Inflation Protected

Highlights Does it still make sense to use historical yield betas for fixed income country allocation? Yes, favoring countries with higher government bond yield betas when global yields are falling, and vice versa, is still an appropriate way to manage fixed income risk – although betas do vary between global bond bull and bear markets. Can inflation breakevens and real yields continue moving in opposite directions? Yes, but that negative correlation will become less intense, especially in the US, with rising inflation expectations eventually becoming the more dominant influence on nominal bond yields. Will inflation breakevens continue to have a strong positive correlation with oil prices? Yes, but only for as long as non-energy inflation remains low and stable, which has made energy prices the only source of inflation variability in most developed countries. Feature Sleepy bond markets got a bit of a jolt over the past couple of weeks, with longer-maturity government bond yields moving higher across the developed markets, led by the US where the 30-year Treasury yield is now back to levels last seen in June. The move higher in US Treasury yields may be a sign that investors are taking the US election polling numbers – which now signal not only a Joe Biden victory on November 3, but also a swing of the US Senate to Democratic Party control – seriously. A so-called “Blue Sweep”, resulting in the full implementation of the Biden policy platform including a massive fiscal stimulus, is potentially bond bearish, and not only for US Treasuries, given the close correlation of US yields to other bond markets. There is a strong correlation between the level of bond yields, and the yield beta, for the major developed market countries. This brief burst of global bond market volatility, stemming from developments in the US, is a reminder that investors should always be aware of the importance of cross-market correlations when making trading and portfolio construction decisions. With that in mind, this week we ask some important questions about the critical correlations across global government bond markets that support our current investment recommendations – and under what conditions they could possibly change. Does It Still Make Sense To Use Historical Yield Betas For Fixed Income Country Allocation? Chart 1Developed Bond Yields Relative To The 'Global' Bond Yield One of the key elements underlying our bond country allocation recommendations is the concept of “yield beta”. Simply put, this is a measure of the sensitivity of changes in individual country bond yields to changes in the overall level of global bond yields. The way we measure yield betas is by using a regression (over a three-year rolling window) of monthly changes for an individual country’s 10-year bond yield on the monthly change of the Bloomberg Barclays Global Treasury index yield for the 7-10 year maturity bucket (as the proxy for the “global” 10-year yield). The regression coefficient on the individual country yield change is the yield beta. There is a strong correlation between the level of bond yields, and the yield beta, for the major developed market countries. Currently, the list of “high-yielders” – with 10-year government bond yields above the benchmark index yield – includes the US, Italy, Canada, Australia and New Zealand (Chart 1). The low-yielders, with 10-year yields below the benchmark index yield, are Germany, France, Spain, the UK and Japan. When we look at the yield betas for that same list of countries, we can also break up the list into high-beta and low-beta bond markets. When we rank the ten countries by their rolling three-year yield betas, the five highest betas belong to the same five countries with the highest yields, and vice versa (Chart 2). This is an intuitive correlation, as countries with higher yield betas are, by definition, more volatile and should require higher yields from investors to compensate for that additional volatility. Chart 2The Higher-Yielding Countries Also Have Higher Yield Betas The yield betas are not stable over time for all countries, however. The US has consistently remained the highest beta market, and Japan the lowest beta market, over the past twenty years. Other countries have seen their yield betas evolve over time. For example, France, Spain and, more recently, the UK have seen their yield betas decline in recent years, while Italy has gone from being low-beta to one of the higher-beta markets. In our view, the evolution of yield betas relates to the “activism” of policymakers in each country. Higher-beta, higher-yield countries also have central banks that move interest rates higher and lower with more frequency compared to the low-beta, low-yield countries. In our view, the evolution of yield betas relates to the “activism” of policymakers in each country. That high-beta group includes bond markets linked to the Federal Reserve, the Bank of Canada, the Reserve Bank of Australia and the Reserve Bank of New Zealand – all central banks that are not shy about aggressively cutting or hiking interest rates. The low-beta markets have central banks that move rates very infrequently, like the European Central Bank and the Bank of Japan. Table 1Yield Betas For The Major Developed Markets One other interesting point on yield betas is that they do vary depending on the overall direction of global bond yields. As a way to show this, we estimated “upside” and “downside” yield betas for the same ten countries shown earlier. Those betas were calculated by sorting the monthly yield changes for all countries by months when the benchmark global bond index yield was rising or falling. Thus, upside yield beta comes from a regression of monthly yield changes for individual countries on changes in overall global bond yields, but only using data for months when global yields increased. The opposite is true for downside beta, where only data from months when the global benchmark index yield declined are used. The individual yield betas – for the overall sample and the upside and downside groupings – are presented in Table 1. One conclusion that comes from breaking up the data this way is that countries that were in the low-beta group when looking at the full set of data have relatively high yield betas during periods of rising global yields, like France and the UK (Chart 3). In addition, when looking at downside betas, US Treasuries have the highest beta, by far, when global yields are falling – with yields for euro area countries having relatively lower betas (Chart 4). Chart 3Yield Betas During Periods Of Rising Global Yields Chart 4Yield Betas During Periods Of Falling Global Yields Our conclusion from this analysis is that yield betas do have a useful role in making country allocation decisions for global fixed income investors. Specifically, adjusting allocations based on a view on the overall direction of global bond yields should help better manage portfolio risk and, potentially, improve returns. Chart 5Italy Has Become High-Beta As Spreads Have Narrowed A final point on Italy – the reason Italy has had such a high yield beta over the past few years is because Italian government bond yields have been driven more by the reduction of Italian sovereign credit risk – including the redenomination risk from a potential Italian exit from the euro (Chart 5). As Italian credit spreads have melted away from the levels reached during the 2011/12 European Debt Crisis, yields have fallen faster than others during periods of falling global yields, and vice versa. Looking ahead, with the ECB continuing to be an aggressive buyer of Italian bonds in its various asset purchase programs, and with the COVID-19 pandemic forcing the European Union into a deeper level of economic co-operation – which now includes grants to Italy – the sovereign risk premium on Italian government debt should continue to narrow. That means Italy will continue to trade as a high-beta market when global yields are falling, and a low-beta market when yields are rising, making Italy an ideal overweight candidate in global bond portfolios. Bottom Line: Favoring countries with higher government bond yield betas when global yields are falling, and vice versa, is still an appropriate way to manage fixed income risk – although betas do vary between global bond bull and bear markets. Can Inflation Breakevens And Real Yields Continue Moving In Opposite Directions? The behavior of real bond yields over the past few months garnered a lot of attention in 2020, particularly the sharp fall in US TIPS yields into deeply negative territory. This has occurred at the same time as a widening of inflation breakevens, which exhibited a deeply negative correlation with real yields. The result: narrow trading ranges for nominal government bond yields in most developed countries, with moves in real yields and inflation breakevens largely offsetting each other. Adjusting allocations based on a view on the overall direction of global bond yields should help better manage portfolio risk and, potentially, improve returns. Looking at the history of real yields and inflation breakevens, periods of a negative correlation between the two are not unusual. In Chart 6, we show the range of historic correlations between 10-year inflation-linked bond yields, and 10-year inflation breakevens, for the US, UK, Germany, France, Italy, Australia, Canada and Japan since 2010. The dark bars represent the range of rolling correlations over a three-year period, while the red diamonds are a more recent correlation over the past thirteen weeks. All countries shown have seen periods of negative correlation, with only Australia and France having the most recent correlation be far lower than the historic experience. Chart 6Negative Real Yield/Breakevens Correlations Are Not Unprecedented So if a negative real yield/inflation breakeven correlation is not that unusual, then what is the cause of it? We see two drivers: the amount of spare capacity in an economy and the central bank policy response to it. We can see this by looking at the data from the countries with the two largest inflation-linked bond markets, the US and UK. In the US, real TIPS yields and inflation breakevens have generally been positively correlated only during Fed tightening cycles, specifically after the Fed has raised the fed funds rate above the rate of realized core inflation (Chart 7). This was the case in the tightening cycles of the mid-2000s and 2016-18. During those episodes, the Fed pushed the real funds rate steadily higher, which also had the effect of pushing real TIPS bond yields higher, even as inflation expectations were stable-to-rising. Looking at the history of real yields and inflation breakevens, periods of a negative correlation between the two are not unusual. The opposite held true during Fed easing cycles since the advent of the TIPS market in the late 1990s, when the Fed always lowered the funds rate below realized inflation. The result was a period of a falling real funds rate, leading to lower real TIPS yields and eventually triggering an increase in inflation breakevens. In other words, the correlation between breakevens and real yields became negative. In the UK, the negative correlation between real index-linked Gilt yields and inflation breakevens has been consistently negative since the 2008 financial crisis (Chart 8). The Bank of England has barely moved policy rates since that crisis, while keeping nominal policy rates below 1% - a level that was consistently below core UK inflation. Thus, the Bank of England has maintained negative real policy rates for the past twelve years, with real Gilt yields declining steadily and inflation breakevens rising – a negative correlation - over that period. Chart 7Fed Policy Influences The US Real Yield/Breakevens Correlation Chart 8A Persistently Negative Correlation Of UK Real Yields & Breakevens   For both the US (Chart 9) and UK (Chart 10), the rolling 3-year correlation between real yields and breakevens has itself been correlated to the unemployment gap, or the difference between the unemployment rate and the full-employment NAIRU rate, over the past two decades. This suggests that the ebbs and flows of labor market slack, and how the Fed and Bank of England have responded to them by easing or tightening monetary policy, also play a role in determining the real yield/breakevens correlation. Chart 9Real Yield/Breakevens Correlation Will Stay Negative In The US Chart 10Real Yield/Breakevens Correlation Will Stay Negative In The UK   In the case of the US, a more extended UK-like period of negative real policy rates and real bond yields is likely if the Fed is to be taken at their word that they will keep rates low to engineer a US inflation overshoot. We suspect that the correlation will not be perfectly negative, as has occurred at times this year, with inflation expectations rising alongside stable-to-falling real TIPS yields as the US economy recovers from the COVID-19 shock – especially if there is a major boost from fiscal stimulus after next month’s elections. Bottom Line: We continue to see a case for inflation breakevens and real yields to stay negatively correlated in the developed economies over at least the next few years, as the labor market slack created by the 2020 COVID-19 global recession is slowly absorbed. That negative correlation will become less intense, especially in the US, with rising inflation expectations eventually becoming the more dominant influence on nominal bond yields. Will Inflation Breakevens Continue To Have A Strong Positive Correlation With Oil Prices? While the negative correlation between real inflation-linked bond yields and real yields has gotten attention this year, the positive correlation between breakevens and oil prices has become familiar to investors over the past several years. That correlation has been persistently high and positive across all developed economies since the 2008 financial crisis. Prior to that, oil prices and inflation breakevens moved together less frequently and, at times, were even uncorrelated (Chart 11). In both the US and euro area, the lack of non-energy inflation is the main reason why breakevens and oil are so correlated. In our view, the reason why breakevens and oil became strongly correlated is relatively straightforward. Since the 2008 crisis and ensuing Great Recession, swings in oil prices have been the main driver of changes in realized inflation, with ex-energy inflation rates staying very low and stable. We can see that in the US, where ex-energy CPI inflation has been broadly stable around 2% for the past decade, even as headline CPI inflation has seen more variability and has even approached 0% after the collapse in oil prices in 2014/15 and 2020 (Chart 12). Chart 11A Persistent Strong Correlation Of Global Breakevens To Oil Chart 12Strong Oil/Breakevens Correlation While US Ex-Energy Inflation Is Low Chart 13Energy Has Become The Only Source Of Euro Area Inflation The same dynamics, only more intense, exist in the euro area. Ex-energy inflation has struggled to stay above 1% over the past decade, leaving changes in energy prices as an even greater determinant of realized headline inflation than in the US (Chart 13). In both the US and euro area, the lack of non-energy inflation is the main reason why breakevens and oil are so correlated. Until there is evidence of a more broad-based move higher in inflation rates outside of oil - which will almost certainly require an extended period of above-trend global growth and accommodative global fiscal and monetary policies - trading inflation breakevens off oil will still be a successful strategy. Bottom Line: Global inflation breakevens will maintain a strong positive correlation to oil prices, but only for as long as non-energy inflation remains low and stable, which has made energy prices the only source of inflation variability in most developed countries Robert Robis, CFA Chief Fixed Income Strategist rrobis@bcaresearch.com Footnotes   Recommendations Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Highlights Q3/2020 Performance Breakdown: Our recommended model bond portfolio outperformed the custom benchmark by +19bps during the third quarter of the year. Winners & Losers: The government bond side of the portfolio outperformed by +10bps, led by overweights in US (+13bps), Canada (+2bps) and Italy (+4bps) that favored allocations to inflation-linked government bonds out of nominals. Spread product generated a similar-sized outperformance (+9bps), led by overweights to US investment grade corporates (+8bps). Portfolio Positioning For The Next Six Months: We continue to prefer keeping aggregate portfolio duration close to benchmark, with only a moderate overweight allocation to spread product versus government bonds, given the lingering uncertainties over the global spread of COVID-19 and near-term US election risk. Instead, we recommend focusing on relative value allocations, favoring countries and sectors that will benefit most in our base case medium-term scenario of slowly improving global growth, reflationary global monetary/fiscal policies, low bond yield volatility and a softening US dollar. Feature As we enter the final quarter of 2020, global financial markets are dealing with many near-term uncertainties related to the upcoming US presidential election, potential next moves in global policy stimulus and, perhaps most worrying, a second wave of coronavirus infections in Europe and the US. That means the “easy money” has been made in global fixed income from the unwind of the blowout in credit spreads, and collapse of government bond yields, seen following the COVID-19 related market turbulence of February and March. Investors should expect substantially lower fixed income returns in the coming months. Relative performance between countries and sectors will be the more dominant influence on bond portfolio returns in the absence of big directional moves in yields or spreads. Alternatively put, expect alpha to win out over beta. This week we are reviewing the performance of the BCA Research Global Fixed Income Strategy (GFIS) model bond portfolio during the third quarter of 2020. We also present our recommended positioning for the portfolio for the next six months. With that in mind, this week we are reviewing the performance of the BCA Research Global Fixed Income Strategy (GFIS) model bond portfolio during the third quarter of 2020. We also present our recommended positioning for the portfolio for the next six months (Table 1), as well as portfolio return expectations for our base case and alternative investment scenarios. Table 1GFIS Model Bond Portfolio Recommended Positioning For The Next Six Months As a reminder to existing readers (and to new clients), the model portfolio is a part of our service that complements the usual macro analysis of global fixed income markets. The portfolio is how we communicate our opinion on the relative attractiveness between government bond and spread product sectors. We do this by applying actual percentage weightings to each of our recommendations within a fully invested hypothetical bond portfolio. Q3/2020 Model Portfolio Performance Breakdown: Another Positive Quarter, Led By Linkers & Corporates Chart of the WeekQ3/2020 Performance: Gains From Both Sides Of The Portfolio The total return for the GFIS model portfolio (hedged into US dollars) in the third quarter was 3.14%, modestly outperforming the custom benchmark index by +19bps (Chart of the Week).1 This is the second consecutive positive quarter, lifting the year-to-date outperformance into positive territory (+12bps) – an impressive accomplishment given the sharp drawdown that occurred during the market volatility of February and March. In terms of the specific breakdown between the government bond and spread product allocations in our model portfolio, the former generated +10bps of outperformance versus our custom benchmark index while the latter outperformed by +9bps. That government bond return includes a substantial gain (+17bps) from inflation-linked bonds, which we added as a new asset class in our model portfolio framework back on June 23.2 In a world of very low bond yields (Table 2), our preference for the relatively higher-yielding government bond markets in the US, Canada and Italy was an important source of outperformance, delivering a combined excess return of +19bps (including inflation-linked bonds). This was only partially offset by the negative active returns from underweights in low-yielding countries such as Germany, France, and Japan (a combined drag of -9bps). Table 2GFIS Model Bond Portfolio Q3/2020 Overall Return Attribution In spread product, our overweights in US investment grade corporates (+8bps), UK investment grade corporates (+3bps) and US Agency CMBS (+4bps) were the main sources of outperformance, while the negative active return from underweighting Euro Area high yield (-2bps) was minimal. Our preference to favor higher-rated US high-yield relative to lower-rated US junk bonds, even as riskier credit rallied, did little damage to portfolio performance, with a combined excess return across all three US junk credit tiers of just -2bps. The moderate outperformance of the model bond portfolio versus the benchmark in Q3 is in line with our cautious recommended stance on what are always the largest drivers of the portfolio returns: overall duration exposure and the relative allocation between government debt and spread product. We have stuck close to benchmark exposures on both, eschewing big directional bets on bond yields or credit spreads while focusing more on relative opportunities between countries and sectors – particularly in sectors most strongly supported by central bank easing actions, like US investment grade corporates. The bar charts showing the total and relative returns for each individual government bond market and spread product sector are presented in Charts 2 & 3. Chart 2GFIS Model Bond Portfolio Q3/2020 Government Bond Performance Attribution Chart 3GFIS Model Bond Portfolio Q3/2020 Spread Product Performance Attribution By Sector Biggest Outperformers: Long US TIPS (+12bps) Overweight US investment grade industrials (+5bps) Overweight US Agency CMBS (+4bps) Overweight UK investment grade corporates (+3bps) Overweight US high-yield Ba-rated corporates (+3bps) Biggest Underperformers: Underweight French government bonds with maturity greater than 10 years (-4bps) Underweight US high-yield B-rated corporates (-2bps) Chart 4 presents the ranked benchmark index returns of the individual countries and spread product sectors in the GFIS model bond portfolio for Q3/2020. Returns are hedged into US dollars (we do not take active currency risk in this portfolio) and 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 Q3/2020 (red for underweight, dark green for overweight, gray for neutral).3 Ideally, we would look to see more green bars on the left side of the chart where market returns are highest, and more red bars on the right side of the chart were returns are lowest. Chart 4Ranking The Winners & Losers From The GFIS Model Bond Portfolio In Q3/2020 The top performing sectors within our model bond portfolio universe in Q3 were well distributed among government bonds and spread products: Italian government bonds (relative index return of +3.8), New Zealand government bonds (+3.0%), EM USD-denominated sovereign (+2.6%), US high-yield corporates (2.4%), Spanish government bonds (+2.3%), and investment grade corporates in the UK (+2%) and US (1.9%). Importantly, we were overweight or neutral all of those markets during the quarter, driven by our main investment themes of “buying what the central banks are buying” and “yield chasing.”4 On the other side, we had limited exposure to the worst performing sectors during Q3, with underweights to government bonds in Germany and Japan, US Agency MBS and euro area high-yield. Cutting our long-standing overweight on UK Gilts to neutral in early August also benefitted the portfolio performance, with Gilts being the worst performer in our model bond universe by far in Q3. Bottom Line: Our model bond portfolio modestly outperformed its benchmark index in the second quarter of the year by +19bps – a positive result driven by our relative positioning that favored higher yielding government debt and spread product sectors directly supported by central bank purchases. Future Drivers Of Portfolio Returns & Scenario Analysis Looking ahead, the performance of the model bond portfolio will be driven by relative positioning across sectors and countries, rather than big directional bets on moves in government bond yields or corporate credit spreads. This is in line with the current strategic investment recommendations of the BCA Research fixed income services. Looking ahead, the performance of the model bond portfolio will be driven by relative positioning across sectors and countries, rather than big directional bets on moves in government bond yields or corporate credit spreads. The overall duration of the portfolio is in line with that of the custom benchmark index (Chart 5), consistent with our strategic investment recommendation to be neutral on exposure to changes in interest rates. With central banks actively seeking to keep policy rates as low as possible until inflation returns – i.e. aiming to push real rates even lower - we expect the negative correlation seen between global inflation breakevens and real bond yields to persist over at least the next 6-12 months. Offsetting moves in both will continue to dampen the volatility of nominal bond yields, as has been the case over the past six months (Chart 6). Chart 5Overall Portfolio Duration Exposure: At Benchmark Central banks aiming for an inflation overshoot and negative real rates will also continue to boost the relative performance of inflation-linked bonds versus nominal equivalents. Chart 6Within Governments, Continue Overweighting Linkers Vs Nominals We see this as a similar environment to the years following the 2008 financial crisis, with central banks keeping rates at 0% while rapidly expanding their balance sheets via quantitative easing and cheap liquidity provision for banks. The result was a multi-year period where linkers outperformed nominal government bonds (Chart 7). Thus, we are maintaining a large core allocation to linkers in the portfolio, focused on US TIPS and inflation-linked bonds in Italy and Canada. Chart 7The Strategic Case For Inflation-Linked Bond Outperformance Chart 8Overall Portfolio Allocation: Moderately Overweight Credit Vs Governments In terms of country allocations on the government bond side of the portfolio, we continue to favor overweights in higher-yielding markets with overall global yield volatility likely to remain subdued. Chart 9Global QE Continues To Support Credit Markets That means overweighting the US, Canada, Australia, Italy and Spain, while underweighting Germany, France and Japan. The UK belongs in that latter list, but we are maintaining a neutral stance on the UK, for now, given the near-term uncertainty surrounding final Brexit negotiations and the surge in new UK COVID-19 cases. Turning to spread product, we are maintaining only a moderate aggregate overweight allocation versus government bonds, equal to 4% of the portfolio (Chart 8). The same aggressive easing of global monetary policy and expansion of central bank balance sheets that is good for relative inflation-linked bond performance also benefits global corporate bonds. The annual rate of growth of the combined balance sheets of the Fed, ECB, Bank of Japan and Bank of England has proven to be an excellent leading indicator of the excess returns of both global investment grade and high-yield corporates over the past decade (Chart 9). With the combined balance sheet now expanding at a 40% pace, corporate bonds are likely to continue to outperform government debt over the next 6-12 months. Thus, our allocation to inflation-linked bonds and corporate credit, both out of nominal government bonds, are both motivated by the same factor – monetary policy reflation. The rally in the lower-rated tiers of the high-yield corporate universe in the US and euro area looks particularly unsustainable, if corporate defaults follow the path of previous recessions in both regions. At the same time, we continue to maintain a cautious stance on allocations to countries and sectors within that overall overweight tilt towards spread product in the model bond portfolio. We prefer to stay relatively up-in-quality within global corporate debt, even with high-yield bonds in the US and Europe offering relatively high spreads using our 12-month breakeven spread metric (Chart 10).5 Chart 10US & European HY Offer Relatively Wide Breakeven Spreads Chart 11US & European HY Offer No Spread Cushion Against Rising Defaults The rally in the lower-rated tiers of the high-yield corporate universe in the US and euro area looks particularly unsustainable, if corporate defaults follow the path of previous recessions in both regions. Our measure of the default-adjusted spread, calculated by taking the option-adjusted spread of the Bloomberg Barclays high-yield index and subtracting default losses, shows that high-yield spreads on both sides of the Atlantic will be dwarfed by expected default losses over the next year, assuming a typical pattern of defaults after recessions (Chart 11). We continue to prefer staying up-in-quality within our recommended corporate allocations, favoring Ba-rated US high-yield over B-rated and Caa-rated credit while also underweighting euro area high-yield relative to euro area investment grade corporates. This strategy lowers the yield of the model portfolio, which is currently in line with that of the custom benchmark index (Chart 12), at the expense of stretching for yields in riskier credit that may not be sustainable over the medium-term. Chart 12Overall Portfolio Yield: At Benchmark Chart 13Overall Portfolio Risk: Moderate At the same time, our measured stance on relative corporate exposure also acts to reduce portfolio risk – a useful outcome as we are targeting a relatively moderate tracking error (relative portfolio volatility versus that of the benchmark) within the model portfolio (Chart 13). Given the near-term uncertainties over the US elections and the potential second wave of COVID-19 in the US and Europe, staying relatively cautious on the usage of the “risk budget” of the portfolio seems prudent. Scenario Analysis & Return Forecasts In past quarterly reviews of our model bond portfolio, we have presented forecasts for the performance of the overall portfolio based off scenario analysis and some simple quantitative model-based predictions of various fixed income sectors. Given the unprecedented nature of the COVID-19 shock, we chose to avoid such model driven forecasts based on historical coefficients and correlations that may not be applicable. As it turns out, we may have been too cautious in that decision. The “risk-factor” models that we have used to forecast future yield changes for global spread product sectors as a function of four major factors - the VIX, oil prices, the US dollar and the fed funds rate (Table 2A) - have actually done a reasonable job of predicting yield changes over the past year. This can be seen in the charts shown in the Appendix on pages 18-20. Only in the case of US Caa-rated high-yield and EM USD-denominated corporates – two sectors where we are underweight given our concerns about valuation - have yields fallen by a far greater amount than implied by our models. Table 2AFactor Regressions Used To Estimate Spread Product Yield Changes Based on how the models have performed in the COVID era, we believe we can use them again to forecast the expected relative returns of the credit side of the model bond portfolio. For the government bond side, we avoid using regression models and instead use a yield-beta driven framework, taking forecasts for changes in US Treasury yields and translating those into changes in non-US bond yields by applying a historical yield beta (Table 2B). Table 2BEstimated Government Bond Yield Betas To US Treasuries Chart 14Risk Factor Assumptions For The Scenario Analysis For our scenario analysis over the next six months, we use a base case scenario plus two alternate “tail risk” scenarios, based on the following descriptions and inputs (Chart 14): Base Case: The US election result is initially uncertain, but a clear winner is determined within a few days. COVID cases continue to increase, but with less severe economic restrictions than during the first wave. Global growth continues to show steady improvement. There will be some additional global fiscal stimulus, with central banks keeping foot on monetary accelerator. There is mild bear steepening of the US Treasury curve with moderate widening of US inflation breakevens. The VIX reaches 25, the USD dollar depreciates by -5%, oil prices climb 10% and the fed funds rate remains at 0%. Based on how the models have performed in the COVID era, we believe we can use them again to forecast the expected relative returns of the credit side of the model bond portfolio. Optimistic Scenario: The US election goes smoothly and a clear winner is declared on election night. The current uptick in global COVID cases does not turn into a full-blown second wave requiring severe economic restrictions. Global growth continues to steadily improve, with additional global fiscal stimulus and central banks staying highly dovish. The US Treasury curve bear steepens as US inflation expectations steadily increase. The VIX falls to 20, the USD dollar depreciates by -7%, oil prices climb 20%, and the fed funds rate stays at 0%. Table 3AGFIS Model Bond Portfolio Scenario Analysis For The Next Six Months Table 3BUS Treasury Yield Assumptions For The 6-Month Forward Scenario Analysis Chart 15US Treasury Yield Assumptions For The Scenario Analysis Pessimistic Scenario: There is a contested US election result taking weeks to resolve, leading to major US social unrest. A full-blown second COVID-19 wave hits the world and severe economic restrictions are implemented. Governments become more worried about debt/deficits and deliver underwhelming stimulus. Central banks do not provide enough additional stimulus to offset the shocks. The US Treasury curve bull-flattens as US inflation breakevens plunge. The VIX soars to 35, the USD dollar rise by 5%, oil prices fall -20%, while the fed funds rate remains at 0%. The excess return scenarios for the model bond portfolio, using the above inputs in our simple quantitative return forecast framework, are shown in Table 3A (the scenarios for the changes in US Treasury yields are shown in Table 3B and Chart 15). The model bond portfolio is expected to deliver an excess return over the next six months of +17bps in the base case and +27bps in the optimistic scenario, but is only projected to underperform by -1bp in the pessimistic scenario. Bottom Line: We continue to prefer keeping aggregate portfolio duration close to benchmark, with only a moderate overweight allocation to spread product versus government bonds, given the lingering uncertainties over the global spread of COVID-19 and near-term US election risk. Instead, we recommend focusing on relative value allocations, favoring countries and sectors that will benefit most in our base case medium-term scenario of slowly improving global growth, reflationary global monetary/fiscal policies, low bond yield volatility and a softening US dollar.   Robert Robis, CFA Chief Fixed Income Strategist rrobis@bcaresearch.com Ray Park, CFA Research Analyst ray@bcaresearch.com Footnotes 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, "How To Play The Revival Of lobal Inflation Expectations", dated June 23 2020, available at gfis.bcaresearch.com. 3 Note that sectors where we made changes to our recommended weightings during Q3/2020 will have multiple colors in the respective bars in Chart 4. 4 Please see BCA Research Global Fixed Income Strategy Weekly Report, "We’re All Yield Chasers Now", dated August 11, 2020, available at gfis.bcaresearch.com. 5 The 12-month breakeven spread measures the amount of spread widening that must take place for a credit product to have the same return over a one-year horizon as a duration-matched position in government bonds. We compare those breakeven spreads to their own history in a percentile ranking to determine the relative attractiveness of a credit product strictly from a spread and spread volatility perspective. Appen dix Appendix Chart 1US Investment Grade Sectors Appendix Chart 2US High-Yield Credit Tiers Appendix Chart 3US MBS & CMBS Appendix Chart 4Euro Area And UK Credit Appendix Chart 5Emerging Markets USD-Denominated Debt Recommendations The GFIS Recommended Portfolio Vs. The Custom Benchmark Index Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Highlights Chart 1Spending Held Up In August The bulk of the CARES act’s income support provisions expired at the end of July and Congress has still not reached consensus on a follow-up package. Unsurprisingly, consumer spending responded by growing much more slowly in August, but at least so far, absolute calamity has been avoided (Chart 1). The failure of consumer spending to collapse has caused some, like St. Louis Fed President Jim Bullard, to question whether more stimulus is even necessary.1 We are less optimistic. The most recent personal income report shows that households still received $867 billion (annualized) of CARES act stimulus in August and the recovery in consumer confidence has been tepid at best (see page 12), suggesting that the savings rate will not drop quickly. We expect Congress to ultimately deliver more fiscal support, which will lead to a bear-steepening Treasury curve and spread product outperformance on a 6-12 month horizon. But continued brinkmanship warrants a more cautious near-term stance. Feature Investment Grade: Overweight Chart 2Investment Grade Market Overview Investment grade corporate bonds underperformed the duration-equivalent Treasury index by 40 basis points in September, dragging year-to-date excess returns down to -394 bps. Last month’s sell-off caused some value to return to the sector. The overall index’s 12-month breakeven spread is back up to its 31st percentile since 1995 and the equivalent Baa spread is at its 38th percentile (Chart 2). Both levels appear somewhat expensive at first blush. However, considering the strong tailwinds from the Fed’s extraordinarily accommodative interest rate policy and emergency lending facilities, we see a lot of room for further spread tightening. Corporate bond issuance was up in August, but nowhere near the extreme levels seen in the spring (panel 4). The fact that the Financing Gap – the difference between capital expenditures and retained earnings – turned negative in the second quarter suggests that firms have sufficient cash to cover their investment needs, and that further debt issuance is unnecessary (bottom panel). At the sector level, we continue to recommend overweight allocations to subordinate bank bonds,2 Healthcare and Energy bonds.3 We also advise underweight allocations to Technology4 and Pharmaceutical bonds.5   Table 3ACorporate Sector Relative Valuation And Recommended Allocation* Table 3BCorporate Sector Risk Vs. Reward* High-Yield: Neutral Chart 3High-Yield Market Overview High-Yield underperformed the duration-equivalent Treasury index by 107 basis points in September, dragging year-to-date excess returns down to -455 bps. Oddly, Ba-rated was the worst performing credit tier on the month and the lowest-rated (Caa & below) credits actually beat the Treasury benchmark by 42 bps. As we wrote last week, this suggests that there remains scope for low-rated junk to sell off in the event of a shock to economic growth expectations.6 Such a development could arise if Congress fails to pass a new stimulus bill. In terms of value, if we assume a 25% recovery rate on defaulted debt and a minimum required spread of 150 bps in excess of default losses, then the High-Yield index is priced for a default rate of 4.8% during the next 12 months (Chart 3). Such a large drop in the default rate would necessitate a rapid economic recovery and we are not yet confident that such a recovery can be achieved. Job Cut Announcements – a variable that correlates tightly with the default rate – ticked higher in September and they remain well above pre-COVID levels (bottom panel). At the sector level, we advise overweight allocations to high-yield Technology7 and Energy bonds.8 We are underweight the Healthcare and Pharmaceutical sectors.9   MBS: Underweight Chart 4MBS Market Overview Mortgage-Backed Securities underperformed the duration-equivalent Treasury index by 14 basis points in September, dragging year-to-date excess returns down to -51 bps. The conventional 30-year MBS index option-adjusted spread (OAS) widened 4 bps on the month, and it continues to trade at a premium compared to other similarly risky sectors. The MBS index OAS is currently 80 bps. This compares to an OAS of 79 bps for Aa-rated corporate bonds, 66 bps for Agency CMBS and 30 bps for Aaa-rated consumer ABS. Despite the OAS advantage, we are concerned that the elevated primary mortgage spread is a warning that refinancing risk could flare during the next few months (Chart 4). Even if Treasury yields are unchanged, a further 50 bps drop in the mortgage rate due to spread compression cannot be ruled out. Such a move would lead to a significant increase in prepayment losses. With that in mind, we are concerned about the low level of expected prepayment losses (option cost) priced into the MBS index (panel 3). A fourth quarter refi wave would undoubtedly send that option cost higher, eating into the returns implied by the OAS. The recent spike in the mortgage delinquency rate does not pose a near-term risk to spreads as it is being driven by households that have been granted forbearance from the federal government (panel 4). The risk for MBS holders only comes into play if many households are unable to resume their regular mortgage payments when the forbearance period expires early next year. But even in that case, further government intervention to either support household incomes or extend the forbearance period would mitigate the risk.   Government-Related: Underweight Chart 5Government-Related Market Overview The Government-Related index underperformed the duration-equivalent Treasury index by 18 basis points in September, dragging year-to-date excess returns down to -313 bps. Sovereign debt underperformed duration-equivalent Treasuries by 99 bps on the month, dragging year-to-date excess returns down to -562 bps. Foreign Agencies underperformed the Treasury benchmark by 13 bps in September, dragging year-to-date excess returns down to -706 bps. Local Authority debt underperformed Treasuries by 4 bps in September, dragging year-to-date excess returns down to -341 bps. Domestic Agency bonds outperformed by 15 bps, bringing year-to-date excess returns up to -39 bps. Supranationals underperformed by 3 bps, dragging year-to-date excess returns down to -12 bps. US dollar weakness is usually a boon for Emerging Market (EM) Sovereign and Foreign Agency returns. However, most of this year’s dollar depreciation has occurred against other Developed Market currencies, not EMs (Chart 5). Added to that, dollar weakness against all trading partners helps US corporate sector profits, and Baa-rated corporate bonds continue to offer a spread pick-up versus EM Sovereigns (panel 4). We looked at EM Sovereign valuation on a country-by-country basis two weeks ago and concluded that Mexican and Russian Sovereigns offer the most compelling risk/reward trade-offs relative to the US corporate sector.10 Of those two countries, Mexican debt offers the best opportunity as the peso is on an appreciating trend versus the dollar. The Russian Ruble has been depreciating versus the dollar, and is vulnerable in the case of a Democratic sweep in November.     Municipal Bonds: Overweight Chart 6Municipal Market Overview Municipal bonds underperformed the duration-equivalent Treasury index by 12 basis points in September, dragging year-to-date excess returns down to -503 bps (before adjusting for the tax advantage). Short-dated municipal bond spreads versus Treasuries were stable in September, but long-maturity spreads widened. The entire Aaa muni curve remains above the Treasury curve, despite municipal debt’s tax-exempt status (Chart 6). Municipal bonds also remain attractively priced relative to corporate bonds across the entire investment grade credit spectrum. Aaa munis offer more after-tax yield than Aaa corporates for investors facing an effective tax rate above 15%. The breakeven effective tax rates for Aa, A and Baa-rated munis are 11%, 13% and 17%, respectively. Extremely attractive valuation causes us to stick with our municipal bond overweight, even as state and local governments face a credit crunch. State & local government payrolls shrank in September and, without federal support, cutbacks will no doubt continue (bottom panel). However, we expect that the combination of austerity measures and all-time high State Rainy Day Fund balances will be sufficient to prevent a wave of municipal ratings downgrades. Treasury Curve: Buy 5-Year Bullet Versus 2/10 Barbell Chart 7Treasury Yield Curve Overview The Treasury curve bull-flattened somewhat in September, though even the 30-year yield only fell 3 bps on the month. The 2/10 and 5/30 Treasury slopes flattened 2 bps and 3 bps, reaching 56 bps and 118 bps, respectively. One easy way to think about nominal Treasury yields is as the market’s expectation of future changes in the fed funds rate.11 With that in mind, the Fed’s recent shift toward a regime of average inflation targeting will likely lead to nominal yield curve steepening on a 6-12 month horizon. That is, the Fed will keep a firm grip on the front-end of the curve but long-maturity yields will rise as investors price-in eventual Fed tightening in response to higher inflation. We recommend positioning for this outcome by owning the 5-year Treasury note and shorting a duration-matched barbell consisting of the 2-year and 10-year notes. This position is designed to profit from 2/10 curve steepening. We expect the economic recovery to be maintained over the next 6-12 months, allowing this steepening to play out. However, we also see near-term risks related to the passage of a follow-up stimulus bill. Those not already invested in steepeners are advised to wait until a deal is struck. Valuation is a concern with our recommended curve steepener, as the 5-year yield is below the yield on the duration-matched 2/10 barbell (Chart 7). However, the 5-year yield looked much more expensive during the last zero-lower-bound period between 2010 and 2013 (bottom 2 panels). We anticipate a return to similar valuation levels. TIPS: Overweight Chart 8TIPS Market Overview TIPS underperformed the duration-equivalent nominal Treasury index by 54 basis points in September, dragging year-to-date excess returns down to -130 bps. The 10-year and 5-year/5-year forward TIPS breakeven inflation rates fell 18 bps and 16 bps on the month. They currently sit at 1.65% and 1.83%, respectively. Core CPI printed a strong +0.4% in August and the large divergence between core and trimmed mean inflation measures leads us to conclude that inflation will continue to rise quickly during the next few months (Chart 8). For this reason, we recommend maintaining an overweight allocation to TIPS versus nominal Treasuries for the time being, even though the 10-year TIPS breakeven rate is no longer cheap according to our Adaptive Expectations Model (panel 2).12 We could see inflation pressures moderating once core and trimmed mean inflation measures re-converge.13 This could give us an opportunity to reduce our exposure to TIPS sometime later this year. We also recommend holding real yield curve steepeners and inflation curve flatteners. With the Fed now officially targeting an overshoot of its 2% inflation goal, we would expect the cost of 2-year inflation protection to rise above the cost of 10-year inflation protection (panel 4). With the Fed also exerting more control over short-dated nominal yields than over long-term ones, this means that short-maturity real yields will come under downward pressure relative to the long end (bottom panel). ABS: Overweight Chart 9ABS Market Overview Asset-Backed Securities outperformed the duration-equivalent Treasury index by 10 basis points in September, bringing year-to-date excess returns up to +63 bps. Aaa-rated ABS outperformed the Treasury benchmark by 7 bps on the month, bringing year-to-date excess returns up to +53 bps. Non-Aaa ABS outperformed by 32 bps, bringing year-to-date excess returns up to +128 bps. Aaa ABS are a high conviction overweight, given that spreads remain elevated compared to historical levels and that the sector benefits from Fed support through the Term Asset-Backed Loan Facility (TALF). However, spreads are even more attractive in non-Aaa ABS (Chart 9) and we recommend owning those securities as well. This is despite the fact that only Aaa-rated bonds are eligible for TALF. We explained our rationale for owning non-Aaa consumer ABS in a June report.14 We noted that stimulus received from the CARES act caused disposable income to increase significantly between February and July. Then, faced with fewer spending opportunities, households used much of that windfall to pay down consumer debt (panel 4). Granted, further income support from fiscal policymakers is needed now that the CARES act’s enhanced unemployment benefits have expired. But given the substantial boost to savings that has already occurred, we are confident that more stimulus will arrive in time to prevent a wave of consumer bankruptcies. Non-Agency CMBS: Neutral Chart 10CMBS Market Overview Non-Agency Commercial Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 63 basis points in September, bringing year-to-date excess returns up to -259 bps. Aaa Non-Agency CMBS outperformed Treasuries by 46 bps on the month, bringing year-to-date excess returns up to -63 bps. Non-Aaa Non-Agency CMBS outperformed by 119 bps, bringing year-to-date excess returns up to -803 bps (Chart 10). We continue to recommend an overweight allocation to Aaa Non-Agency CMBS and an underweight allocation to Non-Aaa CMBS. Our reasoning is simple. Aaa CMBS are eligible for TALF, meaning that spreads can still tighten even as the hardship in commercial real estate continues. Without Fed support, Non-Aaa CMBS will struggle to deal with a climbing delinquency rate (panel 3).15 Agency CMBS: Overweight Agency CMBS underperformed the duration-equivalent Treasury index by 9 basis points in September, dragging year-to-date excess returns down to -12 bps. The average index spread widened 2 bps on the month to 68 bps, well above typical historical levels (bottom panel). The Fed is supporting the Agency CMBS market by directly purchasing the securities as part of its Agency MBS purchase program. The combination of strong Fed support and elevated spreads makes the sector a high conviction overweight. Appendix A: Buy What The Fed Is Buying The Fed rolled out a number of aggressive lending facilities on March 23. These facilities focused on different specific sectors of the US bond market. The fact that the Fed has decided to support some parts of the market and not others has caused some traditional bond market correlations to break down. It has also led us to adopt of a strategy of “Buy What The Fed Is Buying”. That is, we favor those sectors that offer attractive spreads and that benefit from Fed support. The below Table tracks the performance of different bond sectors since the March 23 announcement. We will use this to monitor bond market correlations and evaluate our strategy’s success. Table Performance Since March 23 Announcement Of Emergency Fed Facilities Appendix B: Butterfly Strategy Valuations The following tables present the current read-outs from our butterfly spread models. We use these models to identify opportunities to take duration-neutral positions across the Treasury curve. The following two Special Reports explain the models in more detail: US Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com US Bond Strategy Special Report, “More Bullets, Barbells And Butterflies”, dated May 15, 2018, available at usbs.bcaresearch.com Table 4 shows the raw residuals from each model. A positive value indicates that the bullet is cheap relative to the duration-matched barbell. A negative value indicates that the barbell is cheap relative to the bullet. Table 4Butterfly Strategy Valuation: Raw Residuals In Basis Points (As Of October 2nd, 2020) Table 5 scales the raw residuals in Table 4 by their historical means and standard deviations. This facilitates comparison between the different butterfly spreads. Table 5Butterfly Strategy Valuation: Standardized Residuals (As Of October 2nd, 2020) Table 6 flips the models on their heads. It shows the change in the slope between the two barbell maturities that must be realized during the next six months to make returns between the bullet and barbell equal. For example, a reading of 63 bps in the 5 over 2/10 cell means that we would only expect the 5-year to outperform the 2/10 if the 2/10 slope steepens by more than 63 bps during the next six months. Otherwise, we would expect the 2/10 barbell to outperform the 5-year bullet. Table 6Discounted Slope Change During Next 6 Months (BPs) Appendix C: Excess Return Bond Map The Excess Return Bond Map is used to assess the relative risk/reward trade-off between different sectors of the US bond market. It is a purely computational exercise and does not impose any macroeconomic view. The Map’s vertical axis shows 12-month expected excess returns. These are proxied by each sector’s option-adjusted spread. Sectors plotting further toward the top of the Map have higher expected returns and vice-versa. Our novel risk measure called the “Risk Of Losing 100 bps” is shown on the Map’s horizontal axis. To calculate it, we first compute the spread widening required on a 12-month horizon for each sector to lose 100 bps or more relative to a duration-matched position in Treasury securities. Then, we divide that amount of spread widening by each sector’s historical spread volatility. The end result is the number of standard deviations of 12-month spread widening required for each sector to lose 100 bps or more versus a position in Treasuries. Lower risk sectors plot further to the right of the Map, and higher risk sectors plot further to the left. Chart 11Excess Return Bond Map (As Of October 2nd, 2020)   Ryan Swift US Bond Strategist rswift@bcaresearch.com Footnotes 1https://www.bloomberg.com/news/articles/2020-09-30/fed-s-bullard-says-debate-on-fiscal-aid-can-be-delayed-to-2021?sref=Ij5V3tFi 2 Please see US Bond Strategy Weekly Report, “The Case Against The Money Supply”, dated June 30, 2020, available at usbs.bcaresearch.com 3 Please see US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 1: A Model Of Energy Bond Excess Returns”, dated July 14, 2020 and US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 2: Buy The Dip In High-Yield Energy”, dated July 21, 2020, available at usbs.bcaresearch.com 4 Please see US Bond Strategy Weekly Report, “Take A Look At High-Yield Technology Bonds”, dated June 23, 2020, available at usbs.bcaresearch.com 5 Please see US Bond Strategy Weekly Report, “Assessing Healthcare & Pharma Bonds In A Pandemic”, dated June 9, 2020, available at usbs.bcaresearch.com 6 Please see US Bond Strategy Weekly Report, “Out Of Bullets”, dated September 29, 2020, available at usbs.bcaresearch.com 7 Please see US Bond Strategy Weekly Report, “Take A Look At High-Yield Technology Bonds”, dated June 23, 2020, available at usbs.bcaresearch.com 8 Please see US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 1: A Model Of Energy Bond Excess Returns”, dated July 14, 2020 and US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 2: Buy The Dip In High-Yield Energy”, dated July 21, 2020, available at usbs.bcaresearch.com 9 Please see US Bond Strategy Weekly Report, “Assessing Healthcare & Pharma Bonds In A Pandemic”, dated June 9, 2020, available at usbs.bcaresearch.com 10 Please see US Bond Strategy Weekly Report, “Trading Bonds In A Dollar Bear Market”, dated September 22, 2020, available at usbs.bcaresearch.com 11 For more details on this forecasting framework please see US Bond Strategy Weekly Report, “Positioning For Reflation And Avoiding Deflation”, dated August 11, 2020, available at usbs.bcaresearch.com 12 For more details on our model please see US Bond Strategy Weekly Report, “How Are Inflation Expectations Adapting?”, dated February 11, 2020, available at usbs.bcaresearch.com 13 Please see US Bond Strategy Weekly Report, “Positioning For Reflation And Avoiding Deflation”, dated August 11, 2020, available at usbs.bcaresearch.com 14 Please see US Bond Strategy Weekly Report, “No Holding Back”, dated June 16, 2020, available at usbs.bcaresearch.com 15 For a deeper dive into the outlook for US commercial real estate please see Global Investment Strategy Special Report, “Working From Home, Urban Flight, And Commercial Real Estate Loans: How Bad Can Things Get?”, dated August 28, 2020, available at gis.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification Corporate Sector Relative Valuation And Recommended Allocation

Highlights Bond Yields & Growth: Developed market bond yields have ignored improving cyclical economic data over the past few months, remaining stuck in narrow trading ranges at low levels. That broken correlation will persist until central banks become less concerned about supporting pandemic-ravaged economies and begin worrying more about rising inflation, financial stability or the size of their balance sheets. That shift will not happen anytime soon. Inflation-Linked Trades: Our models suggest US TIPS breakevens are now at fair value. We are taking profits on our tactical long US 10-year inflation breakevens trade for a return of 2.88%. Stay long 10-year breakevens in Italy and Canada until we see further shrinkage in the gap between inflation breakevens and model-implied fair value and watch for a selling opportunity in UK 10-year breakevens. Feature Do bond investors even care about economic growth anymore? This is a valid question to ask, given how government bond yields in the developed markets have stayed in very narrow trading ranges over the past few months, even as economic data has rebounded from the global COVID-19 recession in the first half of 2020. Investors should get used to the current backdrop of rock-bottom interest rates and bond yields, which is unlikely to change anytime soon.  Chart of the WeekBond Yields Are Responding To Inflation, Not Growth For example, the benchmark 10-year US Treasury yield has stayed between 0.65% and 0.75% since June 11, even though the US ISM Manufacturing index rose from 43 in May to 56 in August. Yields are also ignoring the ups and downs of the equity market. The 10-year Treasury yield now sits at 0.66% - the same level as on September 2 even though the NASDAQ equity index has fallen 12% from the all-time peak seen on that day. Our own Global Duration Indicator, comprised of cyclical measures like the global ZEW index and our global leading economic indicator, has surged to the highest level since 2008 (Chart of the Week). Given the usual lead time between broad turns in the Duration Indicator and the level of global bond yields (around 6-9 months), this suggests that yields have bottomed and should soon begin rising. Yet the reality is that the usual factors that typically drive yields higher during a cyclical upturn – namely, rising inflation expectations and a clearly understood signal from central banks that such a move would lead to tighter monetary policy – are not currently in place. Investors should get used to the current backdrop of rock-bottom interest rates and bond yields, which is unlikely to change anytime soon. Four Potential Triggers For A Rise In Bond Yields Chart 2A Breakdown Of The PMI/Yield correlation The breakdown of the positive correlation between growth and bond yields is not just visible in the US. For example, yields on German Bunds and UK Gilts also remain stuck at low levels despite sharp improvements in the German and UK manufacturing PMIs (Chart 2). Yet in China – where there is no zero interest rate policy (ZIRP) or large-scale quantitative easing (QE) programs - bond yields have steadily risen since the China manufacturing PMI bottomed back in April (bottom panel). What could change this backdrop? We see four potential catalysts, ranked below in our own subjective order of importance: Inflation Sustainably Returning Back To Central Bank Targets It may seem obvious, but it still needs to be said – dovish central bank policies are the biggest reason why developed market bond yields have de-linked from economic growth. That includes not only ZIRP or QE, but also forward guidance on future changes in interest rates. Central banks are telling markets they will not raise rates for a period measured in years, and will continue to expand their balance sheets to purchase assets and support bank lending, all in an effort to push undershooting inflation back to policy targets. This is a different message than bond investors have grown accustomed to hearing from central banks, most notably in the US. The Fed is trying to do something that it has never intentionally done before – erode some of its hard-earned inflation fighting credibility. The Fed is trying to do something that it has never intentionally done before – erode some of its hard-earned inflation fighting credibility. The recent shift by the Fed to an Average Inflation Targeting framework – where above-target inflation would be tolerated if inflation was below target for an extended period – is intended to change the perception that the Fed will hike rates preemptively based on a forecast of inflation, as they have done in the past. Chart 3Latest FOMC Projections Justify Years Of 0% Rates The latest set of Fed economic projections is consistent with this new framework (Chart 3): the unemployment rate is forecasted to fall back to the FOMC median estimate of full employment (4.1%) by 2023; headline PCE inflation is also projected to climb back to 2% by 2023; the fed funds rate is projected to stay unchanged near 0% until at least 2023. In many ways, the Fed is trying to atone for the mistakes made while normalizing policy after the extraordinary easing measures taken after the 2008 crisis. From signaling a slowing of QE bond purchases in 2013, to the 250bps of rate hikes and tapering of its balance sheet during 2016-18, the Fed moved aggressively relative to what was actually happening with US inflation. Core PCE inflation only inched above 2% for a few months in 2018 – towards the end of the normalization process - as did market-based inflation measures like TIPS breakevens (Chart 4). The Fed ended up raising the real fed funds rate during that tightening cycle to above its own estimate of neutral (r-star), even with inflation still not close to its target. Unsurprisingly, real US bond yields also rose during that same period, which tightened monetary conditions even further by boosting the value of the US dollar. No wonder US inflation could not stay at the 2% target for very long. This time around, the Fed is sending a much different signal to markets – that it wants to see inflation rise before raising rates, thus keeping real policy rates in negative territory for an extended period. If the Fed is looking for a real world case study of such an approach, it can look across the Atlantic to the Bank of England (BoE). On the surface, the BoE has been acting like a typical inflation-targeting central bank over the past several years, turning more hawkish in its commentary when the UK economy was improving and becoming more dovish when the economy was languishing. Yet since the 2008 crisis, the BoE has kept the Bank Rate in a range of 0.1% to 0.75%, well below realized UK inflation. While it has been difficult for the BoE to attempt to raise rates given the Brexit uncertainty since 2016 – which has also weakened the British pound, helping boost UK inflation - real UK policy rates have now been negative for 12 years (Chart 5). The result: steadily declining UK real bond yields with inflation expectations rising to levels well above the BoE 2% inflation target. Chart 4The Fed Is Trying To Erode Its Hard-Earned Credibility Chart 5Lessons From The BoE On How To Not Be Credible The experience of the ECB provides a cautionary tale for central banks not appearing dovish enough, even when policy settings are already extraordinarily accommodative. The message from central banks on future rate increases – namely, that there will not be any without sustainably higher inflation – must change before bond yields can have any hope of climbing higher. Chart 6Does The ECB Have Any Credibility Left? Inflation expectations have stayed below the ECB’s “just below 2%” target since 2013 (Chart 6), which forced the central bank into cutting nominal rates into negative territory while aggressively expanding its balance sheet through QE and long-term bank liquidity provision (i.e. LTROs). Yet the ECB has always put an expiration date on each of these programs, which sent a message to the markets that the central bank was not fully committed to keeping policy easy until inflation was back to target – however long that would take. In sum, the message from central banks on future rate increases – namely, that there will not be any without sustainably higher inflation – must change before bond yields can have any hope of climbing higher. A Shift From Central Banks To Concerns About Asset Price Bubbles Chart 7When Will CBs Start Worrying About Financial Market Valuations? Policymakers are paying lip service to the notion of the “financial stability” risks inherent in their new promises to keep rates low for a lot longer while intervening in financial markets more aggressively through asset purchase programs. Given the signs of froth in many important asset classes like US equities or global corporate debt, policymakers should at least be somewhat concerned that easy money policies are fueling asset bubbles (Chart 7). A big enough decline could erode confidence and spill over into the real economy, defeating the original purpose of easy money policies. However, given the still fragile state of much of the global economy that remains dependent on fiscal support amid ongoing COVID-19 restrictions, concerns over asset values will take a backseat to maintaining adequate monetary stimulus. Asset bubbles would have to become much larger before a central bank would even consider turning more hawkish to prick them through higher policy rates that would push up bond yields. The Announcement Of A Trustworthy COVID-19 Vaccine That Is Ready For Widespread Distribution Markets have already begun to worry about the “second wave” of the coronavirus that health officials had warned would happen in the cooler autumn months. The development of an effective, and safe, vaccine would thus be a game-changer for financial markets, particularly after the recent surge in new COVID-19 cases in Europe and the still elevated level of new cases in the US (Chart 8). Chart 8A Second Wave Of COVID-19 BCA Research’s Chief Global Strategist, Peter Berezin (a big fan of interesting data sets!), noted in his most recent report that, according to The Good Judgement Project, around 60% of “superforecasters” now expect a vaccine ready for mass distribution to be available by Q1/2021 (Chart 9).1 A vaccine appearing that rapidly – much faster than the usual multi-year process leading to a vaccine declared safe for use – would help boost business and consumer confidence and raise the odds of a return to pre-virus levels of economic activity. Bond yields would likely get a lift, as well, as markets would price in a shorter period of super low policy rates and a faster return of inflation to central bank targets. Yet even if a vaccine is presented to the world by next spring, there is no guarantee that a large enough share of the population will deem the vaccine safe enough to take to ensure “herd immunity”. A recent Economist/YouGuv survey noted that only 36% of American adults would choose to get vaccinated when a COVID-19 vaccine becomes available, 32% would not get vaccinated, while 32% were unsure (Chart 10). Thus, a vaccine would be a bond-bearish development only if it is trusted to be safe to use – the mere announcement of a vaccine will not be enough to declare an “end” to the pandemic. Chart 9High Odds Of A Vaccine In 6-To-12 Months Chart 10Will Enough People Take The Vaccine? Central Banks Slowing QE Purchases Relative To Increased Fiscal Issuance Chart 11Still Room For The Fed, ECB and BoE To Expand QE Right now, it is easy for the major central banks to aggressively expand their balance sheets and provide additional monetary stimulus through asset purchases. Yet there may come a point where a capacity constraint is reached on buying government bonds if it impairs market functionality. That is currently the case in Japan, where the Bank of Japan now owns 49% of the Japanese government bond (JGB) market after years of aggressive QE purchases of JGBs. This has damaged the day-to-day liquidity of JGBs, where there have been instances of days where no single JGB has traded in the secondary market. A move by central banks to buy fewer bonds because they own too many of them could potentially push bond yields higher by worsening the demand/supply balance for government bonds - assuming private investors do not pick up the slack and buy more bonds, of course. Currently, the Fed only owns 22% of the US Treasury market with little evidence suggesting that its purchases are impairing the trading of Treasuries (Chart 11). The BoE and ECB own much larger shares of the UK and euro area government bond markets – 37% and 38%, respectively – suggesting that those central banks are closer to a BoJ-like capacity constraint. However, given the rising budget deficits and surging government bond issuance seen in Europe (and the US) so far in 2020, the odds of a capacity constraint soon being reached that could result in slower QE purchases are low. Bottom Line: Developed market bond yields have ignored improving cyclical economic data over the past few months, remaining stuck in narrow trading ranges at low levels. That broken correlation will persist until central banks become less concerned about supporting pandemic-ravaged economies and begin worrying more about rising inflation, financial stability or the size of their balance sheets. That shift will not happen anytime soon. Reviewing Our Tactical Inflation Breakeven Trades Back in June, we initiated a series of recommended inflation-focused trades in our Tactical Overlay portfolio. Specifically, we went long 10-year inflation breakevens in the US, Italy, and Canada by buying on-the-run inflation-linked bonds and selling government bond futures.2 We chose those trades based on the output of our fundamental valuation models for 10-year inflation breakevens in eight inflation-linked bond (ILB) markets: the US, UK, France, Italy, Japan, Germany, Canada, and Australia. Our fair value models use two inputs for all regions: a) a long-run moving average of headline inflation, representing the medium-term trend that anchors inflation expectations; and b) the annual percentage change of the Brent oil price in local currency terms, which creates shorter-term deviations from the trend to account for moves in oil and currencies. There looks to be little remaining upside to our tactical long TIPS breakeven position. The past few months have seen a sharp rise in global inflation expectations, owing to the extraordinary monetary policy actions taken by the major developed market central banks and recovering growth prospects coming out of the COVID-19 recession. This has led to a convergence between 10-year inflation breakevens and their model-implied fair values in the aforementioned ILB markets (Chart 12). Most notably, breakevens in the US are now at fair value, while breakevens in the UK and Australia are trading above fair value. In the US, 10-year breakeven inflation rates are now back to the long-run average of realized headline inflation, while the -8% decline in the Brent oil price so far this month has lowered the model-implied fair value (Chart 13). Therefore, there looks to be little remaining upside to our tactical long TIPS breakeven position with most of the easy gains following the pandemic-induced collapse having already been realized. Chart 12Global Inflation Breakevens Have Moved Higher Our colleagues over at BCA Research US Bond Strategy have reached a similar conclusion, noting that the Fed’s Jackson Hole announcement of the move to Average Inflation Targeting supercharged the rising trend in TIPS breakevens.3 Chart 13US Breakevens Are At Fair Value Although they also note the likelihood of stronger US CPI prints over the next few months should keep US breakevens well supported heading into year-end. The time horizon for trades that enter our Tactical Overlay portfolio is limited to no longer than six months. Thus, with TIPS breakevens reverting back to fair value after just three months in the trade, we are choosing to take profits on our long 10-year US inflation breakeven trade for a total return of 2.88%. Chart 14UK Breakevens Are Above Fair Value In other ILB markets, UK breakevens are now an intriguing case, and not only for the monetary policy driven interplay between UK real yields and breakevens discussed earlier in this report. The overshoot of UK breakevens relative to our fair value model may be related to growing market speculation that the BoE will move to negative interest rates – an outcome we deem to be unlikely, as we discussed in a recent report.4 Alternatively, the higher breakevens may be a reflection of UK political uncertainty. The risk of a hard Brexit has resurfaced as UK Prime Minister Boris Johnson’s Conservatives have now backed a bill that includes powers for the government to override its withdrawal agreement with the European Union; understandably, this has caused a sell-off in the pound. Within our fundamental fair value framework, the UK 10-year breakeven inflation rate has overshot both the 3-year moving average of headline inflation and the growth of GBP-denominated oil prices, leaving breakevens 0.72 standard deviations expensive (Chart 14). One possible explanation is that markets are pricing in a significant further depreciation in the pound given this resurfacing of Brexit risk. Within our model, GBP/USD impacts the fair value of breakeven inflation via Brent oil prices, which are denominated in local currency terms. Thus, we can back out an implied change in GBP/USD that would make the model-derived fair value breakeven rate equal to the actual 10-year UK inflation breakeven rate, holding all other variables in the model constant. This does produce some extreme results during periods of very rapid moves in UK breakevens, but we can standardize the data to use as an indicator of ILB market-implied views on the currency (Chart 15). With that in mind, pound bearishness in ILB markets is nearing levels where it has historically troughed. A favorable development in Brexit negotiations could cause a reversal in this pound-bearish trend and a sharp downward correction in UK inflation breakevens. We see a potential opportunity to play for narrower UK breakevens if our view on Brexit and negative rates in the UK prove to be correct. On that front, BCA Research’s Chief Geopolitical Strategist, Matt Gertken, sees a no-deal Brexit by year-end as the less likely outcome, with odds of only 35%, given the political calculus that PM Johnson faces with the decision.5 Polls show that the UK public does not support a no-deal Brexit (Chart 16), which would severely hurt a UK economy that remains fragile due to the coronavirus, and would raise the odds of a new independence referendum in Scotland in 2021. Chart 15UK Breakevens Already Discount A Big Fall In GBP Chart 16Only 25% In The UK Think A No-Deal Brexit Is A Good Outcome We will monitor the situation closely in the coming weeks, but we see a potential opportunity to play for narrower UK breakevens if our view on Brexit and negative rates in the UK prove to be correct. Finally, although the majority of the gains from our long inflation breakeven trades in Canada and Italy have likely been realized, there are still some chips left on the table. Canadian breakeven inflation rates have risen in lockstep with Brent prices but have yet to converge with the long-run moving average of inflation (Chart 17). In Italy, the increases in oil prices in euro terms has outstripped the rise in breakevens, pushing up the model-implied fair value and leaving breakevens remain more than one standard deviation under fair value (Chart 18). We will look for the gap between breakevens and fair values to shrink further in these two countries before closing these trades, even though we are substantially in the green on both (see the Tactical Overlay table on page 19). Chart 17Canadian Breakevens Are Just Below Fair Value Chart 18Italian Breakevens Are Well Below Fair Value Bottom Line: Our models suggest US TIPS breakevens are now at fair value. We are taking profit on our tactical long US 10-year inflation breakeven trade for a return of 2.88%. Stay long 10-year breakevens in Italy and Canada until we see further shrinkage in the gap between inflation breakevens and model-implied fair value and watch for a selling opportunity in UK 10-year breakevens.   Robert Robis, CFA Chief Fixed Income Strategist rrobis@bcaresearch.com   Shakti Sharma Research Associate ShaktiS@bcaresearch.com Footnotes 1 Please see BCA Global Investment Strategy Weekly Report, "Pivot To Value", dated September 18, 2020, available at gis.bcaresearch.com. You can also learn more about The Good Judgement Project here: https://goodjudgment.com/about/ 2 Please see BCA Research Global Fixed Income Strategy Weekly Report, "How To Play The Revival Of Global Inflation Expectations", dated June 23, 2020, available at gfis.bcaresearch.com. 3 Please see BCA Research US Bond Strategy Portfolio Allocation Summary, "The Fed’s New Framework Is Bond Bearish…But Not Yet", dated September 8, 2020, available at usbs.bcaresearch.com. 4 Please see BCA Research Global Fixed Income Strategy Weekly Report, "Assessing The Leading Candidates To Join The Negative Rate Club", dated August 26, 2020, available at gfis.bcaresearch.com. 5 Please see BCA Research Geopolitical Strategy Weekly Report, "The End-Game For Trump And Brexit", dated September 18, 2020, available at gis.bcaresearch.com. Recommendations The GFIS Recommended Portfolio Vs. The Custom Benchmark Index Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Highlights Chart 1Permanent Job Losses Still Rising The biggest event in bond markets last month was the Fed’s shift toward a regime of average inflation targeting. Treasuries sold off in the days following the announcement and, overall, the Bloomberg Barclays Treasury index underperformed cash by 111 basis points in August (Chart 1). We view this market reaction as sensible, since it seems clear that the Fed’s new commitment to tolerate an overshoot of its 2% inflation target will be bearish for bonds in the long run. However, for this bond bear market to play out the US economy must first generate some inflation. This will take time. Despite the drop in the headline U3 unemployment rate, August’s employment report showed that permanent job losses continue to rise (bottom panel). This is a clear sign that the economic recovery is not yet on a solid footing. We advise bond investors to keep portfolio duration close to benchmark for the time being. We also recommend several yield curve trades across the nominal, real and inflation compensation curves (see pages 10 & 11). Feature Investment Grade: Overweight Chart 2Investment Grade Market Overview Investment grade corporate bonds outperformed the duration-equivalent Treasury index by 5 basis points in August, bringing year-to-date excess returns up to -356 bps. Spreads on Baa-rated corporate bonds continued their tightening trend through August, even as spreads were roughly flat for bonds rated A and above. As a result, Baa-rated bonds outperformed duration-matched Treasuries by 30 bps on the month while higher-rated credits underperformed. Valuation remains more attractive for the Baa space than for higher-rated credits (Chart 2), but spreads for all credit tiers look cheaper than they did near the end of 2019. Given the Fed’s strong support for the market through both its emergency lending facilities, and now, its extraordinarily dovish forward rate guidance, we see further room for spread compression across all credit tiers. At the sector level, we continue to recommend a focus on high-quality Baa-rated issuers. That is, Baa-rated bonds that are unlikely to face a ratings downgrade during the next 12 months. Subordinate bank bonds are a prime example of debt that falls into this sweet spot.1 We also recommend overweight allocations to Healthcare and Energy bonds2 and underweight allocations to Technology3 and Pharmaceutical bonds.4 Table 3ACorporate Sector Relative Valuation And Recommended Allocation* Table 3BCorporate Sector Risk Vs. Reward* High-Yield: Neutral Chart 3High-Yield Market Overview High-Yield outperformed the duration-equivalent Treasury index by 121 basis points in August, bringing year-to-date excess returns up to -351 bps. All junk credit tiers delivered strong returns in August, but the lowest-rated credits performed best. Caa-rated & below junk bonds outperformed Treasuries by 255 bps on the month compared to 98 bps of outperformance for Ba-rated bonds (Chart 3). The recent strong performance of low-rated junk bonds makes us question whether our focus on the Ba-rated credit tier is overly conservative. If the economy is indeed on a quick road to recovery, then we are leaving some return on the table by avoiding the B-rated and lower credit tiers. However, we aren’t yet confident enough in the economic recovery to move down in quality. Last week’s employment report showed that permanent job losses continue to rise and Congress has still not passed a much needed follow-up to the CARES act. What’s more, current junk spreads imply a very rapid decline in the corporate default rate during the next 12 months, from its current level of 8.4% all the way to 4.4% (panel 3).5  In this regard, August’s steep drop in layoff announcements is a positive development (bottom panel), though job cuts are still running well above pre-pandemic levels. At the sector level, we advise overweight allocations to high-yield Technology6 and Energy7 bonds. We are underweight the Healthcare and Pharmaceutical sectors.8 MBS: Underweight Chart 4MBS Market Overview Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 9 basis points in August, bringing year-to-date excess returns up to -37 bps. The conventional 30-year MBS index option-adjusted spread (OAS) tightened 7 bps in August, but it still offers a small spread pick-up compared to other similarly risky sectors. The MBS OAS of 77 bps is greater than the 75 bps offered by Aa-rated corporate bonds, the 67 bps offered by Agency CMBS and the 35 bps offered by Aaa-rated consumer ABS. Despite the spread advantage, we are concerned that the elevated primary mortgage spread is a warning that refinancing risk could flare later this year  (Chart 4). Even if Treasury yields are unchanged, a further 50 bps drop in the mortgage rate due to spread compression cannot be ruled out. Such a move would lead to a significant increase in prepayment losses. With that in mind, we are concerned about the low level of expected prepayment losses (option cost) priced into the MBS index (panel 3). A fourth quarter refi wave would undoubtedly send that option cost higher, eating into the returns implied by the OAS. The recent spike in the mortgage delinquency rate does not pose a near-term risk to spreads as it is being driven by households that have been granted forbearance from the federal government (panel 4). The risk for MBS holders only comes into play if many households are unable to resume their regular mortgage payments when the forbearance period expires early next year. But even in that case, further government action to either support household incomes or extend the forbearance period could mitigate the risk. Government-Related: Underweight Chart 5Government-Related Market Overview The Government-Related index outperformed the duration-equivalent Treasury index by 31 basis points in August, bringing year-to-date excess returns up to -295 bps. Sovereign debt outperformed duration-equivalent Treasuries by 105 bps on the month, bringing year-to-date excess returns up to -468 bps. Foreign Agencies outperformed the Treasury benchmark by 13 bps in August, bringing year-to-date excess returns up to -694 bps. Local Authority debt outperformed Treasuries by 33 bps in August, bringing year-to-date excess returns up to -337 bps. Domestic Agency bonds outperformed by 8 bps, bringing year-to-date excess returns up to -54 bps. Supranationals outperformed by 5 bps, bringing year-to-date excess returns up to -9 bps. US dollar weakness is usually a boon for Sovereign and Foreign Agency returns. However, most of the dollar’s recent depreciation has occurred against other Developed Market currencies, not Emerging Markets (Chart 5). Added to that, dollar weakness against all trading partners helps US corporate sector profits, and Baa-rated corporate bonds continue to offer a spread pick-up versus EM sovereigns (panel 4). Within the Emerging Market Sovereign space: Turkey, South Africa, Mexico, Colombia and Russia all offer a spread pick-up relative to quality and duration-matched US corporate bonds. Of those attractively priced countries, Mexico stands out as particularly compelling on a risk/reward basis.9 Municipal Bonds: Overweight Chart 6Municipal Market Overview Municipal bonds underperformed the duration-equivalent Treasury index by 19 basis points in August, dragging year-to-date excess returns down to -492 bps (before adjusting for the tax advantage). Municipal bond spreads versus Treasuries have widened during the past month, more so at the long-end than at the short-end, and the entire Aaa muni curve remains above the Treasury curve, despite municipal debt’s tax-exempt status (Chart 6). Municipal bonds also remain attractively priced relative to corporate bonds across the entire investment grade credit spectrum, as we demonstrated in a recent report.10 The Fed reduced the pricing on its Municipal Liquidity Facility (MLF) by 50 basis points last month. Most likely, it felt pressure to act as Congress has still not passed a state & local government aid package. However, the Fed’s move will not have much impact on municipal bond spreads. Even after the reduction, municipal yields continue to run well below the cost offered by the MLF (panel 3). Extremely attractive valuation causes us to stick with our municipal bond overweight, though spreads will widen in the near-term if much needed stimulus doesn’t arrive soon. In the long-run, we remain optimistic that elevated state rainy day funds will help cushion the fiscal blow and lessen the risk of ratings downgrades (bottom panel).    Treasury Curve: Buy 5-Year Bullet Versus 2/10 Barbell Chart 7Treasury Yield Curve Overview The Treasury curve bear-steepened in August. The 2/10 and 5/30 Treasury slopes steepened 14 bps and 22 bps, reaching 58 bps and 121 bps, respectively. One easy way to think about nominal Treasury yields is as the market’s expectation of future changes in the federal funds rate.11 With that in mind, the Fed’s recent shift toward a regime of average inflation targeting will likely lead to nominal yield curve steepening. That is, the Fed will keep a firm grip on the front-end of the curve, but long-maturity yields could rise as investors price-in the possibility that the Fed will have to eventually respond to high inflation by quickly tightening policy. For this reason, we retain a core position in nominal yield curve steepeners. Specifically, we recommend buying the 5-year bullet and shorting a duration-matched 2/10 barbell. This position is designed to profit from 2/10 Treasury curve steepening, which should play out over the next 6-12 months, assuming the economic recovery is sustained. Valuation is a concern with this recommended positioning. The 5-year yield is below the yield on the duration-matched 2/10 barbell (Chart 7), and the 5-year bullet looks expensive on our yield curve models (Appendix B).     However, the 5-year bullet traded at much more expensive levels during the last zero-lower-bound period between 2010 and 2013 (bottom panel). With short rates once again pinned at zero, we expect the 5-year to once again hit extreme levels of overvaluation. TIPS: Overweight Chart 8TIPS Market Overview TIPS outperformed the duration-equivalent nominal Treasury index by 240 basis points in August, bringing year-to-date excess returns up to -76 bps. The 10-year and 5-year/5-year forward TIPS breakeven inflation rates rose 25 bps and 22 bps on the month. They currently sit at 1.67% and 1.78%, respectively. TIPS breakeven inflation rates have moved up rapidly during the past couple months, a trend that was supercharged by the Fed’s Jackson Hole announcement. In fact, the 10-year TIPS breakeven inflation rate is now right around fair value according to our Adaptive Expectations Model (Chart 8).12 TIPS will soon turn expensive if current trends continue. That is, unless stronger CPI inflation sends our model’s fair value reading higher. We place strong odds on the latter occurring during the next few months, with trimmed mean inflation measures still running well above core (panel 3). However, we cautioned in a recent report that inflation is likely to moderate in 2021 after core inflation re-converges with the trimmed mean.13 In addition to our overweight stance on TIPS, we continue to recommend real yield curve steepeners and inflation curve flatteners. With the Fed now officially targeting an overshoot of its 2% inflation goal, we would expect the cost of 2-year inflation protection to rise above the cost of 10-year inflation protection (panel 4). With the Fed also keeping a firmer grip over short-dated nominal yields than over long-dated ones, this means that short-maturity real yields will come under downward pressure relative to the long end (bottom panel). ABS: Overweight Chart 9ABS Market Overview Asset-Backed Securities outperformed the duration-equivalent Treasury index by 30 basis points in August, bringing year-to-date excess returns up to +53 bps. Aaa-rated ABS outperformed the Treasury benchmark by 24 bps on the month, bringing year-to-date excess returns up to +46 bps. Non-Aaa ABS outperformed by 73 bps, bringing year-to-date excess returns up to +95 bps. Aaa ABS are a high conviction overweight, given that spreads remain elevated compared to historical levels and that the sector benefits from Fed support through the Term Asset-Backed Loan Facility (TALF). However, spreads are even more attractive in non-Aaa ABS (Chart 9) and we recommend owning those securities as well. This is despite the fact that only Aaa-rated bonds are eligible for TALF. We explained our rationale for owning non-Aaa consumer ABS in a recent report.14  We noted that the stimulus received from the CARES act caused real disposable personal income to increase significantly between February and July and, faced with fewer spending opportunities, households used that windfall to pay down consumer debt (bottom panel). Granted, further income support from fiscal policymakers is needed now that the CARES act’s enhanced unemployment benefits have expired. But given the substantial boost to savings that has already occurred, we are confident that more stimulus will arrive in time to prevent a wave of consumer bankruptcies.  Non-Agency CMBS: Neutral Chart 10CMBS Market Overview Non-Agency Commercial Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 77 basis points in August, bringing year-to-date excess returns up to -320 bps. Aaa Non-Agency CMBS outperformed Treasuries by 57 bps on the month, bringing year-to-date excess returns up to -108 bps. Non-Aaa Non-Agency CMBS outperformed by 160 bps, bringing year-to-date excess returns up to -1008 bps (Chart 10). We continue to recommend an overweight allocation to Aaa non-agency CMBS and an underweight allocation to non-Aaa CMBS. Our reasoning is simple. Aaa CMBS are eligible for TALF, meaning that spreads can still tighten even as the hardship in commercial real estate continues. Without Fed support, non-Aaa CMBS will struggle to deal with a climbing delinquency rate (panel 3).15 Agency CMBS: Overweight Agency CMBS outperformed the duration-equivalent Treasury index by 38 basis points in August, bringing year-to-date excess returns up to -4 bps. The average index spread tightened 6 bps on the month to 66 bps, still well above typical historical levels (bottom panel). The Fed is supporting the Agency CMBS market by directly purchasing the securities as part of its Agency MBS purchase program. The combination of strong Fed support and elevated spreads makes the sector a high conviction overweight. Appendix A: Buy What The Fed Is Buying The Fed rolled out a number of aggressive lending facilities on March 23. These facilities focused on different specific sectors of the US bond market. The fact that the Fed has decided to support some parts of the market and not others has caused some traditional bond market correlations to break down. It has also led us to adopt of a strategy of “Buy What The Fed Is Buying”. That is, we favor those sectors that offer attractive spreads and that benefit from Fed support. The below Table tracks the performance of different bond sectors since the March 23 announcement. We will use this to monitor bond market correlations and evaluate our strategy’s success. TablePerformance Since March 23 Announcement Of Emergency Fed Facilities​​​​​​​ Appendix B: Butterfly Strategy Valuations The following tables present the current read-outs from our butterfly spread models. We use these models to identify opportunities to take duration-neutral positions across the Treasury curve. The following two Special Reports explain the models in more detail: US Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com US Bond Strategy Special Report, “More Bullets, Barbells And Butterflies”, dated May 15, 2018, available at usbs.bcaresearch.com Table 4 shows the raw residuals from each model. A positive value indicates that the bullet is cheap relative to the duration-matched barbell. A negative value indicates that the barbell is cheap relative to the bullet. Table 4Butterfly Strategy Valuation: Raw Residuals In Basis Points (As Of September 3, 2020) Table 5 scales the raw residuals in Table 4 by their historical means and standard deviations. This facilitates comparison between the different butterfly spreads. Table 5Butterfly Strategy Valuation: Standardized Residuals (As Of September 3, 2020) Table 6 flips the models on their heads. It shows the change in the slope between the two barbell maturities that must be realized during the next six months to make returns between the bullet and barbell equal. For example, a reading of 72 bps in the 5 over 2/10 cell means that we would only expect the 5-year to outperform the 2/10 if the 2/10 slope steepens by more than 72 bps during the next six months. Otherwise, we would expect the 2/10 barbell to outperform the 5-year bullet. Table 6Discounted Slope Change During Next 6 Months (BPs) Appendix C: Excess Return Bond Map The Excess Return Bond Map is used to assess the relative risk/reward trade-off between different sectors of the US bond market. It is a purely computational exercise and does not impose any macroeconomic view. The Map’s vertical axis shows 12-month expected excess returns. These are proxied by each sector’s option-adjusted spread. Sectors plotting further toward the top of the Map have higher expected returns and vice-versa. Our novel risk measure called the “Risk Of Losing 100 bps” is shown on the Map’s horizontal axis. To calculate it, we first compute the spread widening required on a 12-month horizon for each sector to lose 100 bps or more relative to a duration-matched position in Treasury securities. Then, we divide that amount of spread widening by each sector’s historical spread volatility. The end result is the number of standard deviations of 12-month spread widening required for each sector to lose 100 bps or more versus a position in Treasuries. Lower risk sectors plot further to the right of the Map, and higher risk sectors plot further to the left. Chart 11Excess Return Bond Map (As Of September 3, 2020)   Ryan Swift US Bond Strategist rswift@bcaresearch.com Footnotes 1 Please see US Bond Strategy Weekly Report, “The Case Against The Money Supply”, dated June 30, 2020, available at usbs.bcaresearch.com 2 Please see US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 1: A Model Of Energy Bond Excess Returns”, dated July 14, 2020 and US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 2: Buy The Dip In High-Yield Energy”, dated July 21, 2020, available at usbs.bcaresearch.com 3 Please see US Bond Strategy Weekly Report, “Take A Look At High-Yield Technology Bonds”, dated June 23, 2020, available at usbs.bcaresearch.com 4 Please see US Bond Strategy Weekly Report, “Assessing Healthcare & Pharma Bonds In A Pandemic”, dated June 9, 2020, available at usbs.bcaresearch.com 5 We assume a 25% recovery rate and target a spread of 150 bps in excess of default losses. For more details on this calculation please see US Bond Strategy Weekly Report, “No Holding Back”, dated June 16, 2020, available at usbs.bcaresearch.com 6 Please see US Bond Strategy Weekly Report, “Take A Look At High-Yield Technology Bonds”, dated June 23, 2020, available at usbs.bcaresearch.com 7 Please see US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 1: A Model Of Energy Bond Excess Returns”, dated July 14, 2020, and US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 2: Buy The Dip In High-Yield Energy”, dated July 21, 2020, available at usbs.bcaresearch.com 8 Please see US Bond Strategy Weekly Report, “Assessing Healthcare & Pharma Bonds In A Pandemic”, dated June 9, 2020, available at usbs.bcaresearch.com 9 Please see US Bond Strategy Weekly Report, “The Treasury Market Amid Surging Supply”, dated May 12, 2020, available at usbs.bcaresearch.com 10 Please see US Bond Strategy Weekly Report, “Bonds Are Vulnerable As North America Re-Opens”, dated May 26, 2020, available at usbs.bcaresearch.com 11 For more details on this forecasting framework please see US Bond Strategy Weekly Report, “Positioning For Reflation And Avoiding Deflation”, dated August 11, 2020, available at usbs.bcaresearch.com 12 For more details on our model please see US Bond Strategy Weekly Report, “How Are Inflation Expectations Adapting?”, dated February 11, 2020, available at usbs.bcaresearch.com 13 Please see US Bond Strategy Weekly Report, “Positioning For Reflation And Avoiding Deflation”, dated August 11, 2020, available at usbs.bcaresearch.com 14 Please see US Bond Strategy Weekly Report, “No Holding Back”, dated June 16, 2020, available at usbs.bcaresearch.com 15 For a deeper dive into the outlook for US commercial real estate please see Global Investment Strategy Special Report, “Working From Home, Urban Flight, And Commercial Real Estate Loans: How Bad Can Things Get?”, dated August 28, 2020, available at gis.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification Corporate Sector Relative Valuation And Recommended Allocation

Special Report

Highlights Fed Policy Changes: The official shift to an average inflation targeting regime represents a massive structural break relative to how the Fed conducted monetary policy in the past. The main takeaway for investors should be that inflation expectations will carry more weight than ever in the Fed’s thinking, with far less emphasis on estimated measures like the output gap. Investment Implications: The Fed’s new policy framework supports our current US fixed income recommendations: a neutral duration stance; overweighting TIPS versus nominal US Treasuries; positioning for real yield curve steepeners; and overweighting US spread product most directly supported by the Fed’s balance sheet (i.e. investment grade corporates and Ba-rated high-yield). Feature The pandemic forced the Federal Reserve to move its annual Jackson Hole Economic Policy Symposium online this year.  That change deprived policymakers of a late-August vacation in the mountains of Wyoming, but offered the public a rare glimpse at the full proceedings live on YouTube.1 Federal Reserve Chairman Jerome Powell took advantage of that larger audience to announce significant changes to the Fed’s Statement on Longer-Run Goals and Monetary Policy Strategy. Though many of the basic elements of the new strategy were well telegraphed in advance, the adjustments are hugely significant and will shape the conduct of US – and, potentially, global - monetary policy for years to come. This Special Report presents the most important takeaways – and fixed income investment implications - from the Fed’s new approach to setting monetary policy. Say Hello To Average Inflation Targeting The most significant change has to do with how the Fed defines its price stability mandate. In its old Statement, the Fed defined its 2% inflation target as “symmetrical”, meaning that the Fed would be equally concerned if inflation were running persistently above or below the target. In the Fed’s words, communicating this symmetry was enough to “keep longer-term inflation expectations firmly anchored.” The Fed now believes that a more aggressive approach is required to keep inflation expectations anchored. The new Statement reads: In order to anchor longer-term inflation expectations at [2 percent], the Committee seeks to achieve inflation that averages 2 percent over time, and therefore judges that, following periods when inflation has been running persistently below 2 percent, appropriate monetary policy will likely aim to achieve inflation moderately above 2 percent for some time.2 In other words, the Fed’s 2% inflation target is no longer purely forward-looking. It is now dependent on the history of realized US inflation, and thus is now much more like a price level target than an inflation target. We will know that the Fed has seen enough inflation overshooting when long-term expectations are anchored at levels consistent with its 2% inflation target. For example, Chart 1 shows how the headline PCE price index would have evolved since the end of 2007 had it averaged 2% growth per year, exactly equal to the Fed’s target. Starting from today, PCE inflation would need to average 3% for the next seven years, or 2.5% for the next fourteen years, for the index to converge with this target. In other words, if the Fed seeks to achieve average 2% inflation since 2007, we are in for a prolonged period of overshooting the old 2% target. Chart 1An Illustration Of Average Inflation Targeting Notice that we had to make several assumptions in our above example. First, we had to assume that the Fed will seek to achieve average 2% inflation since the end of 2007. The Fed could just as easily choose a different start date for calculating the 2% average. We also assumed that the year-over-year PCE inflation rate never breaks above 3% during the overshooting phase. As of now, we have no sense of whether the Fed would act to make sure that inflation only overshoots 2% by a small amount (say, between 0.5 and 1 percentage point) or whether it would tolerate a larger overshoot. A larger overshoot would potentially be more de-stabilizing, but it would allow the Fed to catch up to its price level target more quickly. We will probably get some more information about these missing details when the Fed translates its new framework into more explicit forward rate guidance (see section titled "Are There Any Additional Changes Coming?" below), but the Fed will still want to retain some flexibility. That is, we shouldn’t expect the Fed to tie its hands with a strict policy rule. This means that the question of how much inflation would prompt any future Fed tightening could linger for some time. Faced with this ambiguity, investors are advised to focus more keenly than ever on inflation expectations (Chart 2). Note that in the above excerpt from the revised Statement on Longer-Run Goals and Monetary Policy Strategy, the explicit goal of average inflation targeting is to “anchor long-term inflation expectations at [2 percent]”. This means that we will know that the Fed has seen enough inflation overshooting when long-term expectations are anchored at levels consistent with its 2% inflation target. We view this “well anchored” level as a range between 2.3% and 2.5% for long-dated TIPS breakeven inflation rates (top two panels). When TIPS breakevens reach those levels, we should expect the Fed to shift toward a more restrictive policy stance. Chart 2The Fed Wants Higher Inflation Expectations How long will it take for TIPS breakevens to reach our target range? We expect it will take quite some time because Fed communications alone cannot drive long-term TIPS breakevens back to target. Rather, inflation expectations tend to follow trends in the actual inflation data, so expectations will only return to well-anchored levels once inflation has risen significantly. Further, long-dated inflation expectations tend to adapt slowly to changes in the actual inflation data. Notice in Chart 3 that the 5-year/5-year forward CPI swap rate correlates much more strongly with the 8-year rate of change in CPI inflation than it does with the 1-year rate of change. This suggests that, most likely, 12-month inflation will have to run above 2% for some time before long-term TIPS breakevens sustainably return to our target range. One way to understand the link between actual inflation and inflation expectations is to look at the distribution of individual inflation forecasts. Chart 4 shows the distribution of 10-year headline CPI inflation forecasts from the Survey of Professional Forecasters from 2004 – a year when inflation expectations were well anchored around 2% – and from August 2020. Notice that a similar proportion of respondents at both points in time expect inflation to be near the Fed’s target, in a range of 2% to 2.5%. The difference is that, in 2004, a large minority of respondents anticipated a significant overshoot of the inflation target. Today, hardly anyone anticipates a significant overshoot, and many expect a significant undershoot. Chart 3Inflation Expectations Adapt Slowly To The Actual Data Chart 4Distribution Of Inflation Forecasts ##br##(2004 & Today) Since market prices can be thought of as a weighted average of the entire distribution of inflation forecasts, it follows that to drive TIPS breakevens higher we need to see investors shift their forecasts from the left tail of the distribution to the right tail. This will only happen if actual inflation rises, and probably only if it stays durably above 2% for a prolonged period. Chart 5shows that the percentage of respondents that expect inflation to average above 3% for the next ten years tends to follow both the long-run inflation rate and the median inflation forecast. Chart 5Few Expect Inflation To Be Above 3% Bottom Line:  The official shift to an average inflation targeting regime represents a massive structural break relative to how the Fed conducted monetary policy in the past. The main takeaway for investors should be that inflation expectations carry more weight than ever in the Fed’s thinking. In particular, we should expect the Fed to move to a more restrictive policy stance only when long-maturity TIPS breakeven inflation rates return to a well-anchored range of 2.3% to 2.5%. Some Key Questions Following The Fed’s Big Shift Does The Phillips Curve Still Matter? The second big change that the Fed made to its official Statement on Longer-Run Goals and Monetary Policy Strategy is in how it views the unemployment rate relative to its “natural” level. Specifically, the change has to do with making estimates of the natural rate of unemployment (NAIRU) less important in the Fed’s reaction function. In its old Statement, the Fed talked about minimizing “deviations of employment from the Committee’s assessments of its maximum level”. The revised Statement talks about mitigating “shortfalls of employment from the Committee’s assessment of its maximum level.” This one word change says a lot about the Fed’s faith in the Phillips curve. In the past, the Fed viewed an unemployment rate below its estimate of NAIRU as a signal that inflation was poised to accelerate. This often led to premature tightening, and over time, a pattern of missing the inflation target to the downside. Now, the Fed is explicitly saying that it only cares about shortfalls of employment from its estimated maximum level. If the labor market appears overheated, the Fed will not take this as a sign that inflation is about to accelerate. Rather, it will wait for the evidence to show up in the actual inflation data. The percentage of respondents that expect inflation to average above 3% for the next ten years tends to follow both the long run inflation rate and the median inflation forecast. This change sends a very clear signal that the Fed will put much less emphasis on expected “Phillips curve effects” in the future than it has in the past. In addition to long-term implications, this change will likely also impact the type of forward rate guidance the Fed provides this year. What’s Missing? It is also interesting to touch on the things that Powell did not mention in his Jackson Hole speech. First, as noted above, Powell provided few details on the length of time over which the Fed will seek to hit average 2% inflation and did not specify any upper limit to the amount of inflation the Fed would tolerate during the overshooting phase. Perhaps more importantly, Powell also did not say much about how the Fed will seek to drive inflation higher, and whether there are additional tools at his disposal that have not yet been rolled out. We think there is good reason for this. In effect, we think the Fed is more or less tapped out in terms of the amount of additional monetary easing it can provide. Negative interest rates have already been ruled out. A Yield Curve Control policy of capping intermediate-maturity bond yields has been discussed, but this policy doesn’t accomplish much beyond what the Fed is already doing with its forward rate guidance. For example, a policy of capping the 2-year Treasury yield at the current level of 0.13% has essentially the same impact on bond prices as convincing the market that the fed funds rate will stay in a range between 0% and 0.25% for the next two years or more. The notion that the Fed is “out of bullets” was hammered home during the final Jackson Hole panel on Friday. The speakers for the panel titled “Post-Pandemic Monetary Policy and the Effective Lower Bound” shifted much of the onus for boosting growth, with policy interest rates at the effective lower bound, toward fiscal policymakers. Given the limitations on the amount of additional easing that the Fed can deliver, the potent impact of the changes announced last week will not really be felt until the economic recovery is further underway. Only once inflation starts to rise will we get a test of the Fed’s resolve to stay on the sidelines. Now that the changes have been enshrined in an official Fed document, we have no doubt that they will follow through. What About The Role Of QE? Chart 6The Future Of QE: Go Big & Go Fast Not every speaker at Jackson Hole, however, felt that central banks had run out of policy options.  Bank of England (BoE) Governor Andrew Bailey gave a speech on Day Two of the conference that focused on the use of central bank balance sheets as a more regular part of policymakers’ toolkits over the next decade with policy rates at the effective lower bound. Bailey noted that the use of quantitative easing (QE) in the future would be less about signaling future central bank intentions on interest rates, or forcing changes to the composition of assets held by the private sector, and would be more about “going big and going fast” to calm financial markets during periods of instability.3 Some past examples of such use of QE include the 2008 Global Financial Crisis, the 2011/12 European Debt Crisis and the 2016 UK Brexit shock (Chart 6). In Bailey’s view, QE will now have to be “state contingent”, based on the nature of the financial market shock and where liquidity (cash) needs are greatest at that time.  In 2008, it was the banking system that needed liquidity, so central banks expanded their balance sheets in ways that got cash directly to the banks – like repos and government bond purchases.  In 2020, the demand for liquidity from the COVID-19 shock came more from non-bank entities, like investment funds or the corporate sector itself.  Therefore, central bank balance sheets had to be used to support loans to the private sector or even buying private assets like corporate debt, on top of the usual QE buying of sovereign debt to help drive down risk-free bond yields. What does that mean for the new policy regime of the Fed?  It means that the type of market intervention we saw earlier this year – with the Fed announcing a variety of measures to support liquidity like corporate bond purchases when markets were not functioning – will become more commonplace during periods of severe market stress.  This is because there cannot be any “emergency” Fed rate cuts to calm markets if the Fed is keeping rates at very low levels to try and make up for past undershoots of its inflation target. Chart 7The Fed Has Room To Do More QE In The Future This also means that the balance sheets of the Fed, and other major global central banks, will likely continue to get larger over time.  Tapering of balance sheets, as the Fed engineered during 2014-19, will become very rare events before inflation expectations are stabilized at policymaker targets.  That does raise issues of capacity constraints for QE programs, as Bailey mentioned in his speech, where the central bank footprint in financial markets becomes so large as to impair market functionality.  That is the case today where the Bank of Japan now owns nearly 50% of all outstanding Japanese government bonds (JGB) and the day-to-day liquidity in the JGB market is extremely challenging for market participants that need to buy and trade JGBs, like Japanese banks and investment funds.  Bailey noted that there was still ample capacity for the BoE to ramp up its buying of UK Gilts (and even UK corporate debt) before the sheer size of its presence became a BoJ-like problem for the UK bond market (Chart 7). The same can be argued in the US, where the Fed only owns a little over 20% of outstanding US Treasuries – the supply of which is growing rapidly thanks to large US budget deficits. Are There Any Additional Changes Coming? As we outlined in a recent US Bond Strategy Webcast, after revising the Statement on Longer-Run Goals and Monetary Policy Strategy, the Fed’s next step will be to provide more explicit guidance about the economic conditions that will have to be in place before it considers lifting the fed funds rate.4 We speculate that this next announcement will occur before the end of the year, possibly at this month’s FOMC meeting, and that the guidance will be similar to the Evans Rule employed in 2012. The Evans Rule was a promise that the Fed would not lift rates at least until the unemployment rate was below 6.5% or inflation was above 2.5%. For the 2020 version of the Evans Rule, policymakers had been debating whether to specify both an unemployment target and an inflation target, as was done in 2012, or whether to specify only an inflation target. With the Fed’s new Statement putting much less emphasis on Phillips curve effects and estimates of NAIRU, it now appears much more likely that the 2020 version of the Evans Rule will have only an inflation trigger, or perhaps an inflation trigger and an inflation expectations trigger. Bottom Line: There are still many lingering unanswered questions about the new Fed strategy, but what we do know is that the Fed will focus more on inflation, rather than forecasts of inflation, when making future interest rate decisions.  The Fed will also likely use its balance sheet more as a market stability tool during times of crisis. Investment Implications Chart 8Financial Conditions The first implication of the Fed’s big shift has to do with the long-run outlook for risk asset prices (corporate bonds, equities and other fixed income spread product). With the Fed committing to give the economic recovery more runway before choking it off, risk asset valuations have been provided with a massive tailwind. In fact, the longer it takes for inflation to move up, the longer the Fed will stay on hold and the more expensive risk asset valuations will become. It is even possible that, if inflation remains subdued for a few more years, risk asset valuations will become so stretched that the Fed might have to exercise its financial stability “out clause”. That is, if the Fed viewed a growing asset bubble as a threat to the economic recovery and/or financial system, it could abandon its inflation target and lift interest rates to deflate that bubble. This out clause is specifically enshrined in the Fed’s Statement on Longer-Run Goals and Monetary Policy Strategy: Moreover, sustainably achieving maximum employment and price stability depends on a stable financial system. Therefore, the Committee’s policy decisions reflect its longer-run goals, its medium-term outlook, and its assessments of the balance of risks, including risks to the financial system that could impede the attainment of the Committee’s goals. We should stress that US financial asset valuations are currently nowhere near expensive enough to prompt this sort of move (Chart 8). However, that picture could change after a few more years of low inflation and zero interest rates. We have been saying since March 2019 that the two most important indicators to watch for gauging the eventual pace of Fed tightening are inflation expectations and financial conditions.5 Last week’s announcement serves to reinforce that view. The Fed could abandon its inflation target and lift interest rates to combat a growing asset bubble. A second investment implication of the Fed’s announcement is that TIPS will continue to outperform nominal US Treasuries until there is an eventual re-anchoring of long-run TIPS breakeven inflation rates in a range between 2.3% and 2.5%. As noted above, this structural investment position could take some time to pan out, and we may even get an opportunity to tactically position for periods of TIPS underperformance if breakevens start to look too high compared to the actual inflation data.6 For now, our models suggest that TIPS breakevens are fairly valued relative to the actual inflation data, and we recommend staying overweight TIPS versus nominal Treasuries as a core allocation in fixed income portfolios. We would also advise investors to enter flatteners along the inflation protection curve (TIPS breakevens or CPI swaps). This recommendation flows directly from the Fed’s announcement. If the Fed is eventually successful at achieving a temporary overshoot of its 2% inflation target, then the cost of short-maturity inflation protection should rise above the cost of long-maturity inflation protection. That is, the inflation protection curve should invert (Chart 9). This would be a stark dislocation compared to the past, but it is a logical one if the Fed is to be attacking its inflation target from above instead of from below. As for nominal Treasury yields, our baseline view is that yields will be flat-to-higher over the next 12 months, with the amount of upside dictated by the pace of economic recovery. The Fed’s extraordinarily dovish monetary policy will keep some downward pressure on nominal yields, but expectations of Fed tightening will eventually infiltrate the long end of the curve. Given that the Fed’s grip is much firmer at the short end of the curve than at the long end, we prefer to play the nominal Treasury curve through yield curve steepeners rather than through outright duration bets (Chart 10). Chart 9Position For Inflation Curve Inversion Chart 10Enter Nominal Curve Steepeners Finally, the level of real yields is perhaps the trickiest to get right in the current environment. The Fed’s dovish policies are clearly meant to push real yields down, but now that those policies have been announced, it may signal that we are near the trough. In fact, real yields actually rose somewhat on Thursday after the Fed’s announcement. As with nominal yields, we prefer to play the real Treasury (TIPS) curve via steepeners (Chart 11). Whether or not the Fed is able to apply further downward pressure on real yields, as long as its policies are viewed as reflationary and the economic recovery is maintained, then the real yield curve has ample room to steepen. Chart 11Enter Real Curve Steepeners Bottom Line: The Fed’s new policy framework supports our current US fixed income recommendations: a neutral duration stance; overweighting TIPS versus nominal US Treasuries; positioning for real yield curve (TIPS) steepeners; and overweighting US spread product most directly supported by the Fed’s balance sheet (i.e. investment grade corporates and Ba-rated high-yield).   Ryan Swift US Bond Strategist rswift@bcaresearch.com   Robert Robis, CFA Chief Fixed Income Strategist rrobis@bcaresearch.com Footnotes 1 https://www.youtube.com/user/KansasCityFed 2 https://www.federalreserve.gov/monetarypolicy/guide-to-changes-in-statement-on-longer-run-goals-monetary-policy-strategy.htm 3 The full text of BoE Governor Bailey’s speech can be found here: https://www.bankofengland.co.uk/speech/2020/andrew-bailey-federal-reserve-bank-of-kansas-citys-economic-policy-symposium-2020 4 https://www.bcaresearch.com/webcasts/detail/338 5 Please see US Bond Strategy Weekly Report, “The New Battleground For Monetary Policy”, dated March 26, 2019, available at usbs.bcaresearch.com 6 This possibility is discussed in US Bond Strategy Weekly Report, “Positioning For Reflation And Avoiding Deflation”, dated August 11, 2020, available at usbs.bcaresearch.com

Highlights Nominal Yields: Nominal Treasury yields will move modestly higher during the next 6-12 months with the increase concentrated at the long-end of the curve. Investors should keep portfolio duration close to benchmark and enter duration-neutral yield curve steepeners. Inflation Compensation: Remain overweight TIPS versus nominal Treasuries for now, but we anticipate getting an opportunity to tactically reverse this position near the end of the year. Investors should also position in flatteners across the inflation compensation curve, as both a near-term and long-term trade. Real Yields: The outlook for the level of real yields is highly uncertain, particularly at the long-end of the curve. However, as long as the reflation trade continues, real yield curve steepeners should perform well whether real yields are rising or falling. US Economy: Another stimulus bill is required in order to extend the economic recovery and prolong the reflation trade in financial markets. The President’s executive orders are not sufficient. The pressure on Congress to reach a compromise deal is high, and we expect one to be announced in the coming days. Feature Chart 1Reflation Pushes Real Yields Lower Market movements during the past couple of months are consistent with an environment of economic reflation. Equities and commodity prices are up, the US dollar is down, spread product has outperformed Treasuries and TIPS breakeven inflation rates have widened. This “reflation trade” is the result of global economic recovery and highly accommodative Fed policy, the latter being particularly important. In fact, Fed policy has been so accommodative that bonds are the one asset class that has so far bucked the broader reflationary trend. Nominal Treasury yields dipped during the past few weeks, as rising inflation expectations were more than offset by plunging real yields (Chart 1). Our base case expectation is that, broadly speaking, the reflation trade will continue. Global economic growth will improve during the next 6-12 months and Fed policy will remain highly accommodative. In this week’s report we consider how to position for that outcome in US rates markets. In the process, we provide trade recommendations for the nominal, real and inflation compensation curves. We also consider the main risk to our reflationary view: The possibility that further US fiscal stimulus is too little or arrives too late. Positioning For Reflation Chart 2More Downside In Short-Maturity Real Yields Back in April, we explained how the Fed’s zero-lower-bound interest rate policy can lead to unusual movements in bond markets, particularly in how real bond yields respond to broader market trends.1  The importance of the zero lower bound is easily seen through the lens of the Fisher Equation – the equation that connects nominal yields, real yields and inflation expectations. Real Yield = Nominal Yield – Inflation Expectations If the Fed is expected to hold the nominal short rate steady for a long period of time, then nominal bond yields won’t move around very much in response to the economy. Necessarily, this means that increases in inflation expectations must be matched by falling real yields. Chart 1 shows how this has played out for 10-year yields, but the dynamic is even more pronounced at the short-end of the curve where the Fed has greater control over nominal rate expectations (Chart 2). With these relationships in mind, we consider the outlooks for the nominal, inflation compensation and real yield curves. Nominal Treasury Curve Chart 3Fed Guidance Has Crushed Nominal Rate Vol As is alluded to above, fed funds rate expectations drive nominal Treasury yields. Treasury yields rise when the market revises its rate expectations up and fall when the market revises its expectations down. But what happens when the Fed signals that the funds rate will stay pinned at its current level, even as inflationary pressures mount? What happens is that nominal bond yields become increasingly insensitive to fluctuations in economic data and rate volatility plunges (Chart 3). Not surprisingly, this decline in rate volatility has been more pronounced at the front-end of the curve than at the long-end (Chart 3, bottom panel). This is because the Fed’s rate guidance exerts more influence over short maturities. The market might be very confident that the fed funds rate will stay at its current level for the next year or two, but it will be less confident about rate expectations five or ten years down the road. The conclusion we draw is that the Fed’s dovish rate guidance will prevent a large increase in nominal bond yields, even as the reflation trade rolls on. But at some point, rising inflation expectations will cause the market to price-in policy firming at the long-end of the curve and long-maturity nominal Treasury yields will move somewhat higher. Historically, nominal bond yields usually move in the same direction as TIPS breakeven inflation rates (Chart 4). Chart 4Nominal Yields And Inflation Expectations Are Positively Correlated While this base case outlook calls for flat-to-slightly higher Treasury yields, we recommend keeping portfolio duration close to benchmark on a 6-12 month investment horizon. The reason for this caution is that significant downside risks to our base case economic scenario remain (see section “Avoiding Deflation” below). Chart 5Bullets Trade Expensive When Rates Are Pinned At Zero Instead, we recommend positioning for the continuation of the reflation trade via duration-neutral yield curve steepeners. The nominal yield curve will respond to global economic recovery by steepening because the market will price-in eventual policy tightening at the long-end of the curve before it prices-in near-term policy tightening at the front-end of the curve. Specifically, we suggest buying the 5-year bullet and shorting a duration-neutral barbell consisting of the 2-year and 10-year notes. This trade is designed to profit from steepening of the 2/10 yield curve.2 The one problem with our proposed trade is that it is not cheap. The 5-year bullet yield is below the 2/10 barbell yield and the 5-year bullet trades as expensive on our yield curve model (Chart 5). However, we note that the 5-year looked much more expensive at the height of the last zero-lower-bound episode in 2012. In today’s similar environment, we anticipate a return to similar valuation levels. Bottom Line: Nominal Treasury yields will move modestly higher during the next 6-12 months with the increase concentrated at the long-end of the curve. Investors should keep portfolio duration close to benchmark and enter duration-neutral yield curve steepeners. Inflation Compensation Curve Chart 6Adaptive Expectations Model Almost by definition, the continuation of the reflation trade means that the cost of inflation compensation will rise (i.e. TIPS breakeven inflation rates will move higher), and we remain positioned for that outcome. However, at least according to our Adaptive Expectations Model, the inflation component of bond yields could have a more difficult time rising going forward. Our model, which is based on several different measures of realized inflation, shows that the 10-year TIPS breakeven inflation rate is more or less at its fair value (Chart 6). In other words, further upside from here is contingent upon rising inflation. Fortunately, rising inflation seems likely during the next few months. Month-over-month headline CPI bottomed in April (Chart 7), the oil price is trending up (Chart 7, panel 2) and core inflation has undershot relative to the trimmed mean (Chart 7, panel 3). All of this suggests that our model’s fair value will move higher during the next few months. Chart 7Inflation Has Bottomed But beyond the near-term snapback that we anticipate, a wide output gap in the United States will prevent inflation from entering a sustainable uptrend as we head into 2021. After all, our Pipeline Inflation Indicator remains deep in deflationary territory (Chart 7, bottom panel).  At some point near the end of this year, we anticipate getting an opportunity to move tactically underweight TIPS versus nominal Treasuries, once breakevens start to look expensive on our model. Our Adaptive Expectations Model shows that the 10-year TIPS breakeven inflation rate is more or less at its fair value. A higher conviction long-run trade relates to the slope of the inflation curve. At present, the 10-year CPI swap rate remains somewhat above the 2-year rate, but we eventually expect this slope to invert (Chart 8). With the Fed explicitly targeting a temporary overshoot of its 2% inflation target, it would make sense for the cost of short-maturity inflation protection to trade above the cost of long-maturity inflation protection. This would mark a significant break from historical trends, but this is also true of the Fed’s new policy approach. Chart 8Inflation Curve Will Invert Bottom Line: Remain overweight TIPS versus nominal Treasuries for now, but we anticipate getting an opportunity to tactically reverse this position near the end of the year. Investors should also position in flatteners across the inflation compensation curve, as both a near-term and long-term trade. Real Yield Curve Chart 9Buy Real Yield Curve Steepeners At the beginning of this report we noted that the combination of stable nominal rate expectations and rising inflation expectations has led to a steep decline in real Treasury yields. This decline has been more severe at the short-end of the curve, which has resulted in real yield curve steepening (Chart 9). At the long-end of the curve, the outlook for the level of real yields is highly uncertain, even under the assumption that the reflation trade continues. If 10-year nominal rate expectations hold steady, then continued reflation will lead to a further decline in the 10-year real yield. However, as discussed above, long-dated nominal rate expectations will eventually follow inflation expectations higher. If that adjustment to long-dated rate expectations outpaces the increase in expected inflation compensation, then the 10-year real yield will move up as well. The outlook for the short-end of the curve is more certain. Two-year nominal rate expectations are unlikely to budge anytime soon. This means that the continuation of the reflation trade will send the cost of 2-year inflation protection higher and the 2-year real yield lower. For this reason, we would rather take a position in real yield curve steepeners than an outright position on the level of real yields. In fact, as long as the reflation trade continues, the real yield curve should steepen whether the absolute level of real yields is rising or falling. It is only in a renewed deflation scare where we would expect the real yield curve to flatten, as occurred back in March. As long as the reflation trade continues, real yield curve steepeners should perform well whether real yields are rising or falling. Bottom Line: The outlook for the level of real yields is highly uncertain, particularly at the long-end of the curve. However, as long as the reflation trade continues, real yield curve steepeners should perform well whether real yields are rising or falling. Avoiding Deflation The first part of this report talked about how to position in rates markets assuming that the global economic recovery remains on track and that the so-called reflation trade continues. While this is our base case scenario, it is by no means a certainty. In fact, this view is contingent upon continued US fiscal stimulus that is sufficient to sustain household income and prevent a snowballing of foreclosures and bankruptcies. March’s CARES act did a more-than-admirable job supporting household income. In fact, disposable household income rose 7.2% in the four month period between March and June compared to the four months that preceded the COVID recession (Chart 10). This is a far greater increase than what was seen in the first four months of the 2008 recession (dashed line in Chart 10, panel 2), despite the fact that the hit to wage compensation has been worse (dashed line in Chart 10, bottom panel). Chart 11A confirms that, without the CARES act, the hit to disposable income would have been substantial. Chart 10Income Well Supported... So Far Chart 11ADisposable Personal Income Growth And Its Drivers I The problem is that the main income supporting provisions of the CARES act have either been paid out or have expired. Chart 11B shows the impact on disposable income of the CARES act’s different provisions. The two most important were: The Economic Impact Payments: The one-time $1200 stimulus checks. The Pandemic Unemployment Compensation Payments: The extra $600 per week that was added to unemployment benefits. Chart 11BDisposable Personal Income Growth And Its Drivers II The Economic Impact Payments have all been delivered, and the Pandemic Unemployment Compensation Payments expired at the end of July. Based on the information that has been released about the ongoing negotiations over a follow-up stimulus bill, we expect that a compromise deal will be large enough to keep disposable income at or above pre-recession levels.3 However, a compromise is proving difficult. Congress’ foot dragging prompted President Trump to announce several executive orders of questionable legality in an attempt to deliver some stimulus. However, even if the executive orders are followed, the boost to household income will be meager without another bill. The President’s executive order to extend the extra unemployment benefits appropriates only $44 billion from the Disaster Relief Fund and asks states to contribute the rest. Many states will be unable to contribute anything, and an extra $44 billion amounts to only 8% of the income support provided by the CARES act. State & local government aid must be addressed in the new stimulus bill. The other urgent area that must be addressed in a follow-up stimulus bill is aid for state & local governments. State & local government spending fell 5.6% (annualized) in the second quarter, as governments have been forced to impose harsh austerity in the face of collapsing tax revenues (Chart 12). This is one area where the Democrats and Republicans are still far apart. The Center on Budget and Policy Priorities estimates that states need $555 billion to close COVID-related budget shortfalls.4 The Democrats’ initial proposal contained $1.13 trillion for states, the Republicans’ initial offer left out state & local government aid altogether. Chart 12State & Local Governments Need A Bailout Bottom Line: Another stimulus bill is required in order to extend the economic recovery and prolong the reflation trade in financial markets. The President’s executive orders are not sufficient. The pressure on Congress to reach a compromise deal is high, and we expect one to be announced in the coming days. Based on the numbers that have been floated, that deal will contain sufficient income support to keep households afloat and the recovery on track. Appendix A: Buy What The Fed Is Buying The Fed rolled out a number of aggressive lending facilities on March 23. These facilities focused on different specific sectors of the US bond market. The fact that the Fed has decided to support some parts of the market and not others has caused some traditional bond market correlations to break down. It has also led us to adopt of a strategy of “Buy What The Fed Is Buying”. That is, we favor those sectors that offer attractive spreads and that benefit from Fed support. The below Table tracks the performance of different bond sectors since the March 23 announcement. We will use this to monitor bond market correlations and evaluate our strategy’s success. Table 1Performance Since March 23 Announcement Of Emergency Fed Facilities Footnotes 1 Please see US Bond Strategy Weekly Report, “Negative Oil, The Zero Lower Bound And The Fisher Equation”, dated April 28, 2020, available at usbs.bcaresearch.com 2 To understand why this trade profits in an environment of yield curve steepening please see US Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com 3 In their initial proposals, House Democrats offered $435 billion in Economic Impact Payments and $437 billion for expanded unemployment benefits. The Senate Republicans offered $300 billion for Economic Impact Payments and $110 billion for expanded unemployment benefits. For context, the CARES act authorized $293 billion for Economic Impact Payments and $268 billion for expanded unemployment benefits. For more details on the ongoing budget negotiations please see Geopolitical Strategy Weekly Report, “A Tech Bubble Amid A Tech War”, dated July 31, 2020, available at gps.bcaresearch.com 4 https://www.cbpp.org/research/state-budget-and-tax/states-continue-to-face-large-shortfalls-due-to-covid-19-effects   Ryan Swift US Bond Strategist rswift@bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification

Highlights Global Bond Yields: The growing divide between falling negative real bond yields and rising inflation expectations in the US and other major developed economies may be a sign of investors pricing in slower long-run potential economic growth in the aftermath of the COVID-19 recession – and, thus, lower equilibrium real interest rates. Stay overweight inflation-linked bonds versus nominal equivalents. Currency-hedged spread product: A broad ranking of currency-hedged global spread product yields, adjusted for volatility and credit quality, shows that the most attractive yields (hedged into USD, EUR, GBP and JPY) are on offer in emerging market USD-denominated investment grade corporates and high-yield company debt in the US and UK. Feature Global bond yields are testing the downside of the narrow trading ranges that have persisted since May. As of last Friday, the yield on the Bloomberg Barclays Global Treasury index was at 0.41%, only 3 basis points (bps) above the 2020 low seen back in March. The 10-year US Treasury yield closed yesterday at 0.56%, only 6bps above the year-to-date low. Chart of the Week Concerns about global growth, with the number of new COVID-19 cases still surging in the US and new breakouts occurring in countries like Spain and Australia, would seem to be the logical culprit for the decline in yields. The first reads on global GDP data for the 2nd quarter released last week were historically miserable, with declines of -33% (annualized) in the US and -10% in the euro area (non-annualized). That represents a very deep hole of lost output, literally wiping out several years of growth. Even with the sharp improvements seen recently in cyclical indicators like global manufacturing PMIs, especially in China and Europe, a return to pre-pandemic levels of global economic output is many years away. Central banks will have no choice but to keep policy rates near 0% for at last the next couple of years, as is the current forward guidance provided by the Fed, ECB and others. Lower global bond yields may simply be reflecting the reality that it will take a long time to heal the economic wounds from the pandemic. However, there may be a more insidious reason why bond yields are falling. Investors may be permanently marking down their expectations for long-term potential economic growth, and equilibrium interest rates, in response to the devastation caused by the COVID-19 recession. Last week, Fitch Ratings lowered its estimates for long-term potential GDP growth, used to determine sovereign credit ratings, by 0.5 percentage points for the US (now 1.4%), 0.5 percentage points for the euro area (now 0.7%) and 0.7 percentage points in the UK (now 0.7%).1 These are declines similar in magnitude to the plunge in the OECD’s potential growth rate estimates seen after the 2009 Great Recession (Chart of the Week). Bond yields in the US and Europe witnessed a fundamental repricing in response, with nominal 5-year yields, 5-years forward breaking 200bps below the 4-6% range that prevailed in the US and Europe during the decade prior to the Great Recession. A similar re-rating of global bond yields to structurally lower levels may now be happening, with investors now believing that central banks will have difficulty raising rates much (if at all) in the future - even after the pandemic has ended. The Message From Declining Negative Real Bond Yields Chart 2The Real Rate/Breakevens Divergence Continues The typical signals about economic growth from government bond yields are now less clear because of the aggressive policy responses to the COVID-19 crisis. 0% policy rates, dovish forward guidance on the timing of any future rate increases, large scale asset purchases (QE), and more extreme measures like yield curve control to peg bond yields, have all acted to suppress the level and volatility of nominal global bond yields. Within those calm nominal yields, however, the dynamic that has been in place since May - rising inflation breakevens and falling real bond yields – is growing in intensity. The 10-year US TIPS real yield is now at a new all-time low of -1.02%, while the 10-year TIPS breakeven is now up to 1.58%, the highest since February before the pandemic began to roil financial markets (Chart 2). Similar trends are evident in most other major developed economy bond markets, with the gap between falling real yields and widening breakevens growing at a notably faster pace in Canada and Australia. More often than not, longer-term real yields tend to move in the same direction as inflation expectations when economic growth is improving. The former responds to faster economic activity, often with an associated pick up in private sector credit demand. At the same time, rising inflation expectations discount higher economic resource utilization (i.e. lower unemployment) and confidence that inflation will start to pick up. A deeply negative correlation between longer-term real yields and inflation expectations is unusual, but not unprecedented. A deeply negative correlation between longer-term real yields and inflation expectations is unusual, but not unprecedented. In Chart 3, we show the range of rolling three-year correlations between 10-year inflation-linked (real) government bond yields and 10-year inflation breakevens in the US, Germany, France, Italy, the UK, Japan, Canada and Australia for the post-crisis period. The triangles in the chart are the latest three-year correlation, while the diamonds are a more recent measure showing the 13-week correlation. There are a few key takeaways from this chart: Chart 3Negative Real Yield/Breakevens Correlations Are Not Unprecedented All countries shown have experienced a sustained period of negative correlation between real yields and inflation breakevens; The correlation has mostly been positive in Australia and has always been negative in Japan; Most importantly, the deeply negative correlations seen over the past three months – with rising breakevens all but fully offsetting falling real yields – are at or below the range of historical experience for all countries shown. Chart 4TIPS Yields May Stay Negative For Some Time In the current virus-stricken world, where many businesses that have closed during the pandemic may never reopen, there will be abundant spare global economic capacity for several years. In the US, measures of spare capacity like the unemployment gap (the unemployment rate minus the full-employment NAIRU rate) have been a reliable leading directional indicator of the long-run correlation between real TIPS yields and TIPS breakevens over the past decade (Chart 4). The surge in US unemployment seen since the spring, which has pushed the jobless rate into double-digit territory, suggests that the current deeply negative correlation between US real yields and inflation breakevens can persist over the next 6-12 months. Given the large increases in unemployment seen in other countries, the negative correlations between real yields and inflation breakevens should also continue outside the US. As for inflation expectations, those remain correlated in the short-run to changes in oil prices and exchange rates in all countries. On that front, there is still some room for breakevens to widen to reach the fair value levels implied by our models.2 A good conceptual way to think about inflation breakevens on a more fundamental level, however, is as a “vote of confidence” in a central bank’s monetary policy stance. If investors perceive policy settings to be too tight, markets will price in slower growth and lower inflation expectations, and vice versa. Every developed market central bank is now setting policy rates near or below 0% - and promising to keep them there until at least the end of 2022. Thus, the trend of rising global inflation breakevens can continue as a reflection of very dovish central banks that will be more tolerant of increases in inflation and not tighten policy pre-emptively. Currently, real 10-year inflation-linked bond yields are below the New York Fed’s estimates of the neutral real short-term rate, or “r-star”, in the US and the UK (Chart 5), as well as in the euro area and Canada (Chart 6).3 In the US and euro area, real yields have followed the broad trend of r-star, but the gap between the two is relatively moderate with r-star estimated to be only 0.5% in the US and 0.2% in the euro zone (where the ECB is setting a negative nominal interest rate on European bank deposits at the central bank – a policy choice that the Fed has been very reluctant to consider). Chart 5Negative Real Bond Yields Are Below R* In The US & UK ... Chart 6... As Well As In The Euro Area & Canada A more interesting study is in the UK where 10yr inflation-linked Gilt yields have fallen below -2.5%, but without the Bank of England implementing any negative nominal policy rates. In the UK, inflation expectations have been relatively high – running in the 2.5-3% range prior to the COVID-19 recession – as the Bank of England has consistently kept overnight interest rates below actual CPI inflation since the 2008 financial crisis. Thus, nominal Gilt yields have stayed relatively low for longer, as real yields and inflation expectations have remained negatively correlated for a long period with the Bank of England maintaining a consistently negative real policy rate. Chart 7Spillovers From Negative TIPS Yields Into Other Assets If the Fed were to do the same in the US, keeping the funds rate very low even as inflation rises, then a similar dynamic could take place where real TIPS yields continue to fall and TIPS breakevens continue to rise as the market prices in a sustained negative real fed funds rate. That may already be happening, with Fed Chair Jerome Powell hinting last week that the Fed is in the process of completing its inflation strategy review – with a shift towards rate hikes occurring only after realized inflation has sustainably increased to the Fed’s 2% target. A forecast of inflation heading to 2% because of falling unemployment will no longer be enough.4 Other factors may be at work depressing real bond yields while boosting inflation expectations, such as the massive QE bond buying programs of the Fed, ECB and other central banks. Yet even QE programs are essentially an aggressive form of forward guidance designed to drive down longer-term bond yields by lowering expectations of future interest rates. In sum, it is increasingly likely that the current phase of negative global real bond yields may become longer lasting if markets believe that equilibrium real policy rates are now negative. Bond investors will expect central banks to sit on their hands and do nothing in that environment, even if inflation starts to increase. This not only has implications for bond markets, but other asset classes as well based on what is happening in the US. The steady decline in the in the 10-year US TIPS yield has boosted the valuation of assets that typically have been considered inflation hedges, like equities and gold (Chart 7). The fall in TIPS yields also suggests that more weakness in the US dollar is likely to come over the next 6-12 months – another reflationary factor that should help lift global inflation expectations and boost the attractiveness of inflation-linked bonds. The current phase of negative global real bond yields may become longer lasting if markets believe that equilibrium real policy rates are now negative. Bottom Line: The growing divide between falling negative real bond yields and rising inflation expectations in the US and other major developed economies may be a sign of investors pricing in slower long-run potential economic growth in the aftermath of the COVID-19 recession – and, thus, lower equilibrium real interest rates. Stay overweight inflation-linked bonds versus nominal equivalents. Searching For Value In Global Spread Product Last week, we looked at the impact of currency hedging on the attractiveness of government bond yields across the developed markets.5 We concluded that US Treasuries still offered superior yields to most other countries’ sovereign bonds, even with the US dollar in a weakening trend and after hedging out currency risk. We also presented a cursory look at the relative attractiveness of the major global spread product categories in that report, but without factoring in any considerations on the relative credit quality or volatility between sectors. This week, we will look at the relative value of global spread products hedged into USD, GBP, EUR and JPY, but after controlling for those credit and volatility risks. We conducted a similar analysis in early 2018,6 ranking the currency-hedged yields for a wide variety of global spread products by the ratio of yields to trailing volatility. This time, instead of looking at the just that simple valuation metric, we use regression models to make a judgment on how under- or over-valued spread products are relative to their “fair value”. To recap the methodology of this analysis, we take the Bloomberg Barclays index yield-to-maturity (YTM) for each spread product category, hedged into the four currencies used in this analysis, and divide it by the annualized trailing volatility of those yields over both short-term (1-year) and long-term (3-year) windows. In order to hedge the yields into each currency, we used the annualized differentials between spot and 3-month forward exchange rates, which is the all-in cost of hedging. We then compare those currency-hedged, volatility-adjusted yields to two measures of risk: the index credit rating and duration times spread (DTS) for each spread product. Table 1 summarizes the attractiveness of each product when hedged into different currencies. The rank is based on the average of four different valuation measures.7 The higher the rank, the more attractive the sector is in terms of yield relative to risk measures such as both short-term and long-term volatilities, credit ratings, and DTS. Table 1Ranking Currency-Hedged, Risk-Adjusted Global Spread Product Yields A few interesting points come from the table: Emerging market (EM) USD-denominated investment grade (IG) corporate debt ranks at or near the top of the rankings, for all currencies; the opposite holds true for EM USD-denominated sovereign bonds Almost all European spread products rank poorly for non-euro denominated investors US & UK high-yield (HY) rank highly for all currencies US real estate related assets (MBS and CMBS) also rank well for all investor groups In general, US products are more attractive than European credit sectors. This is mainly because US spread products offer higher yields than European ones even after accounting for volatility and the weakening US dollar. Almost all European spread products rank poorly for non-euro denominated investors. Chart 8 shows the unhedged YTM on the x-axis and the option-adjusted spread (OAS) on the y-axis (Table 2 contains the abbreviations used in this chart and all remaining charts in this report). Unsurprisingly, the YTM and OAS follow a very tight linear relationship. However, when yields are hedged into different currencies and risk measures are factored in, the result changes. Chart 8Global Spread Product Yields & Spreads Charts 9A to 12B show the details of spread product analysis with different currency hedges and risk factors. To limit the number of charts shown, we show only currency-hedged yields adjusted by long-term trailing volatility (the rankings do not change significantly when using a shorter-term volatility measure). The y-axis in all charts shows the volatility-adjusted yields, while the x-axis shows credit ratings and DTS. Sectors that are close to upper-right in each chart are more attractive (undervalued), while spread products that are close to bottom-left are less attractive (overvalued). Chart 9AGlobal Spread Product Yields, Hedged Into USD, Adjusted For Credit Quality Chart 9BGlobal Spread Product Yields, Hedged Into USD, Adjusted For Duration-Times-Spread Chart 10AGlobal Spread Product Yields, Hedged Into EUR, Adjusted For Credit Quality Chart 10BGlobal Spread Product Yields, Hedged Into EUR, Adjusted For Duration-Times-Spread Chart 11AGlobal Spread Product Yields, Hedged Into GBP, Adjusted For Credit Quality Chart 11BGlobal Spread Product Yields, Hedged Into GBP, Adjusted For Duration-Times-Spread Chart 12AGlobal Spread Product Yields, Hedged Into JPY, Adjusted For Credit Quality Chart 12BGlobal Spread Product Yields, Hedged Into JPY, Adjusted For Duration-Times-Spread Table 2Global Spread Products In Our Analysis An interesting result is that when comparing the three major high-yield products (US-HY, EMU-HY and UK-HY), US-HY is the most attractive in USD terms, but UK-HY is more attractive when hedged into GBP, EUR, and JPY. Another observation is that higher quality bonds such as government-related and agency debt in the US and euro area are overvalued and less attractive given how low their yields are, regardless of their low volatility. The results from this analysis may differ from our current recommendations. For example, we currently only have a neutral recommendation on EM corporates, but based on this analysis, EM corporates offer the most attractive return in USD terms. This analysis is purely based on YTM and traditional risk factors without considering other concerns that could make EM assets riskier such as the spread of COVID-19 in major EM countries. However, these rankings do line up with our major spread product call of overweighting US IG and HY corporate debt versus euro area equivalents. Based on this analysis, EM corporates offer the most attractive return in USD terms.  Bottom Line: A broad ranking of currency-hedged global spread product yields, adjusted for volatility and credit quality, shows that the most attractive yields (hedged into USD, EUR, GBP and JPY) are on offer in emerging market USD-denominated investment grade corporates and high-yield company debt in the US and UK.   Robert Robis, CFA Chief Fixed Income Strategist rrobis@bcaresearch.com   Ray Park, CFA Research Analyst ray@bcaresearch.com Footnotes 1https://www.fitchratings.com/research/sovereigns/coronavirus-impact-on-gdp-will-be-felt-for-years-to-come-27-07-2020 2 Please see BCA Global Fixed Income Strategy Weekly Report, "How To Play The Revival Of Global Inflation Expectations", dated June 23, 2020, available at gfis.bcaresarch.com. 3 We use the French 10-year inflation-linked bond as the proxy for the entire euro area, as this is the oldest inflation-linked bond market in the region and thus has the most data history. 4https://www.wsj.com/articles/fed-weighs-abandoning-pre-emptive-rate-moves-to-curb-inflation-11596360600?mod=hp_lead_pos6 5 Please see BCA Research Weekly Report, “What A Weaker US Dollar Means For Global Bond Investors”, dated July 28, 2020, available at gfis.bcaresarch.com. 6 Please see BCA Global Fixed Income Strategy Weekly Report, "Policymakers Are Now Selling Put Options On Volatility, Not Asset Prices", dated March 6, 2018, available at gfis.bcareseach.com. 7 Hedged YTM/Short-term trailing volatility vs. Credit Rating; Hedged YTM/Long-term trailing volatility vs. Credit Rating; Hedged YTM/Long-term trailing volatility vs. Duration; Hedged YTM/Long-term trailing volatility vs. Duration. Recommendations The GFIS Recommended Portfolio Vs. The Custom Benchmark Index Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Highlights Chart 1How Much Lower For Real Yields? Treasury yields moved lower last month even as the overall bond market priced-in a more reflationary economic environment. Spread product outperformed Treasuries and inflation expectations rose, but nominal bond yields still fell as plunging real yields offset the rising cost of inflation compensation (Chart 1). This sort of market behavior is unusual, but it is also easily explained. The market is starting to believe in the economic recovery, and it is pushing inflation expectations higher as a result. However, it also believes that the Fed will keep the nominal short rate pinned at zero even as inflation rises. Falling real yields result from rising inflation expectations and stable nominal rate expectations. This combination of market moves can’t go on forever. Eventually, inflation expectations will rise enough that the market will price-in policy tightening. This will push real yields higher, starting at the long-end of the curve. However, it’s difficult to know when this will occur, especially with the Fed doing its best to convey a dovish bias. In this environment, we advise investors to keep portfolio duration near benchmark and to play the reflation trade through real yield curve steepeners (see page 11). Real yield curve steepeners will profit in both rising and falling real yield environments, as long as the reflation trade remains intact. Feature Investment Grade: Overweight Chart 2Investment Grade Market Overview Investment grade corporate bonds outperformed the duration-equivalent Treasury index by 177 basis points in July, bringing year-to-date excess returns up to -361 bps. Spreads continue to tighten and investment grade corporate bond valuation is turning expensive, particularly for the highest credit tiers. The 12-month breakeven spread for the overall corporate index has been tighter 29% of the time since 1996 (Chart 2). The similar figure for the Baa credit tier is a relatively cheap 38% (panel 3). With the Fed providing a strong back-stop for investment grade corporates – one that has now officially been extended until the end of the year – we should expect spreads to turn even more expensive, likely returning to the all-time stretched valuations seen near the end of 2019. With that in mind, we want to focus our investment grade corporate bond exposure on high quality Baa-rated bonds. These are bonds that offer greater expected returns than those rated A and above, but that are also unlikely to be downgraded into junk (panel 4). Subordinate bank bonds are prime examples of securities that exist within this sweet spot.1 At the sector level, we also recommend overweight allocations to Healthcare and Energy bonds,2 as well as underweight allocations to Technology3 and Pharmaceutical bonds.4 Table 3ACorporate Sector Relative Valuation And Recommended Allocation* Table 3BCorporate Sector Risk Vs. Reward* High-Yield: Neutral Chart 3High-Yield Market Overview High-Yield outperformed the duration-equivalent Treasury index by 425 basis points in July, bringing year-to-date excess returns up to -466 bps. All junk credit tiers delivered strong returns on the month with the exception of the lowest-rated (Ca & below) bonds (Chart 3). These securities underperformed Treasuries by 267 bps, as a rising default rate weighs on the weakest credits. We are sticking with our relatively cautious stance toward high-yield, favoring bonds only from those issuers that will be able to access the Fed’s emergency lending facilities if need be. This includes most of the Ba-rated credit tier, some portion of the B-rated credit tier, and very few bonds rated Caa & below. We view the Fed back-stop as critically important because junk spreads are far too tight based on fundamentals alone. For example, current market spreads imply that the default rate must come in below 4.5% during the next 12 months for the junk index to deliver a default-adjusted spread consistent with positive excess returns versus Treasuries (panel 3).5 This would require a rapid improvement in the economic outlook. At the sector level, we advise overweight allocations to high-yield Technology6 and Energy7  bonds. We are underweight the Healthcare and Pharmaceutical sectors.8 MBS: Underweight Chart 4MBS Market Overview Mortgage-Backed Securities underperformed the duration-equivalent Treasury index by 2 basis points in July, dragging year-to-date excess returns down to -46 bps. The conventional 30-year MBS index option-adjusted spread (OAS) tightened 12 bps in July, but it still offers a pick-up relative to other comparable sectors. The MBS OAS of 86 bps is greater than the 75 bps offered by Aa-rated corporate bonds (Chart 4), the 47 bps offered by Aaa-rated consumer ABS and the 72 bps offered by Agency CMBS. Despite this spread advantage, we are concerned that the elevated primary mortgage spread is a warning that refinancing risk could flare later this year (bottom panel). Even if Treasury yields are unchanged, a further 50 bps drop in the mortgage rate due to spread compression cannot be ruled out. Such a move would lead to a significant increase in prepayment losses. With that in mind, we are concerned about the low level of expected prepayment losses (option cost) priced into the MBS index (panel 3). A refi wave in the second half of this year would undoubtedly send that option cost higher, eating into the returns implied by the lofty OAS.   Government-Related: Underweight Chart 5Government-Related Market Overview The Government-Related index outperformed the duration-equivalent Treasury index by 77 basis points in July, bringing year-to-date excess returns up to -325 bps. Sovereign debt outperformed duration-equivalent Treasuries by 285 bps on the month, bringing year-to-date excess returns up to -567 bps. Foreign Agencies outperformed the Treasury benchmark by 62 bps in July, bringing year-to-date excess returns up to -706 bps. Local Authority debt outperformed Treasuries by 74 bps in July, bringing year-to-date excess returns up to -368 bps. Domestic Agency bonds underperformed by 4 bps, dragging year-to-date excess returns down to -62 bps. Supranationals outperformed by 5 bps, bringing year-to-date excess returns up to -14 bps. The US dollar’s recent weakness, particularly against EM currencies, is a huge boon for Sovereign and Foreign Agency returns (Chart 5). However, US corporate spreads will also perform well in an environment of improving global growth and dollar weakness and, for the most part, value remains more compelling in the US corporate space (panel 3). Within the Emerging Market Sovereign space: South Africa, Mexico, Colombia, Malaysia, UAE, Saudi Arabia, Qatar, Indonesia, Russia and Chile all offer a spread pick-up relative to quality and duration-matched US corporate bonds. Of those attractively priced countries, Mexico stands out as particularly compelling on a risk/reward basis.9   Municipal Bonds: Overweight Chart 6Municipal Market Overview Municipal bonds outperformed the duration-equivalent Treasury index by 115 basis points in July, bringing year-to-date excess returns up to -473 bps (before adjusting for the tax advantage). Municipal bond spreads versus Treasuries tightened in July, but remain elevated compared to typical historical levels. In fact, both the 2-year and 10-year Aaa Muni yields are above equivalent-maturity Treasury yields, despite municipal debt’s tax exempt status (Chart 6). Municipal bonds are also attractively priced relative to corporate bonds across the entire investment grade credit spectrum, as we demonstrated in a recent report.10 In that report we also mentioned our concern about the less-than-generous pricing offered by the Fed’s Municipal Liquidity Facility (MLF). At present, MLF funds are only available at a cost that is well above current market prices (panel 3). This means that the MLF won’t help push Muni yields lower from current levels. Despite the MLF’s shortcomings, we stick with our overweight allocation to municipal bonds. For one thing, federal assistance to state & local governments will be included in the forthcoming stimulus bill. The Fed will also feel increased pressure to reduce MLF pricing the longer the passage of that bill is delayed. Further, while the budget pressure facing municipal governments is immense, states hold very high rainy day fund balances (bottom panel). This will help cushion the blow and lessen the risk of ratings downgrades. Treasury Curve: Buy 5-Year Bullet Versus 2/10 Barbell Chart 7Treasury Yield Curve Overview The Treasury curve bull flattened in July. The 2/10 and 5/30 Treasury slopes flattened 6 bps and 13 bps, reaching 44 bps and 99 bps, respectively. Unusually, the bull flattening of the Treasury curve that occurred last month was not the result of a deflationary market environment. Rather, the inflation compensation curve bear flattened – the 2-year and 10-year CPI swap rates increased 25 bps and 16 bps, respectively – while the real yield curve underwent a large parallel shift down. It will be difficult for the nominal yield curve to keep flattening if this reflationary back-drop continues. Eventually, rising inflation expectations will pull up real yields at the long-end of the curve. For this reason, we retain our bias toward duration-neutral yield curve steepeners on a 6-12 month horizon. Specifically, we advise going long the 5-year bullet and short a duration-matched 2/10 barbell. In a recent report we noted that valuation is a concern with this positioning.11 The 5-year yield is below the yield on the duration-matched 2/10 barbell (Chart 7), and the 5-year bullet looks expensive on our yield curve models (Appendix B). However, the 5-year bullet traded at much more expensive levels during the last zero-lower-bound period between 2010 and 2013 (bottom panel). With short rates once again pinned at zero, we expect the 5-year to once again hit extreme levels of overvaluation.   TIPS: Overweight Chart 8TIPS Market Overview TIPS outperformed the duration-equivalent nominal Treasury index by 95 basis points in July, bringing year-to-date excess returns up to -309 bps. The 10-year TIPS breakeven inflation rate rose 21 bps on the month to hit 1.56%. The 5-year/5-year forward TIPS breakeven inflation rate rose 18 bps on the month to hit 1.71%. TIPS breakeven inflation rates have moved up rapidly during the past couple of months, and the 10-year breakeven is now within 6 bps of the fair value reading from our Adaptive Expectations Model (Chart 8).12 TIPS will soon turn expensive if current trends continue. That is, unless stronger CPI inflation sends our model's fair value estimate higher. We place strong odds on the latter occurring. Month-over-month core CPI bottomed in April, as did the oil price. In addition, trimmed mean inflation measures suggest that core has room to play catch-up (panel 3). As mentioned on page 1, we continue to recommend real yield curve steepeners as a way to take advantage of the ongoing reflation trade. With the Fed now targeting a temporary overshoot of its 2% inflation goal, we would expect the cost of 2-year inflation protection to eventually trade above the cost of 10-year inflation protection (panel 4). With the Fed also keeping a firmer grip over short-dated nominal yields than over long-dated ones, this means that short-maturity real yields will come under downward pressure relative to the long-end (bottom panel).13   ABS: Overweight Chart 9ABS Market Overview Asset-Backed Securities outperformed the duration-equivalent Treasury index by 25 basis points in July, bringing year-to-date excess returns up to +23 bps. Aaa-rated ABS outperformed the Treasury benchmark by 15 bps on the month, bringing year-to-date excess returns up to +22 bps. Non-Aaa ABS outperformed by 111 bps, bringing year-to-date excess returns up to +22 bps. Aaa ABS are a high conviction overweight, given that spreads remain elevated compared to historical levels and that the sector benefits from Fed support through the Term Asset-Backed Securities Loan Facility (TALF). However, spreads are even more attractive in non-Aaa ABS (Chart 9) and we recommend owning those securities as well. This is despite the fact that only Aaa-rated bonds are eligible for TALF. We explained our rationale for owning non-Aaa consumer ABS in a recent report.14 We noted that the stimulus received from the CARES act caused real personal income to increase significantly during the past four months and, faced with fewer spending opportunities, households used that windfall to pay down consumer debt (bottom panel). Granted, further fiscal stimulus is needed to sustain recent income gains. But we expect the follow-up stimulus bill to be passed soon. Our Geopolitical Strategy service has shown that the new bill will likely contain sufficient income support for households.15   Non-Agency CMBS: Neutral Chart 10CMBS Market Overview Non-Agency Commercial Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 112 basis points in July, bringing year-to-date excess returns up to -395 bps. Aaa CMBS outperformed Treasuries by 43 bps on the month, bringing year-to-date excess returns up to -111 bps. Non-Aaa CMBS outperformed by 256 bps, bringing year-to-date excess returns up to -1042 bps (Chart 10). We continue to recommend an overweight allocation to Aaa non-agency CMBS and an underweight allocation to non-Aaa CMBS. Our reasoning is simple. Aaa CMBS are eligible for TALF, meaning that spreads can still tighten even as the hardship in commercial real estate continues. Without Fed support, non-Aaa CMBS will struggle as the delinquency rate continues to climb (panel 3).16 Agency CMBS: Overweight Agency CMBS outperformed the duration-equivalent Treasury index by 17 basis points in July, bringing year-to-date excess returns up to -42 bps. The average index spread tightened 5 bps on the month to 72 bps, still well above typical historical levels (bottom panel). The Fed is supporting the Agency CMBS market by directly purchasing the securities as part of its Agency MBS purchase program. The combination of strong Fed support and elevated spreads makes the sector a high conviction overweight. Appendix A: Buy What The Fed Is Buying The Fed rolled out a number of aggressive lending facilities on March 23. These facilities focused on different specific sectors of the US bond market. The fact that the Fed has decided to support some parts of the market and not others has caused some traditional bond market correlations to break down. It has also led us to adopt of a strategy of “Buy What The Fed Is Buying”. That is, we favor those sectors that offer attractive spreads and that benefit from Fed support. The below Table tracks the performance of different bond sectors since the March 23 announcement. We will use this to monitor bond market correlations and evaluate our strategy’s success. Table Performance Since March 23 Announcement Of Emergency Fed Facilities Appendix B: Butterfly Strategy Valuations The following tables present the current read-outs from our butterfly spread models. We use these models to identify opportunities to take duration-neutral positions across the Treasury curve. The following two Special Reports explain the models in more detail: US Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com US Bond Strategy Special Report, “More Bullets, Barbells And Butterflies”, dated May 15, 2018, available at usbs.bcaresearch.com Table 4 shows the raw residuals from each model. A positive value indicates that the bullet is cheap relative to the duration-matched barbell. A negative value indicates that the barbell is cheap relative to the bullet. Table 4Butterfly Strategy Valuation: Raw Residuals In Basis Points (As Of July 31, 2020) Table 5 scales the raw residuals in Table 4 by their historical means and standard deviations. This facilitates comparison between the different butterfly spreads. Table 5Butterfly Strategy Valuation: Standardized Residuals (As Of July 31, 2020) Table 6 flips the models on their heads. It shows the change in the slope between the two barbell maturities that must be realized during the next six months to make returns between the bullet and barbell equal. For example, a reading of 57 bps in the 5 over 2/10 cell means that we would only expect the 5-year to outperform the 2/10 if the 2/10 slope steepens by more than 57 bps during the next six months. Otherwise, we would expect the 2/10 barbell to outperform the 5-year bullet. Table 6Discounted Slope Change During Next 6 Months (BPs) Appendix C: Excess Return Bond Map The Excess Return Bond Map is used to assess the relative risk/reward trade-off between different sectors of the US bond market. It is a purely computational exercise and does not impose any macroeconomic view. The Map’s vertical axis shows 12-month expected excess returns. These are proxied by each sector’s option-adjusted spread. Sectors plotting further toward the top of the Map have higher expected returns and vice-versa. Our novel risk measure called the “Risk Of Losing 100 bps” is shown on the Map’s horizontal axis. To calculate it, we first compute the spread widening required on a 12-month horizon for each sector to lose 100 bps or more relative to a duration-matched position in Treasury securities. Then, we divide that amount of spread widening by each sector’s historical spread volatility. The end result is the number of standard deviations of 12-month spread widening required for each sector to lose 100 bps or more versus a position in Treasuries. Lower risk sectors plot further to the right of the Map, and higher risk sectors plot further to the left. Chart 11Excess Return Bond Map (As Of July 31, 2020)   Ryan Swift US Bond Strategist rswift@bcaresearch.com Footnotes 1 Please see US Bond Strategy Weekly Report, “The Case Against The Money Supply”, dated June 30, 2020, available at usbs.bcaresearch.com 2 For our outlook on Energy bonds please see US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 1: A Model Of Energy Bond Excess Returns”, dated July 14, 2020 and US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 2: Buy The Dip In High-Yield Energy”, dated July 21, 2020, available at usbs.bcaresearch.com 3 Please see US Bond Strategy Weekly Report, “Take A Look At High-Yield Technology Bonds”, dated June 23, 2020, available at usbs.bcaresearch.com 4 Please see US Bond Strategy Weekly Report, “Assessing Healthcare & Pharma Bonds In A Pandemic”, dated June 9, 2020, available at usbs.bcaresearch.com 5 We assume a 25% recovery rate and target a spread of 150 bps in excess of default losses. For more details on this calculation please see US Bond Strategy Weekly Report, “No Holding Back”, dated June 16, 2020, available at usbs.bcaresearch.com 6 Please see US Bond Strategy Weekly Report, “Take A Look At High-Yield Technology Bonds”, dated June 23, 2020, available at usbs.bcaresearch.com 7 Please see US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 1: A Model Of Energy Bond Excess Returns”, dated July 14, 2020 and US Bond Strategy Special Report, “The Outlook For Energy Bonds Part 2: Buy The Dip In High-Yield Energy”, dated July 21, 2020, available at usbs.bcaresearch.com 8 Please see US Bond Strategy Weekly Report, “Assessing Healthcare & Pharma Bonds In A Pandemic”, dated June 9, 2020, available at usbs.bcaresearch.com 9 Please see US Bond Strategy Weekly Report, “The Treasury Market Amid Surging Supply”, dated May 12, 2020, available at usbs.bcaresearch.com 10 Please see US Bond Strategy Weekly Report, “Bonds Are Vulnerable As North America Re-Opens”, dated May 26, 2020, available at usbs.bcaresearch.com 11 Please see US Bond Strategy Weekly Report, "Take A Look At High-Yield Technology Bonds”, dated June 23, 2020, available at usbs.bcaresearch.com 12 For more details on our model please see US Bond Strategy Weekly Report, “How Are Inflation Expectations Adapting?”, dated February 11, 2020, available at usbs.bcaresearch.com 13 For more details on our recommended real yield curve steepener trade please see US Bond Strategy Weekly Report, “Negative Oil, The Zero Lower Bound And The Fisher Equation”, dated April 28, 2020, available at usbs.bcaresearch.com 14 Please see US Bond Strategy Weekly Report, “No Holding Back”, dated June 16, 2020, available at usbs.bcaresearch.com 15 Please see Geopolitical Strategy Weekly Report, “A Tech Bubble Amid A Tech War (GeoRisk Update)”, dated July 31, 2020, available at gps.bcaresearch.com 16 We discussed our CMBS outlook in more detail in US Bond Strategy Weekly Report, “No Holding Back”, dated June 16, 2020, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification Corporate Sector Relative Valuation And Recommended Allocation

Highlights Equities and other risk assets face near-term headwinds from the surge in Covid cases in the US Sun Belt and the looming fiscal cliff. We think these problems will be resolved, but the next few weeks could be rough sledding for markets. Government bond yields have moved sideways-to-down since late March even though inflation expectations have rebounded. The resulting decline in real yields has been an important, if rather overlooked, driver of higher equity prices. The failure of government bond yields to rise in line with higher inflation expectations can be attributed to the ongoing dovish shift in monetary policy. Nominal yields are likely to increase modestly over the next two years as growth recovers. However, inflation expectations should rise even more. Hence, real yields may fall further, justifying an overweight position in TIPS and a generally positive medium-term view on equities. As long as there is spare capacity in the economy, fiscal stimulus will not push up real yields. This is because bigger budget deficits tend to raise overall savings, thus creating the resources with which to finance the deficits. Once economies return to full employment in about three years, the fiscal free lunch will end. At that point, the combination of easy monetary and fiscal policies could cause inflation to accelerate. Central banks will welcome higher inflation initially. However, they will eventually be forced to hike rates aggressively if inflation continues to march upwards. When this happens, bond yields will rise sharply, while stocks will tumble. A Curious Divergence Government bond yields have moved sideways-to-down in most developed economies since stocks bottomed in late March (Chart 1). In contrast, inflation expectations have risen. As a result, real yields have declined. In the US, TIPS yields have fallen into negative territory across all maturities (Chart 2). Chart 1Nominal Yields Have Moved Sideways-To-Down, Inflation Expectations Have Risen, And Real Yields Have Declined Chart 2TIPS Yields Have Fallen Into Negative Territory Across The Board The decline in real yields has been one of the unsung drivers of higher equity prices this year. The forward P/E ratios of the major US indices have moved closely in line with real yields (Chart 3). Gold prices have also risen, as they are often wont to do when real yields go down (Chart 4). Chart 3Lower Real Yields Have Lifted Stock Multiple Chart 4Gold Prices Have Risen On The Back Of Falling Real Yields It is fairly uncommon for inflation expectations to rise without a commensurate increase in nominal bond yields (Chart 5). As a rule of thumb, when the economic data surprise to the upside, as has occurred over the past few months, bond yields go up (Chart 6). Chart 5It Is Unusual For Inflation Expectations To Rise Without A Corresponding Increase In Nominal Bond Yields Chart 6Bond Yields Usually Rise When Economic Data Surprise To The Upside An important exception to this rule occurs when monetary policy is becoming more expansionary. Bond yields tend to follow the path of expected policy rates (Chart 7). When central banks guide rate expectations lower, bond yields can fall, even as the reflationary impulse from lower yields delivers an upward kick to inflation projections. Chart 7ABond Yields Tend To Follow The Path Of Expected Policy Rates Chart 7BBond Yields Tend To Follow The Path Of Expected Policy Rates The last time such a divergence between yields and inflation expectations occurred was in early 2019. The stock market crash in late 2018 forced the Fed to abandon its plans to hike rates. Jay Powell’s dovish pivot occurred just three months after he said that rates were “a long way” from neutral. The Fed would go on to cut rates by 75 bps over the course of 2019. Real Yields Could Fall Further Chart 8Inflation Expectations Are Still Quite Depressed In Most Countries The key question for investors is how much longer the pattern of rising inflation expectations and stable bond yields can persist. Our sense is that nominal bond yields will rise modestly over the next few years as growth recovers. However, inflation expectations are likely to rise even more, justifying an overweight position in TIPS relative to nominal bonds. Inflation expectations are still quite depressed in most countries (Chart 8). If global growth rebounds, both actual and expected inflation should edge higher. Chart 9 shows that the US ISM manufacturing index leads core inflation by about 12-to-18 months. Higher oil prices should also lift inflation expectations (Chart 10). Will global growth recover? The answer is “yes” if we are talking about a horizon of 12 months or so. That said, as we discuss below, there are some near-term risks to growth. This implies that equities and other risk assets could trade nervously over the next few weeks.   Chart 9Global Growth Recovery Will Lead To Higher Inflation Down The Line Chart 10Inflation Expectations And Oil Prices Move In Lockstep   Near-Term Risks To Global Growth The two biggest threats to global growth over the coming months are the Covid outbreaks in a number of countries and the possibility that fiscal stimulus will be rolled back, especially in the US, where a “fiscal cliff” is looming. Despite progress in suppressing the virus in Europe, Japan, and most of East Asia, the number of reported daily infections continues to rise globally (Chart 11). In the developed world, the US remains a major hotspot. Although the number of cases appears to have peaked in Arizona, it is still rising in the other Sun Belt states (Chart 12). Among emerging markets, the epicenter has moved from Brazil and Russia to India (Chart 13). Chart 11Despite Progress In Europe, Japan, And Most Of East Asia, The Number Of Covid Infections Continues To Rise Globally Chart 12A Second Wave Is A Key Macro Risk Chart 13BRICs: Covid Leaving No Stone Unturned While efforts to contain the virus will boost growth in the long run, they will weigh on economic activity in the near term. Over half of the US population lives in states that have either reversed or suspended reopening plans (Chart 14). Chart 14Not So Fast Google data on visits to shopping malls, recreation centers, public transport facilities, and office destinations have dipped in recent weeks. The decline in visits has occurred alongside a decrease in the New York Fed’s high-frequency economic activity indicator (Chart 15). Initial unemployment claims also rose this week. At this point, it looks likely that the recovery in US consumer spending will stall in July and August. Chart 15Covid Outbreak Is Weighing On Spending While it is difficult to know what will happen starting in September, our guess is that the pandemic will ebb in the southern states, just like it did in the northeast. This is partly because mask-wearing is becoming more widespread. Back in early March, when most mainstream news sources were tweeting out misinformation such as “Oh, and face masks? You can pass on them,” we noted that both logic and evidence suggest that masks are an effective tool against the virus. Increased testing should also help identify asymptomatic people before they have had the chance to spread the virus to many others. Meanwhile, improved medical care should also help reduce the mortality and morbidity rates from the disease. Just this week, scientists presented the results of a double-blind clinical trial showing that the inhalation of interferon beta, a cytokine used to treat multiple sclerosis, reduced the risk of developing severe Covid symptoms by nearly 80%. Fiscal Cliff Ahead? In addition to the pandemic, investors have to grapple with uncertainty over whether fiscal policy will remain sufficiently accommodative to reflate the economy. Unlike the EU, which managed to cobble together a framework for creating a 750 billion euro pandemic relief fund earlier this week, the US Congress remains deadlocked on the size and complexion of a new stimulus bill. Under current law, US households will stop receiving expanded unemployment benefits at the end of July. These benefits were legislated as part of the original CARES Act and currently total over 4% of GDP. The Paycheck Protection Program for small businesses is also nearly drained, while state and local governments are facing a major cash crunch due to evaporating tax revenues and higher pandemic-related spending needs. We estimate that about $2-to-$2.5 trillion in new stimulus will be necessary to keep fiscal policy from turning unduly restrictive. Senate Majority Leader Mitch McConnell has been floating a number of $1.3 trillion. If McConnell gets his way, risk assets will likely sell off. Our guess is that he will not prevail, however. President Trump favors a larger stimulus bill, as do the Democrats. Critically, more than four out of five voters, both nationwide and in swing states, support extending benefits (Table 1). Thus, there is a high probability that Senate Republicans will agree on a much larger package than what they are currently proposing. Table 1There Is Much Public Support For Fiscal Stimulus Fiscal Stimulus And Bond Yields Could continued fiscal stimulus deplete national savings, leading to significantly higher real yields? For the next few years, the answer is no. National savings depend not just on how much people spend, but on how much they earn. To the extent that fiscal stimulus raises GDP, it also raises national income. For the global economy as a whole, savings must equal investment. If fiscal stimulus in the major economies prompts firms to undertake more investment spending than they would have otherwise, overall savings will rise. How can that be? The answer is that fiscal stimulus raises private savings by more than it reduces government savings when an economy is operating below its full capacity. From the perspective of the bond market, this means that currently, large budget deficits are self-financing. Bigger budget deficits will produce an even bigger pool of private income, allowing the private sector to buy more government bonds.   Indeed, a premature pullback in fiscal support would almost certainly raise real rates by depressing inflation expectations. If that sounds far-fetched, recall that this is precisely what happened in March. Full Employment And Beyond Chart 16Government Debt Levels Have Surged In The Wake Of The Pandemic The fiscal free lunch will end only when economies return to full employment. At that point, bigger budget deficits will no longer be able to raise output since everyone who wants to work will already have found a job. Rather, increased government borrowing will crowd out private-sector investment. National savings will decline. If monetary and fiscal policy stay accommodative, inflation could accelerate. Central banks will probably welcome the initial burst of inflation, since they have been lamenting below-target inflation for many years now. However, if inflation continues to march higher, central banks may get spooked and start talking up the prospect of rate hikes. Higher rates would create a lot of problems for debt-saddled governments (Chart 16). It would not be at all surprising if politicians leaned on central banks to keep rates low. Governments could also end up forcing central banks to buy more debt in order to keep long-term yields from rising. In the extreme case, governments could even force central banks to cap yields. While such measures would prevent bond prices from tumbling, this would be cold comfort for bondholders. If central banks were to keep bond yields below their equilibrium level, inflation would rise even further, thus eroding the purchasing power of the bonds. In the end, central banks would still have to raise rates, probably more than they would have had they acted more swiftly to quell inflation. Investment Conclusions To answer the question posed in the title of this report, yes, bond yields will eventually go up. However, they are not likely to rise very much until inflation reaches intolerably high levels. That point is at least three years away. Despite the near-term risks posed by the pandemic and the looming fiscal cliff, investors should remain overweight equities over a 12-month horizon. Given the run-up in some of the large cap US tech names, we suggest shifting equity exposure to other parts of the stock market. The cyclically-adjusted price-earnings ratio is significantly lower outside the US, implying that international stocks are well placed to outperform their US peers over the coming decade (Chart 17). A weaker dollar should also help non-US stocks as well as the more cyclical equity sectors (Chart 18). Chart 17Non-US Stocks: The Place To Be Over The Coming Decade Chart 18A Weaker Dollar Should Boost Non-US Stocks Along With The More Cyclical Equity Sectors Peter Berezin Chief Global Strategist peterb@bcaresearch.com Global Investment Strategy View Matrix Current MacroQuant Model Scores