises, on average over time. Further, dovish surprises coincide with positive excess returns only 45% of the time (Chart 4C and Chart 4D). Chart 4CJapan: Government Bond Index Excess Return & BoJ Policy Rate Surprises (1994 - Present) Chart 4DJapan: Government Bond Index Total Return & BoJ Policy Rate Surprises (1994 - Present) Table 3Japan: 12-Month Government Bond Index Returns And Rate Surprises (1994 - Present) Looking ahead, given that the BoJ will in all likelihood maintain its ultra-accommodative monetary policy stance in the near future, we do not expect the GGR to become more effective when applied to the Japanese bond market. Bank Of Canada (BoC) Policy Rate Surprises The GGR works relatively well for the Canadian bond market (Chart 5A). Chart 5ABoC Policy Rate Surprise & Yields I Chart 5BBoC Policy Rate Surprise & Yields II We observe a tight correlation between 12-month BoC policy rate surprises and the 12-month change in the Bloomberg Barclays Canada Treasury index yield, especially post-2010 (Chart 5B). Dovish surprises coincide with positive excess returns 81% of the time and 94% of the time if we look at total returns (Chart 5C and Chart 5D). Chart 5CCanada: Government Bond Index Excess Return & BoC Policy Rate Surprises (1993 - Present) Chart 5DCanada: Government Bond Index Total Return & BoC Policy Rate Surprises (1993 - Present) Table 4Canada: 12-Month Government Bond Index Returns And Rate Surprises (1993 - Present) Looking ahead, the BoC will most likely continue to follow the tightening path of the Federal Reserve, admittedly with a lag. However, accelerating inflation at a time when there is no spare capacity in the Canadian economy suggests that the BoC could deliver more rate hikes than are already priced for the next 12 months. As shown in Table 4, hawkish surprises from the BoC do coincide with negative monthly excess returns of -2.8%. Reserve Bank Of Australia (RBA) Policy Rate Surprises The GGR applies extremely well to the Australian bond market (Chart 6A). Chart 6ARBA Policy Rate Surprise & Yields I Chart 6BRBA Policy Rate Surprise & Yields II The 12-month RBA policy rate surprise and the 12-month change in the Bloomberg Barclays Australia Treasury index yield displays the tightest correlation out of all the countries covered (Chart 6B). Dovish surprises coincide with positive excess returns 83% of the time and 96% of the time if we look at total returns (Chart 6C and Chart 6D). Turning to hawkish surprises, they reliably coincide with negative excess returns. Chart 6CAustralia: Government Bond Index Excess Return & RBA Policy Rate Surprises (1994 - Present) Chart 6DAustralia: Government Bond Index Total Return & RBA Policy Rate Surprises (1994 - Present) Table 5Australia: 12-Month Government Bond Index Returns And Rate Surprises (1994 - Present) As can be seen on the bottom panel of Chart 6A, the RBA Monitor has been rapidly falling since 2016 and now stands in the "easier monetary policy" required. However, the RBA will likely have to see a rise in unemployment or a decline in realized inflation before it considers cutting rates, which raises a risk of "hawkish" surprises if the market begins to price in rate cuts. Reserve Bank Of New Zealand (RBNZ) Policy Rate Surprises The GGR works fairly well for Nez Zealand (NZ) government bonds (Chart 7A). Chart 7ARBNZ Policy Rate Surprise & Yields I Chart 7BRBNZ Policy Rate Surprise & Yields II 12-month RBNZ policy rate surprises and the 12-month change in the Bloomberg Barclays NZ Treasury yield exhibit a decent correlation (Chart 7B). Unusually, NZ is the only bond market covered in this report where both dovish and hawkish surprises coincide with positive excess returns on average, although positive episodes are much less frequent for hawkish surprises than for dovish surprises; respectively 55% and 86% (Chart 7C and Chart 7D). Chart 7CNZ: Government Bond Index Excess Return & RBNZ Policy Rate Surprises (2000 - Present) Chart 7DNZ: Government Bond Index Total Return & RBNZ Policy Rate Surprises (2000 - Present) Table 6New Zealand: 12-Month Government Bond Index Returns And Rate Surprises (2000 - Present) Looking ahead, the RBNZ has already provided forward guidance indicating that the Overnight Cash Rate (OCR) will most likely stay flat until 2020 - an assessment that we agree with, so the odds are against any policy surprises over at least the next 6-12 months. Using The Global Golden Rule To Forecast Government Bond Returns The practical application of the GGR is that it can be used as a framework for generating expected changes in yields and calculating total return forecasts for global government bond indices. The strong correlation demonstrated in the previous section between the 12-month policy rate surprises and the 12-month change in the average yield from the government bond indexes allows us to translate our "assumed" policy rate surprise over the next 12 months into expected changes in yields along the curve. With these expected yield changes, we can simply generate expected total returns using the following formula: Expected Total Return = Yield - (Duration*Expected Change In Yield) + 0.5*Convexity*E(DY2) E(DY2) = 1-year trailing estimate of yield volatility It is important to note that we would not give too much importance to what this analysis yields for longer-dated bonds. As shown in the Appendices, once we move into longer government bond maturities, the correlation between the policy rate surprise and the change in yields declines or even becomes non-existent for some countries. This result should not be surprising, as longer-term yields are driven by other factors besides simply changes in interest rate expectations. Inflation expectations, government debt levels and demand from longer-term investors like pension funds all can have a more outsized influence on the path of longer-term bond yields relative to the shorter-end. That results in much more uncertainty when it comes to the total return forecasts for long-dated maturities calculated with this framework. Practically speaking, we are not encouraging our readers to blindly follow that yield and return expectations generated by the GGR, even for bond markets where it clearly seems to be working over time. Rather, the GGR can be integrated in a larger asset-allocation framework for a global fixed-income portfolio by providing one possible set of bond market outcomes. On a total return basis, the results presented below, interpreted alongside the readings on the BCA Central Bank monitors, suggest that investors should be underweight core Euro Area (Germany, France and Italy), Australia and New Zealand while remaining overweight the U.K. and Canada over the next twelve months. As for Japan, given the likelihood that BoJ will leave its policy rate flat, the results hint at a neutral allocation. Jeremie Peloso, Research Analyst jeremie@bcaresearch.com Robert Robis, CFA, Senior Vice President Global Fixed Income Strategy rrobis@bcaresearch.com 1 Please see U.S. Bond Strategy Special Report, "The Golden Rule Of Bond Investing", dated July 24, 2018, available at usbs.bcaresearch.com. 2 Please see Global Fixed Income Strategy Weekly Report, "BCA Central Bank Monitor Chartbook: Divergences Opening Up," dated September 19, 2018, available at gfis.bcaresearch.com. Global Golden Rule: Germany In light of the forward guidance ECB President Mario Draghi has been providing to the markets, it appears that the most likely scenario over the next 12 months is for the ECB to keep interest rates on hold. Based on the strong relationships between 12-month ECB policy rate surprises and 12-month changes in yields along the curve (Appendix A), a flat interest rate scenario would be bond bearish for German government bonds especially at the short end of the curve with the 1-year German yield expected to rise by 16bps (Table 7A). Table 7AGermany: Expected Changes In Bund Yields Over The Next 12 Months (BPs) Using the expected change in yields thus inferred by the policy rate surprise, the German government bond aggregate index is forecasted to return 0.45% over the next 12 months (Table 7B). Table 7BGermany: Government Bond Index Total Return Forecasts Over The Next 12 Months Global Golden Rule: U.K. Markets are currently discounting only 21bps of rate hikes in the U.K. over the next year. Thus, even a scenario where the BoE delivers only a single 25bp rate hike would be bearish for U.K. Gilts, especially at the short-end of the curve. Applying the GGR, 1- and 3-year Gilt yields would be expected to rise by 20bps and 10bps respectively (Table 8A). Table 8AU.K.: Expected Changes In Gilt Yields Over The Next 12 Months (BPs) Interpolating these expected yield changes, the 1-3 year government bond index total return forecast would be 0.46%. On the other hand, if the BoE prefers to keep rates on hold given the uncertainty of the Brexit outcome, that same 1-3 year government bond index is forecasted to deliver 0.97% of total return over the next 12 months (Table 9B). This is our current base case scenario for Gilts. Table 8BU.K.: Government Bond Index Total Return Forecasts Over The Next 12 Months Global Golden Rule: Japan Despite many rumors to the contrary earlier this year, the base case view remains that the BoJ will not change its stance on monetary policy anytime soon. As such, the expected changes in JGB yields under a flat interest rate scenario over the next 12 months are close to zero at the short end of the curve and rather bond bullish at the longer end of the curve; for instance, the 30-year JGB yield would be expected to rally by 9bps (Table 9A). Table 9AJapan: Expected Changes In JGB Yields Over The Next 12 Months (BPs) In that most likely scenario, the Japanese government bond index is forecasted to deliver 0.83% of total return over the next 12 months. In the event that the BoJ surprises the markets by delivering one rate hike of 25bps, it would be bond bearish for JGBs and the total return forecasts for the government bond indices would be negative, regardless of the maturity (Table 9B). Table 9BJapan: Government Bond Index Total Return Forecasts Over The Next 12 Months Global Golden Rule: Canada Will the Bank of Canada follow the footsteps of the Fed? The markets certainly seem to think so, with more than three 25bps rate hikes priced in for next 12 months in the OIS curve. Table 10ACanada: Expected Changes In Government Bond Yields Over The Next 12 Months (BPs) That scenario would be outright bearish for Canadian government bonds, with 1- and 2-year yields rising by 16bps and 21bps, respectively (Table 10A). In terms of total returns, the GGR framework forecasts that with 75bps of rate hikes, the Canadian government bond aggregate index would deliver a positive return of 2.35% (Table 10B). This is because 75bps of hikes are currently discounted in the Canadian OIS curve, thus it would neither be a hawkish nor dovish surprise. Table 10BCanada: Government Bond Index Total Return Forecasts Over The Next 12 Months Global Golden Rule: Australia The RBA Monitor just dipped below the zero line, implying that easier monetary policy is required based on financial and economic data. Table 11A shows that a rate cut delivered by the RBA in the next 12 months would be bond bullish for Aussie yields, especially at the long end of the curve, where the 30-year Aussie bond yield would fall by 34bps. Table 11AAustralia: Expected Changes In Aussie Yields Over The Next 12 Months (BPs) Of all the interest rate scenarios presented in Table 11B, the two rate cut scenarios would return the highest total returns. For instance, the Australian government bond aggregate index would return 2.80% and 3.90% in the event of one and two 25bps rate hikes, respectively. Table 11BAustralia: Government Bond Index Total Return Forecasts Over The Next 12 Months Global Golden Rule: New Zealand Our view is that the Reserve Bank of New Zealand will stay on hold for a while longer, which is broadly the same message conveyed by the RBNZ Monitor being positive, but very close to 0. With that in mind, a flat interest rate scenario appears to be bond bearish for the NZ bond yields, except for the longer end of the curve (Table 12A). Table 12ANew Zealand: Expected Changes In NZ Yields Over The Next 12 Months (BPs) Table 12BNew Zealand: Government Bond Index Total For New Zealand, the government bond aggregate bond index is the only index provided by Bloomberg Barclays, as opposed to the other countries in our analysis where different maturities are given. In the flat interest rate scenario, the total return forecast for the overall index would be of 2.53% over the next 12 months. Appendix A: Germany Chart 1Change In 1-Year German Bund Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 2Change In 2-Year German Bund Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 3Change In 3-Year German Bund Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 4Change In 5-Year German Bund Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 5Change In 7-Year German Bund Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 6Change In 10-Year German Bund Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 7Change In 30-Year German Bund Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Appendix B: France Chart 8Change In 1-Year French OAT Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 9Change In 2-Year French OAT Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 10Change In 3-Year French OAT Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 11Change In 5-Year French OAT Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 12Change In 7-Year French OAT Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 13Change In 10-Year French OAT Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 14Change In 30-Year French OAT Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Appendix C: Italy Chart 15Change In 1-Year Italian Gov't Bond Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 16Change In 2-Year Italian Gov't Bond Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 17Change In 3-Year Italian Gov't Bond Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 18Change In 5-Year Italian Gov't Bond Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 19Change In 7-Year Italian Gov't Bond Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 20Change In 10-Year Italian Gov't Bond Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Chart 21Change In 30-Year Italian Gov't Bond Yield##BR##Vs. 12-Month ECB Policy Rate Surprise Appendix D: U.K. Chart 22Change In 1-Year Gilts Yield##BR##Vs. 12-Month BoE Policy Rate Surprise Chart 23Change In 2-Year Gilts Yield##BR##Vs. 12-Month BoE Policy Rate Surprise Chart 24Change In 3-Year Gilts Yield##BR##Vs. 12-Month BoE Policy Rate Surprise Chart 25Change In 5-Year Gilts Yield##BR##Vs. 12-Month BoE Policy Rate Surprise Chart 26Change In 7-Year Gilts Yield##BR##Vs. 12-Month BoE Policy Rate Surprise Chart 27Change In 10-Year Gilts Yield##BR##Vs. 12-Month BoE Policy Rate Surprise Chart 28Change In 30-Year Gilts Yield##BR##Vs. 12-Month BoE Policy Rate Surprise Appendix E: Japan Chart 29Change In 1-Year Japanese JGB Yield##BR##Vs. 12-Month BoJ Policy Rate Surprise Chart 30Change In 2-Year Japanese JGB Yield##BR##Vs. 12-Month BoJ Policy Rate Surprise Chart 31Change In 3-Year Japanese JGB Yield##BR##Vs. 12-Month BoJ Policy Rate Surprise Chart 32Change In 5-Year Japanese JGB Yield##BR##Vs. 12-Month BoJ Policy Rate Surprise Chart 33Change In 7-Year Japanese JGB Yield##BR##Vs. 12-Month BoJ Policy Rate Surprise Chart 34Change In 10-Year Japanese JGB Yield##BR##Vs. 12-Month BoJ Policy Rate Surprise Chart 35Change In 30-Year Japanese JGB Yield##BR##Vs. 12-Month BoJ Policy Rate Surprise Appendix F: Canada Chart 36Change In 1-Year Canadian Yield##BR##Vs. 12-Month BoC Policy Rate Surprise Chart 37Change In 2-Year Canadian Yield##BR##Vs. 12-Month BoC Policy Rate Surprise Chart 38Change In 3-Year Canadian Yield##BR##Vs. 12-Month BoC Policy Rate Surprise Chart 39Change In 5-Year Canadian Yield##BR##Vs. 12-Month BoC Policy Rate Surprise Chart 40Change In 7-Year Canadian Yield##BR##Vs. 12-Month BoC Policy Rate Surprise Chart 41Change In 10-Year Canadian Yield##BR##Vs. 12-Month BoC Policy Rate Surprise Chart 42Change In 30-Year Canadian Yield##BR##Vs. 12-Month BoC Policy Rate Surprise Appendix G: Australia Chart 43Change In 1-Year Aussie Yield##BR##Vs. 12-Month RBA Policy Rate Surprise Chart 44Change In 2-Year Aussie Yield##BR##Vs. 12-Month RBA Policy Rate Surprise Chart 45Change In 3-Year Aussie Yield##BR##Vs. 12-Month RBA Policy Rate Surprise Chart 46Change In 5-Year Aussie Yield##BR##Vs. 12-Month RBA Policy Rate Surprise Chart 47Change In 7-Year Aussie Yield##BR##Vs. 12-Month RBA Policy Rate Surprise Chart 48Change In 10-Year Aussie Yield##BR##Vs. 12-Month RBA Policy Rate Surprise Appendix H: New Zealand Chart 49Change In 1-Year NZ Yield##BR##Vs. 12-Month RBNZ Policy Rate Surprise Chart 50Change In 2-Year NZ Yield##BR##Vs. 12-Month RBNZ Policy Rate Surprise Chart 51Change In 3-Year NZ Yield##BR##Vs. 12-Month RBNZ Policy Rate Surprise Chart 52Change In 5-Year NZ Yield##BR##Vs. 12-Month RBNZ Policy Rate Surprise Chart 53Change In 7-Year NZ Yield##BR##Vs. 12-Month RBNZ Policy Rate Surprise Chart 54Change In 10-Year NZ Yield##BR##Vs. 12-Month RBNZ Policy Rate Surprise

Highlights Recent estimates by ship trackers put the loss of Iranian exports at close to 1mm b/d as of mid-September vs April levels. This loss is higher (and sooner) than our previous baseline expectation, and prompts us to raise our estimate of lost Iranian oil exports to 1.25mm b/d by November, when U.S. sanctions kick in. Venezuela still is close to collapse, but may avoid a complete meltdown with Chinese companies stepping in to safeguard the $50 billion loaned to the country's oil industry.1 We expect production to fall below 1mm b/d next year - to less than half its end-2016 level. With Fed policy likely to continue tightening into 2019 as oil prices surge, the odds of an equity bear market and recession arriving in 2H19 - vs our 2H20 House view - also increase. Our dominant scenario now includes a supply shock and higher prices in 1Q19, which is followed by a U.S. SPR release and price-induced demand destruction (Chart of the Week). As a result, we are raising the odds of Brent prices reaching or exceeding $100/bbl by as early as 1Q19, and lifting our 2019 forecast to $95/bbl. Energy: Overweight. U.S. refining capacity utilization remains close to 19-year highs. At 97.1% of operable capacity, it is within a whisker of the four-week-moving-average highs of 97.3% recorded in August, driven by strong product demand ex U.S. Base Metals: Neutral. The U.S. Treasury granted permission to Rusal's existing customers to continue signing new contracts with the aluminum producer. The announcement stopped short of a full removal of sanctions, which are set to come into effect on October 23. Precious Metals: Neutral. The strong trade-weighted USD continues to hold gold prices on either side of $1,200/oz. We remain long as a portfolio hedge. Ags/Softs: Underweight. The USDA's Crop Production report forecasts record yields for corn and soybeans - 181.3 and 52.8 bushels/acre, respectively - which continues to weigh on prices. The bean harvest is expected to be a record. Feature Chart of the WeekBCA Ensemble Forecast Lifts Brent To $95/bbl, As Market Tightens With the loss of Iranian exports occurring faster and sooner than expected, and Venezuela remaining on the brink of collapse, senior energy officials from the U.S., Russia and the Kingdom of Saudi Arabia (KSA) are going to great lengths to reassure their domestic consumers everything - particularly on the supply side - is under control. We are inclined to believe their comfort level re global oil supply is inversely proportional to the amount of reassurance they provide their domestic audiences. The more they meet and talk - particularly to the media - the more concerned they are. And right now, they're pretty concerned. Rick Perry, the U.S. Energy Secretary, held a presser in Moscow following his meeting with Alexander Novak, Russia's Energy Minister, saying the U.S., KSA and Russia can lift output over the next 18 months to compensate for the loss of exports from Iran, Venezuela, and other unplanned outages.2 That might be true, but the market's already tightening far faster and far sooner than many analysts expected. Covering a supply shortfall in 18 months does nothing for the market over the next few months, particularly with demand remaining robust (Chart 2) and OECD inventories falling (Chart 3). Since 2017, our factor model shows Brent prices have been supported by two factors acting simultaneously together: Chart 2Fundamentals Support Strong Prices Chart 3Inventory Draws Will Accelerate Strong compliance of OPEC 2.0 members to the coalition's production-cutting agreement, which reduced the OPEC Supply-and-Inventory factor's role, and The pickup in global oil demand, particularly in EM economies, which pushed our Global Demand factor up. These effects were partly counterbalanced by the rise in our non-OPEC Supply factor - driven by strong growth in U.S. shale-oil output - which became the largest negative contributor to price movements. Global demand's been strengthening since the end of 1H17 on the back of stellar EM income growth. This remains the fundamental backdrop to global oil for now. While our base case remains relatively supportive for oil prices, we are raising the odds of a price spike resulting from a supply shock as early as 1Q19 on the back of larger- and faster-than-expected Iranian export losses, and continued declines in Venezuelan production. Should this occur, we believe it would trigger a U.S. SPR release, and produce demand destruction at a rate that could be faster than historical experience would suggest (Table 1). This further tightens balances, and leads us to raise our 2019 forecast for Brent crude oil to $95/bbl on average, up from $80/bbl last month, with WTI trading $6/bbl below that (Chart 4). This forecast is highly conditional, given our assumptions re supply-side variables, a U.S. SPR release, and demand destruction estimates. Table 1BCA Global Oil Supply - Demand Balances (MMb/d) (Base Case Balances) Chart 4BCA's Oil Balances Tighter Oil Balances Tighten As Supply Contracts In our monthly balances update, we are incorporating a sharply accelerated loss of Iranian export barrels to the market, which already is evident. Bloomberg this week reported its tanker-tracking service registered a decline in Iranian exports of close to 1mm b/d between April, when sanctions were announced, and mid-September.3 At this rate, the assessment by Platts Analytics last week that as much as 1.4mm b/d of Iranian exports could be lost by the time U.S. sanctions kick in November 4 appears more likely.4 However, to be conservative, we are building in a loss of 1.25mm b/d in our balances, and have it developing over the July - November period in increments of 250k b/d, instead of the November - February interval we assumed in last month's balances. We will monitor this situation and revise our estimates as new information arrives. Also on the supply side, we are assuming the U.S. SPR releases 500k b/d starting a month after Brent prices go over $90/bbl in March 2019. This is in line with the SPR's enabling legislation, which limits drawdowns to 30mm b/d over a 60-day period, after the President authorizes such action to meet a severe energy supply interruption. Lastly, we continue to carry supply constraints arising from the lack of sufficient take-away capacity to get all of the crude produced in the Permian Basin to refining markets in our models. To wit: We continue to expect 1.2mm b/d of supply growth from the U.S. shales, driven largely by Permian production, vs an earlier expectation of 1.4mm b/d of growth. We expect the Permian to be de-bottlenecked by 4Q19. We expect the Big 3 producers Secretary Perry expects to fill supply gaps in 18 months - the U.S., Russia, and KSA - to produce 10.83mm, 11.4mm and 10.4mm b/d in 2H18, and 11.79mm, 11.43mm and 10.4mm b/d next year, respectively. They will get some help from OPEC's Gulf Arab producers - i.e., the core OPEC producers (Chart 5) - but, supply will continue to fall/stagnate in most of the rest of the world, particularly in offshore producers (Chart 6). Chart 5While Core OPEC Can Increase Supply... Chart 6... 'The Other Guys' Output Stagnates We also note the EIA and IEA have lowered their supply-growth estimates this month. The EIA this month reduced expected U.S. crude production growth by 210k b/d in 2019, and the IEA lowered its estimate of offshore production growth in Brazil from 260k b/d to just 30k b/d this year. These are non-trivial adjustments in a market that was tight prior to the downgrade in supply growth. Still, there are significant marginal disagreements on the supply side among the major data supporters (the EIA, IEA and OPEC), which can be seen in Table 2. Table 2Comparison Of Major Balances Estimates Demand Destruction Likely As Prices Spike In 1Q19 We expect the rate of growth in EM incomes and trade - a proxy for income - to slow slightly this year vs 2017, on the back of a strengthening USD. This will reduce the rate of growth in EM imports and the rate of growth in EM commodity demand, at the margin. However, y/y growth in EM incomes is expected to remain positive over the next 12 months in our baseline scenario, which will keep the level of commodity demand - particularly for oil and industrial metals - robust. This will drive global demand growth of ~ 1.6mm b/d this year, roughly unchanged from last month. Higher prices risk slowing next year's growth. This is where it gets tricky. An oil-supply shock occurring when global demand is strong most likely will produce a price spike, as we've been arguing for the past several weeks.5 This price spike, coupled with continued monetary-policy tightening by the Fed, raises the likelihood of demand destruction globally. Higher oil prices and a stronger USD act as a double-whammy on EM oil demand. The problem we have now is gauging the elasticity of oil demand, particularly in EM. Oil markets are fundamentally different now than at any point in the modern era - roughly beginning in the early 1960s with the ascendance of OPEC - because many big oil-importing EM economies removed or relaxed subsidies following the prices collapse of 2014 - 2016. Prominent among these states are China and India. OPEC states also took advantage of the price collapse to relax or remove subsidies, e.g., KSA.6 The price shock we anticipate, therefore, will be the first in the modern era in which EM consumers - the principal driver of oil demand in the world, accounting for roughly 70% of the demand growth we expect - are exposed directly to higher prices. How quickly they will respond to higher prices is unknown. For this reason, we're introducing what we consider a reasonable first approximation of how EM demand might respond to higher prices and a stronger USD into the scenarios we include in our ensemble forecast (Chart 7). As a first approximation - subject to at least monthly adjustment, as more data become available - we are modeling a 100k b/d loss of demand for every $10/bbl increase in crude oil prices.7 We will continue to iterate on this as new information becomes available. Chart 7Ensemble Scenarios Reflect New Risks Bottom Line: We've raised the odds of a supply shock in the oil markets that takes Brent prices to or through $100/bbl by 1Q19. Should this occur, we expect it will be met by a U.S. SPR release of 500k b/d a month after prices breach $90/bbl. This price spike will set off a round of demand destruction, which we expect will be quicker than history would suggest, given many large EM oil-consuming states have relaxed or eliminated fuel subsidies, leaving their consumers exposed to the price shock. This will be exacerbated by a stronger USD going forward, as the Fed likely looks through the price spike and continues with its policy-rate normalization. In this scenario, a U.S. recession could arrive in 2H19 vs our House view of 2H20 or later. In addition, we would expect an equity bear market to ensue sooner than presently anticipated. We recommend using Brent call spreads to express the view consistent with our research. At tonight's close, we will go long April, May and June 2019 calls struck at $85/bbl and short $90/bbl calls. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com Hugo Bélanger, Senior Analyst Commodity & Energy Strategy HugoB@bcaresearch.com 1 Please see "Venezuela hands China more oil presence, but no mention of new funds," published by reuters.com September 14, 2018. 2 U.S. Energy Secretary Rick Perry made this claim at a press conference after meeting with Russian Energy Minister Alexander Novak last Friday. Please see "Big Three oil states can offset fall in Iran supplies: Perry," published by reuters.com September 14, 2018. 3 Please see "Saudi Arabia Is Comfortable With Brent Oil Above $80," published by bloomberg.com September 18, 2018. 4 Please see "OPEC crude oil production rises to 32.89 mil b/d in Aug as cuts unwind: Platts survey" published by SP Platts Global September 6, 2018. 5 Please see "Oil-Supply Shock, Rising U.S. Rates Favor Gold As A Portfolio Hedge," published by BCA Research's Commodity & Energy Strategy Weekly Report on September 13, 2018. It is available at ces.bcaresearch.com. For a discussion of the effect of a stronger USD on global oil demand, please see "Trade, Dollars, Oil & Metals ... Assessing Downside Risk," published by the Commodity & Energy Strategy August 23, 2018. 6 Please see the Special Focus in the World Bank's January 2018 Global Economic Prospects entitled "With The Benefit of Hindsight: The Impact of the 2014 - 16 Oil Price Collapse," beginning on p. 49. 7 In this simulation, we employ an iterative one-step-ahead forecasting methodology that reduces demand by 100k b/d for every $10/bbl increase in prices. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed in 2017 Summary of Trades Closed in 2018

Highlights In this Weekly Report, we present our semi-annual chartbook of the BCA Central Bank Monitors. The message now conveyed by the Monitors is that divergences between the cyclical pressures faced by the individual central banks are growing larger. This is occurring within some countries, where the growth and inflation indicators are trending in opposite directions. This is also visible across countries, with not every Monitor calling for rate hikes - a significant shift from the coordinated backdrop seen in 2017 (Chart of the Week). Chart of the WeekFrom Convergence To Divergence In The BCA Central Bank Monitors The combined message from the Monitors is that the slower pace of global growth seen in 2018 has not been enough put a serious dent in inflation pressures stemming from a dearth of spare capacity in most major countries. Perhaps that changes if a full-blown U.S.-China trade war develops, or if the tensions in emerging markets spill over more broadly into global financial conditions, but that remains to be seen. Add it all up, and a below-benchmark stance on overall global duration exposure remains appropriate. Feature An Overview Of The BCA Central Bank Monitors Chart 2CB Monitor Divergence = Bond Yield Divergence The BCA Central Bank Monitors are composite indicators designed to measure the cyclical growth and inflation pressures that can influence future monetary policy decisions. The economic data series used to construct the Monitors are not the same for every country, but the list of indicators generally measure the same things (i.e. manufacturing cycles, domestic demand strength, commodity prices, labor market conditions, exchange rates, etc). The data series are standardized and combined to form the Monitors. Readings above the zero line for each Monitor indicate pressures for central banks to raise interest rates, and vice versa. Through the nexus between growth, inflation, and market expectations of future interest rate changes, the Monitors do exhibit broad correlations to government bond yields in the Developed Markets (Chart 2). Our current recommended country allocation for global government bonds reflects the trends seen in the Central Bank Monitors - underweighting countries were the Monitors are most elevated (the U.S., Canada) in favor of regions where the Monitors are lower (Australia, Japan, euro area, New Zealand). In each BCA Central Bank Monitor Chartbook, we include a new chart for each country that we have not shown previously. In this edition, we show the Monitors plotted against the relative returns for each country versus the overall Bloomberg Barclays Global Treasury index (shown inversely in the charts so that a rising line means underperformance versus the benchmark index). Fed Monitor: Still On A Gradual Rate Hike Path Our Fed Monitor remains in the "tight money required" zone, signalling that the cyclical backdrop justifies additional Fed rate hikes (Chart 3A). Resilient U.S. growth, a dearth of spare capacity and an acceleration of both wage growth and core inflation are all consistent with a U.S. economy starting to overheat and requiring tighter monetary policy (Chart 3B). Chart 3AU.S.: Fed Monitor Chart 3BU.S. Inflation On The Rise The growth and inflation components of the Fed Monitor have both accelerated since our last Central Bank Monitor Chartbook was published back in April. Most notably, the inflation component has blasted through the zero line to the highest level since 2008 (Chart 3C). The financial conditions component has retreated from very elevated (i.e. growth-supportive) levels, mostly due to the stronger U.S. dollar but also because of wider corporate credit spreads seen since the start of 2018. Importantly, the financial conditions component has not tightened enough to offset the impact on the Monitor from faster growth and inflation. Chart 3CAll Fed Monitor Components Now Above Zero Recent comments from senior Fed officials (Chair Jay Powell and Governor Lael Brainard) have indicated that the Fed is less confident in its own estimates of the full-employment NAIRU or the appropriate neutral level of the funds rate. Our read on this is that the Fed will instead continue to raise the funds rate at a gradual 25bp per quarter pace until there are signs that U.S. monetary policy has become tight (i.e. an inverted yield curve, wider credit spreads, softer U.S. economic data). Until then, the message sent by the Fed Monitor is to remain underweight U.S. Treasuries with below-benchmark duration, as market pricing of expectations for both the funds rate and inflation remain too low (Chart 3D). Chart 3DU.S. Treasury Underperformance Will Continue - Stay Underweight BoE Monitor: Brexit Uncertainty Trumps Inflation Pressures The BoE Monitor remains in the "tighter money required" zone as it has since late 2015 (Chart 4A). Despite that persistent signal, the BoE has kept monetary policy at highly accommodative levels, only raising the base rate 50bps over the past year. The BoE Monetary Policy Committee remains torn between signs that inflation risks are tilted to the upside and the downside risks to U.K. growth from an uncertain Brexit outcome. The U.K. unemployment rate is well below NAIRU with an output gap that is now estimated to be closed (Chart 4B). Yet realized inflation has peaked, with core inflation drifting back below 2%. Wages are finally starting to grow in real terms, which the BoE cites as an important factor underpinning consumer spending, but the pace remains modest. Chart 4AU.K.: BoE Monitor Chart 4BNo Spare Capacity, Yet Has Inflation Peaked? Looking at the breakdown of our BoE Monitor, both the growth and inflation sub-components of the indicator have recently reaccelerated (Chart 4C). Yet U.K. leading economic indicators continue to decline and dampened business confidence measures reflect the heightened uncertainty over the future relationship between the U.K. and the European Union. Chart 4CBoth Growth & Inflation Components Are Boosting The BoE Monitor The performance of U.K. Gilts has diverged from the Monitor since the 2016 Brexit vote (Chart 4D), as the BoE has been more worried about Brexit than inflation and has stayed accommodative. Stay overweight U.K. Gilts within global government bond portfolios, even with the more bearish signal implied by our BoE Monitor, given the weakening trend in leading economic indicators and persistent Brexit uncertainty. Chart 4DBrexit Uncertainty Preventing More BoE Hikes - Stay Overweight Gilts ECB Monitor: No Pressure To Hike Rates Quickly Post-QE Our European Central Bank (ECB) Monitor has fallen sharply since we last published this Chartbook back in April, and it now sits below the zero line (Chart 5A). The growth deceleration in the first half of the year from the rapid pace seen in 2017 is the main reason for this move, as inflation pressures have not subsided (Chart 5B). Chart 5AEuro Area: ECB Monitor Chart 5BEuro Area At Full Capacity ECB President Mario Draghi noted last week that the plan remains in place to end the net new buying phase of the ECB's Asset Purchase Program at the end of 2018. Policymakers' have grown more confident that their inflation forecasts will be met as most measures of euro area wage growth (and headline inflation) have accelerated to 2% over the past year. It remains to be seen if those expectations are too optimistic, as the growth component of our ECB Monitor remains well below the zero line, while the inflation component is no longer rising (Chart 5C). Chart 5CGrowth Component Dragging Down The ECB Monitor For now, we recommend a neutral stance on core euro area government bonds with an underweight posture on Peripheral sovereign debt as a way to manage these conflicting trends. The overall performance of euro area bonds versus global benchmarks has followed the pace of the ECB's bond-buying since 2015, and not the pressures suggested by our ECB Monitor (Chart 5D), suggesting a bearish stance as the bond buying ends. Yet from a more bullish perspective, the mixed message on growth and lack of immediate pressures on core inflation (still at 1%) imply that the ECB will not deviate from its current dovish forward guidance of no interest rate hikes until at least September 2019. Chart 5DECB Will Not Hike Rates Quickly After QE Ends - Stay Neutral Core European Bonds BoJ Monitor: Too Soon To Consider Policy Changes Our Bank of Japan (BoJ) Monitor has stayed just barely in the "tighter money required" zone since last October, due mostly to growing inflation pressures (Chart 6A). Yet with the Japanese labor market now as tight as it has been in decades, headline and core CPI inflation are only at 0.9% and 0.3% respectively, well below the BoJ's 2% target (Chart 6B). Chart 6AJapan: BoJ Monitor Chart 6BInflation Pressures Slowly Building In Japan Japanese firms appear to finally be reacting to the tightness of the labor market, however, as wage growth has accelerated in recent months. The pick-up in wages has helped boost inflation expectations, both of which are part of the inflation component of the BoJ Monitor that is now at the highest level since 2008 (Chart 6C). However, the growth component just rolled over and now sits at the zero line, as the Japanese economy has lost some momentum. Chart 6CInflation Boosting BoJ Monitor We continue to recommend an overweight stance on JGBs, based on our view that the BoJ will maintain hyper-easy monetary policy settings - especially compared to the rest of the developed markets - until there is much higher realized inflation in Japan. JGBs have indeed been outperforming over the past year, even with the less dovish signal sent by the BoJ Monitor (Chart 6D). Yet the absolute level of the Monitor remains around zero, suggesting that no policy changes should be expected. That means no upward adjustment of the BoJ's 0% yield target on 10-year JGBs or major further reductions in the annual pace of BoJ JGB buying (even though the central bank is hitting capacity constraints as it now owns close to ½ of all outstanding JGBs). Chart 6DBoJ In No Hurry To Turn Hawkish - Stay Overweight JGBs BoC Monitor: Rate Hikes - More To Come The Bank of Canada (BoC) Monitor has stayed in "tighter money required" since the beginning of 2017 and is now well above the zero line (Chart 7A). The BoC has been following our BoC Monitor, hiking rates by a cumulative 100bps since July 2017. Chart 7ACanada: BoC Monitor Chart 7BAn Overheating Canadian Economy? The BoC has been responding to the growing inflation pressure in Canada. There is no evidence that spare economic capacity exists, while realized inflation is near the upper bound of BoC's target range of 1-3% (Chart 7B). There is a growing divergence between the growth and inflation subcomponents of the BoC Monitor, with the latter decelerating over the past several months. That was due to a combination of slowing Chinese import demand and the imposition of trade tariffs on Canada by the Trump administration (Chart 7C). Yet the domestic economy remains in good shape, with the overall indicator from the BoC's Business Outlook Survey at the highest level since 2010. Chart 7CInflation Component Boosting BoC Monitor We continue to recommend an underweight stance on Canadian government bonds, as the relative performance has broadly followed the path of the BoC Monitor over the past three years (Chart 7D). The BoC tends to follow the policy actions of the Fed with a short lag, thus our bearishness on Canadian government bonds is related to our more hawkish views on the Fed. Yet the surge in Canadian inflation, at a time when the economy has no spare capacity, suggests that there are good domestic reasons to expect more rate hikes from the BoC over the next year than what is currently discounted by markets. Chart 7DBoC Not Done Yet - Stay Underweight Canadian Bonds RBA Monitor: Easier Policy Needed The Reserve Bank of Australia (RBA) monitor has rapidly fallen below the zero line for the first time since 2016, and now indicates that easier monetary policy is required (Chart 8A). This stands out from the more stable trajectory of the rest of the BCA Central Bank Monitors. Unlike most other developed countries, there is still excess capacity in the Australian economy. Australia's output gap has not closed while the current unemployment rate is just at the OECD's NAIRU estimate of 5.3%. Headline and core inflation are at the low end of the RBA's 2-3% target and struggling to gain much upward momentum (Chart 8B). Chart 8AAustralia: RBA Monitor Chart 8BMinimal Inflation Pressure In Australia While both the growth and inflation components of the RBA Monitor have fallen, the biggest decline has come from the inflation side (Chart 8C). The sluggishness of Australia's economy is due to the slow growth of consumer spending and a big deceleration in exports related to softer Chinese demand. On inflation, excess labor market slack, with an underemployment rate close to 8.5%, is the main factor explaining soft wage growth and overall sluggish inflation. Chart 8CInflation Component Weighing On RBA Monitor Our highest conviction country allocation call this year has been to overweight Australian Government bonds, and we see no need to change that given the bullish signal from our RBA Monitor (Chart 8D). It would likely take a rise in unemployment, a renewed decline in realized inflation or a big external shock for the RBA to actually cut rates as our Monitor suggests, but the signal is still bullish for Australian debt on a relative basis. Chart 8DRBA A Long Way From A Hike - Stay Overweight Australian Government Bonds RBNZ Monitor: Policy On Hold For A While Longer The Reserve Bank of New Zealand (RBNZ) Monitor is currently just above the zero line, indicating that tighter monetary policy is required (although just barely) (Chart 9A). This is consistent with the mixed messages in the New Zealand economic data. For example, there is no spare capacity in the economy according to estimates of the output and employment gaps, yet both headline and core inflation have decelerated to the lower end of the RBNZ's 1-3% target band (Chart 9B). Chart 9ANew Zealand: RBNZ Monitor Chart 9BNo Spare Capacity In NZ, But No Inflation Either Looking at the components of the RBNZ Monitor, the growth factors have continued to plunge whereas the inflation factors have been increasing (from below zero) since the start of 2018 (Chart 9C). New Zealand's economic growth has slowed because of softer consumer spending and weaker housing activity, the latter of which is related to lower net immigration. Yet business confidence is falling, both the manufacturing and services PMIs have also declined, and export growth has cooled thanks to weaker growth from China and Australia. Meanwhile, the uptick in the inflation components has not yet translated into any broader improvement in realized inflation that would cause the RBNZ to take a more hawkish turn. Chart 9CConflicting Trends Within The RBNZ Monitor We continue to recommend an overweight stance on New Zealand Government Bonds, in line with the bullish signal sent by our RBNZ Monitor (Chart 9D). The RBNZ has already provided forward guidance indicating that the Overnight Cash Rate (OCR) will stay unchanged until 2020, and it will take some time before there is evidence that the recent hook down in inflation is nothing more than a temporary blip. Chart 9DRBNZ To Remain On Hold - Stay Long New Zealand Bonds Robert Robis, CFA, Senior Vice President Global Fixed Income Strategy rrobis@bcaresearch.com Ray Park, CFA, Research Analyst ray@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

Special Report

Following up from our inaugural U.S. Equity Market Indicators Report in early-August 2017, this week we introduce the second part in our Indicators series. In this Special Report we have drilled down to the ten GICS1 S&P 500 sectors (excluding the real estate sector) and have compiled the most important Indicators in four broad categories: earnings, financial statement reported, valuations and technicals. Once again this is by no means exhaustive, but contains a plethora of Indicators - roughly thirty Indicators per sector condensed in seven charts per sector - we deem significant in aiding us in our decision making process of setting/changing a view on a certain sector. The way we have structured this Special Report is by sector and we start with the early cyclicals continue with the deep cyclicals and finish with the defensives. Within each sector we then show the four broad categories. In more detail, the first three charts depict earnings Indicators including our EPS growth model, EPS breadth, profit margins, relative forward EPS and EBITDA growth forecasts and ROE and its deconstruction into its components. The following two charts relate to financial statement Indicators including indebtedness, cash flow growth and capital expenditures. And conclude with one valuation and one technical chart. As a reminder, the charts in this Special Report are also made available through BCA's Analytics platform for seamless continual updates. Due to length constraints, Part III of our Indicators series, expected in mid-October, will introduce a style and size flavor along with cyclicals versus defensives and end with the S&P 500, again highlighting Indicators in these four broad categories. Finally, likely before the end of 2018, we aim to conclude our Indicators series with Part IV that would feature our most sought after Macro Indicators per the ten GICS1 S&P 500 sectors, along with value/growth, small/large and cyclicals/defensives. We trust you will find this comprehensive Indicator chartbook useful and insightful. Anastasios Avgeriou, Vice President U.S. Equity Strategy anastasios@bcaresearch.com Dulce Cruz, Senior Analyst dulce@bcaresearch.com Consumer Discretionary Chart 1Consumer Discretionary: Earnings Indicators Chart 2Consumer Discretionary: Earnings Indicators Chart 3Consumer Discretionary: ROE And Its Components Chart 4Consumer Discretionary: Financial Statement Indicators Chart 5Consumer Discretionary: Financial Statement Indicators Chart 6Consumer Discretionary: Valuation Indicators Chart 7Consumer Discretionary: Technical Indicators Financials Chart 8Financials: Earnings Indicators Chart 9Financials: Earnings Indicators Chart 10Financials: ROE And Its Components Chart 11Financials: Financial Statement Indicators Chart 12Financials: Financial Statement Indicators Chart 13Financials: Valuation Indicators Chart 14Financials: Technical Indicators Energy Chart 15Energy: Earnings Indicators Chart 16Energy: Earnings Indicators Chart 17Energy: ROE And Its Components Chart 18Energy: Financial Statement Indicators Chart 19Energy: Financial Statement Indicators Chart 20Energy: Valuation Indicators Chart 21Energy: Technical Indicators Industrials Chart 22Industrials: Earnings Indicators Chart 23Industrials: Earnings Indicators Chart 24Industrials: ROE And Its Components Chart 25Industrials: Financial Statement Indicators Chart 26Industrials: Financial Statement Indicators Chart 27S&P Industrials: Valuation Indicators Chart 28S&P Industrials: Technical Indicators Materials Chart 29Materials: Earnings Indicators Chart 30Materials: Earnings Indicators Chart 31Materials: ROE And Its Components Chart 32Materials: Financial Statement Indicators Chart 33Materials: Financial Statement Indicators Chart 34Materials: Valuation Indicators Chart 35Materials: Technical Indicators Tech Chart 36Technology: Earnings Indicators Chart 37Technology: Earnings Indicators Chart 38ROE And Its Components Chart 39Technology: Financial Statement Indicators Chart 40Technology: Financial Statement Indicators Chart 41Technology: Valuation Indicators Chart 42Technology: Technical Indicators Health Care Chart 43Health Care: Earnings Indicators Chart 44Health Care: Earnings Indicators Chart 45Health Care: ROE And Its Components Chart 46Health Care: Financial Statement Indicators Chart 47Health Care: Financial Statement Indicators Chart 48Health Care: Valuation Indicators Chart 49Health Care: Technical Indicators Consumer Staples Chart 50Consumer Staples: Earnings Indicators Chart 51Consumer Staples: Earnings Indicators Chart 52Consumer Staples: ROE And Its Components Chart 53Consumer Staples: Financial Statement Indicators Chart 54Consumer Staples: Financial Statement Indicators Chart 55Consumer Staples: Valuation Indicators Chart 56Consumer Staples: Technical Indicators Telecom Services Chart 57Telecom Services: Earnings Indicators Chart 58Telecom Services: Earnings Indicators Chart 59Telecom Services: ROE And Its Components Chart 60Telecom Services: Financial Statement Indicators Chart 61Telecom Services: Financial Statement Indicators Chart 62Telecom Services: Valuation Indicators Chart 63Telecom Services: Technical Indicators Utilities Chart 64Utilities: Earnings Indicators Chart 65Utilities: Earnings Indicators Chart 66Utilities: ROE And Its Components Chart 67Utilities: Financial Statement Indicators Chart 68Utilities: Financial Statement Indicators Chart 69Utilities: Valuation Indicator Chart 70Utilities: Technical Indicator

Special Report

Highlights Cable is cheap on a PPP basis. However, the discount does not reflect a geopolitical risk premium; it reflects the dollar's general expensiveness. In fact, when the British productivity picture is taken into account, the trade-weighted pound's discount appears rather modest. Our model specifically designed to capture the geopolitical risk premia in GBP/USD and EUR/GBP shows that investors are currently pricing in a very rosy political outlook in the U.K., and near certainty that a soft Brexit will materialize. We are not willing to bet that the path toward a soft Brexit will be easy. As a result, we would expect that if the GBP experiences any rebounds, they will prove short-lived, especially as the outlook for global growth outside the U.S. remains murky. Feature This fall will be a tumultuous time for the pound, as the Brexit process goes into full swing ahead of March 2019. While there remain many possible paths that the U.K.'s relationship with the rest of the EU could ultimately take, ranging from a complete reset of the relationship (i.e. a hard Brexit) to no Brexit at all, another unknown needs to be tackled: Is the GBP priced to adequately compensate investors for such heightened uncertainty? In this week's piece, we develop a simple model to try to ascertain whether geopolitical risk premium is currently present in the pound. We conclude that even though the pound seems cheap enough to compensate investors for the high degree of uncertainty surrounding the U.K.'s long-term economic outlook, this picture is deceiving. As a result, BCA remains concerned about the pound's cyclical outlook, especially against the euro. Is The Pound That Cheap? At first glance, it seems obvious that the pound is very cheap. Cable currently trades at a prodigious 20% discount to it purchasing power parity (PPP) estimate (Chart I-1). Such bargain-basement levels must be a reflection of the economic risks surrounding Brexit. Well, perhaps not. First, the pound may be trading at a large discount against the dollar, but the euro also trades well below its PPP fair value. In fact, when using PPP models, it is hard to dissociate the cheapness of the pound from the expensiveness of the U.S. dollar against its trading partners (Chart I-2). Thus, PPP models are not enough to gauge whether or not the pound is adequately compensating investors for inherent geopolitical risk. Chart I-1Is The Pound Cheap... Chart I-2U.S. Dollar And PPP ...Or Is The Dollar Expensive? Second, when one uses a slightly more sophisticated valuation approach, the discount of the pound seems much more muted than when one looks at PPP alone. Based on our proprietary long-term fair value model, the trade-weighted pound exhibits a much more muted discount of only 3% - well within the historical norm (Chart I-3). Chart I-3Pound: A Much Smaller Discount On A Trade-Weighted Basis What explains this disconnect is the U.K.'s poor productivity performance. In the world of exchange rate determination, there is a phenomenon called the Penn effect. It is an empirical observation - one not fully understood under a theoretical lens1 - which shows that countries with higher levels of productivity growth than their trading partners tend to experience an appreciation in their real exchange rates. As Chart I-4 illustrates, the U.K. is on the wrong side of this phenomenon, as its relative productivity has been falling in comparison to its trading partners. This factor has played an important role in dragging down the pound's fair value. This poor productivity performance has also had another pernicious effect: unit labor costs in the U.K. have risen much more sharply than in the majority of its important trading partners (Chart I-5). This hurts the pound's competitiveness and suggests that a simple PPP model based purely on producer prices might be missing the mark for the true fair value of the British currency - further supporting the message of our proprietary long-term valuation model. Chart I-4Negative Penn Effect For The Pound Chart I-5The U.K. Is Uncompetitive Even when these adjustments are taken into account, our model might still be missing the mark due to a very significant problem: All fair value models for the pound are now based on a potentially unrepresentative sample, one where the U.K. was part of the EU. Thus, another exercise is needed to evaluate the pound's geopolitical risk premium. Bottom Line: Based on simple PPP models, cable looks cheap and therefore may already embed a large geopolitical risk premium. However, this conclusion is misleading. A large share of the pound's undervaluation is not GBP-specific and instead simply mirrors the USD's premium to its fair value. Additionally, the U.K.'s poor productivity performance relative to its trading partners already provides an economic justification for a cheap pound. Thus, we need a different exercise to zero in on the degree of geopolitical discount present in the pound. Zeroing In On The Geopolitical Risk In order to assess the degree of political risk priced into the pound, one needs to isolate this risk. The problem is that the traditional variables used to explain exchange rate movements were also greatly affected by the shock following the Brexit vote in June 2016. For example, looking at the behavior of British gilts, the Footsie, consumer confidence or business confidence, one can see very abnormal moves occurring in conjunction with large fluctuations in the pound during the summer of 2016 (Chart I-6). Thus, if one were to regress the pound on these variables, one would not have observed a risk premium, even though the market was clearly very concerned with the geopolitical outlook for the U.K. Chart I-6ATraditional Variables Are Of Little Use To Isolate A Geopolitical Risk Premium (I) Chart I-6BTraditional Variables Are Of Little Use To Isolate A Geopolitical Risk Premium (II) We therefore decided to try to explain the pound's normal behavior using variables that did not experience large abnormal moves in the direct aftermath of the British referendum. Moreover, we wanted to keep the model simple, as simplicity permits us to better understand the pound's deviation from its predicted value. Practically, we settled on the following specification: for GBP/USD, we regressed the pair versus the dollar index and the British leading economic indicator. For EUR/GBP, we regressed the cross against the trade-weighted euro and the U.K. LEI. The reason for using the trade-weighted dollar and euro as explanatory variables is simple: it helps us isolate the pound's movements from the impact of fluctuations in the other leg of the pair. Using the U.K. LEI helps incorporate the immediate outlooks for U.K. growth and U.K. monetary policy into the pound's movement. The remaining error term was mostly a reflection of geopolitical risks. To make sure the exercise was robust, we then tested the out-of-sample performance of the model. Reassuringly, the GBP/USD and EUR/GBP models showed great predictive power out-of-sample (see Appendix), while remaining significant and explaining 80% and 65% of the pairs' variations, respectively. The results of the models are shown in Chart I-7, and they are startling. While the pound did show a geopolitical discount in the second half of 2016 (as evidenced by the abnormally large discount from our fundamental-based model), today the pound's pricing shows an absence of geopolitical risk premium, both against the dollar and against the euro. This corroborates the message from the uncertainty index computed by Baker Bloom and Davis, which shows a very low level of economic policy uncertainty based on language in the press (Chart I-8). Chart I-7ALittle Risk Premium In The Pound (I) Chart I-7BLittle Risk Premium In The Pound (II) This is particularly salient when compared to the euro, where the geopolitical risk premium is currently exaggerated. As Chart I-9 illustrates, the probability of a euro area breakup in the next five years priced into the bond market is at its highest level since the heyday of the euro area crisis in 2011 and 2012. However, this risk is currently overstated as investors have been frightened by the recent Italian elections. Yet, after a tumultuous beginning, the populist Five Star Movement / Lega Nord coalition is backing away from a budget confrontation with Brussels. Giovanni Tria, Italy's minister of finance, wants a 2% budget deficit while Deputy Prime Minister Matteo Salvini is arguing for a 2.9% budget hole - well south of the 6% levels touted during the campaign. Italians realize that life outside the euro area will not be a land of milk and honey. Chart I-8British Political Uncertainty Has Collapsed Chart I-9Investors Are Worries About The Euro Area Instead, the pound's cheapness reflects the weakness in the British LEI. This is a consequence of the deterioration in global economic activity. As Chart I-10 shows, the trade-weighted pound has been more sensitive to EM gyrations than the euro or the dollar. This is because total trade represents a stunning 40% of U.K. GDP, versus 37% for the euro area or 28% for the U.S. The U.K. is therefore highly sensitive to global economic conditions. Moreover, the tightening in global liquidity conditions that has contributed to the deterioration of the global growth outlook is itself particularly negative for the pound. The U.K. runs a current account deficit of 4% of GDP, and as FDI inflows into Great Britain have collapsed, the U.K. now runs a basic balance-of-payments deficit (Chart I-11). As such, it is highly dependent on global liquidity flows to finance its current account deficit. As a result, the recent weakness in the pound is more a function of global economic conditions than Brexit itself. Chart I-10The Pound Has Fallen Because of EM Risks... Chart I-11...And As Global Liquidity Has Tightened Bottom Line: After developing a more precise method for evaluating the size of the geopolitical risk premium embedded in the pound, we arrived at an interesting conclusion: There is currently no evidence of a risk premium at all. Instead, the pound's weakness reflects the expensiveness of the dollar, weakening global growth and deteriorating global liquidity conditions. In fact, it is the euro that currently suffers from an exaggerated geopolitical risk premium, as euro area bonds currently incorporate too-large of a break-up risk premium. Investment Implications Taking into account the thin risk premium embedded in the pound against both the dollar and the euro, the GBP does not have much maneuvering room through the fall season. The problem for the pound is that the 5% net disapproval of Brexit among the British public remains smaller than the cohort of British voters who remain undecided (Chart I-12). This means that domestic politics in the U.K. could remain a source of surprise, especially as Prime Minister Theresa May's polling remains tenuous (Chart I-13). This raises the risk that Hard Brexiters end up controlling 10 Downing Street - despite their status as a minority within Conservative MPs. This also raises the risk that Jeremy Corbyn, whose popularity is rising, could end up as British Prime Minister (Chart I-14). Both of these outcomes are worrisome. The pound is currently pricing in neither the risk of a hard Brexit, nor the risk of the U.K. being controlled by the most leftist government of any G10 nation since the election of Francois Mitterrand in France in 1981. Chart I-12More Undecided Voters Than ##br##Net Brexit Detractors Chart I-13A Risk To ##br##U.K. Stability... Chart I-14...Especially With Mitterand 2.0 ##br##Lurking In The Shadows Moreover, while Germany and EU chief negotiator Michel Barnier seem amenable to keeping the window of negotiations open for the ultimate form of Brexit during the transition period, it remains to be seen what kind of concessions London is willing to make on the free movement of people required to be granted access to the common market in goods. Additionally, the Northern Ireland border remains an unresolved issue. These factors increase the chances that negotiations with the EU will remain difficult. Hence, the implementation of the Chequers White Paper is far from certain, yet the pound currently seems to be priced for an absolute soft Brexit. With the global economic outlook already justifying a lower pound, especially versus the dollar, it therefore seems that the pound today is too risky an investment. It is true that positioning and sentiment in cable are currently very depressed, raising the risk of a short-term rebound (Chart I-15), especially if the EU meeting in Salzburg in two weeks shows an acquiescent EU. However, this will not remove Britain's domestic political problems. Hence, we would be inclined to fade any such rebound. Finally, it is unlikely that the Bank of England will be of much help to the pound either. The British LEI continues to slow, which not only drags the fair value of the pound lower, but also limits how fast the BoE can raise interest rates. Moreover, while British inflation surged as imported goods prices skyrocketed after the GBP plummeted in 2016, domestic prices have remained well behaved (Chart I-16). Thus, as the pass-through to inflation of the previous pound weakness dissipates, British inflation will decelerate further, limiting the upside for interest rates in the process. This combination is only made more binding for the BoE as the government is expected to exert some drag on growth as the British fiscal thrust will subtract 0.4%, 0.2%, and 0.2% to growth in 2018, 2019, and 2020, respectively (Chart I-17). Chart I-15There Is Room For A ##br##Countertrend GBP Rally Chart I-16Little Domestic ##br##Price Pressures Chart I-17Fiscal Drag ##br##Not Over On a six- to nine-month basis, it makes most sense to short the pound against the dollar and the yen, as slowing global growth hurts the pound but also hurts the euro while benefiting the greenback and the yen. However, on a longer-term basis, we would expect the break-up risk premium in the euro area to dissipate, which will boost the cheap euro in the process. This means that on investment horizons beyond one year, being long EUR/GBP still makes sense. Bottom Line: Since this year's pound weakness did not represent a swelling of the GBP's geopolitical risk premium but instead has been a reflection of poor global growth and liquidity, any hiccups in British politics could inflict considerable pain on cable. While the EU negotiations may progress positively, domestic British politics remain a big source of risk that investors are not being compensated to take on. As such, we are inclined to fade any rally in the pound. While the pound could weaken most against the dollar and the yen through the fall months, the longer-term outlook looks riskier against the euro. To be clear, the confidence interval around these views remains wide, as the British political situation remains very fluid. Mathieu Savary, Vice President Foreign Exchange Strategy mathieu@bcaresearch.com Juan Manuel Correa, Senior Analyst juanc@bcaresearch.com Ekaterina Shtrevensky, Research Associate ekaterinas@bcaresearch.com 1 The Balassa-Samuelson effect has been cited as a potential explanation for this observation, but it still does not fully satisfy many theorists. Appendix Chart II-1Out-Of-Sample Testing Of Model (I) Chart II-2Out-Of-Sample Testing Of Model (II) Trades & Forecasts Forecast Summary Core Portfolio Tactical Trades Closed Trades

Highlights 2018 YTD Summary: Investment grade corporate debt in the developed economies has performed poorly so far in 2018, led by lagging returns in Financials and some steepening of credit curves. U.S. credit has outperformed European equivalents. These trends are likely to continue over at least the next six months. Our Sector Portfolios: Our investment grade sector model portfolios have underperformed modestly so far in 2018 (-3bps each in the U.S., euro area & U.K.) - primarily due to our overweight stance on Financials which have performed poorly. Looking Ahead: We are maintaining a neutral level of target spread risk (i.e. duration-times-spread equal that of the benchmark index) in our sector model portfolios for the U.S., euro area and U.K. We will look to reduce that spread risk on signs of a deeper global growth slowdown, which we expect will unfold in 2019. Feature Chart of the WeekReversal Of Fortune The performance of investment grade (IG) corporate bonds in the developed markets, as an asset class, has been underwhelming so far in 2018. Using the total return indices from Bloomberg Barclays, IG corporates in the U.S., euro area and U.K. - the regions with the three largest corporate bond markets among the developed economies - have lost -2.0%, -0.3% and -1.1%, respectively. The numbers do not look much better when shown on an excess return basis versus duration-matched government bonds: U.S. IG -0.8%, euro area -1.2% and the U.K. -1.3%. The sluggish performance for IG corporates is a mirror image of the strong showing in 2017 when looking at credit spreads, which reached very tight levels at the end of last year (Chart of the Week). The 2017 rally left global corporates exposed to any negative shocks, of which there have been many so far in 2018 (the February VIX spike, the Q1 global growth slowdown, intensifying U.S.-China trade tensions, ongoing Fed tightening, a strengthening U.S. dollar, less dovish non-U.S. central banks, Italian politics, emerging market turmoil). Given the more challenging environment for overall corporate bond performance, the role of sector selection as a way to generate alpha, by mitigating losses from beta, is critical. In this Weekly Report, we take a brief look at IG sector performance so far this year and update our sector allocations based on our relative value models for IG corporates in the U.S., euro area and U.K. 2018 YTD Global Corporates Performance: A Down Year The major IG sector groupings for the U.S., euro area and U.K. are presented in Table 1, ranked by the 2018 year-to-date excess returns (all are shown in local currency terms). The overall index return for each region is also shown (highlighted in gray) in the table, to highlight how individual sectors have performed relative to the overall IG index. Table 12018 Year-To-Date Investment Grade Sector Returns For The U.S., Euro Area & U.K. As is always the case with IG corporates, the performance of the broad Financials grouping (which includes banks, insurance companies, REITs, etc.) heavily influences the returns of the overall IG index given the large weighting of Financials within the Corporates index across all three regions. In both the euro area and U.K., the sharp underperformance of Financials seen year-to-date (-1.3% and -1.4%, respectively) has created a somewhat odd situation where the majority of sectors have outperformed the overall index. That could only happen given the large weight of Financials in the euro area index (40%) and U.K. index (43%). Financials are also a big part of the U.S. index (32%), but there is more balance in the U.S. IG index which has helped boost the "beta" return from U.S. corporates. Specifically, the weightings of the top three largest U.S. broad sector groupings - Energy (9%), Technology (8%) and Communications (9%) - are a combined 26% of the overall U.S. IG index. Those three sectors are also among upper tier of the 2018 performance table in the euro area and U.K., but only represent a combined 15% and 8%, respectively, of each region's IG index. The conclusion is that index composition has flattered the performance of U.S. IG corporates versus European equivalents, given the latter's heavier weighting in Financials. The poor performance of Financials can be attributed to flattening global government bond yield curves (which is a negative for banks) and poor returns from global credit, especially in emerging markets (which is a negative for insurers that invest in spread product). We do not anticipate either of those trends reversing anytime soon - particularly the ongoing selloff in emerging market assets - thus Financials are likely to remain a drag on corporate bond performance for at least the next 3-6 months. One other factor that has weighed on overall IG corporate performance has been the steepening of credit spread curves. The gaps between credit spreads for Baa- and A-rated corporates have widened since the end of January, most notably in the euro area and the U.K. where growth has been slower than in the fiscal-policy fueled U.S. economy (Chart 2). With Baa-rated debt now representing one-half of the IG index for the U.S., euro area and U.K. (Chart 3) - a function of rising corporate leverage - continued underperformance of lower quality sectors will negatively impact the future overall returns from IG corporates. Chart 2Spread Curves Are##BR##Steepening In Europe Chart 31/2 Of Investment Grade Corporate Indices##BR##Are Now Baa-Rated Looking ahead, credit investors should be wary of the potential for downgrade risk in their portfolios given the high proportion of Baa-rated debt in the IG benchmark indices. This risk will become more acute when the global business cycle runs out of steam (a 2019 story, at the earliest, in our view). Bottom Line: Investment grade corporate debt in the developed economies has performed poorly so far in 2018, led by lagging returns in Financials and some steepening of credit curves. U.S. credit has outperformed European equivalents. These trends are likely to continue over at least the next six months. Our Corporate Sector Valuation Models: Winners & Losers Our recommended IG sector allocations come from our relative value model, which measures the valuation of each individual sector compared to the overall Bloomberg Barclays corporate bond index for each region. The methodology takes each sector's individual option-adjusted spread (OAS) and regresses it in a panel regression with all other sectors in each region. The dependent variables in the model are each sector's duration, convexity (duration squared) and credit rating - the primary risk factors for any corporate bond. Using the common coefficients from that panel regression, a risk-adjusted "fair value" spread is calculated. The difference between the actual OAS and fair value OAS is our valuation metric used to inform our sector allocation ranking. The latest output from the models can be found in the tables and charts in the Appendix starting on Page 13. We also show the duration-times-spread (DTS) for each sector in those tables, which we use as the primary way to measure the riskiness (volatility) of each sector. The scatterplot charts in the Appendix show the tradeoff between the valuation residual from our model and each sector's DTS. We then apply individual sector weights based on the model output and our desired level of overall spread risk in our recommended credit portfolio. The weights are determined at our discretion and are not the output from any quantitative portfolio optimization process. The only constraints are that all sector weights must add to 100% (i.e. the portfolio is fully invested with no use of leverage) and the overall level of spread risk (DTS) must equal our desired target. That target portfolio DTS is the first decision in our discretionary allocation process, which is informed by our strategic views on corporate credit in each region. For example, if we were recommending an overweight allocation to U.S. IG corporates, then we would target a portfolio DTS that was greater than the index DTS. If we then became a bit more cautious on U.S. corporates, we could reduce the target DTS (spread risk) of our model sector portfolio while maintaining an overall overweight allocation to U.S. corporates versus U.S. Treasuries. That is exactly what we did one year ago, when we began to target a weighted DTS of all our individual sector tilts that was roughly equal to the overall IG corporate index DTS for each region (U.S. euro area, U.K.) while maintaining an overall overweight stance on global corporate credit versus government debt. More recently, we have downgraded our stance on global spread product to neutral, while continuing to favor the U.S. over Europe, in response to growing tensions from emerging markets and the brewing U.S.-China trade war.1 Chart 4Performance Of Our IG Sector Allocations We last presented a performance update for our global IG corporate sector allocations back on April 12th of this year. Since then, our recommended tilts have modestly underperformed the benchmark index in excess return terms by a combined -3bps (Chart 4). This came entirely from the euro area, with both the U.S. and U.K. sector allocations simply matching the benchmark index. Year-to-date, our IG sector allocations have underperformed the benchmark by a combined -9bps in excess return terms, split equally among the U.S., euro area and U.K. This is a result entirely consistent with our long-standing stance to overweight Financials in all three regions, which continue to appear cheap in our valuation framework. Also, an increasing number of sectors had become expensive within that framework, in all three regions, so some portion of that overweight to global Financials was "by default" given that our model portfolios must be fully invested (finding value has been a persistent problem for credit investors over the past year). The return numbers for our U.S. sector allocations can be found in Table 2. Since our last update in April, the best performing sectors (in excess return terms) within our recommended tilts have all been underweights: Pharmaceuticals (+1.2bps), Electric Utilities (+1.1bps), Retailers (+0.6bps), Health Care (+0.6bps), Diversified Manufacturing (+0.5bps) and Chemicals (+0.4bps). These were fully offset, however, by underperformance from our large overweights to Energy (-4.1bps) and Financials (-2.7bps). Table 2U.S. Investment Grade Performance The return numbers for our euro area sector allocations - shown here hedged into U.S. dollars as is the case when we present all our model portfolio returns - can be found in Table 3. Since our last update in April, the best performing sectors (in excess return terms) within our recommended tilts have been underweights to Transportation (+2.0bps) and Electric Utilities (+0.6bps), with underperformance coming from our underweight to Food/Beverage (-2.4bps) and overweight to Life Insurers (-3.1bps). Table 3Euro Area Investment Grade Performance The return numbers for our U.K. sector allocations (again, hedged into U.S. dollars) can be found in Table 4. Since our last update in April, the best performing sectors (in excess return terms) within our recommended tilts have been our underweight to Utilities (+2.0bps) and Consumer Non-Cyclicals (+0.9bps), but this was nearly fully offset by our large overweight to Financials (-2.6bps). Table 4U.K. Investment Grade Performance Despite the underperformance of our sector portfolios year-to-date, the cumulative alpha from the portfolios since we began tracking the performance of the recommendations remains positive (+2bps in the U.S., +9bps in the euro area, +42bps in the U.K.). Bottom Line: Our investment grade sector model portfolios have underperformed modestly so far in 2018 (-3bps each in the U.S., euro area & U.K.) - primarily due to our overweight stance on Financials which have performed poorly. Changes To Our Sector Model Portfolios As mentioned earlier, the first choice we make when determining the recommended sector allocations within our model portfolios is how much spread risk (DTS) to take. For the U.S., euro area and U.K., we have already been maintaining a portfolio DTS that is close to the index DTS since August 2017. After our recent decision to downgrade global spread product allocations to neutral versus government bonds, we do not feel a need to further reduce our spread risk by targeting a below-index DTS. That would likely be our next decision when we wish to get more defensive on credit, which would await evidence that global leading economic indicators are sharply slowing and/or global monetary policy is becoming restrictive. Within that neutral level of spread risk, we are making the following portfolio changes based on the updated output from our valuation models presented in the Appendix Tables on pages 13-18. The goal is to favor sectors that have a DTS close the index DTS but have positive valuation residuals from our model: U.S.: We downgrade Tobacco and Wireless to Neutral; we downgrade Paper to Underweight. Euro Area: We upgrade Transportation, Other Industrials, Natural Gas, Brokerages/Asset Managers and Finance Companies to Overweight; we upgrade Automotive, Retailers and Tobacco to Neutral; we downgrade Wireless to Neutral; we downgrade Diversified Manufacturing & Media Entertainment to Underweight. U.K.: We upgrade Health Care, Transportation and Other Industrials to Overweight; we upgrade Integrated Energy to Neutral; we downgrade Technology & Wireless to Neutral; we downgrade Metals & Mining and Supermarkets to underweight. Robert Robis, Senior Vice President Global Fixed Income Strategy rrobis@bcaresearch.com 1 Please see BCA Global Fixed Income Strategy Weekly Report, "Time To Take Some Chips Off The Table: Downgrade Global Corporate Bond Exposure To Neutral", dated June 26th 2018, available at gfis.bcaresearch.com. Appendix Appendix Table 1U.S. Corporate Sector Valuation And Recommended Allocation* Appendix Chart 1U.S. Corporate Sector Risk Vs. Reward* Appendix Table 2Euro Area Corporate Sector Valuation And Recommended Allocation* Appendix Chart 2Euro Area Corporate Sector Risk Vs. Reward* Appendix Table 3U.K. Corporate Sector Valuation And Recommended Allocation* Appendix Chart 3U.K. Corporate Sector Risk Vs. Reward* Recommendations The GFIS Recommended Portfolio Vs. The Custom Benchmark Index Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Feature GAA DM Equity Country Allocation Model Update The GAA DM Equity Country Allocation model is updated as of August 31, 2018. The quant model has further lifted its U.S. allocation to overweight from neutral, and the U.K. underweight has also been reduced by half. On the other hand, Italy is downgraded and the overweight in Spain, Germany and the Netherlands are all significantly reduced, as shown in Table 1. Table 1Model Allocation Vs. Benchmark Weights As shown in Table 2 and Charts 1, 2 and 3, the overall model underperformed its benchmark by 32 bps in August, largely driven by Level 2 model which underperformed its benchmark by 75 bps. expected, the model did not catch the "Turkey Effect" which drove deep losses in the Italian and Spanish markets. The Level 1 model slightly unperformed its MSCI World benchmark by 2 bps in August. Since going live, the overall model has outperformed its benchmarks by 87bps, driven by the Level 2 outperformance of 260 bps offset by the 5 bps of Level 1 underperformance. Table 2Performance (Total Returns In USD %) Chart 1GAA DM Model Vs. MSCI World Chart 2GAA U.S. Vs. Non U.S. Model (Level 1) Chart 3GAA Non U.S. Model (Level 2) Please see also the website http://gaa.bcaresearch.com/trades/allocation_performance. For more details on the models, please see Special Report, "Global Equity Allocation: Introducing The Developed Markets Country Allocation Model," dated January 29, 2016, available at https://gaa.bcaresearch.com. Please note that the overall country and sector recommendations published in our Monthly Portfolio Update and Quarterly Portfolio Outlook use the results of these quantitative models as one input, but do not stick slavishly to them. We believe that models are a useful check, but structural changes and unquantifiable factors need to be considered too in making overall recommendations. GAA Equity Sector Selection Model Dear Client, The GAA Equity Sector Selection Model has been live since July 2016, and has outperformed the benchmark over this period in line with our back-testing. However, GICS will make significant changes to sector compositions at the end of September, most notably creating a new "Communication Services" sector, dominated by internet-related companies, to replace "Telecommunication Services". However, MSCI has not yet made available the final details of membership or historical performance of the revised sectors. Accordingly, after this update we are temporarily suspending publication of this model until full data is available and we have been able to rebuild the model using the newly constituted sectors. The GAA Equity Sector Selection Mode (Chart 4) is updated as of August 31, 2018. Table 3Allocations Table 4Performance Since Going Live Chart 4Overall Model Performance The model continues to have a negative outlook on global growth and consequently has a net underweight on cyclical sectors. However, the magnitude of this tilt was reduced from 5.8% to 2.8%. The biggest move was a downgrade of consumer staples from a 2.5% overweight to a 1.4% underweight on the back of unfavorable momentum indicators. The only two sectors with favorable momentum are healthcare and technology. Finally, energy stocks also saw a 0.8% boost in its overweight recommendation on the back of attractive valuations. For more details on the model, please see the Special Report "Introducing The GAA Equity Sector Selection Model," dated July 27, 2016, available at https://gaa.bcaresearch.com. Xiaoli Tang, Associate Vice President xiaoliT@bcaresearch.com Aditya Kurian, Senior Analyst adityak@bcaresearch.com