Developed Countries

Special Report

Highlights Global Corporates: The cyclical backdrop – accommodative monetary policies, improving global growth momentum, moderate inflation and subdued volatility – is supportive for the continued outperformance of global corporate bonds over sovereign debt in 2020. Corporate Bond Valuation: Extending a valuation framework we’ve introduced for US corporates to non-US credit – looking at volatility-adjusted spreads relative to both their own history and the “phase” of the monetary policy cycle - we calculate spread targets for non-US corporates in the euro area, UK and Canada. 2020 Opportunities: Current corporate spread levels are furthest above our targets (i.e. cheap) for US high-yield (most notably for Ba- and Caa-rated credit), UK high-yield and UK investment grade. Spreads are furthest below our targets (i.e. expensive) for euro area high-yield (mostly Ba-rated), US investment grade (all credit tiers) and Canadian investment grade. Feature Chart 1Deviations From Corporate Spread Targets One of our main investment themes for 2020 is that accommodative monetary policies and faster economic growth will delay the peak in the aging global credit cycle, giving investors another year of corporate bond outperformance versus sovereign debt in the developed economies. Returns this year will be nowhere near as robust as in 2019, however, given rich valuation starting points for much of the global corporate universe. Against this backdrop, fixed income investors will have to be more selective in allocations by country, sector and credit quality in order to outperform. To that end, in this Special Report we extend a valuation framework for corporate bond spreads first introduced for US corporates by our sister service, US Bond Strategy, to non-US credit. This methodology looks at spreads on a volatility-adjusted basis, allowing comparison of valuations versus their own history and relative to similar stages of past monetary policy cycles. Chart 1 shows the deviations of current benchmark index option-adjusted spreads (OAS) from spread targets derived from our methodology for different countries (the US, euro area, the UK and Canada) and credit quality tiers (investment grade vs. high-yield). Positive deviations imply current spreads are above the targets derived from our framework – in other words, relatively undervalued - and vice-versa. Returns this year will be nowhere near as robust as in 2019, however, given rich valuation starting points for much of the global corporate universe. Against this backdrop, fixed income investors will have to be more selective in allocations by country, sector and credit quality in order to outperform. The conclusions are that there are still opportunities for additional spread tightening from current levels for lower-rated US high-yield and both UK investment grade and high-yield corporates. At the same time, valuations are looking most stretched for euro area high-yield, US investment grade (all credit tiers) and Canadian investment grade. A Brief Word On The Outlook For Global Corporate Credit In 2020 Chart 2Positive Backdrop For Corporate Bonds The backdrop for global corporate bond markets will remain positive in 2020 for three main reasons: Global monetary policies will remain accommodative. Central bankers are now focusing more on boosting soft growth and low inflation expectations. Real policy interest rates in the US, euro area, UK and Canada are already below estimates of neutral like r-star (Chart 2, top panel), and will likely remain so throughout 2020. In the past, periods of credit market underperformance have occurred when monetary policy was restrictive, with real rates above neutral or government bond yield curves that were very flat or inverted (more on that later). Global growth momentum will improve. Recent data releases (global manufacturing PMIs, sentiment surveys like the global ZEW and German IFO) have shown that the 2019 global industrial downturn was in the process of bottoming out during the 4th quarter of the year. Additional improvement is likely in the coming months, based on the steady gains of the BCA Global Leading Economic Indicator (LEI). The elevated level of our global LEI diffusion index – measuring the share of individual country LEIs that are rising and which is itself a leading indicator of both the global LEI and corporate bond returns - suggests that additional outperformance of global corporates versus sovereign bonds is likely within the next 12 months (Chart 2, middle panel). Financial conditions are stimulative. Global equities and credit are off to a strong start in 2020, while market volatility is subdued across a variety of asset classes. For example, the US VIX index is now just above its 2019 low, which is consistent with narrow global corporate bond spreads (Chart 2, bottom panel). That low volatility backdrop – supported by market-friendly central bank policies - is helping keep financial conditions easy enough to lift economic growth, while also boosting investor risk appetite for corporate credit. The overall outlook for global corporate credit is still positive and investors should expect another year of corporate bond outperformance versus sovereign debt in the developed economies. Nonetheless, returns will be lower in 2020 than in 2019 due to expensive valuation starting points. As can be seen from Chart 3, global corporate bond spreads are already fairly tight relative to their long-term historical range. Also, outright index yields in many asset classes, like US high-yield, are now at new all-time lows. We interpret this as a sign that the “easy money” has already been made in being generally long corporate credit versus government bonds. Having the right tools to assess the relative values among differing credit markets will be critical to finding the best investment opportunities in this environment. Chart 3Rich Valuation Starting Points In Corporate Credit We can use the breakeven spread as a valuation tool by looking at the percentile rank relative to its own history, effectively showing the percentage of time that the breakeven spread has been lower in the past. Bottom Line: The cyclical backdrop – accommodative monetary policies, improving global growth momentum, moderate inflation and subdued volatility – is supportive for the continued outperformance of global corporate bonds over sovereign debt in 2020. Valuations are likely to be more of a headwind for corporate bond returns, though. Using Breakeven Spreads As A Credit Valuation Tool As a reminder to existing readers (and to new clients), one of our main valuation tools for credit instruments is the 12-month breakeven spread. That is, the amount of spread widening required for corporate bond returns to break even with a duration-matched position in government bond securities over a 12-month horizon. It can be approximated by dividing the OAS of a bond (or a benchmark bond index) by its duration. More specifically, we can use the breakeven spread as a valuation tool by looking at the percentile rank relative to its own history, effectively showing the percentage of time that the breakeven spread has been lower in the past. We find this valuation tool to be superior to others for two main reasons: (i) using the breakeven spread rather than the average index OAS allows us to control for the changing average duration of the benchmark bond indices; and (ii) the percentile rank is often a better representation of credit spreads than the spread itself.1 BCA Research US Bond Strategy and Global Fixed Income Strategy have both regularly shown the percentile rankings of US investment grade and high-yield breakeven spreads as part of our discussion of US corporate bond markets. We have never produced such rankings for non-US credit, until now. InCharts 4- 7, we show those percentile ranks relative to history for credit in the US and, for the first time, the euro area, UK, Canada, Japan, Australia and Emerging Markets US dollar denominated corporates. We also provide the breakeven spread historical percentile ranks for each individual credit tier in the Appendix charts on pages 13-16. Chart 4US: Corporate Bond Breakeven Spreads Chart 5Euro Area: Corporate Bond Breakeven Spreads Chart 6UK: Corporate Bond Breakeven Spreads Using these charts, we can gauge which markets offer the best (or worst) level of spread, adjusted by its own volatility and compared to its own history. The most attractive corporate credit spreads on a volatility-adjusted basis are: US high-yield (mostly Caa-rated and B-rated) Japan investment grade (mostly Baa-rated and A-rated) Canada Aaa-rated UK high-yield (excluding financials) Chart 7Other Countries: Corporate Bond Breakeven Spreads The least attractive corporate credit spreads on a volatility-adjusted basis are: US investment grade (all credit tiers) UK Aaa-rated Canada Baa-rated Euro Area high-yield Using this metric, US Caa-rated junk bonds look most “undervalued”, with a volatility-adjusted spread in the upper 20% of all observations. Chart 8 displays the current breakeven spread historical percentile ranks across countries and credit quality, with high-yield markets shown in red. Using this metric, US Caa-rated junk bonds look most “undervalued”, with a volatility-adjusted spread in the upper 20% of all observations. While this chart provides a quick overview of which corporate bond markets are cheap/expensive with respect to their own history, it does not allow for comparisons of the relative cheapness between markets. To do this, we need to find a way to convert the percentile rankings into some measure of a “fair value” credit spread. Chart 8Global Corporate Bond 12-Month Breakeven Spreads By Percentile Rank (%) Using Monetary Policy Cycles To Determine Corporate Spread Targets Our colleagues at BCA Research US Bond Strategy have come up with a novel approach for determining spread targets for US corporate credit, based on the breakeven spread percentile rankings.2 Essentially, the stance of US monetary policy, as measured by the slope of the US Treasury curve, is used to predict changes in the US credit cycle, helping to determine “cyclical” spread targets relative to the stance of monetary policy. The first step of this process is to group corporate bond excess returns (vs government debt) into buckets defined by the following “phases” of the US monetary policy cycle, measured by the yield differential between 10-year and 3-year Treasuries: Phase 1: from the end of the previous recession until the slope goes below 50 bps. Phase 2: from the time that the slope crosses below 50 bps until it inverts. Phase 3: from the time that the yield curve first inverts to the start of the next recession. Recessionary periods are not included in these phases, as all corporate credit exhibits the worst returns during those episodes. That is because economic growth and downgrade/default risks, and not the state of monetary policy, are the driving factor behind credit spread moves during recessions. Chart 9 shows the history of the US corporate bond markets broken down into “curve-defined” cycles.3 Dating back to 1974, the earliest date for investment grade bond index data, there have been five such cycles. Chart 9US Corporate Bond Performance And The Yield Curve In Charts 10-12, we show the same phases for the euro area, the UK and Canada, using their own government bond yield curves to determine the phase of the monetary policy cycle in the same fashion as was done for the US.4 Once the phases of the monetary policy cycle are defined, we can then calculate corporate bond excess returns during each phase. Chart 10Euro Area: Corporate Bond Performance And The Yield Curve Chart 11UK: Corporate Bond Performance And The Yield Curve Chart 12Canada: Corporate Bond Performance And The Yield Curve Table 1 shows the average corporate bond annualized excess returns under each phase across every cycle that can be defined with available data. Excess returns tend to be highest in Phase 1, quite low but still positive in Phase 2, and usually turn negative during Phase 3, once the yield curve has inverted. Table 1Corporate Bond Annualized Excess Returns* (%) Under Each Phase Of The Cycle Currently, we are in Phase 2 in the US, euro area and UK, with yield curves that are relatively flat but still positively sloped. Historically, such periods have generated positive excess returns for corporate debt versus duration-matched government bonds, although of far smaller magnitudes compared to Phase 1 periods. Given our expectation that the Fed, ECB and Bank of England will maintain a dovish bias throughout 2020, we expect the no shift from Phase 2 for the US, euro area and the UK that would hurt corporate bond excess returns in those countries. With the Canadian yield curve now slightly inverted, however, Canada is now in Phase 3. This raises the risk that the recent strong outperformance of Canadian investment grade corporate bonds could end if the Bank of Canada does not deliver the monetary easing currently discounted in the Canadian yield curve. How We Determine Corporate Spread Targets Having defined the three phases of the monetary policy cycle, we then re-calculate our corporate bond breakeven spread percentile ranks within each phase. We then back-out a spread target for each credit tier by taking the median 12-month breakeven spread seen in similar monetary policy environments, as determined by the slope of the yield curve.5 Finally, we convert those “median” breakeven spreads into OAS targets using the current benchmark index duration and credit rating distribution. We are assuming that a reasonable spread target for any corporate bond market is determined by adjusting for both spread volatility AND the monetary policy cycle. So, essentially, we are assuming that a reasonable spread target for any corporate bond market is determined by adjusting for both spread volatility AND the monetary policy cycle. Charts 13-16 show the index OAS and their respective targets for the US (both investment grade and high-yield), euro area (both investment grade and high-yield), the UK (both investment grade and high-yield excluding financials), and Canada (only investment grade). Further, the spread targets for each individual credit tier are provided in the Appendix on pages 17-19. Chart 13US: Corporate Bond Spread Targets Chart 14Euro Area: Corporate Bond Spread Targets Chart 15UK: Corporate Bond Spread Targets Chart 16Canada: Corporate Bond Spread Targets For example, our spread target for US B-rated high-yield is 227bps, which is 80bps below the current index OAS. From the charts, we can make the following conclusions about the relative attractiveness of current spread levels: The largest deviations from our spread target (i.e. potentially most undervalued) are: US high-yield (mostly Caa-rated and B-rated) UK high-yield (excluding financials) The lowest deviations from our spread target (i.e. potentially most overvalued) are: Euro Area high-yield (mostly Ba-rated) Canada investment grade US investment grade (all credit tiers) This framework is an interesting way to derive corporate bond value, by adjusting for both the volatility and monetary policy backdrop. Of course, there are other factors that are more difficult to quantify that can keep spreads too tight or too wide versus these fair value levels, like investor risk tolerance or risk premia for political uncertainty. In terms of factors that are quantifiable, however, this spread target methodology is a useful way to get a sense of the richness or cheapness of global corporate debt. In terms of factors that are quantifiable, however, this spread target methodology is a useful way to get a sense of the richness or cheapness of global corporate debt. We will regularly update these targets in future BCA Research Global Fixed Income Strategy reports. Bottom Line: Current corporate spread levels are furthest above our targets (i.e. cheap) for US high-yield (most notably for Ba- and Caa-rated credit), UK high-yield and UK investment grade. Spreads are furthest below our targets (i.e. expensive) for euro area high-yield (mostly Ba-rated), US investment grade (all credit tiers) and Canadian investment grade.   Jeremie Peloso Research Analyst jeremiep@bcaresearch.com Robert Robis, CFA Chief Fixed Income Strategist rrobis@bcaresearch.com Footnotes 1 This is because credit spreads often tighten to very low levels and then remain tight for an extended period. Thus, by showing the percentage of time that a given spread has been tighter than its current level, the percentile rank gives a better sense of this pattern than the actual spread. 2 Please see US Bond Strategy Special Report, “2019 Key Views: Implications For US Fixed Income”, dated December 11, 2018, available at usbs.bcaresearch.com. 3 Note that the Treasury curve used for this analysis is the spread between the 10-year Treasury and the 3-year Treasury yield. The more widely-followed 2-year Treasury was not used as there was more historical data available for the 3-year maturity. 4 Note that there are fewer cycles to analyze for these countries due to the shorter available history of corporate bond market data outside the US. 5 For more details on the spread targets please see US Bond Strategy Weekly Report, “Paid To Wait”, dated February 26, 2019, available at usbs.bcaresearch.com. Appendix Chart 1AUS: Investment Grade Breakeven Spreads Chart 1BUS: High-Yield Breakeven Spreads Chart 1CEuro Area: Investment Grade Breakeven Spreads By Credit Tiers Chart 1DEuro Area: High-Yield Breakeven Spreads By Credit Tiers Chart 1EUK: Investment Grade Breakeven Spreads Chart 1FCanada: Investment Grade Breakeven Spreads Chart 1GJapan: Investment Grade Breakeven Spreads Chart 2AUS: Investment Grade Spread Targets Chart 2BUS: High-Yield Spread Targets Chart 2CEuro Area: Investment Grade Spread Targets By Credit Tiers Chart 2DEuro Area: High-Yield Spread Targets By Credit Tiers Chart 2EUK: Investment Grade Spread Targets By Credit Tiers Chart 2FCanada: Investment Grade Spread Targets By Credit Tiers

Special Report

Highlights Macro winds are slowly changing, compelling investors to take a second look at highly cyclical sectors such as industrials. In this Special Report we highlight that the S&P industrials index is trading at a nearly 10% discount to the market on a forward P/E basis, meanwhile a list of welcoming omens has appeared on the horizon which will serve as a foundation for the relative share price outperformance. Increasing odds of a modest rebound in the US manufacturing PMI Stabilization in EM in general and Chinese economy in particular BCA’s House View of a weaker dollar Healthy industry operating metrics Table 1 Feature While US manufacturing remains in recession, investors have already positioned for a V-shaped recovery according to the spectacular run-up in the S&P 500. However, drilling beneath the surface is revealing. Instead of hypersensitive industrials equities sniffing out a recovery in the manufacturing sector, the SPX’s advance has been solely driven by tech mega caps, and thus leaving in the dust all their deep cyclical peers. Why? Because the industrials complex has borne the brunt of the trade war (Chart 1). However, our expectation remains for a natural healing of the economy sometime in the first half of the year, and a rotation out of tech equities into capital goods stocks. Industrials equities will be among the first beneficiaries of this improving macro backdrop. Specifically, four key themes underpin our bullish stance on the S&P industrials index: Increasing odds of a modest rebound in the US manufacturing PMI Stabilization in EM in general and Chinese economy in particular BCA’s House View of a weaker dollar Healthy industry operating metrics Chart 1Trade War Echoes The ISM Will Soon Turn The Corner ISM’s manufacturing PMI survey is still below the 50 boom/bust line, but much of the macro pessimism is likely already reflected in the data, as it is set to bottom by the end of Q1 2020 and rebound into Q2 2020. To be clear, we are not expecting a jump into the high 50s, but rather look for a more balanced recovery into the low 50s. Nevertheless, it is a 5-point rise from the current recessionary level, and the S&P industrials index will cheer a relief in macro data. Charts 2 & 3 clearly depict that the ISM manufacturing PMI is set to improve in the coming months. First, our demand/supply proxy for the overall US economy (comprising of retail and industrial production data) is signaling that industrial production will likely bottom by mid-2020, despite Boeing’s ails (Chart 2, bottom panel). Industrial production is also known to lag PMI by roughly three months, which translates into PMI bottoming in Q1 2020, if history at least rhymes. Moreover, the bond market is also sending a similar message (Chart 2, third panel). As a reminder, BCA’s House View calls for a sell-off in the bond market to a range of 2.25-2.5% for this year. Chart 2Long Term And … Chart 3… Short Term Drivers Are Positive Both coincident and short-term leading economic variables also emit a positive signal. Lumber prices have recovered sharply over the past year (Chart 3, second panel), and given their close correlation with the ISM manufacturing PMI, the move underscores that the bottoming process in the latter has already begun. Moreover, survey internals have made a modest turn on a three-month moving average basis, suggesting that the path of least resistance is to the upside (Chart 3, bottom panel). On the political front, our sister Geopolitical Strategy service has been right on Trump having to retreat and to agree on a “ceasefire” deal as the 2020 elections are looming. Following the October and December tariff “postponement/cancelation” coupled with lower chances of any tariff hikes at all until the 2020 election will likely provide a further boost to the “soft” survey data. As a reminder, the increase in trade policy uncertainty over 2018-2019 has been a large contributor to data deterioration (Chart 1). EM And Chinese Green Shoots A rising share of international sales for the S&P industrials index originates from the EM, as those countries are responsible for a large percent of world total commodity consumption and construction activity. The EM manufacturing PMI has done a good job at tracing the S&P industrials relative share price performance, and the current divergence between the two can serve as yet another catalyst for a rally (Chart 4, top panel). Most importantly, China is also enjoying green shoots. BCA’s Chinese credit & fiscal impulse has been grinding higher over the past year foreshadowing a further rebound in EM data, and consequently benefiting US industrials. Finally, Chinese infrastructure spending that managed to climb out of negative territory will, at the margin, further boost industrials end-demand (Chart 4, middle & bottom panels). Chart 4International Arena Improving The Dollar Is Petering Out Switching gears to the greenback, more good news is in store for the S&P industrials index. Forty-four percent of sales are international for the S&P industrials sector, making it sensitive to FX fluctuations. BCA’s House View for 2020 calls for a weaker USD and should it be proven correct, industrials P&Ls will enjoy positive currency translation tailwinds. Chart 5 shows a newly created dollar model first published by our sister The Bank Credit Analyst service, heralding a softer greenback in the coming months. The real trade-weighted US dollar has been a key driver for the S&P industrials’ sales ever since 1975 as they remained in positive territory even during the recent manufacturing recession. A turn in the US dollar will reignite sales growth and underpin the relative share price ratio (Chart 6, bottom panel). Chart 5The Softening US Dollar … In addition, a depreciating US dollar has been synonymous with a relative multiple expansion phase, and a definitive fall in the dollar will likely serve as a catalyst to unlock excellent value in bombed out relative valuations (Chart 6, third panel). Looking at sales from a different angle, our industrials sector exports proxy has a tight inverse correlation with the USD, and is likely to hook higher given that the US dollar is flat on a year-over-year basis (Chart 6, second panel). Chart 6… Holds The Key … Chart 7… And So Do Our EPS Models Finally, our relative earnings growth model does an excellent job at encapsulating all the profit drivers and is signaling that an earnings-led outperformance period looms (Chart 7). Enticing Operating Metrics Drilling down and away from macro and toward industry-level data is instructive. First, industrials forward earnings breadth – defined as the net number of sub-sectors with higher forward earnings revisions – has bounced off extreme lows signaling that the sell-off in relative share prices is reaching exhaustion. Historically, similar sharp bounces have marked previous price reversals (Chart 8, second panel). Chart 8Selling Climax? Chart 9More Welcoming News Expanding profit margins since mid-2018 also reflect positive sector dynamics as the industry managed to shrug off falling selling prices at the expense of labor (Chart 8, third & fourth panels). Importantly, the CRB raw industrials index is tracing a bottom and a depreciating US dollar will assist in orchestrating a recovery in industrials PPI (Chart 9). With regard to balance sheet health, interest coverage and net debt-to-EBITDA ratios are not sounding alarm bells as they are comparable to the broad market. Industrials are also steadily pumping out healthy free cash flow numbers, which should be a welcoming sign for investors (Chart 10). Despite all these tailwinds, sell-side analysts are still overly pessimistic on relative long-term profit prospects (Chart 11, fourth panel). Chart 10Good B/S All Around Chart 11No Red Flags Neither BCA composite Valuations nor Technical Indicators caution against overweighting the S&P industrials index (Chart 11, second and third panels). In fact, on a forward P/E basis, the sector is trading at a nearly 10% discount to the broad market making it an historically “cheap” addition to one’s portfolio (Chart 11, bottom panel). Risks To Monitor There are a few key risks to our view. First, we treat the current de-escalation in the trade war as temporary. Should Trump get reelected in 2020 for another term, all of the pre-election constraints will be lifted, likely allowing him to get back to his tariff hawkishness. Second, the US economy has already suffered too much damage making it vulnerable to an external shock. Were a black swan to materialize, the dollar would skyrocket putting our industrials view offside. Finally, should the Chinese government miscalculate the amount of stimulus required to reignite their economy, US industrials will again be among the first ones to suffer the consequences via the final-demand link. Put simply, any combination of these risks would slay animal spirits and postpone capex-led US industrials sector recovery. Bottom Line: Increasing chances of a rebound in the ISM manufacturing PMI coupled with green shoots in the EM, expectations for a weaker US dollar and sound industry level data, argue for an above benchmark allocation in the S&P industrials index. Please stay tuned for an upcoming Special Report on how to position within the industrials complex.     Arseniy Urazov Research Associate ArseniyU@bcaresearch.com  

Please note that next week’s US Bond Strategy Weekly Report will be replaced by a Special Report on Commercial Real Estate that was produced jointly with our US Investment Strategy team. That report will be published on Monday instead of Tuesday. Highlights Duration: Financial markets have taken the Fed’s dovish guidance on board, and the asset prices that are most sensitive to monetary policy are rallying strongly. If we follow the typical pattern, as was the case in 2015/16, accommodative policy will soon lead to a rebound in our preferred global growth indicators and higher bond yields. Keep portfolio duration low. Credit: The macro environment for corporate bonds remains attractive, but investors should favor high-yield bonds – particularly Caa-rated and energy debt – where spreads still have room to narrow. Yield Curve: Barbelled Treasury portfolios still make sense in the current macro environment. Specifically, we recommend that investors overweight a duration-matched 2-year/30-year barbell and underweight the 5-year bullet. Feature Bond yields have mostly trended sideways during the past few weeks, even as the S&P 500 surged. The result is that a wide gulf has opened up between the equity and bond markets (Chart 1). At times like this it becomes popular to ask whether the stock market or bond market is “right”. That is, are equities bound to sell off and re-converge with bonds? Or, will the stock market pull bond yields higher? We agree with our Global Investment Strategy team that the risk of a near-term equity sell-off is high.1 But we also think that both the equity and bond markets are responding rationally to an economic environment characterized by abundant central bank liquidity and global growth that has yet to convincingly rebound. Tech stocks are responsible for the bulk of the recent rally. To see why, we can take a look at the relative performance of different equity sectors. Technology stocks are responsible for the bulk of the recent rally, while defensive sectors have performed in-line with the benchmark index and cyclical sectors have lagged (Chart 2). This is consistent with an environment of depressed global growth and plentiful central bank liquidity. Chart 1Stocks Versus##br## Bonds Chart 2Cyclical (or Growth Sensitive) Sectors Have Lagged ... Many technology firms trade off the promise of large cash flows that will only be delivered in the distant future. In a sense, we can think of these stocks as long duration assets whose prices are very sensitive to the discount rate. The Fed’s highly accommodative interest rate guidance is the main reason for the tech sector’s outperformance. In contrast, cyclical equity sectors – like materials, industrials and energy – are less sensitive to Fed policy and more geared toward global economic growth. These sectors have lagged because global growth has yet to put in a decisive bottom. Like cyclical equity sectors, Treasury yields are also most sensitive to trends in global growth. In fact, the 10-year Treasury yield closely tracks the relative performance of cyclical versus defensive equity sectors (Chart 3). Commodity prices are also consistent with this picture (Chart 4). Gold has rallied sharply, something that often results from a shift toward more dovish monetary policy, while the growth-sensitive CRB Raw Industrials commodity index has only just begun to hook up. Historically, bond yields only rise when gains in the CRB index start to outpace gains in gold (Chart 4, bottom panel). Chart 3... Consistent With Bond Yields Chart 4The CRB/Gold Ratio But we can’t think of monetary policy and global growth as completely separate issues. They tend to follow each other in a pattern explained by our Fed Policy Loop (Chart 5). Applying the Loop to the current environment, we see that the Fed eased policy after growth weakened last year and financial markets are currently responding to this shift in monetary conditions. The most interest rate sensitive assets – e.g. tech stocks and gold – are rallying. This represents an easing of financial conditions that will eventually lead to a rebound in global growth indicators. It is only when those global growth indicators increase that US bond yields will rise. Chart 5The Fed Policy Loop On that note, we also see signs that the economy is transitioning from the ‘Asset Price Inflation’ section of the Loop to the ‘Stronger Economic Growth’ section. The US ISM Manufacturing PMI is currently downbeat at 47.2, but it should be at 50.8 according to a model based on regional Fed manufacturing surveys (Chart 6). Further, the ISM non-Manufacturing index is well above 50 and moving higher (Chart 6, panel 2). Finally, industrial production growth is nowhere near as weak as it was in 2016, even though the PMI is lower (Chart 6, bottom panel). Chart 6ISM Will Soon Trough Bottom Line: Financial markets have taken the Fed’s dovish guidance on board, and the asset prices that are most sensitive to monetary policy are rallying strongly. If we follow the typical pattern, as was the case in 2015/16, accommodative policy will soon lead to a rebound in our preferred global growth indicators and higher bond yields. Keep portfolio duration low. Stay Long Junk It’s still early, but corporate bonds have so far not joined in with this year’s equity rally. Year-to-date, the investment grade corporate bond index is only up 8 bps versus Treasuries (Chart 7). High-yield bonds have fared better. They have outperformed duration-matched Treasuries by 48 bps so far this year, and the segments of the junk market that were most beaten down in 2019 are leading the charge. Caa-rated junk bonds have outperformed Treasuries by 108 bps so far in 2020.  The energy sector has also fared well since December, and is up a decent 43 bps versus Treasuries in January. Chart 7Corporate Bond Returns Chart 8Favor HY Over IG We see the divergence between investment grade and high-yield returns continuing during the next few months, due to large differences in valuation. The investment grade corporate index spread is well below our cyclical target, while the high-yield index spread still looks cheap (Chart 8).2  High-yield’s attractiveness is mostly due to Caa-rated securities which underperformed dramatically in 2019 even as junk bonds overall delivered solid returns (Chart 8, bottom panel). As we discussed in a recent report, the underperformance of Caa-rated debt was in large part due to weakness in the shale oil sector.3 The yield curve is no longer deeply inverted out to the 5-year maturity point. Bottom Line: Corporate bonds will deliver solid returns as the economy transitions from the ‘Asset Price Inflation’ stage to the ‘Stronger Economic Growth’ stage of our Fed Policy Loop. However, relative valuation dictates that returns will concentrated in high-yield, especially Caa-rated and energy debt. Finding The Best Spot On The Yield Curve We have been recommending that investors run barbelled Treasury portfolios for some time, favoring the long and short ends of the curve at the expense of the belly (5-year/7-year). However, the shape of the curve has changed a lot since the 2/10 slope briefly inverted last August. Specifically, the curve is no longer deeply inverted out to the 5-year maturity point (Chart 9A). In light of this shift, it is worth considering whether our recommended curve positioning still makes sense. First, we take a look at the 12-month rolling yield for each point on the Treasury curve (Chart 9B). The 12-month rolling yield equals each security’s coupon return plus rolldown return. It is essentially the return you would earn in each maturity if the yield curve stayed completely unchanged during the next 12 months. Despite recent curve shifts, we still see a significant pick-up in rolling yield beyond the 5-year maturity point, as was the case last August. Chart 9APar Coupon Yield Curve Chart 9B12-Month Rolling Yield Curve But yield pick-up is just one consideration. We also need to think about how the shape of the curve will change during the next 6-12 months. One way to do this is to look at a sample of recent data – we use the past six months – and calculate how sensitive each point on the Treasury curve has been to changes in our 12-month Fed Funds Discounter.4 That is, if the market moves to price-in fewer Fed rate cuts during the next 12 months, as we expect, how should we expect each point on the Treasury curve to respond? To answer this question, Chart 10 shows how sensitive weekly changes in each Treasury yield have been to changes in our Discounter during the past six months. Chart 10Risk & Reward Along The Treasury Curve The first thing we notice is that the 5-year yield is the most sensitive to changes in our Discounter and the 2-year yield is the least sensitive. The 20-year and 30-year yields are relatively insulated from changes in the Discounter, and offer the greatest rolling yields. The second and third panels of Chart 10 show how these sensitivities change if we consider increases and decreases in our Discounter differently. Here we see that maturities from 5-20 years have been similarly sensitive to increases in the Discounter during the past six months. Meanwhile, the 5-year yield has been most sensitive to declines in the Discounter. The 2-year yield is not sensitive at all to a rising Discounter, but is fairly exposed to a falling Discounter. In general, since we expect the Discounter to move up during the next 6-12 months, the 2-year note looks like the safest place to camp out. Meanwhile, the 30-year bond looks attractive in terms of its yield pick-up per unit of sensitivity. The 2-year yield is least sensitive to changes in our Fed Funds Discounter. Another approach we can take is to look at how different parts of the yield curve respond to “risk on” and “risk off” market environments. To do this, we classify months as “risk on” if both the stock-to-bond total return ratio rises and the high-yield index spread tightens. Conversely, we classify months as “risk off” if both the stock-to-bond total return ratio falls and the high-yield index spread widens. Chart 11A shows the cumulative changes in different yield curve slopes since 2010 during “risk on” months only. The chart shows that, recently, “risk on” financial market behavior has coincided with the yield curve steepening out to the 7-year/10-year part of the curve, and then flattening beyond the 10-year point. Similarly, Chart 11B shows that “risk off” months have recently coincided with yield curve flattening out to the 7-year/10-year part of the curve, and steepening beyond that. Chart 11ASlope Changes In "Risk On" Environments Chart 11BSlope Changes In "Risk Off" Environments In other words, if recent correlations hold, a “risk on” environment during the next few months would cause the 7-year and 10-year yields to rise the most, while the 2-year and 30-year yields would have less upside. Investment Conclusions We expect economic growth to strengthen during the next 6-12 months, leading to “risk on” financial market behavior and a rising Fed Funds Discounter. Based on this view and our analysis of rolling yields and curve sensitivities, we conclude that a barbelled Treasury portfolio still makes the most sense. We want to be overweight the 2-year note because it should have less upside in a “risk-on” environment, and overweight the 30-year bond to get some extra yield pick-up while taking less risk than in the 5-year, 7-year or 10-year notes. In general, we want to avoid the 5-year, 7-year and 10-year maturities. According to our yield curve models, all three of those maturities look expensive relative to a duration-matched 2/30 barbell (Chart 12).5 Chart 12Butterfly Spread Fair Value Models If we wanted to get even more precise, we could note that a duration-matched 2/30 barbell offers 5 bps of yield pick-up compared to the 5-year note, only 1 bp of yield pick-up relative to the 7-year note and about the same yield as the 10-year note. To split hairs, those extra few basis points give us a slight preference for being short the 5-year bullet compared to the 7-year and 10-year notes, though we would prefer to avoid all three. Bottom Line: Barbelled Treasury portfolios still make sense in the current macro environment. Specifically, we recommend that investors overweight a duration-matched 2-year/30-year barbell and underweight the 5-year bullet.   Ryan Swift US Bond Strategist rswift@bcaresearch.com Footnotes 1 Please see Global Investment Strategy Weekly Report, “Time For A Breather”, dated January 10, 2020, available at gis.bcaresearch.com 2 For details on how we arrive at our spread targets please see US Bond Strategy Weekly Report, “The Value In Corporate Bonds”, dated February 19, 2019, available at usbs.bcaresearch.com 3 Please see US Bond Strategy Weekly Report, “Caa-Rated Bonds: Warning Sign Or Buying Opportunity?”, dated November 26, 2019, available at usbs.bcaresearch.com 4 Our 12-month Fed Funds Discounter measures the 12-month change in the fed funds rate that is currently priced into the overnight index swap curve. 5 For details on our yield curve models please see US Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification

Special Report

Highlights Strikes result from divergent perceptions of bargaining power: A strike reflects a negotiating failure, and negotiations fail when parties cannot agree on which side has the stronger position, typically because at least one of them overestimates its leverage. Once a strike starts, broad macro factors influence the outcome: Labor market slack, economic concentration, trends in labor relations law and regulations, and the gap between labor’s and management’s fortunes (if extreme) are the key macro determinants of negotiating leverage. Key factors that have bolstered management for decades are poised to reverse: Legal and regulatory trends have little room to improve from management’s perspective, and the gap between management’s and labor’s share of rewards is ripe for narrowing. A union resurgence is a low-probability, high-impact event: We view the potential for labor to gain the upper hand over management as a low-probability event that would have a significant impact on markets if it did come to pass, so it merits a close look. Feature We concluded Part 1 by highlighting two consistent themes from the modern history of the US labor movement: successful strikes beget strikes, and employees are unlikely to make gains if judges and government officials are disposed to favor management. In this installment, we will focus on the factors that encourage strikes and the non-government determinants of their success. Part 3 will focus on public opinion, elected officials and the judiciary. Our ultimate goal is to evaluate relative bargaining power, and the potential for organized labor to push wages significantly higher, upending the risk-friendly status quo. The Origin Of Strikes Strikes (and lockouts) occur when labor and management cannot reach a mutually acceptable settlement, often because at least one side overestimates its bargaining power. It is easy to agree when labor and management hold similar views about each side’s relative power, as when both perceive that one of them is considerably stronger. In that case, a settlement favoring the stronger side can be reached fairly quickly, especially if the stronger side exercises some restraint and does not seek to impose terms that the weaker side can scarcely abide. Restraint is rational in repeated games like employer-employee bargaining, and when both parties recognize that relative bargaining positions are fluid, they are likely to exercise it. History shows that the pendulum between labor and management swings, albeit slowly, as societal views evolve1 and the business cycle fluctuates. As a general rule, management will have the upper hand during recessions, when the supply of workers exceeds demand, and labor will have the advantage when expansions are well advanced, and capacity tightens. A high unemployment rate broadly favors employers, and a low unemployment rate favors employees. Neither the number of work stoppages (Chart 1, top panel), nor the number of workers involved (Chart 1, middle panel) correlates very well with the unemployment gap (Chart 1, bottom panel), in the Reagan-Thatcher era, however, as work stoppages have dwindled almost to zero. Chart 1Swamped By The Legal And Regulatory Tide Game theory is better equipped than simple regression models to offer insight into the origin of strikes. We posit a simple framework in which each side can hold any of five perceptions of its own bargaining power, resulting in a total of 25 possible joint perceptions. Management (M) can believe it is way stronger than Labor (L), M >> L; stronger than Labor, M > L; roughly equal, M ≈ L; weaker than Labor, L > M; or way weaker than Labor, L >> M. Labor also holds one of these five perceptions, and the interaction of the two sides’ perceptions establishes the path negotiations will follow. The fur flies when each party thinks the other should make the bulk of the concessions: labor negotiations over the next couple of years could be interesting. Limiting our focus to today’s prevailing conditions, Figure 1 displays only the outcomes consistent with management’s belief that it has the upper hand. For completeness, the exhibit lists all of labor’s potential perceptions, but we deem the two in which labor is feeling its oats (circled) to be most likely, given the success of recent high-profile strikes.2 Management’s confidence follows logically from four decades of victories, but may prove to be unfounded if its power has already peaked. Figure 1The Eye Of The Beholder Strike outcomes turn on which side has overestimated its leverage. The broad factors we use to assess leverage are overall labor market slack; economic concentration; regulatory and legal trends; and the sustainability of either side’s accumulated advantage, which we describe as the labor-management rubber band. Other factors that matter on a case-by-case basis, but are beyond the scope of our analysis, include industry-level slack, a labor input’s susceptibility to automation, and the degree of labor specialization/skill involved in that input. For these micro-level factors, a given group of workers’ leverage is inversely related to the availability of substitutes for their input. Labor Market Slack Chart 2Surprise Wage Retracements Despite muted wage growth (Chart 2), the labor market is demonstrably tight. The unemployment rate is at a 50-year low, the broader definition of unemployment is at the lowest level in its 26-year history, and the prime-age employment-to-population ratio is back to its 2001 levels, having surpassed the previous cycle’s peak (Chart 3). The job openings rate is high, indicating that demand for workers is robust, and so is the quits rate, indicating that employers are competing vigorously to meet it. The NFIB survey’s job openings and hiring plans series (Chart 4) echo the JOLTS findings. Chart 3Prime-Age Employment Is At An 18-Year High ... Chart 4... But There Are Still Lots Of Help Wanted Signs The lack of labor market slack decisively favors workers’ negotiating position. It is a sellers’ market when demand outstrips supply, and labor victories tend to be self-reinforcing. Successful strikes beget strikes, and management volunteers concessions as labor peace becomes a competitive advantage during strike waves. Given that the crisis-driven damage to the labor force participation rate has healed as the gap between the actual part rate (Chart 5, solid line) and its demographically-determined structural proxy has closed (Chart 5, dashed line), the burden of proof rests squarely with those who argue that there is an ample supply of workers waiting to come off the sidelines. Chart 5The Labor Force Participation Gap Has Closed Economic Concentration Chart 6Less Competition = More Power The trend toward economic concentration (Chart 6) has endowed the largest companies with greater market power, as evidenced by surging corporate profit margins. The greater the concentration of employment opportunities in local labor markets, the more closely they resemble monopsonies.3 Unfortunately for labor, monopsonies restrain prices just as monopolies inflate them. As our Bank Credit Analyst colleagues have shown,4 there is a robust inverse relationship between employment concentration and real wages (Chart 7). Chart 7One Huge Buyer + Plus Multiple Small Sellers = Low Prices Economic concentration has been a major driver of management’s Reagan-Thatcher era dominance. Sleepy to indifferent antitrust enforcement has helped businesses capture market power, and it will continue to prevail through 2024 unless the Democrats take the White House in November. The silver lining for workers is that concentration could have the effect of promoting labor organization in services, where unions have heretofore made limited progress. The only way for employees to combat employers’ monopsony power is to organize their way to becoming a monopoly supplier of labor. Regulatory And Legal Trends Over the last four decades, unions have endured a near-constant drubbing from state capitols, federal agencies and the courts, as union and labor protections have been under siege from all sides. Since the air traffic controllers’ disastrous strike, labor’s regulatory and legal fortunes have most closely resembled the competitive fortunes of the Harlem Globetrotters’ beleaguered opposition. But the regulatory and legal tide has been such a huge benefit for management since the beginning of the Reagan administration that it cannot continue to maintain its pace. Employees and employers need each other, and their tether can only be stretched so far before it starts pulling them back together. Investors seem to assume that it will, however, to the extent that they think about it at all. It stands to reason that employers may be similarly complacent. We will look more closely at the presidential election and its potential consequences in Part 3, but labor concerns and inequality are capturing more attention, even among Republicans. With Republicans’ inclination to side with business only able to go in one direction, the chances are good that it has peaked. The Labor-Management Rubber Band For all of the romantic allure of labor’s battles with management in the Colosseum era, employees and employers have a deeply symbiotic relationship. One can’t exist without the other, and pursuing total victory in negotiations is folly. Even too many incremental wins can prove ruinous, as the UAW discovered to its chagrin in 2008. A half-century of generous compensation and stultifying work rules saddled Detroit automakers with a burden that would have put them out of business had the federal government not intervened. Table 1Average Salaries Of Public School Teachers By State We think of labor and management as being linked by a tether with a finite range. Since neither side can thrive for long if the other side is suffering, the tether pulls the two sides closer together when the gap between them threatens to become too wide. When labor does too well for too long at management’s expense, profit margins shrink and the company’s viability as a going concern is threatened. When management does too well, deteriorating living standards drive the best employees away, undermining productivity and profitability. Before the low-paying entity’s work force becomes a listless dumping ground for other firms’ castoffs, it may rise up and strike out of desperation. Teachers’ unions might have appeared to be setting themselves up for a fall in 2018 by illegally striking in staunchly conservative West Virginia, Oklahoma and Arizona, but desperate times call for desperate measures. Per the National Education Association’s data for the 2017-18 academic year, average public school teacher pay in West Virginia ranked 50th among the 50 states and the District of Columbia, Oklahoma ranked 49th and Arizona ranked 45th (Table 1). Adjusting the nominal salaries for cost disparities across states, West Virginia placed 41st, Oklahoma 44th and Arizona 48th. Given that real teacher salaries had declined by 8% and 9% since 2009-10 in West Virginia and Arizona, respectively, the labor-management rubber band had stretched nearly to the breaking point. Consolidating The Macro Message Parties to negotiations derive leverage from the availability of substitutes. When alternative employment opportunities are prevalent, workers have a lot of leverage, because they can credibly threaten to avail themselves of them. Teaching is a skill that transfers easily, and every state has a public school system, so teachers in low-salary states have a wealth of ready alternatives. The converse is true for low-salary states; despite much warmer temperatures, it is unlikely that teachers from top-quintile states will be willing to take a 25-33% cost-of-living-adjusted pay cut to decamp to Arizona (Table 2). Table 2Cost Of Living-Adjusted Public School Teacher Salaries By State It is easy to see from Figure 2 why management has had the upper hand. Economic concentration and the legal and regulatory climate have increasingly favored it for decades. The immediate future seems poised to favor labor, however, as the legal and regulatory climate cannot get materially better for employers, and the labor-management rubber band has become so stretched that some sort of mean reversion is inevitable. We have high conviction that labor’s one current advantage, a tight labor market, will remain in its column over the next year or two. On a forward-looking basis, the macro factors as a whole are poised to support labor. Figure 2Macro Drivers Of Negotiating Leverage​​​​​​​ Takeaways After discussing this Special Report series with clients, we realized our views could easily be misinterpreted. We are not calling for an imminent union revival that drives wages higher across industries. To be clear, we think it is more likely than not that the labor movement in the United States will remain weak relative to its 1950s to 1970s heyday. We do think, however, that the probability that unions could rise up to exert the leverage that accrues to workers in a tight labor market is considerably larger than the great majority of investors perceive. Alpha – market-beating return – arises from surprises. An investor captures excess returns when s/he successfully anticipates something that the consensus does not. If the disparity involves a trivial outcome, then any excess return is likely to be trivial, but if the outcome is significant, the investor who zigged when the rest of the market zagged stands to separate him/herself from the pack. Management has been in the driver's seat, but the factors that have kept it there have a high risk of reversing. We think the outcome of a shift in leverage from employers to employees would be very large indeed. We would expect that aggregate wage gains of 4% or higher would quickly drive the Fed to impose restrictive monetary policy settings, eventually inducing the next recession and the end of the bull markets in equities, credit and property. A union revival may be a low-probability event, but it would have considerable impact on markets and the economy. Given our conviction that the probability, albeit low, is much greater than investors expect, we think the subject is well worth sustained attention. We will examine public opinion and its effect on elected officials and the courts in Part 3, which will conclude our examination of labor-management dynamics. We will publish that installment on February 3rd; next week we will publish a joint US Investment Strategy – US Bond Strategy Special Report on commercial real estate, lead-authored by Jennifer Lacombe.   Doug Peta, CFA Chief US Investment Strategist dougp@bcaresearch.com Footnotes 1 We will discuss public opinion, and its impact on elected officials and courts, in Part 3. 2 Please see the January 13, 2020 US Investment Strategy Special Report, “Labor Strikes Back, Part 1: An Investor’s Guide To US Labor History,” available at www.bcaresearch.com. 3 A monopsony is a market with a single buyer, akin to a monopoly, which is a market with only one seller. 4 Please see the July 2019 Bank Credit Analyst Special Report, “The Productivity Puzzle: Competition Is The Missing Ingredient,” available at bcaresearch.com.

Overweight The S&P software index has gone parabolic. SPX returns are extremely concentrated as we showed on Monday’s Weekly Report, with the S&P software & services GICS2 sector being responsible for 18% of the broad market’s gains since late-2018 (see bar chart below). We are participating in this rally via sustaining an overweight stance in the S&P software index – a positon that is currently up nearly 40% since inception in relative terms. However, we are compelled to raise our trailing stop to 32% (from 27% previously) as this concentrated nature of returns is making us uneasy. Should it get triggered, it will have a domino effect on our portfolio. The move to a benchmark S&P software index allocation will push the broad S&P technology sector to underweight, and consequently tilt the portfolio to a modest defensive over cyclical bent. Bottom Line: Remain overweight the S&P software index, but tighten the trailing stop to the 32% relative return mark. The ticker symbols for the stocks in this index are: BLBG – S5SOFT: MSFT, ADBE, CRM, ORCL, INTU, NOW, ADSK, ANSS, SNPS, CDNS, NLOK, FTNT, CTXS.