High-Yield

Highlights The current global trade downtrend has primarily been due to a contraction in Chinese imports. The latter reflects weakness in China's domestic demand in general and capital spending in particular. The current global manufacturing and trade downturns will prove to be drawn out. Several important markets have already experienced technical breakdowns, and a few others are at risk of doing so. EM domestic bonds and EM credit markets could be the last shoe to drop in this EM selloff. Steel, iron ore and coal prices, will all deflate further due to supply outpacing demand in China. Feature In our report last week, we argued that the odds of a liquidation phase in EM are growing. This week’s report continues exploring this theme, offering additional rationale and evidence of a pending breakdown in EM. Trade Tariffs: The Wrong Focus? The media and many investors seem to be solely focused on the impact of U.S. tariffs against imports from China. Yet these tariffs have not been the primary cause of the ongoing global manufacturing and trade recessions. It appears that the headlines and many investors are looking at individual trees and ignoring the forest. Chart I-1Chinese Imports Are Worse Than Exports Global trade contraction and China’s growth slump are not solely due to the trade tariffs imposed by the U.S. but rather stem from weakening domestic demand in China. Chart I-1 illustrates that Chinese aggregate exports are faring much better than imports. If the imposed tariffs were the main culprit behind both weakness in Chinese growth and global trade, mainland exports would have registered a far-greater hit by now than imports. However, they have not yet done so. This entails that U.S. tariffs have so far not had a substantial impact on Chinese and global manufacturing. The key point we would like to emphasize is that the current global trade downtrend has primarily been due to a contraction in Chinese imports. In turn, the accelerating decline in mainland imports is a reflection of relapsing domestic demand in China. The latter has been instigated by lethargic money/credit impulses owing to the government’s 2017-2018 deleveraging campaign and its reluctance to undertake an economy-wide irrigation type stimulus. What’s more, the recent RMB depreciation will likely intensify the Chinese import contraction already underway, as the same amount of yuan will buy less goods priced in U.S. dollars than before (Chart I-2). Given the majority of goods and commodities procured by mainland companies are priced in dollars, suppliers will receive fewer dollars, and their revenue derived from sales to and in China will continue to shrink (Chart I-3). Chart I-2RMB Depreciation Will Depress China's Purchases From Rest Of The World Chart I-3China Is In A Recession From Perspective Of Its Suppliers   We do not deny that the trade war has prompted a deterioration in sentiment among Chinese businesses and consumers as well as multinational companies, which in turn has dented both their spending and global trade. We do not see these issues reversing anytime soon. If the imposed tariffs were the main culprit behind both weakness in Chinese growth and global trade, mainland exports would have registered a far-greater hit by now than imports. Chart I-4EM EPS Are Contracting Even though U.S. President Donald Trump is flip-flopping on tariffs and their implementation, barring a major deal between the U.S. and China, business sentiment worldwide will not improve on a dime. In brief, delaying some import tariffs from September to December is unlikely to promote an imminent global trade recovery. The confrontation between the U.S. and China is profoundly not about trade: it is a geopolitical confrontation for global hegemony that will last years if not decades. Businesses in China and CEOs of multinational companies realize this, and they will not change their investment plans on Trump’s latest tweet delaying some tariffs. For now, we do not detect signs of an impending growth turnaround in China’s domestic demand and global trade. Therefore, China-related risk assets, commodities and global cyclicals are at risk of breaking down. Economic Rationale The global trade and manufacturing recession will linger for a while longer, and a recovery is not in the offing: The business cycle in EM/China continues to downshift. Consistently, corporate earnings are already or soon will be contracting in EM, China and the rest of emerging Asia (Chart I-4). EM corporate EPS contraction is broad-based (Chart I-5A and I-5B). The recent declines in oil and base metals prices entail earnings shrinkage for energy and materials companies (Chart I-5B, bottom two panels). Chart I-5AEM EPS Contraction Is Broad Based Chart I-5BEM EPS Contraction Is Broad Based   China’s monetary and fiscal stimulus has not yet been sufficient to revive capital spending in general and construction activity in particular (Chart I-6). Chinese household spending is also exhibiting little signs of recovery (Chart I-7). Chart I-6China: Building Construction Is Dwindling Chart I-7China: Consumer Spending Has Not Yet Recovered   Domestic demand continues to deteriorate, not only in China but also in other emerging economies, as we documented in our July 25 report. In EM ex-China, imports of capital goods and auto sales are contracting (Chart I-8). High-frequency freight data point to ongoing weakness in shipments in both the U.S. and China (Chart I-9). Chart I-8EM Ex-China: Domestic Demand Is Depressed Bottom Line: The current global manufacturing and trade downturns will prove to be drawn out, and investors should be wary of betting on an impending recovery. This is BCA’s Emerging Markets Strategy view and is different from BCA’s house view which is anticipating an imminent global business cycle recovery. Chart I-9Global Freight Does Not Signal Recovery   Breakdown Watch Financial market segments sensitive to the global business cycle have been splintering at the edges. These cracks appear to be proliferating to the center and will render considerable damage to aggregate equity indexes. EM corporate EPS contraction is broad-based. We explained our rationale behind using long-term moving averages to identify significant breakouts and breakdowns in last week’s report. We also highlighted the numerous breakdowns that have already transpired. Today, we supplement the list: EM equity relative performance versus DM has fallen below its previous lows (Chart I-10, top panel). Crucially, emerging Asian stocks’ relative performance versus DM has clearly breached its 2015-2016 lows (Chart I-10, bottom panel). The KOSPI and Chinese H-share indexes have broken below their three-year moving averages (Chart I-11, top two panels). Chart I-10EM Equities Relative Performance Has Broken Down Chinese bank stocks in particular have been responsible for dragging China’s H-share index lower (Chart I-11, bottom panel). In addition, Chinese small-cap stocks dropped below their December low, as have copper prices and our Risk-On versus Safe-Haven currency ratio1 (Chart I-12). Finally, German chemical and industrial share prices such as BASF, Siemens and ThyssenKrupp have decisively broken down (Chart I-13). Chart I-11Breakdowns In Korea And China...   Chart I-12...In Commodities Space As Well Chart I-13German Manufacturing Stocks Are In Free Fall   This implies that Germany’s manufacturing slowdown is not limited to the auto sector but rather is pervasive. Besides, these companies are greatly exposed to China/EM demand, and their share prices simply reflect the ongoing slump in China/EM capital spending. There are several other market signals that are at a critical technical juncture, and their move lower will confirm our downbeat view on global growth and cyclical markets. In particular: The global stocks-to-U.S. Treasurys ratio has dropped to a critical technical line (Chart I-14, top panel). Failure to hold this defense line would signal considerable downside in global cyclical assets. Similarly, the Chinese stock-to-bond ratio – calculated using total returns of both the MSCI China All-Share index and domestic government bonds – has plunged. The path of least resistance for this ratio might be to the downside (Chart I-14, bottom panel). Given China is the epicenter of the global slowdown, this ratio is of vital importance. The lack of recovery in this ratio signifies lingering downside growth risks. Finally, global cyclical sectors’ relative performance versus defensive ones is sitting on its three-year moving average (Chart I-15). A move lower will qualify as a major breakdown and confirm the absence of a global manufacturing and trade recovery. Chart I-14Global Stocks-To-Bonds Ratio: Sitting On Edge Chart I-15Global Cyclicals Versus Defensives: At A Critical Juncture   Bottom Line: Several important markets have already experienced technical breakdowns, and a few others are at risk of doing so. All in all, these provide us with confidence in maintaining our downbeat stance on EM risk assets and currencies. EM Bonds: The Last Shoe To Drop? Although EM share prices are back to their December lows, EM local currency and U.S. dollar bonds have done well this year, benefiting from the indiscriminate global bond market rally. However, there are limits to how far and for how long the performance of EM domestic and U.S. dollar bonds can diverge from EM stocks, currencies and commodities prices (Chart I-16). EM domestic bond yields have plunged close to the 2013 lows they touched prior to the Federal Reserve’s ‘Taper Tantrum’ selloff (Chart I-17, top panel). That said, on a total return basis in common currency terms, the GBI EM domestic bond index has not outperformed U.S. Treasurys, as shown in the bottom panel of Chart I-17. Chart I-16Which Way These Gaps Will Close? Chart I-17EM Domestic Bonds: Poor Risk-Reward Profile   Looking forward, EM exchange rates remain critical to the returns of this asset class. With the GBI EM local currency bond index’s yield spread over five-year U.S. Treasurys at about 400 basis points, EM currencies have very little room to depreciate before foreign investors begin experiencing losses. We believe that further RMB depreciation, commodities prices deflation and EM exports contraction all bode ill for EM exchange rates. Consequently, we expect EM local bonds to underperform U.S. Treasurys of similar duration over the next several months. German chemical and industrial share prices such as BASF, Siemens and ThyssenKrupp have decisively broken down. Finally, the euro has begun rapid appreciation versus EM currencies. This will erode EM local bonds’ returns to European investors and trigger a period of outflows. Within this asset class, our overweights are Mexico, Russia, Central Europe, Chile, Korea and Thailand, while we continue to recommend underweight positions in the Philippines, Indonesia, Turkey, South Africa, Brazil, Argentina and Peru within an EM local currency bond portfolio. As to EM credit space (hard currency bonds), these markets are overbought, and investors positioning is heavy. EM currency depreciation and lower commodities prices typically herald widening spreads. Argentina has a large weight in the EM credit indexes, and the crash in Argentine markets could be a trigger for outflows from this asset class. Technically speaking, there are already several negative signposts. The excess returns on EM sovereign and corporate bonds seem to have rolled over, having failed to surpass their early 2018 highs (Chart I-18). Besides, EM sovereign CDS spreads are breaking out (Chart I-19, top panel). Chart I-18EM Credit Markets Is Toppy Chart I-19EM Credit Space Is Entering Selloff   Finally, there are noticeable cracks in the emerging Asian corporate credit market. The price index of China’s high-yield property bonds – that account for a very large portion not only of the Chinese but also the emerging Asian corporate bond universes – has petered out at an important technical resistance level (Chart I-19, bottom panel). Further, the relative total return of emerging Asia’s investment-grade corporate bonds against their high-yield peers is correlated with Asia corporate spreads, and presently points to wider spreads (Chart I-20). The rationale is that periods when safer parts of the credit universe outperform the riskier ones are usually associated with widening credit spreads. China’s property market remains vulnerable as the central authorities in Beijing have not provided much housing-related stimulus in the current downtrend. Furthermore, companies in this space are overleveraged, generate poor cash flow and have limited access to credit. The euro has begun rapid appreciation versus EM currencies. This will erode EM local bonds’ returns to European investors and trigger a period of outflows. Overall, Chinese property developers will affect the EM credit space in two ways. First, their credit spreads will likely continue to shoot up, generating investor anxiety and outflows from this asset class. Second, reduced investment by debt-laden and cash-strapped property developers will inflict pain on industrial and materials companies in Asia and beyond. We discuss the outlook for steel, iron ore and coal, which are very exposed to Chinese construction, in the section below. Bottom Line: For asset allocators, we recommend underweighting EM sovereign and corporate credit versus U.S. investment grade, a strategy we have been advocating since August 16, 2017 (Chart I-21). For dedicated portfolios, the list of our overweights and underweights, as always, is presented at the end of the report (page 21). Chart I-20Emerging Asian Corporate Spreads Will Widen Chart I-21Favor U.S. Investment Grade Versus EM Overall Credit   As for EM domestic bonds, we continue to recommend betting on yield declines in select countries without taking on currency risk. These include Korea, Chile, Mexico and Russia. We will warm up to this asset class in general when we alter our negative EM currency view. Arthur Budaghyan Chief Emerging Markets Strategist arthurb@bcaresearch.com Chinese Steel, Iron Ore And Coal Markets: Renewed Deflation Chart II-1Is Deflation In Steel And Coal Back? Unlike 2015 when steel, iron ore and coal prices collapsed, in the current downturn they have so far held up reasonably well. They have begun falling only recently (Chart II-1). Even though we do not anticipate a 2015-type Armageddon in steel, iron ore and coal prices, they will deflate further due to supply outpacing demand in China. For both steel and coal, the pace of “de-capacity” reforms in China has diminished considerably, with declining shutdowns of inefficient capacity and rising advanced capacity, as we argued in a couple of reports last year.  This has led to a faster growth in supply, while demand has been dwindling with weak economic growth. Lower steel, iron ore and coal prices will harm Chinese and global producers along with their respective countries.2 Steel And Iron Ore First, both crude steel and steel products output will likely grow at a pace of 5-7% (Chart II-2). As the 2016-2020 steel de-capacity target (150 million tons capacity reduction) was already achieved by the end of 2018, the scale of further shutdowns will be limited. In addition, collapsing graphite electrode prices reflect an increased supply of this material. This along with more availability of scrap steel will facilitate the continuing expansion of cleaner technology (electric furnace (EF)) steel capacity and their output in China. The newly added EF steel capacity is planned at about 21 million tons in 2019 (representing 1.8% of official aggregate steel production capacity), slightly lower than the 25 million tons in 2018. Second, we expect steel products demand to grow at 3-5%, slightly weaker than output. Construction accounts for about 55% of Chinese final steel demand, with about 35% stemming from the property market and 20% from infrastructure. The automotive sector contributes about 10% of final Chinese steel demand. All of these end markets are weak and do not yet show signs of revival (Chart II-3). Chart II-2Steel Production In China Chart II-3No Recovery In Chinese Demand   Concerning iron ore price, we expect more downside than in steel. Supply disruptions among Brazilian and Australian producers were the main cause for the significant rally in iron ore prices this year. Evidence is that these producers have already resumed their output recovery. Current iron ore prices are still well above marginal production costs of major global iron ore producers. Besides, ongoing large currency depreciation in commodity producing countries will push down their marginal production costs in U.S. dollars terms. This will encourage further supply.  As China has increased its use of scrap steel in its crude steel production, the country’s iron ore demand has not grown much. In fact, imports of this raw material have contracted (Chart II-4) As scrap steel prices are currently very low relative to the price of imported iron ore (Chart II-5), steel producers in China will continue to use scrap steel instead of iron ore. Chart II-4China's Imports Of Iron Ore Have Been Shrinking Chart II-5Scrap Steel Is A Cheap Substitute For Iron Ore   Coal Chart II-6Coal Output Is Rising, But Beijing's Goal To Reduce Its Usage Chinese coal prices will also be under downward pressure. First, coal output growth will likely slow but will still stand at 2-4% down from a current 6% level (Chart II-6, top panel). The government has set a production goal of 3900 million tons for 2020. Given last year’s output of 3680 million tons, this implies only a 2.9% annual growth rate this year and the next. Second, the demand for both thermal coal and coking coal will likely weaken. They account for 80% and 20% of total coal demand, respectively. About 60% of Chinese coal is used to generate thermal power. As the country continues to promote the use of clean energy, thermal power output growth will likely slow further. Increasing the nation’s reliance on clean energy is an imperative strategic objective for Beijing. Given that thermal coal still accounts for a whopping 70% of electricity production, China will maintain its effort on reducing coal in its energy mix (Chart II-6, bottom panel). In the same vein, the government will continue to replace coal with natural gas in home heating. Finally, Chinese coal import volumes are likely to decline as the nation is increasingly relying on its domestic sources. In particular, the strategic Menghua railway construction will be completed in October. It will be used to transport the commodity from large producers in the north to the coal-deficit provinces in the south. This will reduce the nation’s coal imports, as the transportation cost of shipping domestic coal to the southern power plants will become more competitive than imported coal. Macro And Investment Implications First, companies and economies producing these commodities will face deflationary pressures. These include - but are not limited to - Indonesia, Australia, Brazil and South Africa, as well as steel producers around the world. Second, the RMB depreciation will allow China to gain further market share in the global steel market. In fact, China’s share of global steel output has been rising (Chart II-7, top panel). The bottom panel of Chart II-7 shows that steel production in the world excluding China have actually come to a grinding halt at a time when mainland producers have enjoyed high output growth. Global steel stocks have broken down and global mining equities are heading into a breakdown (Chart II-8). Chart II-7China Has Been Gaining A Share In Global Steel Market Chart II-8Breakdown In Steel And Mining Stocks   Finally, we remain bearish on commodities and other global growth sensitive currencies. In particular, we continue shorting the following basket of EM currencies against the U.S. dollar: ZAR, CLP, COP, IDR, MYR and KRW. Ellen JingYuan He, Associate Vice President ellenj@bcaresearch.com   Footnotes 1          Average of CAD, AUD, NZD, BRL, CLP & ZAR total return (including carry) indices relative to average of JPY & CHF total returns. 2      This is BCA’s Emerging Markets Strategy view and is different from BCA’s house view. Equities Recommendations Currencies, Fixed-Income And Credit Recommendations

Highlights Duration: Hawkish trade policy will continue to weigh on bond yields for at least the next few months, but a rebound in global economic growth should take hold before the end of the year. Ultimately, a growth rebound will lead to higher bond yields on a 12-month horizon, but the timing is difficult and investors should keep portfolio duration close to benchmark for the time being. High-Yield: The Fed’s accommodative policy stance and the likelihood of a global growth recovery argue for maintaining an overweight allocation to corporate credit. Within that allocation, junk bonds should outperform investment grade due to much more attractive valuations. 10-Year Treasury Yield: The current shock to global economic growth is of a similar magnitude to the one that occurred in 2015/16. However, wage and inflationary pressures are higher now than they were back then. This means that the 10-year Treasury yield will not re-visit the 2016 trough of 1.37%, and is probably already close to its floor. Feature Regular readers will be aware of our Fed Policy Loop framework for analyzing the wiggles in financial markets. The Loop works as follows: Step 1: A dovish shift in Fed policy leads to a favorable market reaction, easing financial conditions. Step 2: Easier financial conditions suggest to the Fed that economic growth will strengthen in the future. The Fed can therefore respond by adopting a more hawkish policy stance. Step 3: The Fed’s hawkish policy shift leads to a negative market reaction, tightening financial conditions. Step 4: Tighter financial conditions suggest to the Fed that economic growth will weaken in the future. The Fed is forced to ease monetary policy at the margin. Return to Step 1 But it appears that BCA readers aren’t the only ones aware of the Fed Policy Loop. President Trump has also been exploiting the two-way relationship between Fed policy and financial conditions as he escalates his trade war with China. Chart 1 illustrates how this has been working. Step 1 of the Fed policy loop continues to function exactly as described above. However, the last few times that financial conditions have eased, the President has seized the opportunity to ratchet up trade tensions. Much like the Fed, the President reasons that periods of easier financial conditions are when the economy and financial markets can best handle a negative shock. The fall-out is that financial conditions tighten in response to the hawkish trade announcement, and the Fed is forced to respond to tighter financial conditions by turning even more dovish. The end result is that the part of the Fed Policy Loop labeled “Hawkish Fed” is by-passed. Without that step it is impossible for bond yields to rise (Chart 2). Chart 2The Back-Drop Of The Interrupted Fed Policy Loop Our Geopolitical Strategy service provided a comprehensive breakdown of U.S./China trade negotiations in last week’s report.1 The overall message is that the 2020 election is the President’s main constraint. He views hawkish trade policy as a winning issue, but only insofar as it can be accomplished without a significant decline in the stock market or economic activity. Faced with that constraint, the President will continue to interrupt the Fed Policy Loop, and the Fed will continue to do its job by adopting a more dovish monetary policy to offset possible trade shocks. At present, this means that another rate cut is likely in September. Against the back-drop of the “interrupted” Fed Policy Loop, Treasury yields can only move higher if global economic growth strengthens. In that case, the policy loop will remain operative, but at an overall higher level of yields. With that in mind, while hawkish trade policy will continue to weigh on bond yields for at least the next few months, a rebound in global economic growth should take hold before the end of the year. This will lead to higher bond yields on a 12-month horizon. Still Tracking The 2015/16 Roadmap In our research, we have repeatedly pointed out the similarities between the 2015/16 episode of flagging global growth and the current period. Specifically, we continue to witness weak manufacturing data – both in the U.S. and abroad – but a resilient service sector and strong labor market. Much like in 2015/16, we expect that the shifts toward easier monetary policy in the U.S. and more accommodative credit conditions in China will eventually put a floor under the global manufacturing cycle. The Fed will continue to do its job by adopting a more dovish monetary policy to offset possible trade shocks. At present, this means that another rate cut is likely in September. Case in point, even as President Trump has tightened global financial conditions at the margin through his hawkish trade policy, overall global financial conditions have eased since the beginning of the year (Chart 3). In 2016, easier financial conditions eventually led to upturns in crucial measures of global growth such as the Goldman Sachs Current Activity Indicator (Chart 3, top panel), the Global Manufacturing PMI (Chart 3, panel 2), and the CRB Raw Industrials index (Chart 3, bottom panel). The same dynamic should play out this time around. It’s likely that the main reason why global growth has not responded as quickly as it did in 2016 is that Chinese policy easing has not been as rapid (Chart 4). Our China Investment Strategy service’s Li Keqiang Leading Indicator – a composite measure of money and credit indicators designed to lead Chinese economic activity – has clearly bottomed, but has not yet surged as it did in 2015/16. However, Chinese policy easing continues to ramp up, a process that will continue in the months ahead. The most recent indication of this trend was China’s decision to de-value its currency versus the U.S. dollar, causing the exchange rate to jump above the important psychological threshold of 7 yuan per dollar (Chart 4, bottom panel). China took similar measures to de-value its currency in August 2015, a move that initially roiled markets but eventually helped usher in a rebound in global growth. Chart 3The 2015/2016 Scenario Has Yet To Play Out... Chart 4...As Long As China Does Not Stimulate More When it comes to strategy, we remain confident that global growth is close to a trough, but admit that timing the rebound is difficult. One indicator that should help with timing is the ratio between the CRB Raw Industrials index and Gold (Chart 5). This ratio is tightly correlated with the 10-year Treasury yield, and will only rise when the perceived improvement in global growth – proxied by the CRB index – starts to outpace the perceived dovish tilt to Fed policy – proxied by the rising gold price. Chart 5Keep Tracking The CRB / Gold Ratio In light of these difficulties with timing, we recommend that investors keep portfolio duration close to benchmark, but position for a rebound in global growth by maintaining an overweight allocation to credit risk and by running a heavily barbelled Treasury portfolio, overweighting the long and short ends of the curve while avoiding the 5-year and 7-year maturities. The barbell strategy increases average portfolio yield, and also avoids the part of the yield curve that will suffer the most when yields rise. Take Credit Risk In Junk As mentioned above, we recommend that investors maintain an overweight allocation to corporate credit versus Treasuries, despite our recent shift to benchmark duration.2 This is particularly true for high-yield bonds, where spreads are very attractive. Charts 6A and 6B show one of our favorite ways of looking at corporate bond spreads. The charts show the 12-month breakeven spread for each credit tier as a percentile rank relative to history.3 We show each credit tier individually to control for the time-varying average credit rating of the overall indexes. Similarly, we show breakeven spreads instead of the average option-adjusted spreads to control for the time-varying average duration of the bond indexes. Chart 6A shows the following valuation for investment grade credit tiers: Throughout history, Aaa credits have been more expensive than they are today only 13% of the time. Aa credits have been more expensive than they are today 19% of the time. A-rated credits have been more expensive 20% of the time. Baa credits have been more expensive 33% of the time. Chart 6B shows that the corresponding valuation for high-yield is much more compelling: Ba credits have been more expensive than today 55% of the time. B credits have been more expensive 81% of the time. Caa credits have been more expensive 84% of the time. Chart 6AInvestment Grade Breakeven Spreads Chart 6BHigh-Yield Breakeven Spreads In general, this way of looking at spreads shows that investment grade credits are quite expensive, while high-yield credits are either fairly valued or cheap. However, there is one more adjustment we can make to get an even better picture of corporate bond value. Adjusting For The Phase Of The Cycle A useful tool for cyclical portfolio allocation is to split the cycle into three phases based on the slope of the yield curve (Chart 7). We define the three phases as: Chart 7The Three Phases Of The Cycle Phase 1: From the end of the last recession until the 3/10 Treasury slope flattens to below 50 bps. Phase 2: When the 3/10 slope is between 0 bps and +50 bps. Phase 3: From when the 3/10 slope inverts until the start of the next recession. We have previously discussed the implications of the different phases for bond portfolio allocation in more depth.4 This week, we simply want to point out that credit spreads tend to be tighter during Phase 2 of the cycle, when monetary policy has tightened, but not by enough to cause a surge in corporate defaults. The recent surge in investment grade net debt-to-EBITDA likely reflects the shift toward a greater concentration of Baa-rated issuers. With this cyclical decomposition in mind, we can calculate the median breakeven spread for each credit tier in past Phase 2 periods and use that as a spread target for this cycle. We then convert our breakeven spread targets into average option-adjusted spread targets using current index duration. Charts 8A and 8B show how far each credit tier’s spreads are from target. The message is quite clear. Outside of Aaa, investment grade credits are more or less fairly valued, while high-yield credits appear very cheap. Chart 8AInvestment Grade Spread Targets Chart 8BHigh-Yield Spread Targets One might reasonably challenge this approach to corporate bond valuation by noting that, outside of looking at credit tiers individually, we have not taken fundamental credit quality trends into account. That is, we have made no adjustment for the fact that the credit quality of a Ba-rated issuer might be worse today than in prior cycles. We are skeptical that fundamental credit metrics matter more than the phase of the monetary policy cycle when it comes to corporate bond spread forecasting.5 However, this point of view is still worth exploring, especially considering that net debt-to-EBITDA for the median corporate bond issuer is quite elevated compared to history (Chart 9). Note that we have not attempted to maintain consistent weightings between the different credit tiers in the bottom-up samples shown in Chart 9. This means that the recent surge in investment grade net debt-to-EBITDA likely reflects the shift toward a greater concentration of Baa-rated issuers. Nonetheless, the net debt-to-EBITDA ratio of the median junk issuer is clearly worse than during the past two recoveries. But even if we take this into account by looking at the ratio between the junk index 12-month breakeven spread and the median net debt-to-EBITDA, we see that the ratio is still close to its historical median (Chart 10). In other words, at current spread levels junk investors appear reasonably compensated for the elevated median net debt-to-EBITDA ratio Chart 9Elevated Corporate Leverage Chart 10Favor Junk Bonds Bottom Line: The Fed’s accommodative policy stance and the likelihood of a global growth recovery argue for maintaining an overweight allocation to corporate credit. Within that allocation, junk bonds should outperform investment grade due to much more attractive valuations. Close To The Floor Chart 11Now Vs. Mid-2016 In a prior report we walked through the process of creating a macroeconomic fair value model for the 10-year Treasury yield, with a focus on describing the different independent variables that might be included in such a model, and the rationale for each one.6 This week, we focus on two vital macroeconomic variables and use them to demonstrate why the 10-year Treasury yield is unlikely to re-visit its mid-2016 trough of 1.37%. The two main variables we focus on are (i) the pace of economic growth, and (ii) the size of the output gap. All else equal, a stronger pace of economic growth leads to expectations for a higher policy rate in the future and a higher 10-year Treasury yield today. However, it is not just the pace of growth that matters. The same rate of economic growth generates more inflationary pressure when the output gap is small than when it is large. This means that bond yields should be higher when the output gap is smaller (or more specifically, less negative). We have found that the Global Manufacturing PMI is probably the indicator of economic growth that correlates best with the 10-year Treasury yield. Similarly, measures of wage growth – and to a lesser extent core inflation – tend to give the best read on the output gap. With that in mind, we can see how these factors look today relative to when the 10-year yield troughed at 1.37% in mid-2016 (Chart 11). Global economic growth looks slightly worse, but not dramatically so. The Global Manufacturing PMI is at 49.3 today. It troughed at 49.9 in 2016. If this were the only variable that mattered, we might reason that the 10-year yield should be below 1.37% already. But we also need to consider that wage growth and inflation are both much higher than in 2016. Average hourly earnings are growing at a year-over-year rate of 3.2%, compared to a rate of 2.8% when the 10-year troughed in 2016. Similarly, the Atlanta Fed’s measure of median wage growth is up to 3.7% for the un-weighted sample and 3.9% for the sample that is weighted to more closely match the demographic characteristics of the overall population (Chart 11, panel 3). It’s true that core PCE inflation is running below where it was in mid-2016, but the trimmed mean measure is much higher (Chart 11, bottom panel). The core PCE inflation measure also has a strong track record of converging toward the trimmed mean, a process we expect is playing out again. The core PCE inflation measure also has a strong track record of converging toward the trimmed mean, a process we expect is playing out again. Bottom Line: The current shock to global economic growth is of a similar magnitude to the one that occurred in 2015/16. However, wage and inflationary pressures are higher now than they were back then. This means that the 10-year Treasury yield will not re-visit the 2016 trough of 1.37%, and is probably already close to its floor.   Ryan Swift, U.S. Bond Strategist rswift@bcaresearch.com Footnotes 1 Please see Geopolitical Strategy Weekly Report, “The Rattling Of Sabers”, dated August 9, 2019, available at gps.bcaresearch.com 2 Please see U.S. Bond Strategy Portfolio Allocation Summary, “Underinsured”, dated August 6, 2019, available at usbs.bcaresearch.com 3 The 12-month breakeven spread is the basis point widening required on a 12-month horizon for each credit tier to break even with a duration-matched position in Treasuries. 4 Please see U.S. Bond Strategy Special Report, “2019 Key Views: Implications For U.S. Fixed Income”, dated December 11, 2018, available at usbs.bcaresearch.com 5 Please see U.S. Bond Strategy / Global Fixed Income Strategy Special Report, “The Risk From U.S. Corporate Debt: Theory And Evidence”, dated April 23, 2019, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Weekly Report, “Bond Kitchen”, dated April 9, 2019, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification

Highlights Chinese economic growth slowed in June & July, but at a more moderate pace than had been the case earlier this year. The housing market is a notable exception, which appeared in June to slow in a broad-based fashion. The near-term (0-3 month) outlook is bearish for China-related assets, and investors should bet on further weakness in the RMB. However, investors should remain cyclically bullish (i.e., over a 6-12 month time horizon) in anticipation of an eventual reflationary response from Chinese policymakers that will boost domestic demand enough to offset the intensifying drag from weak external demand. Feature Tables 1 and 2 on pages 2 and 3 highlight key developments in China’s economy and its financial markets over the past month. On the growth front, coincident measures of economic activity suggest that China’s economy experienced “controlled weakness” in June and July: growth continued to slow, but at a more moderate pace than had been the case in late-2018 and early-2019. The housing market appeared to be the exception to this relative stability; all 10 of the core housing indicators that we track decelerated in June, suggesting that a moderation in housing-related activity was broad-based. This implies that a further slowdown in construction is likely over the coming months, barring a meaningful pickup in sales. Table 1China Macro Data Summary Table 2China Financial Market Performance Summary Within financial markets, all of the key developments have occurred over the past several trading days, in response to President Trump’s threat last week to further hike U.S. import tariffs at the beginning of September. Both investable and domestically-listed Chinese stocks have significantly underperformed the global benchmark, as have Hong Kong stocks in response to intensifying protests in the city. A sharp decline in the RMB and the U.S. designation of China as a currency manipulator have unnerved Chinese and global investors, and our bias is to expect even further weakness in the yuan. The near-term outlook remains bearish for China-related assets, as we see the selloff over the past week as the beginning of a financial market riot point that will force policymakers, both in China and the U.S., to address the economic weakness that a full-tariff scenario will entail. The near-term outlook remains bearish for China-related assets, as we see the selloff over the past week as the beginning of a financial market riot point that will force policymakers, both in China and the U.S., to address the economic weakness that a full-tariff scenario will entail. Still, investors should remain cyclically bullish in anticipation of an eventual reflationary response from Chinese policymakers that will boost domestic demand enough to offset the drag from weak external demand. Otherwise stated, we expect Chinese relative performance to trend lower in the near-term, but to be higher 12-months from today. Investors should also continue to hedge the inherent RMB exposure from a long US$ cyclical position in Chinese stocks with a long USD-CNH position. In reference to Tables 1 and 2, we provide below several detailed observations concerning developments in China’s macro and financial market data: Our leading indicator for the Li Keqiang Index is now in a clear uptrend, most recently led by a meaningful improvement in monetary conditions and credit growth. Chart 1Our Leading Indicator Is Now In A (Moderate-Strength) Uptrend Chart 2Money, And Still-Modest Credit Growth, Are Holding Back Our Leading Indicator The Li Keqiang index (LKI) rose moderately in June after a significant decline in May, but remains in a downtrend (Chart 1). The June increase was driven entirely by a pickup in electricity production (which had nearly contracted in May); bank loan growth and rail cargo volume both decelerated. The takeaway for investors is that while the Chinese economy did not slow meaningfully further in June, the pace of growth remained tepid, suggesting the economic activity remains vulnerable to a further increase in U.S. import tariffs. Our leading indicator for the LKI is now in a clear uptrend, most recently led by a meaningful improvement in monetary conditions and credit growth (Chart 2). However, the magnitude of the rise in the indicator is being held back by growth in the money supply, which has only slightly accelerated over the past few months, as well as a “half strength” recovery in credit. Our view is that Chinese policymakers are likely to wait for further economic weakness before allowing money & credit growth to significantly overshoot, which increases the odds of a continued market riot in the short-term. Chart 3Decelerating House Price Appreciation Is Coming All 10 of the housing indicators shown in Table 1 decelerated in June, suggesting that a moderation in housing-related activity was broad-based. Our BCA 70-city diffusion index for (YoY) house prices has an excellent track record at leading inflection points in overall price growth (Chart 3), and is currently suggesting that house price appreciation is at risk of falling back to mid-2018 levels (which would imply a 5-6 percentage point deceleration). Continued weakness in floor space sold continues to suggest that the ongoing pace of housing construction is unsustainable; we expect a further moderation in floor space started over the coming several months barring a meaningful pickup in sales. Both the Caixin and official manufacturing PMI for China rose in July, including the official new export orders component (which we have been closely following). However, the survey was taken prior to President Trump’s renewed tariff threat last week, and we expect the July gains to reverse in August barring a major de-escalation in the conflict. Both investable and domestically-listed Chinese stocks have significantly underperformed the global benchmark over the past week due to President Trump’s threat to impose tariffs on all remaining imports from China. We noted in our May 29 weekly report that a financial market riot point remained likely over the coming few months,1 and we explicitly recommend an underweight position in Chinese stocks for the remainder of 2019 in last week’s report.2 Still, investors should remain cyclically bullish in anticipation of an eventual reflationary response from Chinese policymakers that will boost domestic demand enough to offset the drag from weak external demand. Investors who are already positioned in favor of Chinese stocks should stay long, despite the likelihood of further near-term losses. ​​​​​​​Investors should remain cyclically bullish in anticipation of an eventual reflationary response from Chinese policymakers that will boost domestic demand enough to offset the drag from weak external demand. Chart 4Intensifying Protests Have Weighed On Hong Kong's Relative Equity Performance The MSCI Hong Kong index has also significantly underperformed the global benchmark since late-July, in response to intensifying protests in the city (Chart 4). The protests have been driven by underlying socio-economic factors as well as Beijing’s encroachment on traditional political liberties. However, Hong Kong has no real alternative to Beijing’s sovereignty, and the unrest should gradually die down as long as the imposition of martial law is avoided. Nonetheless, Hong Kong’s stock market is likely to remain under pressure in the interim; for now, we recommend that investors stay underweight versus China and Taiwan.​​​​​​​ The sector performance within China’s investable and domestically-listed equity markets over the past month has largely been along cyclical / defensive lines. In the investable market, consumer staples, health care, financials, information technology, communication services, and utilities have all outperformed, in contrast to energy, materials, industrials, consumer discretionary, and real estate stocks. The pattern has been similar in the domestic market, with two exceptions: modest staples underperformance, and material underperformance of comm services. Real estate stocks have been among the worst performers in both markets over the past month, possibly in response to the deteriorating housing market data that we highlighted above. China’s 3-month repo rate has fallen approximately 20 bps over the past month, and is now back close to its one-year low. We continue to believe that a cut to the benchmark lending rate is unlikely in the near-term, but could occur in Q4 if economic conditions in China weaken materially further.​​​​​​​ Chinese onshore corporate spreads increased modestly over the past month, but have not yet risen to a new high for the year. The uptrend in spreads that began in late-May does reflect renewed risks to the Chinese economy from a further increase in U.S. import tariffs, but annualizing the most recent pace of onshore corporate defaults suggests that onshore bond spreads are still much too high. Our long China onshore corporate bond trade continues to register gains in local currency terms (Chart 5), and we recommend that investors stick with a long/overweight currency-hedged stance. ​​​​​​​Our bias is to bet on further RMB weakness until policymakers aggressively ramp up their reflationary efforts. The yuan weakened sharply this week, with the U.S. dollar breaking above 7 versus both the onshore and offshore RMB (Chart 6). This is the weakest level for the currency since the global financial crisis, and the decline has clearly occurred in response to last week’s tariff threat. We noted in our May 15 report that a meaningful decline in the exchange rate would likely be required in order to stabilize the outlook for earnings & the economy,3 and we recommended at that time that investors should hedge the inherent RMB exposure from a long US$ cyclical position in Chinese stocks by opening a long USD-CNH trade. It is difficult to forecast how much further the RMB is likely to fall, but our bias is to bet on further weakness until policymakers aggressively ramp up their reflationary efforts. Stay tuned. Chart 5Despite Ongoing Default Concerns, Onshore Corporate Bonds Are Winners Chart 6Weakest RMB In A Decade, And Further Declines Are Likely   Jonathan LaBerge, CFA, Vice President Special Reports jonathanl@bcaresearch.com   Jing Sima China Strategist jings@bcaresearch.com 1      Please see China Investment Strategy Weekly Report, “Waiting For The Pain,” dated May 29, 2019. 2      Please see China Investment Strategy Weekly Report, “Threading A Stimulus Needle (Part 2): Will Proactive Fiscal Policy Lose Steam?” dated July 24, 2019. 3      Please see China Investment Strategy Weekly Report, “Simple Arithmetic,” dated May 15, 2019.   Cyclical Investment Stance Equity Sector Recommendations

Highlights Chart 1Keep Tracking The CRB / Gold Ratio The Fed cut rates by 25 basis points last week, a move that Chairman Powell described as an “insurance” cut meant to counter the risks from trade tensions and global growth weakness. Powell also described the move as a “mid-cycle adjustment to policy” and not “the beginning of a lengthy cutting cycle”. We agree with the Fed’s “mid-cycle” view of the U.S. economy and think an extended cutting cycle is unwarranted, but the market clearly disagrees. Long-end yields fell on Powell’s remarks and fell further as U.S. / China trade tensions re-escalated during the past few days. The 2015/16 period continues to be a good roadmap for the current environment, and we expect the next big move in Treasury yields will be higher. The timing of that move, however, is highly uncertain. Our political strategists expect an increase in saber-rattling between the U.S. and China in the coming months, and bond yields will not rise until either trade tensions ease and/or the global growth data recover. We recommend a tactical neutral allocation to portfolio duration, but expect to switch back to below-benchmark when those conditions are met. The CRB / Gold ratio will continue to be a good guide for the 10-year yield (Chart 1). Feature Investment Grade: Overweight Chart 2Investment Grade Market Overview Investment grade corporate bonds outperformed the duration-equivalent Treasury index by 63 basis points in July, bringing year-to-date excess returns up to +432 bps. Corporate spreads widened somewhat following Jerome Powell’s perceived hawkishness at last week’s FOMC meeting, but that spread widening will prove fleeting. The Fed remains committed to keeping monetary policy accommodative and that means doing everything it can to prevent a significant tightening of financial conditions.1 The soaring price of gold is the strongest indicator of the Fed’s dovishness, and it is also a buy signal for corporate credit (Chart 2). In terms of valuation, Baa-rated securities offer the most value in investment grade corporate bond space. Baa spreads remain 7 bps above our cyclical target.2 Conversely, Aa and A-rated spreads are 3 bps and 4 bps below target, respectively (panel 4). Aaa spreads are 16 bps below target (not shown). The Fed’s Senior Loan Officer Survey for Q2, released yesterday, showed that commercial & industrial (C&I) lending standards eased for the second consecutive quarter. C&I loan demand continued to contract, but less aggressively than its recent pace (bottom panel). Easing lending standards usually coincide with spread tightening, and vice-versa.  High-Yield: Overweight Chart 3High-Yield Market Overview High-Yield outperformed the duration-equivalent Treasury index by 66 basis points in July, bringing year-to-date excess returns up to +673 bps. The average index option-adjusted spread tightened 6 bps in July, then widened 26 bps in the first two days of August. At 397 bps, it is currently well above the cycle-low of 303 bps. We see more potential for spread tightening in high-yield than in investment grade. Within investment grade, only Baa-rated spreads appear cheap. However, in high-yield, Ba-rated spreads are 71 bps above our target (Chart 3), B-rated spreads are 142 bps above our target (panel 3) and Caa-rated spreads are 298 bps above our target (not shown).3 Junk spreads also offer reasonable value relative to expected default losses. The current Moody’s baseline forecast calls for a default rate of 2.9% over the next 12 months, not far from our own projection.4 This would translate into 238 bps of excess spread in the High-Yield index, after adjusting for default losses (panel 4). This is comfortably above zero, and only just below the historical average of 250 bps. As noted on page 3, C&I lending standards have now eased for two consecutive quarters and job cut announcements are off their highs (bottom panel). Both trends are supportive of lower default expectations in the future. MBS: Neutral Chart 4MBS Market Overview Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 43 basis points in July, bringing year-to-date excess returns up to +32 bps. The conventional 30-year zero-volatility spread tightened 10 bps on the month, consisting of a 9 bps tightening in the option-adjusted spread (OAS) and a 1 bp decline in the compensation for prepayment risk (option cost). Falling mortgage rates hurt MBS in the first half of this year, as lower rates led to an increase in refi activity that drove MBS spreads wider (Chart 4). In fact, the conventional 30-year index OAS moved all the way back to its pre-crisis mean, before tightening last month (panel 3). However, as we noted in a recent report, the nominal 30-year MBS spread remains very tight, at close to one standard deviation below its historical mean.5 The mixed valuation picture means we are not yet inclined to augment MBS exposure, especially given the recent downleg in Treasury yields that could spur another small jump in refis. However, we are equally disinclined to downgrade MBS, given our view that Treasury yields are close to a trough. All in all, we expect the next big move in the MBS/Treasury basis will be a tightening, as global growth improves and mortgage rates rise. However, valuation is not sufficiently attractive to warrant more than a neutral allocation.   Government-Related: Underweight Chart 5Government-Related Market Overview The Government-Related index outperformed the duration-equivalent Treasury index by 30 basis points in July, bringing year-to-date excess returns up to +164 bps. Sovereign debt outperformed duration-equivalent Treasuries by 68 bps on the month, bringing year-to-date excess returns up to +490 bps. Local Authorities outperformed the Treasury benchmark by 31 bps, bringing year-to-date excess returns up to +244 bps. Meanwhile, Foreign Agencies outperformed by 49 bps, bringing year-to-date excess returns up to +153 bps. Domestic Agencies outperformed by 6 bps in July, bringing year-to-date excess returns up to +31 bps. Supranationals outperformed by 7 bps on the month, bringing year-to-date excess returns up to +36 bps. Sovereign debt remains very expensive relative to equivalently rated U.S. corporate credit (Chart 5). While the sector would benefit if the Fed’s dovish pivot eventually results in a weaker dollar, U.S. corporate bonds would still outperform in that scenario given the more attractive starting point for spreads. We continue to recommend an underweight allocation to Sovereigns. Unlike the debt of most other countries, Mexican sovereign bonds continue to trade cheap relative to U.S. corporates (bottom panel). While this remains an attractive option from a valuation perspective, the President’s on again/off again tariff threats make it a risky near-term proposition. Municipal Bonds: Neutral Chart 6Municipal Market Overview Municipal bonds outperformed the duration-equivalent Treasury index by 102 basis points in July, bringing year-to-date excess returns up to +58 bps (before adjusting for the tax advantage). The average Aaa-rated Municipal / Treasury yield ratio fell 8% in July, and currently sits at 78% (Chart 6). The ratio is more than one standard deviation below its post-crisis mean, and even below the 81% average that prevailed in the late stages of the previous cycle, between mid-2006 and mid-2007. We noted the strong outperformance of municipal bonds in our report two weeks ago, and recommended cutting exposure from overweight to neutral, based on how expensive the bonds have become.6 In that report we noted that Aaa-rated Municipal / Treasury yield ratios for 2-year, 5-year and 10-year maturities were all more than one standard deviation below average pre-crisis levels. Only 20-year and 30-year Aaa-rated municipal bonds continue to look cheap, and we recommend that investors focus muni exposure on that segment of the market. Fundamentally, state & local government balance sheets remain in decent shape and a material increase in ratings downgrades is unlikely any time soon (bottom panel). Our shift to a more cautious stance is driven purely by valuation, and not any immediate concern for municipal bond credit quality. Treasury Curve: Maintain A Barbell Curve Positioning Chart 7Treasury Yield Curve Overview The Treasury curve bear-flattened in July, before undergoing a roughly parallel shift down of about 30 bps in the first two days of August, following the FOMC meeting and news about the escalation of the U.S./China trade war. As we go to press, the 2/10 Treasury slope stands at 16 bps, 9 bps flatter than at the end of June. The 5/30 slope is currently 76 bps, exactly equal to its end-of-June level. Our 12-month Fed Funds Discounter is currently -78 bps (Chart 7). This means that the market is priced for roughly three more 25 basis point rate cuts during the next year. While we have shifted to a tactically neutral duration stance because of the uncertainty surrounding the timing of the next move higher in yields, three rate cuts on a 12-month horizon still seems excessive given the underlying strength of the U.S. economy. For this reason we are inclined to maintain a barbelled position across the Treasury curve, and also to stay short the February 2020 fed funds futures contract. The February 2020 contract is priced for three rate cuts spread over the next four FOMC meetings. A short position continues to make sense. On the yield curve, our butterfly spread models continue to show that barbells look cheap relative to bullets (see Appendix B). Further, the 5-year and 7-year yields will rise the most when the market prices-in a more hawkish path for the policy rate. Investors should favor the long-end and short-end of the curve, while avoiding the belly (5-year and 7-year). TIPS: Overweight Chart 8Inflation Compensation TIPS outperformed the duration-equivalent nominal Treasury index by 43 basis points in July, bringing year-to-date excess returns up to +71 bps. The 10-year TIPS breakeven inflation rate rose 8 bps in July to reach 1.77%, before falling back to 1.67% in the first few days of August (Chart 8). The 5-year/5-year forward TIPS breakeven inflation rate followed a similar path and currently sits at 1.88%. As we have noted in recent research, FOMC members are monitoring long-dated inflation expectations and are committed to keeping policy easy enough to “re-anchor” them at levels consistent with the Fed’s 2% target.7 In the long-run, this will support a return of long-dated TIPS breakeven inflation rates (both 10-year and 5-year/5-year forward) to our 2.3% - 2.5% target range. However, for breakevens to move higher, investors will also need to see evidence that realized inflation can be sustained near 2%. On that note, the core PCE deflator grew at an annualized rate of 2.48% during the past three months. However, the 12-month rate of change remains at 1.5%. The 12-month trimmed mean PCE inflation rate is currently running at 2%, exactly equal to the Fed’s target. In a recent report we noted that 12-month core PCE inflation has a track record of converging toward the trimmed mean.8 We see continued upside in core inflation over the remainder of the year, and therefore recommend an overweight allocation to TIPS versus nominal Treasuries.  ABS: Underweight Chart 9ABS Market Overview Asset-Backed Securities outperformed the duration-equivalent Treasury index by 8 basis points in July, bringing year-to-date excess returns up to +59 bps. The index option-adjusted spread for Aaa-rated ABS tightened 3 bps on the month. It currently sits at 31 bps, well below the pre-crisis mean of 64 bps (Chart 9). In addition to poor valuation, the sector’s credit fundamentals are shifting in a negative direction. Household interest payments continue to trend up, suggesting a higher delinquency rate going forward (panel 3). Meanwhile, the Fed’s Senior Loan Officer Survey for Q2, released yesterday, showed a continued tightening in lending standards for both credit cards and auto loans. Tighter lending standards usually coincide with rising delinquencies (bottom panel). On the bright side, stronger demand for both credit cards and auto loans was reported for the first time since the fourth quarter of 2016. All in all, the combination of poor value and deteriorating credit quality leads us to recommend an underweight allocation to consumer ABS.       Non-Agency CMBS: Neutral     Chart 10CMBS Market Overview Non-Agency Commercial Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 42 basis points in July, bringing year-to-date excess returns up to +234 bps. The index option-adjusted spread for non-agency Aaa-rated CMBS tightened 6 bps on the month. It currently sits at 64 bps, below average pre-crisis levels but above levels seen in 2018 (Chart 10). The macro outlook for commercial real estate looks somewhat unfavorable, with lenders tightening standards (panel 4) amidst falling demand (bottom panel). However, on a positive note, commercial real estate prices recently accelerated and are now much more consistent with current CMBS spreads (panel 3). Despite the mixed fundamental picture, CMBS still offer excellent compensation compared to other similarly-rated fixed income sectors.9 Agency CMBS: Overweight   Agency CMBS outperformed the duration-equivalent Treasury index by 26 bps in July, bringing year-to-date excess returns up to +119 bps. The index option-adjusted spread tightened 3 bps on the month and currently sits at 47 bps. The Excess Return Bond Map in Appendix C shows that Agency CMBS offer high potential return compared to other low-risk spread products. An overweight allocation to this defensive sector remains appropriate. Appendix A - The Golden Rule Of Bond Investing We follow a two-step process to formulate recommendations for bond portfolio duration. First, we determine the change in the federal funds rate that is priced into the yield curve for the next 12 months. Second, we decide – based on our assessments of the economy and Fed policy – whether the change in the fed funds rate will exceed or fall short of what is priced into the curve. Most of the time, a correct answer to this question leads to the appropriate duration call. We call this framework the Golden Rule Of Bond Investing, and we demonstrated its effectiveness in the U.S. Bond Strategy Special Report, “The Golden Rule Of Bond Investing”, dated July 24, 2018, available at usbs.bcaresearch.com. Chart 11 illustrates the Golden Rule’s track record by showing that the Bloomberg Barclays Treasury Master Index tends to outperform cash when rate hikes fall short of 12-month expectations, and vice-versa. Chart 11The Golden Rule's Track Record At present, the market is priced for 78 basis points of cuts during the next 12 months. We anticipate fewer rate cuts over that time horizon, and therefore anticipate that below-benchmark portfolio duration positions will profit. We can also use our Golden Rule framework to make 12-month total return and excess return forecasts for the Bloomberg Barclays Treasury index under different scenarios for the fed funds rate. Excess returns are relative to the Bloomberg Barclays Cash index. To forecast total returns we first calculate the 12-month fed funds rate surprise in each scenario by comparing the assumed change in the fed funds rate to the current value of our 12-month discounter. This rate hike surprise is then mapped to an expected change in the Treasury index yield using a regression based on the historical relationship between those two variables. Finally, we apply the expected change in index yield to the current characteristics (yield, duration and convexity) of the Treasury index to estimate total returns on a 12-month horizon. The below tables present those results, along with 95% confidence intervals. Excess returns are calculated by subtracting assumed cash returns in each scenario from our total return projections. Appendix B - Butterfly Strategy Valuation 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: U.S. Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com U.S. 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 August 2, 2019) 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 August 2, 2019) 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 +55 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 55 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 U.S. fixed income market. The Map employs volatility-adjusted breakeven spread analysis to show how likely it is that a given sector will earn/lose money during the subsequent 12 months. The Map does not incorporate any macroeconomic view. The horizontal axis of the Map shows the number of days of average spread widening required for each sector to lose 100 bps versus a position in duration-matched Treasuries. Sectors plotting further to the left require more days of average spread widening and are therefore less likely to see losses. The vertical axis shows the number of days of average spread tightening required for each sector to earn 100 bps in excess of duration-matched Treasuries. Sectors plotting further toward the top require fewer days of spread tightening and are therefore more likely to earn 100 bps of excess return.   Ryan Swift, U.S. Bond Strategist rswift@bcaresearch.com Jeremie Peloso, Research Analyst jeremiep@bcaresearch.com Footnotes 1 Please see U.S. Bond Strategy / Global Fixed Income Strategy Weekly Report, “The Fed’s Got Your Back”, dated June 25, 2019, available at usbs.bcaresearch.com 2 For more details on how we arrive at our spread targets please see U.S. Bond Strategy Weekly Report, “The Value In Corporate Bonds”, dated February 19, 2019, available at usbs.bcaresearch.com 3 For more details on how we arrive at our spread targets please see U.S. Bond Strategy Weekly Report, “The Value In Corporate Bonds”, dated February 19, 2019, available at usbs.bcaresearch.com 4 Please see U.S. Bond Strategy Special Report, “Assessing Corporate Default Risk”, dated March 19, 2019, available at usbs.bcaresearch.com 5 Please see U.S. Bond Strategy Weekly Report, “The Long Awkward Middle Phase”, dated July 2, 2019, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Weekly Report, “A Message To The TIPS Market”, dated July 23, 2019, available at usbs.bcaresearch.com 7 Please see U.S. Bond Strategy Weekly Report, “A Message To The TIPS Market”, dated July 23, 2019, available at usbs.bcaresearch.com 8 Please see U.S. Bond Strategy Weekly Report, “Hedge Near-Term Credit Exposure”, dated May 28, 2019, available at usbs.bcaresearch.com 9 Please see U.S. Bond Strategy Weekly Report, “The Search For Aaa Spread”, dated March 12, 2019, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification Corporate Sector Relative Valuation And Recommended Allocation

Special Report

Highlights China’s infrastructure investment growth rate could rebound moderately from its current nominal 3% pace, but will remain well below the double-digit rate it has registered for most of the past decade.  A lack of funding for local governments and their financing vehicles will somewhat cap the upside in infrastructure fixed-asset investment (FAI) in the next six to nine months. Special bond issuance will be insufficient to ensuring a major recovery in infrastructure spending. Investors should tread cautiously on infrastructure plays in financial markets. Feature Chart I-1Chinese Infrastructure Investment: Double-Digit Growth Again? Nominal infrastructure investment growth in China has slowed from over 15% in 2017 to 3% currently (Chart I-1). This is the weakest growth rate since 2005 excluding the late 2011-early 2012 period. Over the past decade, each time the Chinese economy experienced a considerable slowdown, infrastructure construction was ramped up to revive growth. Infrastructure spending growth skyrocketed in 2009 and was also boosted in 2012. In 2015-2016, it was not allowed to decelerate with the issuance of nearly RMB 2 trillion of special infrastructure bonds. This time the government has also reacted. Since mid-2018, the Chinese authorities have dramatically raised local governments’ special bonds balance limits, prompted local governments to front-load their issuance this year, and also encouraged the private sector to participate in public-private partnership (PPP) infrastructure projects. Will Chinese infrastructure FAI growth accelerate over the next six to nine months from its current nominal 3% pace to double digits? The short answer is no. We believe Chinese infrastructure investment growth could rebound moderately in the next six to nine months, but will still remain below the double-digit growth seen in the past and well below the 18% average growth of the past 15 years. For purposes of this report, the composition of “infrastructure” includes three categories – (1) Transport, Storage and Postal Service, (2) Water Conservancy, Environment & Utility Management, and (3) Electricity, Gas & Water Production and Supply. Chart I-2 presents the breakdown of the nominal infrastructure FAI by category. Funding Constraint Preceding both the 2011-2012 and 2018 infrastructure investment slumps, the Chinese central government increased its scrutiny on local government debt and tightened funding conditions for infrastructure projects. As a result, all three categories of infrastructure spending experienced a sharp deceleration (Chart I-3). Overall, financing and qualitative limitations that Beijing imposes on local government infrastructure spending hold the key to the outlook. We believe Chinese infrastructure investment growth could rebound moderately in the next six to nine months, but will still remain below the double-digit growth seen in the past and well below the 18% average growth of the past 15 years. Looking forward, without a considerable recovery in available financing, there will be no meaningful rebound in Chinese infrastructure investment and construction activity. For now, we are not very optimistic on financing. Chart I-4 shows the breakdown of the major funding sources of Chinese infrastructure investment. All of them are likely to face considerable funding constraints over the next six to nine months. Chart I-3Chinese Infrastructure Investment Growth Has Decelerated Across The Board   1. Self-Raised Funds Self-raised funds contribute nearly 60% of overall infrastructure funding. They include net local government special bond issuance, PPP financing and government-managed funds’ (GMFs) revenues excluding proceeds from special bond issuance. A. Local government special bond issuance, which is exclusively used to fund infrastructure projects, has been the major source of financing for local governments in the past 12 months. The authorities significantly boosted net local government bond issuance to RMB 1.2 trillion in the first six months of this year from only RMB 361 billion in the same period in 2018. However, the amount of special bond issuance in the second half of this year will unlikely be significant enough to boost infrastructure FAI greatly. First, the central government has not only set a limit on the aggregate local government special bond balance, but it also set limits for each of the 31 provinces/provincial-level cities.1 In the past three years, nearly all provinces did not use up their special bond issuance quotas. This resulted in an outstanding aggregate amount of special bonds of only about 85% of the limit.2 In both 2017 and 2018, local governments were left with RMB 1.1 trillion special bond issuance quota unused for that year. Second, based on the limit on outstanding amount special bonds set by the central government for the end of 2019, local governments could issue another RMB 0.8-1 trillion of special bonds in the second half of this year. In comparison, in 2018, the issuance was heavily concentrated in the second half of the year with RMB 1.6 trillion. Our estimate shows there will be only RMB 400-600 billion increase in net total special bond issuance in 2019 versus 2018.3 This will translate into a merely 2-3% growth in Chinese infrastructure investment. Third, net local government special bond issuance made up only 15% of overall infrastructure FAI over the past 12 months. Hence, there is still a huge financing gap to be filled (Chart I-5). B. Public-private partnerships (PPP) are unlikely to meet the financing shortage either. PPPs have become an important financing model for Chinese local governments to fund infrastructure investments since 2014. Nevertheless, to control rising local government debt risks, the central government has tightened regulations on PPP projects since early last year. A series of tightened rules have resulted in a sharp deceleration in both PPP investment and overall infrastructure investment growth. Consequently, PPPs contributions to total infrastructure FAI have plunged from over 30% in 2017 to 10% currently (Chart I-6). Chart I-5Special Bond Issuance Accounted For Only 15% Of Infrastructure FAI Chart I-6Public-Private Partnerships: Too Small To Meet The Financing Shortage   So far, the rules on PPP projects on local governments remain tight. In March, the central government tightened its rule on local government participation in PPP projects. The new rule states that, if a local government has already spent more than 5% of its overall general expenditures on PPP projects excluding sewage and waste disposal PPP projects, it will not be allowed to invest in any new PPP projects. Before March, the threshold was over 10%. In early July, the National Development and Reform Commission (NDRC) demanded all PPP projects undertake a thorough feasibility study. The NDRC emphasized that PPP projects that do not follow standard procedures will not be allowed. Chart I-7Government-Managed Funds: Headwinds From Falling Land Sales C. Government-managed funds (GMF) excluding special bond issuance accounts, which contribute about 15% of overall infrastructure financing, are also facing constraints. According to the country’s Budget Law, the GMF budget refers to the budget for revenues and expenditures of the funds raised for specific developmental objectives. In brief, GMFs constitute de-facto off-balance-sheet government revenues and spending. Land sales by local governments are one major revenue source for GMFs. Contracting property floor space sold is likely to depress real estate developers’ land purchases, further reducing local governments’ revenues from selling land (Chart I-7). This will curb local governments’ ability to finance their infrastructure projects through GMFs. 2. Domestic Loans Domestic loans contribute to about 15% of overall infrastructure financing. Infrastructure projects are generally long term in nature. Presently, the impulse of non-household medium- and long-term (MLT) lending has stabilized but has not yet improved (Chart I-8). While not all of MLT loans are used for infrastructure, sluggish MLT lending reflects commercial banks’ reluctance to finance infrastructure projects. We believe a decelerating economy, mounting local government debt, and often-low returns on infrastructure projects will continue to constrain loan funding of infrastructure projects from both banks and the private sector. 3. General Government Budget The general government budget (which includes central and local governments) accounts for about 15% of overall infrastructure financing. The general budget is also facing headwinds from declining revenue due to recent tax cuts and lower corporate profit growth (Chart I-9). Chart I-8Sluggish Medium/Long-Term Bank Lending Chart I-9Government General Budget: Large Deficit   Bottom Line: Funding constraints will likely linger, making any recovery in Chinese infrastructure investment growth moderate over the next six to nine months. Local government special bonds will not be a game-changer. Their net issuance accounted for only 15% of overall infrastructure FAI over the past 12 months. While local governments could issue another RMB 0.8-1 trillion of special bonds in the second half of 2019, it would be well below the RMB 1.4 trillion of special bond issuance that was rolled out in the second half of 2018. FAI In Transportation: In Nominal Terms… The transportation sector accounts for about 31% of total Chinese infrastructure investment. It includes railway, highway, urban public transit, air and water transport. Table I-1 shows the 13th five-year (2016-2020) transportation investment plan released by the government in February 2017,4 which excludes urban public transit. The authorities planned to invest RMB 15 trillion in the transportation sector over the five-year period between 2016 and 2020, with highways accounting for over half of the investment, followed by railways (23%), air transportation (4.3%) and water transportation (3.3%). The table also shows our calculation of the realized investment amount in these four sub-sectors for the period of January 2016 to June 2019. Local government special bonds will not be a game-changer. Their net issuance accounted for only 15% of overall infrastructure FAI over the past 12 months. Table I-1 suggests the remaining FAI for the transportation sector for the July 2019 to December 2020 period will be considerably smaller than the FAI amount over the past 18 months. This entails a major drag on infrastructure investment at least over the next 18 months. It is important to emphasize that this is conditional on the central planners in Beijing sticking to their five-year plan for infrastructure FAI. As of now, there has been no announcement of revisions to these five-year FAI targets. Bottom Line: China has already completed the overwhelming majority of its planned transportation FAI for 2016-2020. Consequently, without revisions to the targets and budgets by central planners in Beijing, transportation investment will likely contract year-on-year over the next 18 months. …And Real Terms Table I-2 summarizes the 2020 targets for major Chinese infrastructure development (urban rail transit, railway, highway and airport) in real terms. Chart I-10Transportation 2020 Targets: Not Far Away In real terms, the annual growth of transportation infrastructure will likely be 4.2% in both 2019 and 2020. We illustrated in the previous section that the five-year budget plan had been front-loaded, leaving a very small budget for transportation investment over the next 18 months. This may suggest that without considerably exceeding the budget, transportation infrastructure will fail to achieve the 4.2% annual growth in real terms both this year and next. In brief, more funding should be dispatched/allowed by the central planners in Beijing for infrastructure FAI not to shrink. Second, urban rail transit, high-speed railways, highways and airports will reach their respective 2020 targets, while non-high-speed railway construction will likely be a little bit off its 2020 target. Third, based on the 2020 targets, urban rail transit will enjoy very fast growth over the next one and a half years. Fourth, the growth of high-speed railways and highways will be very low, at around 1-2% in real terms (Chart I-10). Finally, while the number of airports will increase at a faster pace, their contribution to overall infrastructure investment will remain insignificant as they only account for about 1.4% of overall infrastructure investment. Bottom Line: In real terms, transport infrastructure growth will likely be only about 4% over the next six to nine months. Future Infrastructure Investment Focus Urban rail transit, environmental management and public utility management will likely be the major driving forces for Chinese infrastructure investment over the next 18 months. Urban rail transit line length will likely register fast growth of around 10% over the next six to nine months. As the central government enforces increasingly stringent rules on environmental protection, investment in environmental management will likely experience continued growth acceleration (Chart I-11). China has already completed the overwhelming majority of its planned transportation FAI for 2016-2020. Consequently, without revisions to the targets and budgets by central planners in Beijing, transportation investment will likely contract year-on-year over the next 18 months. Meanwhile, as the country’s urbanization continues and more townships and city suburbs become urbanized,5 public utility management investment will also grow moderately. Public utility management investment, contributing a massive 45% of overall infrastructure investment, includes sewer systems, sewer treatment facilities, waste treatment and disposal, streetlights, city roads construction, parks, bridges and tunnels in the city. Investment Implications Investors should not hold their breath expecting a major upswing in infrastructure FAI and a major rally in related financial markets. Chinese steel demand is sensitive to construction of railways and urban rail transit lines (Chart I-12, top panel). In turn, mainland cement demand is dependent on highway construction (Chart I-12, bottom panel). Chart I-11Environment Management: Will Continue Booming Chart I-12Chinese Infrastructure Spending Will Moderately Boost Steel & Cement Demand...   Chart I-13...And Steel & Cement Prices At The Margin The infrastructure sector accounts for about 10-15% of total Chinese steel use, and about 30-40% of Chinese cement consumption. Nevertheless, given that we believe Chinese infrastructure spending will only have a moderate recovery, the positive effect on steel and cement prices will be muted as well (Chart I-13). The same holds true for spending on industrial machinery, equipment, chemicals and various materials. Notably, risks to this baseline scenario of a muted recovery are to the downside because of the lack of funding. Barring a substantial increase in the special bond issuance quota this year or a major credit binge, infrastructure FAI growth could in fact stall. Ellen JingYuan He, Associate Vice President ellenj@bcaresearch.com   Footnotes 1  Please note that the central government only set the special bond balance limit (not the quota) for local governments. The often-cited “quota” in the news is derived by calculating the difference between the current limit and the previous year’s limit. The “quota” used in this report is the difference between the current special bond balance limit and the actual special bond balance of the previous year end. 2  At the end of 2018, Chinese special bond balance was RMB 7.4 trillion, only 85.8% of the special bond balance limit of RMB 8.6 trillion. This ratio was 84.6% in 2017 and 85.5% in 2016. On average, the ratio was 85.3% in the past three years. 3  Given that the central government is aiming to somewhat stimulate infrastructure spending by increasing special bond issuance, we assume special bond balance at the end of 2019 to reach 88%-90% of the limit (RMB 10.8 trillion) that it has set for 2019. This will be higher than the 85% average of the past three years. In turn, this means that the special bond balance at the end of this year will likely be RMB 9.5-9.7 trillion. Since the balance at the end of last year was RMB 7.4 trillion, this results that net special bond issuance will be around RMB 2.1-2.3 trillion in 2019. Given the net special bond issuance last year was RMB 1.7 trillion, it follows that there will only be a RMB 400-600 billion increase in total special bond issuance in 2019 versus 2018. 4  Please see www.gov.cn/xinwen/2017-02/28/content_5171576.htm, published February 28, 2017, by the Chinese central government website. 5  Please see Emerging Markets Strategy/China Investment Strategy Special Report “Industrialization-Driven Urbanization In China Is Losing Steam,” dated January 2, 2019, available on ems.bcaresearch.com

Special Report

Highlights Mutual Funds & ETFs: The liquidity mismatch between easily tradeable mutual fund shares and the less liquid underlying corporate bonds makes it possible for negative feedback loops to emerge between fund flows and corporate bond spreads. The growing presence of open-ended mutual funds and ETFs in the corporate bond market should be seen as a risk that could exacerbate future periods of spread widening, leading to worse economic outcomes. BBB Securities: The large amount of outstanding BBB debt could lead to fire sales from corporate bond holders with investment grade-only mandates when the debt is downgraded to junk. However, in contrast to the negative feedback loop that can be generated by mutual fund flows, the evidence shows that the negative price pressure from fallen angel fire sales is fleeting. Leveraged Loans: The rapid surge in leveraged loans has been partially offset by reduced high-yield issuance, helping mitigate a potentially destabilizing rise in all riskier corporate debt. At the same time, bank exposure is focused on the safest CLO tranches, limiting the potential systemic risks from bank losses. Feature In April, we published a Special Report that investigated whether corporate debt poses a risk to the U.S. economy.1 That report focused on what economic theory and empirical evidence say about the relationship between corporate debt and future economic growth. We arrived at the following conclusions: The empirical evidence decisively shows that private (household + business) debt helps predict future economic outcomes. Some evidence shows that household debt is more important than corporate debt in this regard. In contrast to mainstream economic theory, the level of private debt-to-GDP does not help predict future economic outcomes. Rather, it is rapid private debt growth that is linked to more severe economic downturns. Ebullient credit market sentiment is also shown to predict weak economic growth. Tight credit spreads should be viewed as a warning sign, similar to rapid private debt growth. In this follow-up Special Report, we consider three credit market developments that are unique to this cycle: The large ownership stake of open-ended mutual funds and ETFs in the U.S. corporate bond market. The elevated amount of BBB-rated debt outstanding relative to other credit tiers. The rapid issuance of leveraged loans. All three of these developments could mediate the relationship between corporate debt and economic growth, potentially increasing risks to the economy. We consider each factor in turn. 1. Fund Flows One unique feature of the current cycle is that open-ended mutual funds and ETFs own a much larger share of outstanding corporate bonds than in the past. Back in 1990, insurance companies and pension funds were the largest holders of corporate debt, controlling 54% of the market. Meanwhile, open-ended funds owned a paltry 3%. Since then, fund ownership has surged to 16%, mostly at the expense of financial institutions, insurance companies and pensions. Foreign holdings of U.S. corporate bonds have also increased during this period, from 13% of the market to 28% (Charts 1 & 2). Chart 2Mutual Funds Now An Important Market Player Why Does Fund Ownership Matter? We focus on fund ownership of corporate bonds because it has been theorized that flows into and out of open-ended mutual funds can have a similar impact on market prices as leverage, amplifying price moves in either direction. As described in a 2014 paper by Feroli, Kashyap, Shoenholtz, and Shin:2 [W]hen asset flows for certain fixed income securities are high, prices persistently rise and a feedback loop emerges. High flows lead to rising prices, which attract more flows, which further raises prices. Obviously, the proposed feedback loop also works in reverse: Outflows cause prices to decline, and lower prices lead to further outflows. This sort of feedback loop is unique to mutual funds. Insurance companies and pension funds, for example, do not experience investor capital flight in response to a near-term price drop. This makes the larger presence of mutual funds in the corporate bond market potentially destabilizing. Fund ownership has surged to 16%, from a paltry 3% back in 1990. Why Do Fund Flows Behave This Way? Mutual fund shares are much more liquid than the corporate debt securities they hold. As described in a 2017 paper by Goldstein, Jiang and Ng:3 When [mutual] fund investors redeem their shares, they get the net asset value as of the day of redemption. The fund then has to conduct costly liquidation that hurts the value of the shares for investors who keep their money in the fund. Hence, the expected redemption by some investors increases the incentives for others to redeem. In other words, during times of stress, mutual fund investors have an incentive to withdraw their money before other fund shareholders get the chance. Otherwise, they could be stuck holding a basket of illiquid corporate bonds. This bank-run like behavior is well documented for corporate and municipal bond funds, though it appears not to exist for funds that traffic in more liquid instruments, such as Treasuries and equities. In fact, when Goldstein et al looked at how flows into and out of individual corporate bond and equity funds respond to past fund performance, they found that the Flow-Performance curve for an individual corporate bond fund exhibits a pronounced concave shape. Meanwhile, the same curve for an individual equity fund is convex (Chart 3). This means that corporate bond mutual fund shareholders are quick to redeem their shares in response to poor fund performance, while equity fund shareholders are more inclined to stand pat. On the flipside, positive fund performance leads to large equity fund inflows, but doesn’t attract capital to corporate bond funds to the same extent. The above results apply to individual funds, but Goldstein et al also performed the same analysis for corporate bond funds in the aggregate. That is, rather than measuring whether investors sold a particular corporate bond mutual fund in response to its poor performance, they measured whether investors exited the corporate bond mutual fund space altogether in response to poor corporate bond performance. Interestingly, they found a very similar result (Chart 4). Investors are inclined to exit the corporate bond space entirely following periods of poor performance. Meanwhile, they found no relationship between aggregate equity fund flows and performance. Investors might switch between different equity funds in response to recent performance trends, but they don’t exit the asset class altogether.   These results provide a clear indication for why the large presence of corporate bond mutual funds might be destabilizing. Corporate bond fund investors are quick to flee the space during periods of poor performance. For more liquid securities, such as equities and Treasuries, a large mutual fund presence in the market is not a concern, since flows do not respond as aggressively to price shocks. Empirical Evidence For The Flow-Performance Feedback Loop The evidence presented above shows that fund flows respond to performance, but for the theorized feedback loop between fund flows and corporate bond prices to exist, we also need evidence that fund flows impact corporate bond performance. In that regard, a 2019 Banque de France working paper examines the impact of aggregate flows into French corporate bond funds on the yields of the underlying securities.4 It finds that not only do flows impact yields contemporaneously, but also that outflows have a larger influence on yields than inflows. Using a different methodology, a 2015 paper by Hoseinzade finds no material impact of fund flows on underlying corporate bond yields, but for an interesting reason.5 The paper confirms that corporate bond fund shareholders demonstrate bank-run like behavior in response to poor performance, but also argues that “bond fund managers hold a significant level of liquid assets, allowing them to manage redemptions without excessively liquidating corporate bonds.” Chart 5Funds Deploy Cash Before Selling Bonds It’s true that corporate bond mutual funds often hold significant allocations to cash and U.S. Treasuries, and Hoseinzade shows that fund managers tend to discharge their most liquid holdings first, before attempting to sell corporate bonds. This result lines up with our casual observation. Chart 5 shows the aggregate liquid asset (cash and Treasury) holdings of corporate bond mutual funds. It is apparent that they tend to fall during periods of spread widening. We also note that corporate bond mutual funds, in aggregate, currently hold about 6% of their assets in liquid securities. This buffer can probably withstand a sizeable shock, but liquid assets fell from similar levels into negative territory during each of the past two recessions. One other factor that could help break the feedback loop between fund flows and prices is the institutional ownership of corporate bond mutual funds. Goldstein et al find that mutual funds mostly owned by institutional investors exhibit less of a feedback loop between flows and performance. That is, large institutional investors are less likely to redeem their shares in response to a bout of poor performance. While we don’t have data on corporate bond mutual fund ownership specifically, Federal Reserve data reveal that insurance companies and pension funds own a significantly larger proportion of outstanding mutual fund shares than in the 1990s, though less than they did in the mid-2000s (Chart 6). Note that Chart 6 shows data for all mutual funds, including equity funds, Treasury funds, etc… Chart 6Institutional Ownership Of Mutual Funds We conclude that there is enough evidence of a feedback loop between fund flows and corporate bond prices that we should be wary of the growing presence of open-ended mutual funds and ETFs in the corporate bond space. Cash holdings and institutional ownership can help mitigate negative flow/performance feedback loops to some extent, but probably shouldn’t be counted on in the event of a severe shock. What’s The Economic Impact? In our corporate debt Special Report from April, we postulated that changes in corporate bond spreads might, themselves, cause an economic downturn, rather than simply reflect one. The mechanism is summarized nicely by Lopez-Salido, Stein and Zakrajsek (2016):6 [a] sentiment-driven widening of credit spreads amounts to a reduction in the supply of credit, especially to lower credit-quality firms. It is this reduction in credit supply that exerts a negative influence on economic activity. With that in mind, in the current environment it seems very possible that an initially sentiment-driven credit spread widening could be exacerbated by outflows from corporate bond mutual funds. A larger shock to credit spreads leads to a larger reduction in credit supply and a more severe economic impact. Aggregate liquid asset holdings of corporate bond mutual funds tend to fall during periods of spread widening. Bottom Line: The liquidity mismatch between easily tradeable mutual fund shares and the less liquid underlying corporate bonds makes it possible for negative feedback loops to emerge between fund flows and corporate bond spreads. The growing presence of open-ended mutual funds and ETFs in the corporate bond market should be seen as a risk that could exacerbate future periods of spread widening, leading to worse economic outcomes. 2: BBB Debt Outstanding Chart 7The Large Amount Of BBB Debt It has been widely reported that an unusually large amount of outstanding corporate bonds are rated BBB, the lowest credit rating that is still considered investment grade. In fact, the par value of BBB-rated securities now makes up 50% of the Bloomberg Barclays Investment Grade index, up from 21% in 1990 (Chart 7). The amount of outstanding BBB securities is more than double the par value of the Bloomberg Barclays High-Yield index, and BBBs represent 41% of the total combined par value of the investment grade and high-yield indexes. The reason to be wary about the large amount of outstanding BBB debt is that when the credit cycle turns and ratings downgrades start to occur, a larger than normal amount of debt will be downgraded from investment grade into high-yield. When that happens, any investors with an investment grade-only mandate will be forced to sell. The concern is that such forced selling could set off a negative feedback loop very similar to the one discussed in the first section. An added layer of risk comes from the fact that in addition to investment grade-only mutual funds, insurance companies and pension funds – who still control 35% of the corporate bond market (see Chart 2 on page 3) – are often burdened with larger capital costs for high-yield debt. This means that a very large pool of investors could be impacted by a spate of BBB downgrades. In terms of the potential market impact, a 2010 paper by Ellul, Jotikasthira and Lundblad investigated fire sales of downgraded corporate bonds induced by regulatory constraints.7 The authors found that insurance companies often engage in the forced selling of bonds that have been recently downgraded into high-yield. Further, the downgraded bonds experience negative near-term price pressure from the fire sale, but that pressure tends to reverse after a few months. The finding that the negative price pressure is fleeting is important. In contrast to the negative feedback loop that can be generated by mutual fund flows, BBB securities can only be downgraded to high-yield once. In other words, once the initial fire sale of fallen angel debt takes place, there is no mechanism to force the downward price pressure to continue.8 Bottom Line: The large amount of outstanding BBB debt could lead to fire sales from corporate bond holders with investment grade-only mandates when the debt is downgraded to junk. However, in contrast to the negative feedback loop that can be generated by mutual fund flows, the evidence shows that the negative price pressure from fallen angel fire sales is fleeting. 3. Leveraged Loans The rapid growth of leveraged loans – lending made to below investment-grade borrowers - over the past couple of years has caught the attention of global central banks and financial regulators. That concern is understandable, as it would be a dereliction of duty for any policymaker or regulator who lived through the 2008 financial crisis to not consider the potential risks to financial stability and future economic growth from a surge in lower quality lending. This is especially true given the increase in the number of securitized instruments linked to leveraged loans – collateralized loan obligations, or CLOs – which evokes memories of the toxic subprime mortgage products that helped trigger the 2008 crisis. Although as the Fed’s Vice Chair for Supervision, Randal Quarles, recently noted, the financial media has been overplaying the leveraged loan story in such a way that it felt like “the Earth must be getting hit by an asteroid.” The BoE estimates that the CLOs would have to suffer a loss more than twice as severe as seen during the 2008 financial crisis for the AAA-rated piece of CLOs issued in 2018 to incur losses. The leveraged loan and CLO markets can be opaque. However, based on the information we do have from credible sources like central banks, the IMF and the BIS, some conclusions can be made about the potential economic risks from the rapid build-up of U.S. leveraged loans: Leveraged loan expansion has been partially offset by high-yield contraction. Chart 8More Leveraged Loans, Less Junk Bonds Based on estimates from the BIS and IMF, there are around $1.4 trillion in U.S. leveraged loans outstanding, which is greater than the $1.2 trillion U.S. high-yield bond market (Chart 8). That is an all-time high in the dollar amount of leveraged loans, as well as for the share of all lower-rated corporate debt accounted for by loans. The annual growth rate of U.S. leveraged loans is now a whopping 29% - the fastest pace seen since 2007. Yet the growth of the total amount of leveraged loans plus high-yield bonds is a much lower 12%. While that is still a large number, it is below the peak growth rates seen during the past fifteen years. This is because the amount of high-yield bonds outstanding has been modestly contracting since 2015. Much of that run-up in leveraged loan growth has been to satiate the demand for loans created by private equity funds and, more importantly, CLOs. The strong risk appetite from investors resulted in a notable deterioration in lending standards, with loans coming out at higher leverage multiples (debt/EBITDA) and with reduced investor covenant protection. Yet since lower-rated companies were not ramping up high-yield bond issuance at the same time, the economic stability risks from a rapid run-up in total riskier borrowing are lower, on the margin. The ownership structure of leveraged loans (and CLOs) is diverse enough to not create systemic problems. To date, the Bank of England (BoE) has compiled the most detailed estimates of the ownership breakdown of both leveraged loans and CLOs.9 In Chart 9, we have recreated a chart from the BoE’s July 2019 Financial Stability Report, which colorfully shows the ownership of global leveraged loans and CLOs. The way to read the chart is that each square represents a 1% share of the estimated $3.2 trillion of global leveraged loans and CLOs. The split in the chart is 75% loans and 25% CLOs (CLO ownership is shown on the right side of the thick dotted line). The biggest category of leveraged loan investor is what the BoE titled “U.S. and other global banks”, a group that represents 38% of total loans and CLOs. European banks own 12%, U.K. banks own 3% and Japanese banks own 3% (entirely through CLOs), thus bringing the global bank exposure to 56% of all leveraged loan instruments. While that sounds like a large number, the majority of that is in the form of revolving credit facilities – effectively, overdraft facilities to lower-rated borrowers. Revolving credit facilities are typically less risky than leveraged loans, because credit facilities have greater covenant protection and even more seniority in terms of creditor claims on borrower assets. The BoE estimates that 40% of all global leveraged loans and CLOs are owned by non-bank investors. This includes pension funds, insurance companies and investment funds (mutual funds and ETFs). Chart 9 shows how much more diverse the investor base is for CLOs than for other leveraged loans. It suggests that any future potential losses from CLOs will be distributed more evenly within the financial system, rather than being concentrated in the banks. Chart 10Leveraged Loan Losses Are Typically Lowered Compared To Junk Even within the bank holdings of CLOs, the systemic risks are lessened. The BoE noted that the increased amount of subordination (i.e. lower-rated tranches) of more recent CLO deals helps protect the senior tranches from losses. According to the BoE, the AAA-rated piece of a representative sample of CLOs issued in 2018 was 63%; this compares to 70% for CLOs issued in 2006.10 Furthermore, the central bank estimates that the CLOs would have to suffer a loss more than twice as severe as seen during the 2008 financial crisis for the AAA-rated piece of CLOs issued in 2018 to incur losses. That would be an extraordinary outcome, given how 2008 generated losses on leveraged loans that were over twice as bad as the previous worst year in 2002 (Chart 10). Potential losses from AAA tranches are important from a financial stability perspective. The BoE estimates that 40% of all CLOs are owned by global banks (including a large 13% share from yield-chasing Japanese banks). These banks tend to focus on safer AAA-rated CLO tranches. The demand for leveraged loan products is volatile, but that might actually be a good thing for economic stability. The surge in leveraged loans over the past two years has not only been related to demand from private equity funds and CLOs. U.S. retail investors have also been big buyers of mutual funds and ETFs linked to the leveraged loan market, as a way to seek out higher credit returns against a backdrop of Fed rate hikes. Chart 11Fed Rate Expectations Drive The Demand For Loans Vs Bonds Leveraged loans are floating rate instruments. Thus, they are more desirable than fixed-rate high-yield corporate debt when short-term interest rates are rising. This is seen in Chart 11, where we show net flows into the largest U.S. junk bond and leveraged loan ETFs. These flows are plotted with the JP Morgan survey of bond investor duration positioning (top panel) and our Fed Funds Discounter that measures the market-implied expected change in the fed funds rate over the next year (bottom panel). The conclusion is obvious – there was very strong retail demand for floating-rate leveraged loans over fixed-rate junk bonds during 2016-18 when expected rate hikes justified defensive duration positioning. In 2019, the tables have turned. The Fed is more dovish, rate cuts are now expected, investors have been adding duration exposure, and demand for leveraged loan funds has plunged while high-yield bond funds have been seeing inflows. The exodus from all leveraged loan funds has been historically large, with Lipper reporting that there were 33 straight weeks of outflows to July 3, 2019, for a total of $32 billion.11 Already, that reduced demand for leveraged loans has translated into sharply reduced issuance of new U.S. CLOs, which was 73% lower in the first half of 2019 versus the same period in 2018 (Chart 12). At the same time, high-yield bond issuance was up 20% in the first six months of 2019 versus 2018. The reduced demand for leveraged loans has also shifted the balance of power back to lenders, as the share of U.S. leveraged loans that have been issued with limited covenant protection (“cov-lite”) has plunged from 72% in 2018 to around 40% (Chart 13). Chart 12Lower-Rated Issuance Is "Self-Regulating" Chart 13Reduced Covenant-Lite Issuance So Far In 2019     This is a critical point on the potential stability risks from leveraged loans – the market for those loans is “self-regulating”, based on final demand from investors who “toggle” between floating rate and fixed rate credit instruments. This helps limit the growth in overall corporate indebtedness, helping to put off the date when credit booms turn into future credit busts. Ryan Swift, U.S. Bond Strategist rswift@bcaresearch.com Robert Robis, CFA, Chief Fixed Income Strategist rrobis@bcaresearch.com Footnotes 1      Please see U.S. Bond Strategy / Global Fixed Income Strategy Special Report, “The Risk From U.S. Corporate Debt: Theory And Evidence”, dated April 23, 2019, available at usbs.bcaresearch.com 2     https://www.nowpublishers.com/article/DownloadEBook/9781680834864?format=pdf 3     http://finance.wharton.upenn.edu/~itayg/Files/bondfunds-published.pdf 4     https://ideas.repec.org/p/bfr/banfra/706.html 5     https://pdfs.semanticscholar.org/5a60feab84a7d10de084abfce414b5888d5586e2.pdf 6     https://www.nber.org/papers/w21879 7     https://pdfs.semanticscholar.org/55a4/8602b17bc7e7f8428695ab6a3ef2c87756ab.pdf 8      Corporate bonds that are downgraded from investment grade to high-yield are called fallen angels. 9      The Financial Stability Board, the international body that monitors and makes recommendations on the global financial system, is due to publish a comprehensive analysis of the ownership structure of the leveraged loan market in the autumn of 2019. 10     For a more detailed description of this analysis, see pages 28 & 29 of the Bank of England’s July 2019 Financial Stability report, which can be found here: https://www.bankofengland.co.uk/financial-stability-report/2019/july-20… 11     https://www.spglobal.com/marketintelligence/en/news-insights/latest-news-headlines/leveraged-loan-news/leveraged-loan-fund-withdrawal-streak-hits-record-33-weeks-totaling-32b  

Highlights Chart 1Looks Like 2016 & 1998 The Treasury market continues to price-in a recession-like outcome for the U.S. economy, embedding 83 basis points of Fed rate cuts over the next 12 months. But last week’s economic data challenge that narrative. First, the ISM Non-Manufacturing PMI held above 55 in June, even as its Manufacturing counterpart plunged toward the 50 boom/bust line (Chart 1). This divergence between a strong service sector and weak manufacturing sector is more reminiscent of prior mid-cycle slowdowns in 2016 and 1998 than of any pre-recession period. Second, nonfarm payrolls added 224k jobs in June, a strong rebound from the 72k added in May and enough to keep the 12-month growth rate at a healthy 1.5% (bottom panel). Still-low inflation expectations provide sufficient cover for the Fed to cut rates later this month, likely by 25 bps. But beyond that, continued strong economic data could prevent any further easing. Keep portfolio duration low and stay short the February 2020 fed funds futures contract. Feature Investment Grade: Overweight Chart 2Investment Grade Market Overview Investment grade corporate bonds outperformed the duration-equivalent Treasury index by 144 basis points in June, bringing year-to-date excess returns up to +368 bps. We removed our recommendation to hedge near-term corporate credit exposure after the Fed’s clear dovish pivot at the June FOMC meeting.1  At that time, we also noted that the surging gold price, weakening trade-weighted dollar and outperformance of global industrial mining stocks were all signaling that corporate spreads have peaked (Chart 2). Of our “peak credit spread” indicators, only the CRB Raw Industrials index has yet to turn the corner. The macro environment supports tighter spreads. But in the investment grade space, value only looks attractive for Baa-rated securities. Baa spreads remain 7 bps above our target (panel 3), while Aa and A-rated spreads are 1 bp and 4 bps below, respectively (panel 4). Aaa bonds are even more expensive, with spreads 19 bps below target (not shown).2  Investors should focus their investment grade corporate bond exposure on Baa-rated securities. Our measure of gross leverage – total debt over pre-tax profits – jumped in Q1, as corporate debt grew at an annualized pace of 8.5% while corporate profits contracted by an annualized 18% (bottom panel). Leverage will likely rise again in Q2, as profit growth will almost certainly remain weak, but should then level-off as global growth recovers. High-Yield: Overweight Chart 3High-Yield Market Overview High-Yield outperformed the duration-equivalent Treasury index by 154 basis points in June, bringing year-to-date excess returns up to +603 bps. The average index option-adjusted spread tightened 56 bps on the month. At 366 bps, it remains well above the cycle-low of 303 bps. As with investment grade credit, we removed our recommendation to hedge near-term exposure following the June FOMC meeting (see page 3). Further, we see the potential for much more spread tightening in high-yield than in investment grade. Within investment grade, only the Baa credit tier carries a spread above our target. In High-Yield, Ba-rated spreads are 42 bps above our target (Chart 3), B-rated spreads are 108 bps above our target (panel 3) and Caa-rated spreads are 263 bps above our target (not shown).3  Junk spreads also offer reasonable value relative to expected default losses. The current Moody’s baseline forecast calls for a default rate of 2.7% over the next 12 months, not far from our own projection.4 This would translate into 224 bps of excess spread in the High-Yield index, after adjusting for default losses (panel 4). This is comfortably above zero, and only just below the historical average of 250 bps. We will continue to monitor job cut announcements, which have moderated so far this year (bottom panel), and C&I lending standards, which remain in net easing territory, to assess whether our default expectations need to be revised. MBS: Neutral Chart 4MBS Market Overview Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 2 basis points in June, bringing year-to-date excess returns up to -11 bps. The conventional 30-year zero-volatility spread widened 1 bp on the month, as a 4 bps widening in the option-adjusted spread (OAS) was partially offset by a 3 bps decline in the compensation for prepayment risk (option cost). Falling mortgage rates hurt MBS in the first half of this year, as lower rates led to an increase in refi activity that drove MBS spreads wider (Chart 4). In fact, the conventional 30-year index OAS has risen all the way back to its average pre-crisis level (panel 3). However, as we noted in last week’s report, the nominal 30-year MBS spread remains very tight, at close to one standard deviation below its historical mean.5 The mixed valuation picture means we are not yet inclined to augment our recommended allocation to MBS, especially given the favorable environment for corporate bonds, where expected returns are higher. We are equally disinclined to downgrade MBS, given that refi activity could be close to peaking. All in all, we expect that the next move in the MBS/Treasury basis will be a tightening, as global growth improves and mortgage rates rise in the second half of the year. However, valuation is not sufficiently attractive to warrant more than a neutral allocation. Government-Related: Underweight Chart 5Government-Related Market Overview The Government-Related index outperformed the duration-equivalent Treasury index by 26 basis points in June, bringing year-to-date excess returns up to +133 bps. Sovereign debt outperformed duration-equivalent Treasuries by 208 bps on the month, bringing year-to-date excess returns up to +419 bps. Local Authorities underperformed the Treasury benchmark by 6 bps, dragging year-to-date excess returns down to +213 bps. Meanwhile, Foreign Agencies underperformed by 26 bps, dragging year-to-date excess returns down to +103 bps. Domestic Agencies underperformed by 4 bps in June, dragging year-to-date excess returns down to +25 bps. Supranationals outperformed by 1 bp on the month, bringing year-to-date excess returns up to +28 bps. Sovereign debt remains very expensive relative to equivalently rated U.S. corporate credit (Chart 5). While the sector would benefit if the Fed’s dovish pivot results in a weaker dollar, U.S. corporate bonds would still outperform in that scenario, given the more attractive starting point for spreads. We continue to recommend an underweight allocation to Sovereigns. Unlike the debt of most other countries, Mexican sovereign bonds continue to trade cheap relative to U.S. corporates (bottom panel). While this remains an attractive option from a valuation perspective, the President’s on again/off again tariff threats make it a risky near-term proposition. Municipal Bonds: Overweight Chart 6Municipal Market Overview Municipal bonds underperformed the duration-equivalent Treasury index by 73 basis points in June, dragging year-to-date excess returns down to -44 bps (before adjusting for the tax advantage). The average Aaa-rated Municipal / Treasury yield ratio rose 2% in June, and currently sits at 81% (Chart 6). The ratio is close to one standard deviation below its post-crisis mean, but exactly equal to the average that prevailed in the late stages of the previous cycle, between mid-2006 and mid-2007. Recent muni underperformance has been broad-based across the entire maturity spectrum, but long-end (20-year and 30-year) yield ratios continue to look attractive relative to the rest of the curve. 20-year and 30-year Aaa-rated yield ratios are more than one standard deviation above their respective pre-crisis averages. Meanwhile, 10-year, 5-year and 2-year Aaa yield ratios are very close to average pre-crisis levels. State & local government balance sheets are in decent shape and a material increase in ratings downgrades is unlikely (bottom panel). We therefore recommend an overweight allocation to municipal bonds, but with a preference for 20-year and 30-year Aaa-rated securities. We showed in a recent report that value declines sharply if you move into shorter maturities or lower credit tiers.6 Treasury Curve: Maintain A Barbell Curve Positioning Chart 7Treasury Yield Curve Overview The Treasury curve bull-steepened in June, alongside a large drop in our 12-month Fed Funds Discounter from -75 bps to -90 bps (Chart 7). June’s bull-steepening was reversed last week, as the strong employment report caused our discounter to jump back up to -83 bps, resulting in a bear-flattening of the Treasury curve. All in all, the 2/10 Treasury slope steepened 6 bps in June, then flattened 8 bps in the first week of July. It currently sits comfortably above zero at 17 bps. The 5/30 slope steepened 11 bps in June, then flattened 6 bps last week. It currently sits at 70 bps. In last week’s report we reviewed the case for barbelling your U.S. bond portfolio.7 That is, favoring the short and long ends of the yield curve while avoiding the 5-year and 7-year maturities. This positioning continues to make sense. Not only does the barbell increase the average yield of your portfolio, but our butterfly spread models all show that barbells are cheap relative to bullets (see Appendix B). The 5-year and 7-year yields will also rise more than long-end and short-end yields when the market eventually moves to price-in fewer Fed rate cuts. In addition to our recommended barbell positioning, we advocate keeping a short position in the February 2020 fed funds futures contract. That contract is currently priced for a fed funds rate of 1.69% next February, the equivalent of three 25 basis point rate cuts spread over the next five FOMC meetings. The Fed is unlikely to deliver that much easing. TIPS: Overweight Chart 8Inflation Compensation TIPS underperformed the duration-equivalent nominal Treasury index by 11 basis points in June, dragging year-to-date excess returns down to +28 bps. The 10-year TIPS breakeven inflation rate fell 5 bps on the month and currently sits at 1.69% (Chart 8). The 5-year/5-year forward TIPS breakeven inflation rate fell 4 bps on the month and currently sits at 1.83%. As we have noted in recent research, FOMC members are monitoring long-dated inflation expectations and are committed to keeping policy easy enough to “re-anchor” them at levels consistent with the Fed’s 2% target.8 In the long-run, this will support a return of long-dated TIPS breakeven inflation rates (both 10-year and 5-year/5-year forward) to our 2.3% - 2.5% target range. However, for breakevens to move higher, investors will also need to see evidence that realized inflation can be sustained near 2%. On that note, the core PCE deflator grew at a healthy 2.3% (annualized) clip in May, following an even higher 3% (annualized) rate in April. However, it has only grown 1.6% during the past year. 12-month trimmed mean PCE is running almost exactly in line with the Fed’s target at 1.99%. In a recent report we noted that 12-month core PCE inflation has a track record of converging toward the trimmed mean.9   ABS: Underweight Chart 9ABS Market Overview Asset-Backed Securities underperformed the duration-equivalent Treasury index by 13 basis points in June, dragging year-to-date excess returns down to +51 bps. The index option-adjusted spread for Aaa-rated ABS widened 9 bps on the month, moving back above its minimum pre-crisis level (Chart 9). At 36 bps, the spread remains well below its pre-crisis mean of 64 bps. In addition to poor valuation, the sector’s credit fundamentals are shifting in a negative direction. Household interest payments continue to trend up, suggesting a higher delinquency rate going forward (panel 3). Meanwhile, the Fed’s Senior Loan Officer Survey revealed that average consumer lending standards tightened in Q1 for the second consecutive quarter. Tighter lending standards usually coincide with rising consumer delinquencies (bottom panel). Loan officers also reported slowing demand for credit cards for the fifth consecutive quarter, and slowing auto loan demand for the third consecutive quarter. Second quarter data will be made available in early August, but current trends are not promising. The combination of poor value and deteriorating credit quality leads us to recommend an underweight allocation to consumer ABS. Non-Agency CMBS: Neutral Chart 10CMBS Market Overview Non-Agency Commercial Mortgage-Backed Securities underperformed the duration-equivalent Treasury index by 4 basis points in June, dragging year-to-date excess returns down to +191 bps. The index option-adjusted spread for non-agency Aaa-rated CMBS widened 2 bps on the month. It currently sits at 68 bps, below its average pre-crisis level but above levels seen in 2018 (Chart 10). The macro outlook for commercial real estate looks somewhat unfavorable, with lenders tightening standards (panel 4) amidst falling demand (bottom panel). However, on a positive note, commercial real estate prices recently accelerated and are now much more consistent with current CMBS spreads (panel 3). Despite the mixed fundamental picture, CMBS still offer excellent compensation relative to other similarly-rated fixed income sectors.10  Agency CMBS: Overweight Agency CMBS outperformed the duration-equivalent Treasury index by 3 basis points in June, bringing year-to-date excess returns up to +93 bps. The index option-adjusted spread widened 1 bp on the month and currently sits at 50 bps. The Excess Return Bond Map in Appendix C shows that Agency CMBS offer high potential return compared to other low-risk spread products. An overweight allocation to this defensive sector remains appropriate. Appendix A - The Golden Rule Of Bond Investing We follow a two-step process to formulate recommendations for bond portfolio duration. First, we determine the change in the federal funds rate that is priced into the yield curve for the next 12 months. Second, we decide – based on our assessments of the economy and Fed policy – whether the change in the fed funds rate will exceed or fall short of what is priced into the curve. Most of the time, a correct answer to this question leads to the appropriate duration call. We call this framework the Golden Rule Of Bond Investing, and we demonstrated its effectiveness in the U.S. Bond Strategy Special Report, “The Golden Rule Of Bond Investing”, dated July 24, 2018, available at usbs.bcaresearch.com. Chart 11 illustrates the Golden Rule’s track record by showing that the Bloomberg Barclays Treasury Master Index tends to outperform cash when rate hikes fall short of 12-month expectations, and vice-versa. Chart 11The Golden Rule's Track Record At present, the market is priced for 83 basis points of cuts during the next 12 months. We do not anticipate any rate cuts during this timeframe, and therefore recommend that investors maintain below-benchmark portfolio duration. We can also use our Golden Rule framework to make 12-month total return and excess return forecasts for the Bloomberg Barclays Treasury index under different scenarios for the fed funds rate. Excess returns are relative to the Bloomberg Barclays Cash index. To forecast total returns we first calculate the 12-month fed funds rate surprise in each scenario by comparing the assumed change in the fed funds rate to the current value of our 12-month discounter. This rate hike surprise is then mapped to an expected change in the Treasury index yield using a regression based on the historical relationship between those two variables. Finally, we apply the expected change in index yield to the current characteristics (yield, duration and convexity) of the Treasury index to estimate total returns on a 12-month horizon. The below tables present those results, along with 95% confidence intervals. Excess returns are calculated by subtracting assumed cash returns in each scenario from our total return projections. Appendix B - Butterfly Strategy Valuation 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: U.S. Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com U.S. 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 5, 2019) 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 5, 2019) 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 +56 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 56 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 U.S. fixed income market. The Map employs volatility-adjusted breakeven spread analysis to show how likely it is that a given sector will earn/lose money during the subsequent 12 months. The Map does not incorporate any macroeconomic view. The horizontal axis of the Map shows the number of days of average spread widening required for each sector to lose 100 bps versus a position in duration-matched Treasuries. Sectors plotting further to the left require more days of average spread widening and are therefore less likely to see losses. The vertical axis shows the number of days of average spread tightening required for each sector to earn 100 bps in excess of duration-matched Treasuries. Sectors plotting further toward the top require fewer days of spread tightening and are therefore more likely to earn 100 bps of excess return. Ryan Swift, U.S. Bond Strategist rswift@bcaresearch.com Footnotes 1 Please see U.S. Bond Strategy / Global Fixed Income Strategy Weekly Report, “The Fed’s Got Your Back”, dated June 25, 2019, available at usbs.bcaresearch.com 2 For more details on how we arrive at our spread targets please see U.S. Bond Strategy Weekly Report, “The Value In Corporate Bonds”, dated February 19, 2019, available at usbs.bcaresearch.com 3 For more details on how we arrive at our spread targets please see U.S. Bond Strategy Weekly Report, “The Value In Corporate Bonds”, dated February 19, 2019, available at usbs.bcaresearch.com 4 Please see U.S. Bond Strategy Special Report, “Assessing Corporate Default Risk”, dated March 19, 2019, available at usbs.bcaresearch.com 5 Please see U.S. Bond Strategy Weekly Report, “The Long Awkward Middle Phase”, dated July 2, 2019, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Weekly Report, “Full Speed Ahead”, dated April 16, 2019, available at usbs.bcaresearch.com 7 Please see U.S. Bond Strategy Weekly Report, “The Long Awkward Middle Phase”, dated July 2, 2019, available at usbs.bcaresearch.com 8 Please see U.S. Bond Strategy Weekly Report, “The New Battleground For Monetary Policy”, dated March 26, 2019, available at usbs.bcaresearch.com 9 Please see U.S. Bond Strategy Weekly Report, “Hedge Near-Term Credit Exposure”, dated May 28, 2019, available at usbs.bcaresearch.com 10  Please see U.S. Bond Strategy Weekly Report, “The Search For Aaa Spread”, dated March 12, 2019, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification Corporate Sector Relative Valuation And Recommended Allocation