Debt Trends

Highlights Why is the gap between the stock market and the economy so wide?: It is well established that stocks can diverge considerably from fundamentals in the near term, but lately it is as if the stock tables and the front-page headlines are from entirely different newspapers. It may be because the virus poses much less of a threat to the owners of equities than the general populace: More affluent households are more readily able to work from home and to practice social distancing. They also have access to better medical care. With the S&P 500 having hit technical resistance, however, the gap may be nearing its upper limit: Large-caps have run in place since retracing half of their peak-to-trough losses, and the next Fibonacci resistance level is only another 5% higher. Where are the shoddy loans?: During the expansion, corporations were able to borrow on prodigally easy terms. If banks aren't holding the loans, who is? Feature That’s New York’s future, not mine – “Hold On” (Reed) For someone who entered the business as a sell-side trader, it is a matter of course that prices can diverge from fundamentals. The trading desk had a one-day horizon, and the traders necessarily made their way on price signals while barely considering fundamentals. Though the junior traders had been exposed to dividend discount models at their fancy colleges, the ones who lasted recognized they weren’t relevant to the desk’s mission. Trading the daily flow required accepting that new news can have a dramatically larger effect on stocks in the here and now than it would on the lifetime stream of earnings available to common shareholders. Long-run fair value might solely turn on the fundamentals, but animal spirits hold sway over any given tick. The sudden stop imposed by stay-at-home orders has made backward-looking economic data nearly irrelevant, but the sizable upward surprises in unemployment claims should not be ignored. Our Global Investment Strategy colleagues showed last week just how difficult it is for even severe near-term shocks to materially alter the present value of aggregate future earnings.1 Furthermore, the market effects of negative earnings shocks are inherently self-limiting at the margin because they tend to be accompanied by lower interest rates, driving up the equity risk premium and making stocks more attractive relative to “safe” fixed income alternatives. Bear markets coincide with recessions, though, as near-term earnings expectations are revised lower and animal spirits droop (Chart 1). Given that the recession just begun is expected to be the worst since the Great Depression, one would expect that equities would be stumbling in search of a bottom as investors remained fearful of taking on risk. Chart 1Joined At The Hip Joined At The Hip Joined At The Hip They have instead been acting like the S&P 500 found that bottom on March 23rd, when the index completed a 35% peak-to-trough decline in just 23 sessions. It then proceeded to gain 28.5% over the next eighteen sessions. Some retracement is to be expected after a sudden, sharp move, and the S&P 500 has only recovered half of the ground that it lost. It certainly priced in a great deal of bad news on the way down, but the data have been worsening, and investors have been forced to give up on the notion of a swift economic recovery. Why are stocks rising when economic projections are being downwardly revised and good virus news has been few and far between? We ourselves have been barely glancing at backward-looking economic data releases that merely confirm the well-understood fact that draconian social distancing measures have wrung much of the life out of the economy. The degree to which job losses have outrun consensus forecasts stands out nonetheless. Aggregate initial unemployment claims over the last five weeks have exceeded consensus expectations by 5.5 million (Table 1). Even though the forecasts have caught up to the situation on the ground, the claims data suggest that unemployment is now pushing 20%, a worst-case-scenario level that is far above the first forecasts that incorporated the effects of stay-at-home orders. Claims may well have peaked, but they’re still an order of magnitude higher than normal, and they are not finished exerting upward pressure on the unemployment rate. Table 1Job Losses Have Been Worse Than Expected Dichotomy Dichotomy Meanwhile, COVID-19 data have yet to provoke much optimism. The rate of US infections has yet to come down to Italy’s level (Chart 2), and hopes that remdesivir might prove to be a wonder drug were dashed late last week. Clients are increasingly asking us why the stock market is traveling such a dramatically different path than the economy and the virus. How could stocks have plunged at a record rate as the coronavirus drew a bead on the United States, but surged after crippling social distancing measures were put in place? Chart 2The US Has Fallen Behind Italy's Pace Dichotomy Dichotomy A Tale Of Two Boroughs The simplest answer is that the Fed’s response was swifter and more far-reaching than expected. Ditto Congressional actions, and we expect that DC will continue to deploy its fiscal firepower to try to shield households and businesses from the worst of the effects of the anti-virus measures. We believe the monetary and fiscal efforts will make a difference, and do not think it’s a coincidence that equities turned around the week of March 23rd, which began with the Fed’s rollout of a formidable new arsenal and ended with the passage of the CARES Act. But the market action has not accounted for the shift from expectations of a V-bottom to talk of Us, Ls and Ws. Two articles published a week apart in The New Yorker vividly illustrated a demographic virus gap. The first looked at COVID-19 from the perspective of financial professionals at hedge funds and other sophisticated investment aeries.2 Although the views of the investors in the profile shifted with the tide of the incoming data, they were generally of the mind that the health threat was being dramatically overhyped. One retired hedge fund manager boasted about his and his family’s non-stop early March air travel between New York, London and a Wyoming ski resort. The second article followed an emergency room resident at Elmhurst, a publicly funded hospital in a working-class Queens neighborhood, which has been described as the epicenter of the outbreak in several local media reports.3 “‘It’s become very clear to me what a socioeconomic disease this is,’” he said. “‘Short-order cooks, doormen, cleaners, deli workers – that is the patient population here. Other people were at home, but my patients were still working. A few weeks ago, when they were told to socially isolate, they still had to go back to an apartment with ten other people. Now they are in our cardiac room dying.’” Stock ownership is largely reserved to the affluent, with the top percentile of households owning 53% of equities as of the end of 2019, and the rest of the top decile owning another 35% (Chart 3). For households in the top decile, maintaining a healthy distance from the virus isn’t that difficult. Knowledge workers equipped with a laptop and a reliable internet connection can work from anywhere, unlike the Elmhurst patients in low-skilled service positions who have to work onsite. The tonier precincts of Manhattan feel nearly deserted, with their residents having decamped for second homes in lower-density areas. Perhaps it's because the Fed's attempts to shore up the economy have far more personal relevance for investors than the spread of the virus. There are no comprehensive data series on virus infections and outcomes by zip code, which would facilitate analysis of the link between household wealth and COVID-19, but New York state reports age-adjusted fatality rates in four racial/ethnic categories. In New York state ex-New York city, which has lesser extremes of wealth than the city itself, the cross-category disparities are striking (Chart 4). Race/ethnicity is far from an ideal proxy for inequality, but it is fair to conclude that financial market participants have a sound basis for being more sanguine about the virus than the overall population. Assuming that more affluent households will be able to remain out of the virus’ reach, the dichotomy can persist for as long as the economic impacts do not become so bad that investors cannot reasonably look through them. Chart 3Demographics Drive Stock Ownership ... Dichotomy Dichotomy Chart 4... And COVID-19 Fatalities Dichotomy Dichotomy Technical Resistance Back on the trading desk, technical analysis was the go-to tool for traders pricing large blocks of stock in real time. Following sizable moves, the Fibonacci sequence provided a popular method for assessing how far a stock might retrace its steps before resuming its course. The most widely used Fibonacci retracement levels are 38% and 62%, and 50%, a round number exactly between the two, has also become an anticipated stopping point. From the February 19 closing high of 3,386.15 to the March 23 closing low of 2,237.40, the S&P 500 lost 1,148.75 points. The 38%, 50% and 62% retracement levels are 2,673.93, 2,811.78 and 2,949.63, respectively. The S&P paused at the 38% level for just two days before breaking through it decisively, but it’s had more trouble making its way through 2,812, failing to hold above it for more than a day or two at a time (Chart 5). Should it escape 2,812, the 2,950 level waits just 5% higher. Chart 5Fibonacci Retracement Levels For The S&P 500 Fibonacci Retracement Levels For The S&P 500 Fibonacci Retracement Levels For The S&P 500 We are fundamental investors who do not get hung up on technical levels, though they can become self-fulfilling prophecies if enough participants are following them. Given the popularity of Fibonacci retracement, it is possible that a critical mass of short-term investors may view 2,812 and 2,950 as preferred levels for exiting long positions in the S&P. Our bigger near-term concern is that it is hard to see US equities making much more headway while the virus and ongoing distancing measures have the potential to cause investors to revise their fundamental expectations lower and/or lose a little bit of their policy-fueled nerve. Who's Left Holding The Bag? Multiple commentators have expressed alarm at the post-2008 increase in corporate debt, especially given anecdotal reports that lending covenants had been loosened dramatically. If the banks don’t hold the debt, as we’ve argued, who does, and could a wave of virus-inspired defaults cause larger problems in the financial system? The Fed’s fourth quarter Flow of Funds report, published last month, provides some clues, but does not answer the question definitively. As we saw in higher frequency data on aggregate banking system exposures, bank loans to nonfinancial corporations grew modestly (3.2% annualized) since December 31, 2008. Nonfinancial corporations borrowed in the bond market at double that rate (6.2% annualized). Foreign loans, powered by near doubling in 2017 and 2018, grew at an annualized 13.4% pace, and are four times as large as they were at the end of 2008. Finance company loans have shrunk, and trade payables grew at a modest 2% rate. (Chart 6). Chart 6Debt Risks Are Pretty Well Diffused Dichotomy Dichotomy Publicly available data from Preqin on the capital raised by direct lending funds suggests that their impact has been modest, accounting for only about a quarter of outstanding bank loans if every dollar they’ve raised is currently deployed. Demand for leveraged loans, senior floating-rate debt issued to high-yield borrowers, was occasionally intense as investors sought protection from rising rates. The desire for duration protection has faded as rates have plunged to new lows, but ETFs and CLOs were eager buyers at points during the last expansion. In a Special Report published last summer, our US Bond Strategy and Global Fixed Income Strategy services concluded that the ownership of leveraged loans is diffuse enough that credit strains are unlikely to pose a systemic threat. They were also encouraged that leveraged loans and high yield corporate bonds act as substitutes, keeping one another in check as investor preferences for fixed and floating instruments wax and wane. They also noted that leveraged loan lending standards had tightened last year, with a reduced share of covenant-lite loans being issued, though standards have eased again since they published their report (Chart 7). Chart 7Covenant Protections Have Eroded Covenant Protections Have Eroded Covenant Protections Have Eroded Chart 8Diverse Corporate Bond Ownership Will Help Mitigate The Effect Of Defaults Dichotomy Dichotomy There is no way around the fact that high yield corporate bondholders (Chart 8), owners of CLO tranches rated below AAA and leveraged loan holders face elevated credit losses as the broad economic shutdown provokes a wave of defaults in instruments without Fed support. We expect that the default losses will be spread out across enough constituents that they will not become worryingly concentrated, but they may contribute to a further erosion of risk appetites.       Doug Peta, CFA Chief US Investment Strategist dougp@bcaresearch.com Footnotes 1 Please see the April 23, 2020 Global Investment Strategy Weekly Report, "Could The Pandemic Actually Raise Stock Prices?" available at gis.bcaresearch.com. 2 Paumgarten, Nick. "The Price of a Pandemic." The New Yorker, April 20, 2020, pp. 20-24. The article, relaying traders’ conversations, contains some profanity. 3 Galchen, Rivka. "The Longest Shift." The New Yorker, April 27, 2020, pp. 20-26. The article, relaying ER conversations, contains some profanity.

Highlights The Chinese economy is recovering at a slower rate than the equity market has priced in. There is a high likelihood of negative revisions to Q2 EPS estimates and an elevated risk of a near-term price correction in Chinese stocks.  We expect a meaningful pickup in credit growth in H1 to improve domestic demand gain tractions in H2. This supports our overweight stance on Chinese stocks in the next 6-12 months, in both absolute and relative terms. There is still a strong probability that the yield curve will flatten, and the 10-year government bond yield may even dip below 2% in the wake of disappointing economic data in Q2. But our baseline scenario suggests the 10-year government bond yield should bottom no later than Q3 of this year. Feature This week’s report addresses pressing concerns from clients in China’s post-Covid-19 environment. China’s economy contracted by 6.8% in Q1, the largest GDP growth slump since 1976. Furthermore, the IMF’s baseline scenario projects a 3% contraction in global economic growth in 2020, with the Chinese economy growing at a mere 1.2%.1 This dim annual growth outlook means that the contraction in China’s economy will likely extend to Q2, dragging down corporate profit growth. In our April 1st report2 we recommended that investors maintain a neutral stance on Chinese stocks in the next three months due to uncertainties surrounding the pandemic, the oversized passive outperformance in Chinese stocks, and heightened risks for further risk-asset selloffs. On a 6- to 12-month horizon, however, we have a higher conviction that Chinese stocks will outperform global benchmarks. Our view is based on a decisive shift by policymakers to a “whatever it takes” approach to boost the economy. We believe that the speed of China’s economic recovery in the second half of 2020 will outpace other major economies.  Q: China’s economy is recovering ahead of other major economies. Why did you recently downgrade your tactical call on Chinese equities from overweight to neutral relative to global stocks? A: China’s economy is recovering, but it is recovering at a slower rate than the equity market has fully priced in (Chart 1A and 1B). We believe the likelihood of negative revisions to Q2 EPS estimates is high, and the risk of a near-term price correction in Chinese stocks remains elevated.  Chart 1AElevated Chinese Equity Outperformance Relative To Global Stocks Elevated Chinese Equity Outperformance Relative To Global Stocks Elevated Chinese Equity Outperformance Relative To Global Stocks Chart 1BChinese Stocks Largely Ignored Weakness In Domestic Economy Chinese Stocks Largely Ignored Weakness In Domestic Economy Chinese Stocks Largely Ignored Weakness In Domestic Economy The lackluster March data suggests that the pace of China’s economic recovery in April and even May will likely disappoint, weighing on the growth prospects for Q2’s corporate earnings (Chart 2). Chart 2EPS Growth Estimates Likely To Capitulate In Q2 EPS Growth Estimates Likely To Capitulate In Q2 EPS Growth Estimates Likely To Capitulate In Q2 The work resumption rate in China’s 36 provinces jumped sharply between mid-February and mid-March. However, since that time, the resumption rate among large enterprises has hovered around 80% of normal capacity (Chart 3). Chart 3Work Resumption Hardly Improved Since Mid-March Three Questions Following The Coronacrisis Three Questions Following The Coronacrisis The flattening of the work resumption rate curve is due to a lack of strong recovery in demand. Chart 4So Far No Strong Recovery In Domestic Demand So Far No Strong Recovery In Domestic Demand So Far No Strong Recovery In Domestic Demand The flattening of the resumption rate curve is due to a lack of strong recovery in demand. Although there was a surge in Chinese imports in crude oil and raw materials, the increase was the result of China taking advantage of low commodity prices. This surge cannot be sustained without a pickup in domestic demand. The March bounce back in domestic demand from the manufacturing, construction, and household sectors has all been lackluster (Chart 4). External demand, which growth remained in contraction through March, will likely worsen in Q2 (Chart 5). Exports shrunk by 6.6% in March, up from a deep contraction of 17.2% in January-February. Export orders can take more than a month to be processed, therefore, March’s data reflects pent-up orders from the first two months of the year. The US and European economies started their lockdowns in March, so Chinese exports will only feel the full impact of the collapse in demand from its trading partners in April and May. The work resumption rate will advance only if the momentum in domestic demand recovery increases to fully offset the collapse in external demand. The current 83% rate of work resumption implies that industrial output growth in April will remain in contraction on a year-over-year basis (Chart 6). Chart 5External Demand Will Worsen In Q2 External Demand Will Worsen In Q2 External Demand Will Worsen In Q2 Chart 6Will Q2 Industrial Output Growth Remain In Contraction? Will Q2 Industrial Output Growth Remain In Contraction? Will Q2 Industrial Output Growth Remain In Contraction? Although we maintain a constructive outlook on Chinese risk assets in the next 6 to 12 months, the short-term picture remains volatile in view of the emerging economic data. As such, we recommend investors to maintain short-term hedges for risk asset positions. Q: China’s policy response to mitigate the economic blow from COVID-19 has been noticeably smaller than programs rolled out in key developed economies, especially the US. Why do you think such measured stimulus from China warrants an overweight stance on Chinese stocks in the next 6-12 months relative to global benchmarks? A: It is true that the size of existing Chinese stimulus, as a percentage of the Chinese economy, is smaller than that has been announced in the US. But this is due to a different approach China is taking in stimulating its economy. In addition, both the recent policy rhetoric and PBoC actions suggest a large credit expansion is in the works. This will likely overcompensate the damage on China’s aggregate economy, and generate an outperformance in both Chinese economic growth and returns on Chinese risk assets in the next 6 to 12 months. China’s policy responses have an overarching focus on stimulating new demand and investment, which is a different approach from the programs offered by its Western counterparts. In the US, the combination of fiscal and monetary stimulus amounts to 11% of GDP as of April 16, with almost all policy support targeted at keeping companies and individuals afloat. In comparison, China’s policy response accounts for a mere 1.2% of its GDP.3  However, this direct comparison understates the enormous firepower in the Chinese stimulus toolkit, specifically a credit boom. As noted in our February 26 report,4 China has largely resorted to its “old economic playbook” by generating a huge credit wave to ride out the economic turmoil. Our prediction of the policy shift towards a significant escalation in stimulus was confirmed at the March 27 Politburo meeting. Moreover, the April 17 Politburo meeting reinforced a “whatever it takes” policy shift with direct calls on more forceful central bank policy actions, a first since the global financial crisis in 2008.5 Since 2008, the overnight repo rate’s breaking into the IORR-IOER corridor has been a reliable indicator leading to impressive credit upcycles. The PBoC’s recent aggressive easing measures have pushed down the interbank repo rate below the central bank’s interest rate on required reserves (IORR). The price for interbank borrowing is now near the lower range of the rate corridor, between the IORR and the interest rate on excess reserves (IOER).  Since 2008, the overnight repo rate’s breaking into the IORR-IOER corridor has been a reliable indicator leading to impressive credit upcycles (Chart 7).  Such credit super cycles, in turn, have led to both economic booms and an outperformance in Chinese stocks. Chart 7Another Credit Super Cycle Is In The Works Another Credit Super Cycle Is In The Works Another Credit Super Cycle Is In The Works Chart 8Financial Conditions Were Extremely Tight In 2011-2014 Financial Conditions Were Extremely Tight In 2011-2014 Financial Conditions Were Extremely Tight In 2011-2014 The 2012-2015 cycle was an exception to the relationship between the overnight interbank repo rate, credit growth and Chinese stock performance. A steep pickup in credit growth in 2012 coincided with a leap in the overnight interbank repo rate, and the credit boom did not help boost demand in the real economy or improve Chinese stock performance. This is because corporate borrowing was severely curtailed by high lending rates during a four-year monetary tightening cycle from 2011 to 2014 (Chart 8). The credit boom during that cycle was largely driven by explosive growth in short-term shadow-bank lending and wealth management products (WMP), and did not channel into the real economy.6 We do not think such an extreme phenomena will replay under the current circumstances. Monetary stance will likely remain tremendously accommodative through the end of the year to facilitate a continuous rollout of medium- to long-term bank loans and local government bonds. Chinese financial institutions’ “animal spirits” may have been unleashed. But under the scrutiny of the Macro-Prudential Assessment Framework and the New Asset Management Rules,7 the "animal spirits" are unlikely to run up enough risks to prompt the PBoC to prematurely tighten liquidity conditions in the interbank market. Marginal propensity in China is pro-cyclical, which tends to lag credit cycles by 6 months. Chart 9Marginal Propensity In China Is Pro-Cyclical Marginal Propensity In China Is Pro-Cyclical Marginal Propensity In China Is Pro-Cyclical Both corporate and household marginal propensity, a measure of the willingness to spend, will pick up as well. Marginal propensity is pro-cyclical, which tends to lag credit cycles by 6 months (Chart 9). In other words, when interest rates are low and credit growth improves, corporates and households tend to spend more.  The meaningful expansion in credit growth, which started in Q1 and will sustain in the coming two to three quarters, will help corporate and household spending gain tractions in H2. This constructive view on Chinese stimulus and economic recovery supports our overweight stance on Chinese stocks in the next 6-12 months, in both absolute and relative terms.  Q: The yield curve in Chinese government bonds has steepened following PBoC’s aggressive monetary easing announcements. Has the Chinese 10-year bond yield bottomed?   A: No, we do not think the 10-year bond yield has bottomed. There is probability the 10-year government bond yield may briefly dip below 2% in Q2. However, barring a multi-year global economic recession, we think the 10-year government bond yield will bottom no later than Q3 this year. Chart 10A Wide Gap Between The Long and Short A Wide Gap Between The Long and Short A Wide Gap Between The Long and Short The short end of the yield curve dropped disproportionally compared with the long end, following the PBoC’s announcement to place its first IOER cut since 2008 (Chart 10). This led to a rapid steepening in the yield curve. While our view supports a flattening of the yield curve in Q2 and even a 50bps drop in the 10-year government bond yield, we think that the capitulation will be brief. In order for the 10-year government bond yield to remain below 2% for an extended period of time, the market needs to believe one or more of the following will happen: The pandemic will cause a multi-year global economic recession, preventing the PBoC from normalizing its policy stance in the foreseeable future. The duration and depth of the economic impact from the pandemic are still moving targets. Our baseline scenario suggests that the Chinese economic recovery will pick up momentum in H2 this year. The PBoC will not normalize its policy stance even when the economy has stabilized. The PBoC has a track record as a reactive central bank rather than a proactive one. Still, during each of the past three economic and credit cycles, the PBoC has started to normalize its interest rate on average nine months following a bottom in the business cycle (Chart 11). The tightening of interest rate even applied to the prolonged economic downturn and deep deflationary cycle in 2015/16 (Chart 12).    Chart 11The 'Old Faithful' PBoC Policy Normalization Pattern The 'Old Faithful' PBoC Policy Normalization Pattern The 'Old Faithful' PBoC Policy Normalization Pattern Chart 12Policy Normalized Even After A Long Economic Downturn Policy Normalized Even After A Long Economic Downturn Policy Normalized Even After A Long Economic Downturn Chart 132008 Or 2015? 2008 Or 2015? 2008 Or 2015? How the yield curve has historically behaved also depended on the market’s expectations on the speed of the economic recovery, and the timing of the subsequent monetary policy normalization. The yield curved spiked in the wake of substantial monetary easing and pickup in credit growth, in both 2008 and 2015 (Chart 13). While in 2008 the yield curve moved in lockstep with the 3-month SHIBOR with a perfect reverse correlation, in the 2015/16 cycle the yield curve spiked initially but quickly flattened. The long end of the yield curve capitulated as soon as the market realized the economic slowdown was a prolonged one. The 10-year government bond yield, after trending sideways in early 2016, only truly bottomed after the nominal output growth troughed in Q1 2016 (Chart 13, bottom panel). Will the yield curve behave like in 2008, or more like in 2015 in this cycle? We think it will be somewhere in between. The current economic cycle bottomed in Q1, but the economy is only recovering slowly and we expect a U-shaped economic recovery rather than a 2008-style V-shaped one.  At the same time, our baseline scenario does not suggest the current environment will evolve into a 4-year deflationary cycle as in the 2012-2016 period. Therefore, we expect the low interest rate environment to endure for another two to three quarters before the PBoC starts to reverse its policy stance back to the pre-COVID-19 range. As such, the yield on 10-year government bonds will fall, possibly by as much as 50bps, when the economic data disappoint in Q2 and more rate cuts are forthcoming.  But it will bottom when the economic recovery starts to gain traction in H22020 and the market starts to price in a subsequent monetary policy normalization.  When growth slows and debt rises sharply, the PBoC will need to join its western counterparts to permanently maintain an ultra-low interest rate policy to accommodate its high debt level. We acknowledge the fact that China’s potential output growth is trending down (Chart 14).  But it has been trending downwards since 2011. A structurally slowing rate of economic growth has not prevented the PBoC from cyclically raising its policy rate. Hence, unless we see evidence that the pandemic is meaningfully lowering China’s potential growth on par with growth rates in the DMs, our baseline scenario does not support a structural ultra-low interest rate environment in China. China’s debt-to-GDP ratio will most likely rise substantially this year, given that the credit impulse will gain momentum and GDP will grow very modestly. However, this rapid rise in the debt-to-GDP ratio will most likely not be sustained beyond this year. Even if we assume that credit impulse will account for 40% of GDP in 2020 (the same magnitude as in 2008/09), a sharp reversal in the output gap in 2021, as predicted by IMF,8 will flatten the debt-to-GDP ratio curve (Chart 15).  Moreover, following every credit super cycle in the past, Chinese authorities have put a brake on the debt-to-GDP ratio. Chart 14China's Potential Growth Is Likely To Trend Lower... China's Potential Growth Is Likely To Trend Lower... China's Potential Growth Is Likely To Trend Lower... Chart 15...But Has Not Stopped PBoC From Flattening The Debt Curve ...But Has Not Stopped PBoC From Flattening The Debt Curve ...But Has Not Stopped PBoC From Flattening The Debt Curve   All in all, while we see a high possibility for the 10-year government bond yield to fall in Q2, the decline will be limited in terms of duration. Jing Sima China Strategist jings@bcaresearch.com   Footnotes   1IMF World Economic Outlook, April 2020 2Please see China Investment Strategy Weekly Report "Investing During A Global Pandemic," dated April 1, 2020, available at cis.bcaresearch.com 3IMF, Policy Responses To COVID-19 https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19#U 4Please see China Investment Strategy Weekly Report "China: Back To Its Old Economic Playbook?" dated February 26, 2020, available at cis.bcaresearch.com 5“Stable monetary policy must become more flexible” and “use RRR reductions, lower interest rates, re-lending and other measures to preserve adequate liquidity and guide the loan prime rate downwards.” Statements from Xi Jinping, April 17, 2020 Politburo Meeting. http://www.gov.cn/xinwen/2020-04/17/content_5503621.htm  6 Bankers’ acceptances - short-term debt instruments guaranteed by commercial banks - swelled by 887% between end-2008 and 2012. The outstanding amount of WMPs jumped from 1.7 trillion RMB in 2009 to more than 9 trillion RMB by H12013. In contrast, the amount of RMB-denominated bank loans increased by only 67% during the same period. 7The Macro-Prudential Assessment Framework and the New Asset Management Rules were implemented in 2016 and 2018, respectively. They are designed to create additional restrictions to curb shadow-bank lending and broaden the PBoC’s oversight on banks’ WMP holdings. 8The April IMF World Economic Outlook predicts a 1.2% Chinese GDP growth in 2020 and a 9.2% GDP growth in 2021. Cyclical Investment Stance Equity Sector Recommendations

Highlights In mainstream EM, foreign currency debt restructuring is more likely to occur among corporates than governments. Thus, dedicated EM credit investors should overweight mainstream EM sovereign credit and underweight EM corporate debt. Urgency among EM companies and banks to hedge their large foreign currency liabilities will continue exerting downward pressure on EM exchange rates. Ongoing currency depreciation and the lack of buyers of last resort for EM credit underpin the following strategy: short EM sovereign and corporate credit / long US investment-grade corporate credit. Feature Scope And Focus Of Analysis This report re-visits the issue of EM foreign currency debt, assessing EM debt vulnerability. This report focuses on mainstream EM (countries included in Table 1), excluding Gulf countries and frontier markets. Frontier markets like Argentina, Ecuador, Egypt, Ukraine, Lebanon and sub-Saharan African countries occupy somewhat idiosyncratic positions and are therefore not part of this report. Gulf countries on the other hand, are extremely leveraged to oil prices and, unlike mainstream EMs they have currency pegs warranting a separate analysis.1 Chart 1Favor EM Sovereign Against EM Corporate Credit Favor EM Sovereign Against EM Corporate Credit Favor EM Sovereign Against EM Corporate Credit Among mainstream EM countries, public debt restructuring is not imminent and in the majority of cases is unlikely. However, there is a growing risk of foreign currency debt restructuring among EM companies and banks. Hence, we make a new investment recommendation: overweight mainstream EM sovereign credit / underweight EM corporate debt (Chart 1). We also reiterate the short EM sovereign and corporate credit / long US investment-grade corporate credit strategy. In this report, foreign currency debt is defined as the sum of foreign debt securities (i.e., foreign currency bonds) and foreign currency loans. In short, foreign currency debt measures foreign currency borrowing of companies, banks and governments. These statistics do not include foreign holdings of local currency bonds and equities or any other local currency liability of residents to foreigners. Overall, the level of foreign currency debt is pertinent in assessing EM debt vulnerability originating from exchange rate depreciation. Table 12 offers comprehensive foreign currency debt statistics for each individual country and EM as a whole. It details foreign currency debt by type of borrower - the government, corporates and banks – and also reveals the breakdown between foreign debt securities and foreign currency loans for each segment. Table 1EM FX Debt: Who Owes How Much EM: Foreign Currency Debt Strains EM: Foreign Currency Debt Strains Chart 2EM FX Debt Has Doubled Since 2008 EM FX Debt Has Doubled Since 2008 EM FX Debt Has Doubled Since 2008 The foreign currency debt of Chinese companies and banks is quite substantial relative to other EM countries. Hence, including China in the EM aggregates would materially affect these EM aggregates. We thus focus our analysis on EM ex-China and present China’s numbers separately. Since early 2009, EM ex-China aggregate foreign currency debt has doubled to about $3 trillion (Chart 2). Furthermore, this $3 trillion EM ex-China foreign currency debt is split as follows in terms of borrower type: non-financial corporates ($1.25 billion), banks ($846 billion) and governments ($878 billion). Government Foreign Currency Debt Among mainstream EM countries, foreign currency government debt is not vulnerable to restructuring or default. The reason is that the foreign currency debt burden of governments is low, having declined dramatically in the last decade. Table 2 illustrates that the share of local currency government debt is by far greater than the foreign currency debt in each EM country. Table 2EM Public Debt: Local Currency Exceeds FX Debt EM: Foreign Currency Debt Strains EM: Foreign Currency Debt Strains In the past 10 years, EM governments have deliberately replaced their foreign currency debt with local currency debt. Search for yield by international fixed-income investors has facilitated this debt swap: enormous foreign demand for EM domestic bonds has allowed EM governments to issue a considerable amount of local currency bonds. Chart 3EM Foreign Exchange Reserves Are Large EM Foreign Exchange Reserves Are Large EM Foreign Exchange Reserves Are Large In addition, mainstream EM countries, with exception of Turkey and South Africa, hold large foreign currency reserves (Chart 3). Lately, several mainstream EM countries have gained a new defense tool from the Federal Reserves – US dollar swap lines. EM central banks’ swap lines with the Fed are primarily intended to instill confidence among investors in financial markets. They could be used to fend off short-term speculative attacks on EM currencies. Nevertheless, they cannot alleviate insolvency problems. We will elaborate more about these swap lines with the Fed in another report this week. As to local currency public debt, the odds of debt restructuring are also low. First, the majority of EM countries have low aggregate public debt burdens as a share of the GDP (Table 2). Second, the majority of these nations have flexible currency regimes. This means that their central banks control the printing press. In the worst-case scenario - when investors become reluctant to own EM local currency government bonds, EM central banks can buy those bonds in both the secondary or primary markets if needed. In short, EM central banks can resort to a form of quantitative easing, i.e., purchasing local currency government bonds that would amount to public debt monetization. The wild card in this case will be the exchange rate – the currencies could depreciate substantially amid public debt monetization by central banks. Given that government liabilities in foreign currencies have declined substantially, exchange rate depreciation will not be a constraint for policymakers’ ability to monetize local currency debt. Remarkably, in the past two months amid the global indiscriminate selloff, central banks in several EM countries have begun purchasing government bonds or have stated that they will do so if required. This has created a precedent that will be used in future. One country that has large local currency government debt is Brazil. We have previously argued that Brazil requires robust nominal GDP growth to climb out of a public debt trap. With the COVID-19 crisis, the outbreak for its public debt has worsened considerably. Without the central bank monetizing public debt, it will be difficult for Brazil to escape rising government debt strains and, ultimately, local currency debt restructuring. In short, the cost of avoiding local currency public debt restructuring in Brazil could be large currency depreciation. Bottom Line: In mainstream EM, neither foreign currency nor local currency government debt face an imminent risk of restructuring. Public debt restructuring and defaults are occurring in Argentina and among frontier markets like Ecuador, Lebanon and a few sub-Saharan nations that are beyond the scope of this report. If local currency government bond markets become anxious about public debt sustainability, EM central banks could purchase government paper. If done on large scale, this will cause further currency depreciation. Corporate Foreign Currency Debt From a macro perspective, there are presently some pre-conditions that herald rising odds of foreign currency debt restructuring among EM corporates and banks: (1) rapid and massive foreign currency debt built up in the past 10-15 years; (2) substantial plunge in corporate revenues; and (3) massive currency depreciation. Taken together, these create fertile ground for debt restructuring by some corporate debtors. Foreign currency debt of companies and banks in mainstream EM ex-China countries has swelled in the past 10 years reaching $2.1. Bonds account for about $1.4 trillion while foreign currency loans account for the remaining $0.7 trillion. The global recession brought about by the COVID-19 pandemic is producing a collapse in EM companies’ local currency revenues and exports. Notably, EM ex-China exports were contracting even before the COVID-19 outbreak and they are currently crashing (Chart 4). Chart 4EM Exports & Corporate Credit Spreads EM Exports & Corporate Credit Spreads EM Exports & Corporate Credit Spreads Chart 5Commodities Prices And Currencies Drive EM Credit Spreads Commodities Prices And Currencies Drive EM Credit Spreads Commodities Prices And Currencies Drive EM Credit Spreads The top panel of Chart 5 illustrates EM corporate credit spreads (inverted) correlate with commodities prices. Hence, plunging commodities prices entail growing foreign currency debt stress for EM companies and banks. Finally, EM ex-China currencies have depreciated substantially making foreign currency debt more expensive to service (Chart 5, bottom panel). Please refer to Box 1 attesting that for EM debtors with US dollar liabilities, EM exchange rate depreciation is worse than that of higher US bond yields.     Box 1 What Is More Imperative For EM FX Debt: Exchange Rates Or Interest Rates? EM debtors with dollar debt are much more vulnerable to an appreciating dollar than rising US interest rates. Table 3 illustrates this point using the following hypothetical simulation: We consider a conjectural Brazilian debtor with $1,000 in debt with five years remaining to maturity, and a starting point exchange rate of 4 BRL per USD. In our example, a 5% depreciation in local currency against the dollar boosts the overall debt burden by 200 BRL (please refer to row 2 of Table 3). This does not include the rise in local currency costs of interest payments. It reflects only the increased burden of principal. Table 3A Hypothetical Simulation: FX Debt Burden Is More Sensitive To Exchange Rate Than Borrowing Costs EM: Foreign Currency Debt Strains EM: Foreign Currency Debt Strains An equivalent rise in debt servicing costs in local currency will require a 100-basis-point increase in US dollar borrowing costs. In brief, US dollar rates should rise by 100 basis points for interest payments to increase by BRL 200 over a five-year period, the time remaining to maturity. This simulation reveals that a 5% dollar appreciation versus local currency is as painful as a 100 basis points rise in US dollar rates and is more burdensome if the cost of coupon payments is accounted for. Provided there are higher odds of 5% currency depreciation in many EMs than a 100-basis-point rise in US dollar borrowing costs, we infer that EM FX debtors’ creditworthiness is more sensitive to exchange rates than to US Treasury yields. As the bottom panel of Chart 5 above clearly demonstrates, EM corporate and sovereign credit spreads correlate strongly with EM exchange rates. Consequently, the trend in EM exchange rates versus the US dollar is much more important for EM credit spreads than fluctuations in US bond yields. As to the currency composition of EM FX debt, about 82% of EM external debt is in US-dollar terms. Bottom Line: So long as EM currencies depreciate against the greenback, EM FX debt stress will mount, and EM corporate and sovereign credit spreads will widen. This will occur irrespective of whether US Treasury yields rise or drop.   If the bear market in commodities persists and/or EM currencies depreciate further – which is our baseline scenario, defaults on and restructuring of foreign currency debt among EM companies and banks are probable. One avenue to avoid corporate defaults is for the government to guarantee or assume the banks’ and companies’ foreign currency liabilities. It is probable because many of these borrowers are large entities with close links to their governments. However, governments will step in only after a debtor is on the brink default and its credit spreads are very wide. Briefly put, investors should be careful not to bet too early on government backstops of EM corporates’ and banks’ foreign currency debt. Identifying which corporate issuers could default or restructure debt involves bottom-up analysis that is beyond the scope of the macro research that BCA specializes in. An important question is what portion of corporate foreign currency liabilities have these debtors already hedged? Unfortunately, there are no macro data to answer this question either. Judging by the magnitude and speed of EM currency depreciation we have seen in the past two months, odds are that they have already partially hedged their exchange rate risk. Yet, given the sheer size of foreign currency liabilities, it is hard to imagine that corporates and banks have hedged all of them. Below we analyze each countries’ ability to service its foreign currency debt from a macro perspective. Vulnerability Assessment From a macro standpoint, foreign debt servicing vulnerability can be measured by foreign debt obligations (FDOs) and foreign funding requirements (FFRs). Chart 6EM FDOs And FFRs (Annualized) EM FDOs And FFRs (Annualized) EM FDOs And FFRs (Annualized) FDOs are the sum of debt expiring in the next 12 months, and interest as well as amortization payments over the next 12 months. FDO data are available until Q3 of 2019 (Chart 6, top panel). Hence, using this latest datapoint is pertinent to gauging the ability of individual countries to service their foreign debt over the coming six months. FFRs are the sum of FDOs in the next 12 months and current account balance (Chart 6, bottom panel). It measures the amount of foreign capital inflows required in the next 12 months for a country to cover any shortfalls in its balance of payment dynamics. Exports Coverage Of FDO: This measure compares annualized US dollar export revenues available to each country to its foreign debt service obligations in the next 12 months (Chart 7). The most vulnerable countries according to this measure are Brazil, Colombia, Turkey and Peru. On the other hand, Russia, Mexico, India & Korea have higher exports-to-FDO ratios. Chart 7Exports-To-Foreign Debt Obligations Ratio EM: Foreign Currency Debt Strains EM: Foreign Currency Debt Strains Foreign Exchange Reserves-to-FFRs Ratio: These metrics compare the size of foreign exchange reserves held by each nation’s central bank to its FFRs in the next 12 months (Chart 8). By this measure, Chile, Colombia, Turkey, Indonesia and Mexico have large FFRs relative to their central bank foreign exchange reserves. Meanwhile, Russia, Korea and Thailand fare well on this metric. Chart 8FX Reserves-To-Foreign Funding Requirements EM: Foreign Currency Debt Strains EM: Foreign Currency Debt Strains On the whole, Chart 9 is a scatter plot combining both FDO and FFR measures to determine the most and least vulnerable EMs. The most vulnerable EMs are Brazil, Turkey, Colombia and Chile. Meanwhile, Russia, Korea, India and the Philippines are the least vulnerable. Chart 9EM FX Debt And Currency Vulnerability EM: Foreign Currency Debt Strains EM: Foreign Currency Debt Strains Investment Recommendations So long as EM currencies depreciate against the greenback, EM foreign currency debt stress will mount, and EM corporate and sovereign credit spreads will widen. We remain bearish on EM currencies. They usually trade with the global business cycle and the latter remains in free fall. We continue recommending shorting a basket of the following currencies versus the US dollar: BRL, CLP, ZAR, IDR, PHP and KRW. There will likely be no imminent restructuring or default on public debt in mainstream EM countries, outside frontier markets like Argentina, Ecuador, Lebanon and sub-Saharan African countries. However, there could be meaningful credit stress among EM corporate issuers. Consequently, dedicated EM credit investors should overweight mainstream EM sovereign credit and underweight EM corporate debt. We continue to recommend underweighting EM sovereign and corporate credit versus US investment-grade corporate credit (Chart 10). Not only is the Fed buying US investment-grade and some high-yield bonds but US companies will also benefit from the substantial fiscal stimulus. In EM, corporates and banks lack such support. Crucially, in contrast to US corporates, EM issuers also suffer from currency depreciation. Within the EM sovereign credit universe, our overweights are Russia, Mexico, Peru, Thailand and Malaysia. Underweights include South Africa, Brazil, Indonesia, the Philippines and Turkey. The rest warrant a neutral allocation within an EM sovereign credit portfolio. Finally, within corporate credit, we reiterate our long-standing recommendation of long Asian investment-grade corporates / Asian short high-yield corporate (Chart 11). We continue recommending shorting a basket of the following currencies versus the US dollar: BRL, CLP, ZAR, IDR, PHP and KRW. Chart 10Remain Underweight EM Credit Versus US IG Credit Remain Underweight EM Credit Versus US IG Credit Remain Underweight EM Credit Versus US IG Credit Chart 11Long Asian IG Corporate / Short Asian HY Corporate Long Asian IG Corporate / Short Asian HY Corporate Long Asian IG Corporate / Short Asian HY Corporate     Andrija Vesic Associate Editor andrijav@bcaresearch.com Arthur Budaghyan Chief Emerging Markets Strategist arthurb@bcaresearch.com   Footnotes 1We will publish a report on Saudi Arabia in the coming weeks. 2We have compiled data on foreign currency securities issued by non-financial companies and banks from Bloomberg. Bloomberg data accounts for the nationality of debt issuers. For instance, a US dollar bond issued by a Brazilian corporate subsidiary or a shell company located in the Cayman Islands is counted as Brazilian foreign corporate debt, rather than a Cayman Island debt security. For foreign loans, we use the Bank of International Settlements (BIS) datasets on Banking Statistics.

Highlights Investment Grade: Investors should overweight investment grade corporate bonds relative to a duration-matched position in Treasury securities, with a particular focus on bonds that are eligible for the Fed’s purchase programs. High-Yield: Caution is still warranted in the high-yield market. At current levels, spreads do not adequately compensate investors for the coming default cycle. We would recommend buying high-yield if the average index spread rises to a range of 1075 bps – 1290 bps. Fed Purchases: Fed corporate bond purchases will cause investment grade spreads to tighten, particularly out to the 5-year maturity point. However, the program won’t stop the coming onslaught of ratings downgrades. High-Yield Sectors: The Energy, Transportation, Capital Goods, Consumer Cyclical and Consumer Noncyclical sectors are all highly exposed to the looming default cycle. Financials and Utilities look like the best places to hide out. Feature Chart 1Will The Fed's Corporate QE Mark The Top In Spreads? Will The Fed's Corporate QE Mark The Top In Spreads? Will The Fed's Corporate QE Mark The Top In Spreads? The COVID pandemic and associated recession have already caused turmoil in financial markets and prompted a policy response from the Federal Reserve that is unprecedented in its aggressiveness. US investment grade and high-yield corporate spreads widened 280 bps and 764 bps, respectively, to start the year. Then, they tightened by 78 bps and 179 bps, respectively, after the Fed announced it is stepping into the corporate bond market for the first time (Chart 1). Clearly, this is a challenging time for corporate bond investors. But sifting through all the noise, we think there are three key questions to stay focused on: How will the Federal Reserve’s support for the corporate bond market impact spreads? At what level do spreads fully discount the looming default cycle? What sectors within the corporate bond market are most/least at risk of experiencing large-scale defaults? What Can The Fed Hope To Accomplish By Buying Corporate Debt? As part of its package of monetary policy stimulus measures to combat the US COVID-19 recession, the Fed has undertaken a dramatic new step to try and lower borrowing costs for US businesses – the outright buying of US investment grade corporate bonds. The main details of these new programs are as follows: The Fed will purchase investment grade corporate bonds, loans and related exchange-traded funds (ETFs) as part of these programs. Bonds can be purchased in the primary (newly-issued) and secondary markets. The purchases will not be held on the Fed’s balance sheet. Instead, two off-balance sheet Special Purpose Vehicles (SPVs), one for primary market purchases and one for secondary market purchases, will buy the bonds. Both SPVs are initially funded by the US Treasury and will be levered up via loans from the Fed. The primary market SPV will buy newly-issued bonds with credit ratings as low as BBB- and maturities of four years or less.  Eligible issuers are US businesses with material operations in the United States; that list of companies may be expanded in the future. Eligible issuers do not include companies that are expected to receive direct financial assistance from the US government (i.e. no buying of bonds from companies getting bailout funds). The secondary market SPV will buy bonds with maturities of up to five years and credit ratings as low as BBB-, with a buying limit of 10% of the entire stock of eligible debt of any single company. This secondary market SPV will also buy investment grade bond ETFs, up to 20% of the outstanding shares of any single ETF. Through the primary market facility, any eligible company can “borrow” from the Fed, through bond purchases or direct loans, an amount greater than its maximum outstanding debt (bonds plus loans) on any day over the past twelve months. Specifically: 140% of all debt for AAA-rated issuers, 130% for AA-rated issuers, 120% for A-rated issuers and 110% for BBB-rated issuers. Since those percentages are all greater than 100, this effectively means that the Fed will allow eligible companies to potentially roll over their entire stocks of debt through this program, plus some net new borrowing. With the primary market facility, issuers can even defer interest payments on the funds borrowed from the Fed for up to six months, with the interest payments added to the final repayment amount (any company choosing this option cannot do share buybacks or make dividend payments). These programs are set to run until September 30 of this year, with an option to extend as needed. The Fed’s new initiatives represent a new step for the central bank, providing direct lending to any company that needs it. The Fed had to do this through off-balance-sheet SPVs, since direct buying of corporates is not permitted under the Federal Reserve Act. With this structure, it is technically the US Treasury department that bears the initial credit risk through its seed funding of each SPV. The BoJ was the first of the major central banks to start buying corporate bonds. This structure is different than the recent corporate bond QE programs of the European Central Bank (ECB), Bank of England (BoE) and Bank of Japan (BoJ), where the credit risk was directly taken onto the central bank balance sheets. But from an investment perspective, the difference in structure between the Fed’s corporate bond buying program and that of other central banks is nothing more than a technicality. It is still worthwhile to see if any lessons can be learned from these other countries.     The Corporate Bond Buying Experience Of Other Central Banks The BoJ was the first of the major central banks to start buying corporate bonds, in a program that began in February 2009 and continued until October 2012. The program initially involved only the purchase of very high-quality corporate debt (rated A or higher) and only for maturities up to one year. The pool of eligible bonds was later increased to allow for lower credit quality (rated BBB or higher) and longer maturities (up to three years). The BoJ ended up buying a total of 3.2 trillion yen (US$30 billion) of bonds during that program, representing nearly 50% of total Japanese investment grade nonfinancial debt (Chart 2). Credit spreads tightened modestly over the life of the program, particularly for the shorter maturity debt that the BoJ was directly buying.1 Research from the BoJ concluded that the corporate bond buying did improve liquidity for the bonds that were eligible for the program, although there was no discernable pickup in overall Japanese corporate bond issuance.2 The BoE started its Corporate Bond Purchase Scheme (CBPS) in August 2016, as part of a package of stimulus measures to cushion the economic blow from the UK’s stunning vote to leave the European Union. The CBPS bought £10bn of UK nonfinancial investment grade corporate bonds over a period of 18 months, with ratings as low as BBB-. This was a relatively modest share of all eligible nonfinancial bonds (4.7%), but UK credit spreads did tighten over the life of the program (Chart 3). The BoE’s own research has determined that the spread tightening was due to lower downgrade/default risk premiums, and that the program triggered a surge in investment grade issuance in the weeks and months following its launch.3 Chart 2The BoJ's Corporate Bond Buying Experience The BoJ's Corporate Bond Buying Experience The BoJ's Corporate Bond Buying Experience Chart 3The BoE's Corporate Bond Buying Experience The BoE's Corporate Bond Buying Experience The BoE's Corporate Bond Buying Experience The ECB announced its Corporate Sector Purchase Program (CSPP) in March 2016, with the actual bond purchases beginning three months later. This was an expansion of the ECB’s overall Asset Purchase Program that had previously been focused on government debt. Like the BoJ and BoE programs, only nonfinancial debt of domestic euro area companies rated BBB- or higher was eligible. The ECB did buy bonds across a wide maturity spectrum of 1-30 years. The ECB’s purchases in the first 18 months of the CSPP were sizeable, between €60-80bn per month, reaching a cumulative total of nearly 20% of the stock of eligible bonds (Chart 4). This not only drove credit spreads tighter for bonds in the CSPP, but also pushed spreads lower for bonds that were not directly purchased by the ECB, like bank debt. The ECB described this as evidence of a strong “portfolio balance effect”, where investors who sold their bonds to the central bank ended up redeploying the proceeds into other parts of the euro area corporate bond market.4  One major difference between the ECB CSPP and the BoJ and BoE programs was that the ECB could conduct the necessary purchases in the primary market, if necessary. This represented a major new source of funding for smaller euro area companies that did not previously issue corporate bonds, preferring to get most of their debt financing through bank loans. As evidence of this, the year-over-year growth rate of euro area corporate bond issuance soared from 2.5% to 10% in the first year of the CSPP (Chart 5). Chart 4The ECB's Corporate Bond ##br##Buying Experience The ECB's Corporate Bond Buying Experience The ECB's Corporate Bond Buying Experience Chart 5ECB Primary Market Buying Spurred A Boom In Issuance ECB Primary Market Buying Spurred A Boom In Issuance ECB Primary Market Buying Spurred A Boom In Issuance Investment Conclusions Applying these lessons to the US, the first conclusion we reach is that Fed corporate bond purchases will tighten spreads for eligible securities. In this case, eligible securities include all investment grade rated US corporate bonds with maturities less than five years. In effect, the Fed’s primary market facility could be thought of as adding an agency backing to these eligible bonds since the Fed has effectively guaranteed that this debt can be rolled over and that bond investors will be made whole. It’s noteworthy that last week saw a record amount of new investment grade corporate bond issuance as firms rushed to take advantage of the program.    Second, we should see some positive knock-on effects on spreads of ineligible investment grade securities, i.e. investment grade corporate bonds with maturities greater than five years. The impact will not be as large as for eligible securities, but since many of the same issuers operate at both ends of the curve, long-maturity spreads will benefit at the margin from any reduction in interest expense for the issuer. Third, any trickle-down effects to high-yield spreads will be much smaller. No high-yield issuers can benefit from the program, and while the Fed could eventually open up its facilities to include high-yield debt, we wouldn’t count on it. We suspect the moral hazard of “bailing out the junk bond market” would simply be a step too far for the Federal Reserve. We should see some positive knock-on effects on spreads of ineligible securities. In sum, we would advocate an overweight allocation to US investment grade corporate bonds today – especially on securities eligible for the Fed’s programs. We do not recommend a similar overweight stance on US high-yield, where spreads will continue to fluctuate based on the fundamental default outlook (see section titled “Assessing The Value In High-Yield” below). Can The Fed Re-Steepen US Credit Spread Curves And Prevent Ratings Downgrades? Prior to the Fed’s announcement of the new programs, the US investment grade corporate spread curve had become inverted, with shorter maturity spreads exceeding longer maturity ones. This has historically been a harbinger of increased investment grade downgrades and high-yield defaults (Chart 6). With the Fed’s new programs focusing on bonds with maturities of up to five years, the Fed’s buying can potentially lead to a re-steepening of the investment grade spread curve by driving down shorter maturity spreads. Chart 6Inverted US Credit Spread Curves Are Flashing An Ominous Message Inverted US Credit Spread Curves Are Flashing An Ominous Message Inverted US Credit Spread Curves Are Flashing An Ominous Message Already, the investment grade spread curve has begun to disinvert in the first week of the Fed’s programs (Chart 7). At the same time, the bond rating agencies are moving aggressively to adjust credit opinions in light of the US recession. Already, downgrades from Moody’s and S&P are outpacing upgrades by a 3-1 ratio year-to-date – a pace not seen since the depths of the financial crisis, according to Bloomberg.5  Chart 7The Fed's New Programs Are Already Helping Disinvert Investment Grade Spread Curves The Fed's New Programs Are Already Helping Disinvert Investment Grade Spread Curves The Fed's New Programs Are Already Helping Disinvert Investment Grade Spread Curves The Fed’s actions should be successful at re-steepening the investment grade credit curve. However, we doubt that they will have much impact on ratings decisions. While the Fed can reduce borrowing costs and prevent default by rolling over maturing debt for investment grade issuers, this has a relatively minor impact on corporate balance sheet health. In fact, the Fed's programs will only improve balance sheet health for firms that just roll over existing debt loads and don’t take on any new debt. Any firm that takes on new debt during this period will come out of the crisis with more leverage than when it entered. All else equal, that should warrant a downgrade. Bottom Line: Fed corporate bond purchases will cause investment grade spreads to tighten, particularly out to the 5-year maturity point. However, the program won’t stop the coming onslaught of ratings downgrades. Assessing The Value In High-Yield What Kind Of Default Cycle Is Already “In The Price”? High-yield debt may not benefit from the Fed’s corporate bond-buying programs. But, as in every other cycle, there will come a time when spreads discount the full extent of future default losses. At that point it will be appropriate to increase allocations to the sector. Our Default-Adjusted Spread will guide us as we make that determination. Our Default-Adjusted Spread is the excess spread available in the Bloomberg Barclays High-Yield index after subtracting realized default losses. Specifically, we calculate the Default-Adjusted Spread as: Index OAS – [Default Rate x (1 – Recovery Rate)] The default and recovery rates apply to the 12-month period that follows the index spread reading. For example, the Default-Adjusted Spread for January 2019 uses the index OAS from January 2019 and default losses incurred between February 2019 and January 2020. Table 1 shows that there is a strong link between the Default-Adjusted Spread and excess High-Yield returns relative to duration-matched Treasuries. Specifically, we see that losses are a near certainty if the Default-Adjusted Spread is negative and that return prospects are poor for spreads below 150 bps. A Default-Adjusted Spread above its historical average of 250 bps is an obvious buying opportunity, while a spread above 400 bps virtually guarantees strong returns. Table 1The Default-Adjusted Spread & High-Yield Excess Returns Trading The US Corporate Bond Market In A Time Of Crisis Trading The US Corporate Bond Market In A Time Of Crisis This helps clarify the task at hand. We must make an assumption about what the default and recovery rates will be for the next 12 months, then apply those assumptions to the current index spread. The resulting Default-Adjusted Spread will tell us if High-Yield bonds are worth a look. Table 2 shows the Default-Adjusted Spread that results from different combinations of default and recovery rates.6 For example, a 10% default rate and 35% recovery rate together imply a Default-Adjusted Spread of 271 bps, suggesting an attractive buying opportunity. Table 2Default-Adjusted Spread (BPs) Given Different Assumptions For Default And Recovery Rates Trading The US Corporate Bond Market In A Time Of Crisis Trading The US Corporate Bond Market In A Time Of Crisis What Sort Of Default Cycle Should We Expect? To answer this question we turn to Table 3. Table 3 lists periods since the mid-1980s when the default rate rose above 4%, along with several factors that influence the level of default losses: The magnitude of the economic downturn, proxied by the worst year-over-year real GDP growth reading recorded during that timeframe. The duration of the economic downturn, measured as the number of quarters from the peak to trough in real GDP. Nonfinancial corporate leverage – measured as total debt divided by book value of equity – at the cycle peak. Table 3A Brief History Of Default Cycles Trading The US Corporate Bond Market In A Time Of Crisis Trading The US Corporate Bond Market In A Time Of Crisis Alongside these determining factors, the table also shows the peak 12-month default rate seen during the cycle and the recovery rate that occurred alongside it. First, we notice a strong relationship between the magnitude of the economic shock and the peak default rate. Meanwhile, corporate leverage does a better job explaining the recovery rate. Notice that recoveries were greater in 2008 than in 2001, despite 2008’s larger economic shock. Turning to the current situation, our base case assumption is that we will see severe economic contraction in Q1 and Q2 of this year followed by some recovery in the third and fourth quarters. All told, 2020 annual GDP growth could be close to the -3.9% seen in 2008, though the duration of the peak-to-trough economic shock will be only two quarters instead of six.7 Based on the historical comparables listed in Table 3, this sort of economic shock could generate a peak default rate somewhere between 11% and 13%. As for recoveries, nonfinancial corporate leverage is currently higher than during any of the prior episodes in our study. It follows that the recovery rate will be very low, perhaps on the order of 20%-25%. Turning back to Table 2, we see that our default and recovery rate assumptions imply a Default-Adjusted Spread somewhere between -119 bps and +96 bps. This is too low to be considered a buying opportunity. A Default-Adjusted Spread above its historical average of 250 bps is an obvious buying opportunity, while a spread above 400 bps virtually guarantees strong returns. Table 4 flips this analysis around and shows the option-adjusted-spread on the Bloomberg Barclays High-Yield index that would generate a Default-Adjusted Spread of 250 bps based on different assumptions for the default and recovery rates. Recall that we consider a Default-Adjusted Spread of 250 bps or above as a buying opportunity. Using the aforementioned default and recovery rate assumptions, we would see a buying opportunity in high-yield if the average index spread rose to a range of 1075 bps – 1290 bps, or above. As of Friday’s close, the index option-adjusted spread was 921 bps. Table 4High-Yield Index Spread (BPs) That Would Imply A Buying Opportunity* In Different Default Loss Scenarios Trading The US Corporate Bond Market In A Time Of Crisis Trading The US Corporate Bond Market In A Time Of Crisis Bottom Line: High-yield spreads do not discount the full extent of the looming default cycle and will not benefit from the Fed’s asset purchase programs. Investors should stay cautious on high-yield for now and look to increase allocations when the average index spread moves into a range of 1075 bps to 1290 bps. Which High-Yield Sectors Are Most Exposed? Even during a period of large-scale defaults, sector and firm selection are vital in the high-yield bond market. In fact, you could argue that sector selection becomes even more important during a default cycle, as some sectors bear the brunt of default losses while others skate through relatively unscathed. To wit, Chart 8plots the 12-month trailing speculative grade default rate alongside a diffusion index that shows the percentage of 30 high-yield industry groups – as defined by Moody’s Investors Service – that have a trailing 12-month default rate above 4%. Even at the peaks of the default cycles during the last two recessions, only 47% and 63% of industry groups experienced significant default waves. Chart 8Sector Selection Is Vital In A Default Cycle Sector Selection Is Vital In A Default Cycle Sector Selection Is Vital In A Default Cycle To help identify which sectors are most at risk during the current default cycle, we consider how the 10 main high-yield industry groups, as defined by Bloomberg Barclays, stack up on three crucial credit metrics: The share of firms rated Caa Growth in par value of debt outstanding since the last recession Change in the median firm’s net debt-to-EBITDA ratio since the last recession8 Charts A1-A10 in the Appendix show how the three credit metrics for each industry group have evolved over time. In the remainder of this report we compare the sectors against each other across each of the above three dimensions. Note that Box 1 provides a legend for the sector name abbreviations used in Charts 9, 10 and 11. Box 1Sector Abbreviations Trading The US Corporate Bond Market In A Time Of Crisis Trading The US Corporate Bond Market In A Time Of Crisis Chart 9OAS Versus Share Of Caa-Rated Debt Trading The US Corporate Bond Market In A Time Of Crisis Trading The US Corporate Bond Market In A Time Of Crisis Chart 10OAS Versus Debt Growth Trading The US Corporate Bond Market In A Time Of Crisis Trading The US Corporate Bond Market In A Time Of Crisis   Chart 11OAS Versus Net Debt-To-EBITDA Trading The US Corporate Bond Market In A Time Of Crisis Trading The US Corporate Bond Market In A Time Of Crisis Share Of Caa-Rated Debt Even during a large default cycle the bulk of default losses will be borne by firms rated Caa and below. In Chart 9, we see that if we ignore the outlying Technology, Transportation and Energy sectors, there is a fairly linear relationship between credit spreads and the share of firms rated Caa in each sector. Transportation and Energy currently trade at very wide spreads because those sectors’ revenues are heavily impacted by the current crisis. Technology spreads remain low because, despite the high percentage of Caa-rated debt, the sector has one of the lower net debt-to-EBITDA ratios (see Chart A6). All in all, Chart 9 suggests that Capital Goods, Communications, Consumer Cyclicals and Consumer Noncyclicals all carry a large proportion of low-rated debt. In contrast, Financials and Utilities appear much safer. Debt Growth Another good way to assess which sectors are most likely to experience defaults is to look at which sectors added the most debt during the economic recovery (Chart 10). On that note, the rapid levering-up of the Energy sector clearly sticks out. Beyond that, Capital Goods, Consumer Noncyclicals and Technology also added significant amounts of debt during the recovery. In contrast, the Utilities sector actually reduced its debt load. Change In Net Debt-to-EBITDA Finally, it’s important to note that simply adding debt does not necessarily put a sector at greater risk of default if earnings are rising even more quickly. For this reason we also look at recent trends in net debt-to-EBITDA (Chart 11). Here, we see that wide spreads in Energy and Transportation are justified by large increases in net debt-to-EBITDA. Conversely, Financials and Communications have seen improvement. Bottom Line: Based on a survey of three important credit metrics: The Energy, Transportation, Capital Goods, Consumer Cyclical and Consumer Noncyclical sectors are all highly exposed to the looming default cycle. In contrast, Financials and Utilities look like the best places to hide out. Appendix Chart A1Basic Industry Credit Metrics Basic Industry Credit Metrics Basic Industry Credit Metrics Chart A2Capital Goods Credit Metrics Capital Goods Credit Metrics Capital Goods Credit Metrics Chart A3Consumer Cyclical Credit Metrics Consumer Cyclical Credit Metrics Consumer Cyclical Credit Metrics Chart A4Consumer Non-Cyclical Credit Metrics Consumer Non-Cyclical Credit Metrics Consumer Non-Cyclical Credit Metrics Chart A5Energy Credit Metrics Energy Credit Metrics Energy Credit Metrics Chart A6Technology Credit Metrics Technology Credit Metrics Technology Credit Metrics Chart A7Transportation Credit Metrics Transportation Credit Metrics Transportation Credit Metrics Chart A8Communications Credit Metrics Communications Credit Metrics Communications Credit Metrics Chart A9Utilities Credit Metrics Utilities Credit Metrics Utilities Credit Metrics Chart A10Financial Institutions Credit Metrics Financial Institutions Credit Metrics Financial Institutions Credit Metrics     Ryan Swift US Bond Strategist rswift@bcaresearch.com Robert Robis, CFA Chief Fixed Income Strategist rrobis@bcaresearch.com Jeremie Peloso Senior Analyst jeremiep@bcaresearch.com Footnotes  1 The March 2011 earthquake and tsunami in Japan created a lot of short-term credit spread volatility, but even then, shorter-maturity bonds saw less spread widening than the overall index. 2 https://www.imes.boj.or.jp/research/papers/english/18-E-04.pdf 3  https://www.bankofengland.co.uk/quarterly-bulletin/2017/q3/corporate-bond-purchase-scheme-design-operation-and-impact 4 The ECB described this effect in a 2018 report that can be accessed here: https://www.ecb.europa.eu/pub/pdf/other/ecb/ebart201803_02.en.pdf 5  https://www.bloomberg.com/news/articles/2020-03-26/s-p-moody-s-cut-credit-grades-at-fastest-pace-since-2008-crisis 6 Calculations are based on the index spread as of market close on Friday March 27. 7 For more details on BCA’s assessment of the economic outlook please see Global Investment Strategy Second Quarter 2020 Strategy Outlook, “World War V”, dated March 27, 2020, available at gis.bcaresearch.com 8 Median net debt-to-EBITDA is calculated from our bottom-up sample of high-yield firms that consists of all the firms in the Bloomberg Barclays High-Yield index for which data are available. Data are retrieved on a quarterly basis and the sample is adjusted once per year based on changes in the composition of the Barclays indexes. As of Q2 2019, this sample includes 354 companies.

Highlights Global investors have come to accept the secular stagnation narrative as described by Larry Summers in November 2013, and have gravitated to the only available real time estimate of the real neutral rate of interest: the Laubach & Williams (“LW”) “R-star” estimate. With this apparent visualization of secular stagnation as a guide, many investors have concluded that monetary policy ceased to be stimulative last year and that recent Fed rate cuts will be of limited benefit to economic activity even once economic recovery takes hold unless inflation meaningfully accelerates (thus pushing real rates lower for any given nominal Fed funds rate). This report revisits the “LW” R-star estimate in detail, and demonstrates why the estimation is almost certainly wrong, at least over the past two decades. We also outline an inferential approach that investors can use to monitor where the neutral rate is in real time and whether it is rising or falling. The core conclusion for investors is that US Treasury yields reflect a “low rates forever” view with much higher certainty than is analytically warranted and thus appear to be anchored by a false narrative. While bond yields may not rise significantly in the near-term, investors should avoid dogmatic medium-to-longer term views about yields as they may rise meaningfully over a cyclical and secular horizon once a post-COVID-19 expansion takes hold. Feature Over the past several weeks financial markets have moved rapidly to price in a global recession stemming from the COVID-19 outbreak. As financial market participants began to turn to policy makers for support, eyes focused first on the Federal Reserve, and then fiscal authorities. Earlier this week, the ECB joined the party and announced aggressive further measures of its own. When responding to the Fed’s return to the lower bound and its other recent monetary policy decisions, many market participants have expressed the view that the Fed is largely impotent to deal with a global pandemic. There are three elements to this view. The first is that interest rate cuts are ill equipped to stimulate domestic demand if quarantine measures or other forms of “social distancing” are in effect. The second element is that the Fed has only been capable of delivering a fraction of the reduction in interest rates compared to what has occurred in response to previous contractions. The third aspect of this view is that because the neutral rate of interest is so much lower now than it was in the past, Fed rate cuts will not be as stimulative as they were before. Chart II-1Monetary Policy Ceased To Be Stimulative Last Year, According To The LW R-star Estimate Monetary Policy Ceased To Be Stimulative Last Year, According To The LW R-star Estimate Monetary Policy Ceased To Be Stimulative Last Year, According To The LW R-star Estimate While we at least partly agree with the first and second elements of this view, we feel strongly that the third is flawed. Global investors have come to accept the secular stagnation narrative as described by Larry Summers in November 2013,1 and have gravitated to the only available real time estimate of the neutral rate of interest: the Laubach & Williams (“LW”) “R-star” estimate. This time series, which is regularly updated by the New York Fed,2 suggests that the real fed funds rate reached neutral territory in the first quarter of 2019 (Chart II-1). With this apparent visualization of secular stagnation as a guide, many investors have concluded that monetary policy ceased to be stimulative last year and that recent Fed rate cuts will be of limited benefit to economic activity even beyond the near term unless inflation meaningfully accelerates (thus pushing real rates lower for any given nominal Fed funds rate). In this Special Report we revisit the “LW” R-star estimate in detail, and demonstrate why the estimation is almost certainly wrong, at least over the past two decades. Our analysis does not reveal a precise alternative estimate of the neutral rate, although we do provide some inferential perspective on how investors may be able to monitor where the neutral rate is in real time and whether it is rising or falling. However, the core insight emanating from our report, particularly for US fixed income investors, is that US Treasury yields reflect a “low rates forever” view with much higher certainty than is analytically warranted and thus appear to be anchored by a false narrative. While bond yields may not rise significantly in the near-term, this underscores that they have the potential to rise meaningfully over a cyclical and secular horizon once economic activity recovers. As such, we caution fixed-income investors against dogmatic medium-to-longer term views about bond yields, as their potential to rise may be larger than many investors currently expect. Demystifying The LW R-star Estimate The LW estimate of the neutral rate of interest has gained credibility for three reasons. First, as noted above, the evolution of the series fits with the secular stagnation narrative re-popularized by Larry Summers. Second, the series is essentially sponsored by the Federal Reserve even if it is not officially part of the Fed’s forecasting framework, as its two creators are long-time Fed employees (Thomas Laubach is a director of the Fed’s Board of Governors, and John Williams is the current President of the New York Fed). But, in our view, there is a third important reason that global investors have accepted the LW R-star estimate of the neutral rate of interest: the methodology used to generate the estimate is extremely technically complex, and thus is difficult for most investors to penetrate. Much of the technical complexity of the LW estimate is centered around the use of a statistical procedure called a Kalman filter (“KF”). Simply described, the KF is an algorithm that tries to estimate an unobservable variable based on 1) an idea of how the unobservable variable might relate to an observable variable (the “measurement equation”), and 2) an idea of how the unobservable variable might change through time (the “transition equation”). Through a repeated process of simulating the unobserved variable based on a set of assumptions, the KF is able to compare predicted results to actual results on an observation-by-observation basis, and use that information to generate ever more reliable future estimates of the unobserved variable (Chart II-2). Chart II-2A Very Simplified Overview Of The Kalman Filter Algorithm April 2020 April 2020 We acknowledge that a full technical treatment of the Kalman Filter as it relates to the LW estimate of the neutral rate of interest is beyond the scope of this report, and we provide a more technical overview in Box II-1. But what emerges from a detailed analysis of the model is that the Kalman Filter jointly estimates R-star, potential GDP growth, potential GDP, and the variable “z”, the determinants of R-star that are not explained by potential GDP growth. As we will highlight in the next section, this joint estimation of these four variables is a crucial aspect of the model, because a valid estimate of R-star necessitates a valid estimate of the remaining variables. BOX II-1 A Technical Overview Of The Laubach & Williams R-star Model Chart Box II-1 shows that there are three sets of formulas involved in the LW estimation: the “law of motion” for the neutral rate of interest, two measurement equations, and three transition equations. The law of motion for the neutral rate is fairly simple: R-star is a function of trend real GDP growth, as well as “other factors” represented by the variable “z”. Laubach & Williams note that z “captures factors such as households’ rate of time preference”. The measurement equations are also fairly straightforward. First, the (unobservable) output gap is a function of lagged values of itself as well as the lagged real Fed funds rate gap (relative to the unobservable neutral rate). Second, inflation is a function of lagged values of itself, past values of the output gap, relative core import prices, and lagged relative imported oil prices (the latter two variables are included to capture potential supply shocks to inflation). Note that this second measurement equation is required for the model to work, as it relates the unobservable output gap to observable inflation. As presented in Chart II-2, the three transition equations are present to simulate how the unobservable variables might move through time. Potential growth and potential output are a random walk, and “z” from the law of motion follows either a random walk or an autoregressive process. Chart Box II-1The Laubach & Williams R-star Model April 2020 April 2020 Debunking The LW R-star Estimate Before criticizing the LW estimate of the neutral rate of interest, it is important for us to note that we have the utmost respect for the Federal Reserve and its research methods. We fully acknowledge that the LW R-star estimation is rooted in solid economic theory, and we have identified no technical errors in the setup of the LW model. Nevertheless, valid analytical efforts sometimes lead to problematic real-world results, and there are two key reasons to believe that the Kalman filter in the LW model is almost certainly misspecifying R-star, at least in terms of its estimate over the past two decades. The first reason relates to the sensitivity of the model to the interval of estimation (the period over which R-star is estimated). Chart II-3 presents the range of quarterly estimates of R-star since 2005, along with the difference between the high and low end of the range in the second panel. The chart shows that while previous estimates of R-star have generally been stable for values ranging between the early-1980s and 2006/2007, pre-1980 estimates have varied quite substantially and we have seen material revisions to the estimates over the past decade. Q1 2018 serves as an excellent example: in that quarter R-star was estimated to be 0.14%; today, the Q1 2018 R-star estimate sits at 0.92%. Chart II-3Since 2005, There Has Been Some Instability In The LW R-star Estimates Since 2005, There Has Been Some Instability In The LW R-star Estimates Since 2005, There Has Been Some Instability In The LW R-star Estimates However, Table II-1 and Chart II-4 highlight the real instability of the Kalman filter estimation by demonstrating the effect of varying the starting point of the model (please see Box II-2 for a brief description of how our estimation of R-star using the LW approach differs slightly from the original procedure). Laubach & Williams originally estimated R-star beginning in Q1 1961; Table II-1 shows what happens to today’s estimate of R-star simply by incrementally varying the starting point of the model from Q1 1958 to Q4 1979. Table II-1Alternative Current LW Estimates Of R-star By Model Starting Point April 2020 April 2020 Chart II-4Alternative Starting Points Produce Wildly Different Estimates Of R-star Today April 2020 April 2020 BOX II-2 The Laubach & Williams R-star Model With Simplified Inflation Expectations To proxy inflation expectations in their model, Laubach & Williams use a “forecast of the four-quarter-ahead percentage change in the price index for personal consumption expenditures excluding food and energy (“core PCE prices”) generated from a univariate AR(3) of inflation estimated over the prior 40 quarters”. The authors note that a simplified measure of expectations, a 4-quarter moving average of quarterly annualized core inflation, does not materially alter their results. For the sake of parsimony we use this simplified measure in our analysis. We find that the effect shifts the current estimate of R-star only slightly (+10 basis points), and that the historical differences between our version of the 1961 estimation and the official series are indeed minor. The table highlights that the model fails to even generate a result in a majority of the cases (only 39 out of 88 of the model runs were error-free). In addition, Chart II-4 shows that of the successful estimates of R-star using the LW procedure and alternate starting dates of the model, the estimate of R-star today varies from -2% (in one case) to +2%. Excluding the one extremely negative outlier results in an effective estimate range of 0% to 2%, but the key point for investors is that this range is massive and underscores that the original model’s estimate of R-star today is heavily and unduly influenced by the interval of estimation. Investors should also note that of all of the alternative estimates of R-star today shown in Chart II-4, the estimate using the original interval is very much on the low end of the distribution. The second (and most important) reason to believe that the LW estimate is misspecifying R-star is that the output gap estimate generated by the model is almost certainly invalid, at least over the past two decades. Chart II-5presents the LW output gap estimate alongside an average of the CBO, OECD, and IMF estimates of the gap; panel 1 shows the official current LW output gap estimate, whereas panel 2 shows the range of output gap estimates that are generated using the different estimation intervals highlighted in Table II-1 and Chart II-4. Chart II-5The LW Output Gap Estimates, Upon Which R-star Depends, Have Been Wrong For Two Decades The LW Output Gap Estimates, Upon Which R-star Depends, Have Been Wrong For Two Decades The LW Output Gap Estimates, Upon Which R-star Depends, Have Been Wrong For Two Decades Given that the Kalman filter in the LW model jointly determines R-star and the output gap (by way of estimating potential output via estimating potential GDP growth) and that these estimates are dependent on each other, Chart II-5 highlights that in order to believe the LW R-star estimate investors must believe three things: That the US economy was chronically below potential in the late-1990s when the unemployment rate was below 5%, real GDP growth averaged nearly 5%, and the equity market was booming, That output exceeded potential in 2004/2005 by a magnitude not seen since the late-1970s / early-1980s despite an average unemployment rate, That the 2008/2009 US recession was not particularly noteworthy in terms of its deviation from potential output, and that the economy had returned to potential output by 2010/2011 when the unemployment rate was in the range of 8-9%. Chart II-6The US Economy Was Definitely Not At Full Employment In 2010 The US Economy Was Definitely Not At Full Employment In 2010 The US Economy Was Definitely Not At Full Employment In 2010 While we do not believe any of these three statements, the third is especially unlikely. Chart II-6 highlights that the economic expansion from 2009 – 2020 was the weakest on record in the post-war era in terms of average annual real per capita GDP growth. To us, this is a clear symptom of a chronic deficiency in aggregate demand, and that it is essentially unreasonable to argue that the economy was operating at full employment prior to 2014/2015. This means that the Kalman filter is generating incorrect and unreliable estimates of the output gap, which means in turn that the filter’s estimation of R-star is almost assuredly wrong. How Can Investors Tell What The Neutral Rate Is? An Inferential Approach Table II-2 presents the sensitivity of the original Q1 1961 LW estimate of R-star to a series of counterfactual scenarios for inflation, real GDP growth, nominal interest rates, and import and oil prices since mid-2009. While these scenarios do not in any way improve the validity of the LW R-star estimate, they do help clarify the theoretical basis of the model and they help reveal how investors may infer whether the neutral rate of interest is higher or lower than prevailing market rates, and whether it is rising or falling. Table II-2Sensitivity Of Current LW R-star Estimate To Counterfactual Scenarios (2009 - Present) April 2020 April 2020 Chart II-7Core Import Price Growth Has Been Weak On Average During This Expansion Core Import Price Growth Has Been Weak On Average During This Expansion Core Import Price Growth Has Been Weak On Average During This Expansion Table II-2 highlights that today’s estimate of R-star using the original LW approach is mostly sensitive to our counterfactual scenarios for growth and interest rates, but not inflation or oil prices. Shifting down import price growth also has a meaningful effect on R-star, but since core import price growth has been particularly weak over the past several years (Chart II-7), it seems unreasonable to suggest that they have been abnormally high and thus “explain” a low R-star estimate today. Table II-2 essentially highlights that the entire question of the neutral rate of interest over the past decade, and the core contradiction that led to the re-emergence of the secular stagnation thesis, can effectively be boiled down to the following simple question: “Why hasn’t US economic growth been stronger this cycle, given that interest rates have been so low?” Based on the (hopefully uncontroversial) view that interest rates influence economic activity and that economic activity influences inflation, we propose the following checklist for investors to ask themselves in order to not only determine the answer to this important question, but to help identify whether R-star in any given country is likely higher or lower than existing policy rates at any given point in time. Are interest rates above or below the prevailing level of economic growth? Are interest rates rising or falling, and how intensely? Are there identifiable non-monetary shocks (positive or negative) that appear to be influencing economic activity? Is private sector credit growth keeping pace with economic growth? Are debt service burdens in the economy high or low? The first question reflects the most basic view of R-star, which is that the real neutral rate of interest should be equal to, or at least closely related to, the potential growth rate of the economy, ceteris paribus. Questions 2 through 5 attempt to determine whether ceteris paribus holds. In terms of how the answers to these questions relate to identifying the neutral rate, consider two economies, “Economy A” and “Economy B” (Chart II-8). Economy A has broadly stable or slightly rising interest rates that are well below prevailing rates of economic growth (questions 1 & 2), no obvious beneficial shocks to domestic demand from fiscal policy or other factors (question 3), and strong private sector credit growth that is perhaps above or strongly above the current pace of GDP growth (question 4). Chart II-8'Economy A', Versus 'Economy B' April 2020 April 2020 Inferentially, it would seem that interest rates in this hypothetical economy are below R-star today. Question 5 is in our list because the more that active private sector leveraging occurs (thus pushing up debt burdens), the more that we would expect R-star in the future to fall. This is because debt payments as a share of income cannot rise forever, and we would expect that the capacity of economy A’s central bank to raise interest rates in the future are negatively related to economy A’s private sector debt service burden today. Now, imagine another economy (“Economy B”) with interest rates well below average rates of economic growth, an interest rate trend that is flat-to-down, no identifiable non-monetary policy shocks that are restricting aggregate demand, persistently sluggish credit growth, and high private sector debt service burdens in the past. If economy B is growing (even sluggishly) and not in the middle of a recession, it would seem that prevailing interest rates are below R-star, but not significantly so. In this scenario it would seem reasonable to conclude that R-star in economy B has fallen non-trivially below its potential growth rate, and that interest rate increases are likely to move monetary policy into restrictive territory earlier than otherwise would be the case. Is The United States “Economy B”? From the perspective of some investors, our description of economy B above perfectly captures the experience of the US over the past decade: an extremely low Fed funds rate, sluggish to weak growth and inflation, all the result of a huge build-up in leverage and debt service burdens during the last economic cycle. We do not doubt that R-star fell in the US for some period of time during the global financial crisis and in the early phase of the economic recovery. But we doubt that it is as low today as the secular stagnation narrative would imply, in large part because it ignores several important aspects concerning questions 2 through 5 noted above. Chart II-9Fiscal Austerity Has Been A Serious Non-Monetary Shock To Aggregate Demand Fiscal Austerity Has Been A Serious Non-Monetary Shock To Aggregate Demand Fiscal Austerity Has Been A Serious Non-Monetary Shock To Aggregate Demand Non-monetary shocks to the US and global economies: Over the past 12 years, there have been at least five deeply impactful non-monetary shocks to both the US and global economies that have contributed to the disconnect between growth and interest rates: 1) a prolonged period of US household deleveraging from 2008-2014, 2) the euro area sovereign debt crisis, 3) fiscal austerity in the US, UK, and euro area from 2010 – 2012/2014 (Chart II-9), 4) the US dollar / oil price shock of 2014, and 5) the recent trade war between the US and China. Several of these shocks have been policy-driven, and in the case of austerity the negative consequences of that policy has led to a lasting change in thinking among fiscal authorities (outside of Japan) that is unlikely to reverse in the near-future. Chart II-10Recent Trends In US Private Sector Leverage Do Not Suggest R-star Is Very Low Recent Trends In US Private Sector Leverage Do Not Suggest R-star Is Very Low Recent Trends In US Private Sector Leverage Do Not Suggest R-star Is Very Low Private sector credit growth: Chart II-10 highlights the extent of household deleveraging noted above by showing the growth in total household liabilities over the past decade alongside income growth. Panel 2 shows the leveraging trend of firms, as represented by the nonfinancial corporate sector debt-to-GDP ratio. Chart II-10 underscores two points: the first is that while US household sector credit contracted for several years following the global financial crisis, it is now growing again and has largely closed the gap with income growth. The second point is that the nonfinancial corporate sector has clearly leveraged itself over the course of the expansion, arguing that interest rates have not in any way been restrictive for businesses. While it is true that firms have largely leveraged themselves to buy back stock instead of significantly increasing capital expenditures, in our view this reflects the fact that US consumer demand was impaired for several years due to deleveraging. We doubt that firms would have altered their capital structures to this degree if they did not view interest rates as extremely low. Debt service burdens: Chart II-11 highlights that US household debt service burdens were at very elevated levels prior to the financial crisis, suggesting that the neutral rate did fall for some time following the recession. But today, the debt burden facing households is the lowest it has been in the past 40 years due to both rate reductions and deleveraging, arguing against the view that household debt levels will structurally weigh on interest rates in the years to come. Chart II-12 shows that the picture is different for nonfinancial corporations, as the substantial leveraging noted above has indeed raised debt service burdens for firms. However, the nonfinancial corporate sector debt service ratio remains 400 basis points below early-2000 levels when excess corporate sector liabilities had a clear impact on the economy, suggesting that the Fed’s capacity to raise interest rates still exists following the onset of economic recovery if corporate sector credit growth does not rise sharply relative to GDP over the coming 6-12 months. Chart II-11The Debt Burden Facing US Households Is At A Record Low The Debt Burden Facing US Households Is At A Record Low The Debt Burden Facing US Households Is At A Record Low Chart II-12Businesses Have Levered Up Their Balance Sheets, But There Is Still Room For Rates To Rise Businesses Have Levered Up Their Balance Sheets, But There Is Still Room For Rates To Rise Businesses Have Levered Up Their Balance Sheets, But There Is Still Room For Rates To Rise   The intensity of recent interest rate changes: Finally, many investors have pointed to sluggish housing activity over the past three years as evidence of a low neutral rate. However, Chart II-13 highlights that the rise in the 30-year US mortgage rate from late-2016 to late-2018 was one of the largest two-year changes in US history, and Chart II-14 shows that the growth in household mortgage credit did not fall below its trend during this period until Q4 2018, when the US stock market fell 20% from its high in response to the economic consequences of the US/China trade war. Chart II-14 also shows that mortgage credit growth responded sharply to a recent reduction in interest rates. All in all, Charts II-13 & II-14 cast doubt on the notion that the level of mortgage rates over the past three years reached restrictive territory. Chart II-13Mortgage Rates Rose Very Significantly From Late-2016 To Late-2018 Mortgage Rates Rose Very Significantly From Late-2016 To Late-2018 Mortgage Rates Rose Very Significantly From Late-2016 To Late-2018 Chart II-14A Record Rise In Mortgage Rates Did Not Crack The Housing Market A Record Rise In Mortgage Rates Did Not Crack The Housing Market A Record Rise In Mortgage Rates Did Not Crack The Housing Market   Investment Conclusions In the face of a global pandemic and an attendant global recession this year, the idea of eventual Fed rate hikes and the notion that the US economy will be able to tolerate them likely seems preposterous to many investors. We agree that over the coming 6-12 months US Treasury yields are unlikely to rise; even at current levels of the 10-year Treasury yield, we are reluctant to call a trough. Chart II-15US 10-Year Treasurys Are Mostly Priced For A Repeat Of The Past Decade US 10-Year Treasurys Are Mostly Priced For A Repeat Of The Past Decade US 10-Year Treasurys Are Mostly Priced For A Repeat Of The Past Decade However, Chart II-15highlights that over a long-term time horizon, the bond market is now essentially priced for a repeat of the ten-year path of the Fed funds rate following the global financial crisis. While some investors will view this as a reasonable expectation in the face of what they see as a persistent and unexplainable gap between growth and interest rates over the past decade, we think this gap is explainable and we highly doubt that a pandemic with minimal mortality risk to the working age population and the young will cause the US economy to be afflicted with active consumer deleveraging lasting 4 to 6-years, substantial and wide-ranging fiscal austerity, persistently rising trade tariffs, and sharply lower oil prices. So while we agree that the US economy will be substantially cyclically affected by COVID-19, US Treasury yields reflect a “low rates forever” view with much higher certainty than is analytically warranted and thus appear to be anchored by a false narrative. As such, we caution fixed-income investors against dogmatic medium-to-longer term views about bond yields, as their potential to rise following the upcoming recession may be larger than many investors currently believe.   Jonathan LaBerge, CFA, Vice President Special Reports jonathanl@bcaresearch.com Footnotes 1  "IMF Fourteenth Annual Research Conference in Honor of Stanley Fischer," Washington DC, November 8, 2013. 2  "Measuring the Natural Rate of Interest," Federal Reserve Bank of New York.

Dear Client, This week, I provided an update through a webcast on the economic and financial market outlook in the era of the COVID-19 outbreak. You can access the webcast here. In lieu of our regular report this week, we are sending you a Special Report from my colleague Jonathan LaBerge. Jonathan shows why the most widely cited estimate of the US neutral rate of interest, the Laubach & Williams estimate of “R-star”, is very likely wrong and that the true neutral rate may be higher than many investors believe. While bond yields may not rise significantly in the near-term, this underscores that they have the potential to rise meaningfully over a cyclical and secular horizon once a post-COVID-19 expansion takes hold. I hope you find the report insightful. Please note that next week we will be publishing our quarterly Strategy Outlook. Best regards, Peter Berezin, Chief Global Strategist Highlights Global investors have come to accept the secular stagnation narrative as described by Larry Summers in November 2013, and have gravitated to the only available real time estimate of the real neutral rate of interest: the Laubach & Williams (“LW”) “R-star” estimate. With this apparent visualization of secular stagnation as a guide, many investors have concluded that monetary policy ceased to be stimulative last year and that recent Fed rate cuts will be of limited benefit to economic activity even once economic recovery takes hold unless inflation meaningfully accelerates (thus pushing real rates lower for any given nominal Fed funds rate). This report revisits the “LW” R-star estimate in detail, and demonstrates why the estimation is almost certainly wrong, at least over the past two decades. We also outline an inferential approach that investors can use to monitor where the neutral rate is in real time and whether it is rising or falling. The core conclusion for investors is that US Treasury yields reflect a “low rates forever” view with much higher certainty than is analytically warranted and thus appear to be anchored by a false narrative. While bond yields may not rise significantly in the near-term, investors should avoid dogmatic medium-to-longer term views about yields as they may rise meaningfully over a cyclical and secular horizon once a post-COVID-19 expansion takes hold. Feature Over the past several weeks financial markets have moved rapidly to price in a global recession stemming from the COVID-19 outbreak. As financial market participants began to turn to policy makers for support, eyes focused first on the Federal Reserve, and then fiscal authorities. Earlier this week, the ECB joined the party and announced aggressive further measures of its own. When responding to the Fed’s return to the lower bound and its other recent monetary policy decisions, many market participants have expressed the view that the Fed is largely impotent to deal with a global pandemic. There are three elements to this view. The first is that interest rate cuts are ill equipped to stimulate domestic demand if quarantine measures or other forms of “social distancing” are in effect. The second element is that the Fed has only been capable of delivering a fraction of the reduction in interest rates compared to what has occurred in response to previous contractions. The third aspect of this view is that because the neutral rate of interest is so much lower now than it was in the past, Fed rate cuts will not be as stimulative as they were before. Chart 1Monetary Policy Ceased To Be Stimulative Last Year, According To The LW R-star Estimate Monetary Policy Ceased To Be Stimulative Last Year, According To The LW R-star Estimate Monetary Policy Ceased To Be Stimulative Last Year, According To The LW R-star Estimate While we at least partly agree with the first and second elements of this view, we feel strongly that the third is flawed. Global investors have come to accept the secular stagnation narrative as described by Larry Summers in November 2013,1 and have gravitated to the only available real time estimate of the neutral rate of interest: the Laubach & Williams (“LW”) “R-star” estimate. This time series, which is regularly updated by the New York Fed,2 suggests that the real fed funds rate reached neutral territory in the first quarter of 2019 (Chart 1). With this apparent visualization of secular stagnation as a guide, many investors have concluded that monetary policy ceased to be stimulative last year and that recent Fed rate cuts will be of limited benefit to economic activity even beyond the near term unless inflation meaningfully accelerates (thus pushing real rates lower for any given nominal Fed funds rate). In this Special Report we revisit the “LW” R-star estimate in detail, and demonstrate why the estimation is almost certainly wrong, at least over the past two decades. Our analysis does not reveal a precise alternative estimate of the neutral rate, although we do provide some inferential perspective on how investors may be able to monitor where the neutral rate is in real time and whether it is rising or falling. However, the core insight emanating from our report, particularly for US fixed income investors, is that US Treasury yields reflect a “low rates forever” view with much higher certainty than is analytically warranted and thus appear to be anchored by a false narrative. While bond yields may not rise significantly in the near-term, this underscores that they have the potential to rise meaningfully over a cyclical and secular horizon once economic activity recovers. As such, we caution fixed-income investors against dogmatic medium-to-longer term views about bond yields, as their potential to rise may be larger than many investors currently expect. Demystifying The LW R-star Estimate The LW estimate of the neutral rate of interest has gained credibility for three reasons. First, as noted above, the evolution of the series fits with the secular stagnation narrative re-popularized by Larry Summers. Second, the series is essentially sponsored by the Federal Reserve even if it is not officially part of the Fed’s forecasting framework, as its two creators are long-time Fed employees (Thomas Laubach is a director of the Fed’s Board of Governors, and John Williams is the current President of the New York Fed). But, in our view, there is a third important reason that global investors have accepted the LW R-star estimate of the neutral rate of interest: the methodology used to generate the estimate is extremely technically complex, and thus is difficult for most investors to penetrate. Much of the technical complexity of the LW estimate is centered around the use of a statistical procedure called a Kalman filter (“KF”). Simply described, the KF is an algorithm that tries to estimate an unobservable variable based on 1) an idea of how the unobservable variable might relate to an observable variable (the “measurement equation”), and 2) an idea of how the unobservable variable might change through time (the “transition equation”). Through a repeated process of simulating the unobserved variable based on a set of assumptions, the KF is able to compare predicted results to actual results on an observation-by-observation basis, and use that information to generate ever more reliable future estimates of the unobserved variable (Chart 2). Chart 2A Very Simplified Overview Of The Kalman Filter Algorithm Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis We acknowledge that a full technical treatment of the Kalman Filter as it relates to the LW estimate of the neutral rate of interest is beyond the scope of this report, and we provide a more technical overview in Box 1. But what emerges from a detailed analysis of the model is that the Kalman Filter jointly estimates R-star, potential GDP growth, potential GDP, and the variable “z”, the determinants of R-star that are not explained by potential GDP growth. As we will highlight in the next section, this joint estimation of these four variables is a crucial aspect of the model, because a valid estimate of R-star necessitates a valid estimate of the remaining variables. Box 1 A Technical Overview Of The Laubach & Williams R-star Model Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Debunking The LW R-star Estimate Before criticizing the LW estimate of the neutral rate of interest, it is important for us to note that we have the utmost respect for the Federal Reserve and its research methods. We fully acknowledge that the LW R-star estimation is rooted in solid economic theory, and we have identified no technical errors in the setup of the LW model. Nevertheless, valid analytical efforts sometimes lead to problematic real-world results, and there are two key reasons to believe that the Kalman filter in the LW model is almost certainly misspecifying R-star, at least in terms of its estimate over the past two decades. The first reason relates to the sensitivity of the model to the interval of estimation (the period over which R-star is estimated). Chart 3 presents the range of quarterly estimates of R-star since 2005, along with the difference between the high and low end of the range in the second panel. The chart shows that while previous estimates of R-star have generally been stable for values ranging between the early-1980s and 2006/2007, pre-1980 estimates have varied quite substantially and we have seen material revisions to the estimates over the past decade. Q1 2018 serves as an excellent example: in that quarter R-star was estimated to be 0.14%; today, the Q1 2018 R-star estimate sits at 0.92%. Chart 3Since 2005, There Has Been Some Instability In The LW R-star Estimates Since 2005, There Has Been Some Instability In The LW R-star Estimates Since 2005, There Has Been Some Instability In The LW R-star Estimates   However, Table 1 and Chart 4 highlight the real instability of the Kalman filter estimation by demonstrating the effect of varying the starting point of the model (please see Box 2 for a brief description of how our estimation of R-star using the LW approach differs slightly from the original procedure). Laubach & Williams originally estimated R-star beginning in Q1 1961; Table 1 shows what happens to today’s estimate of R-star simply by incrementally varying the starting point of the model from Q1 1958 to Q4 1979. Table 1Alternative Current LW Estimates Of R-star By Model Starting Point Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Chart 4Alternative Starting Points Produce Wildly Different Estimates Of R-star Today Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Box 2 The Laubach & Williams R-star Model With Simplified Inflation Expectations Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis The table highlights that the model fails to even generate a result in a majority of the cases (only 39 out of 88 of the model runs were error-free). In addition, Chart 4 shows that of the successful estimates of R-star using the LW procedure and alternate starting dates of the model, the estimate of R-star today varies from -2% (in one case) to +2%. Excluding the one extremely negative outlier results in an effective estimate range of 0% to 2%, but the key point for investors is that this range is massive and underscores that the original model’s estimate of R-star today is heavily and unduly influenced by the interval of estimation. Investors should also note that of all of the alternative estimates of R-star today shown in Chart 4, the estimate using the original interval is very much on the low end of the distribution. The second (and most important) reason to believe that the LW estimate is misspecifying R-star is that the output gap estimate generated by the model is almost certainly invalid, at least over the past two decades. Chart 5 presents the LW output gap estimate alongside an average of the CBO, OECD, and IMF estimates of the gap; panel 1 shows the official current LW output gap estimate, whereas panel 2 shows the range of output gap estimates that are generated using the different estimation intervals highlighted in Table 1 and Chart 4. Given that the Kalman filter in the LW model jointly determines R-star and the output gap (by way of estimating potential output via estimating potential GDP growth) and that these estimates are dependent on each other, Chart 5 highlights that in order to believe the LW R-star estimate investors must believe three things: That the US economy was chronically below potential in the late-1990s when the unemployment rate was below 5%, real GDP growth averaged nearly 5%, and the equity market was booming, That output exceeded potential in 2004/2005 by a magnitude not seen since the late-1970s / early-1980s despite an average unemployment rate, That the 2008/2009 US recession was not particularly noteworthy in terms of its deviation from potential output, and that the economy had returned to potential output by 2010/2011 when the unemployment rate was in the range of 8-9%. Chart 5The LW Output Gap Estimates, Upon Which R-star Depends, Have Been Wrong For Two Decades The LW Output Gap Estimates, Upon Which R-star Depends, Have Been Wrong For Two Decades The LW Output Gap Estimates, Upon Which R-star Depends, Have Been Wrong For Two Decades While we do not believe any of these three statements, the third is especially unlikely. Chart 6 highlights that the economic expansion from 2009 – 2020 was the weakest on record in the post-war era in terms of average annual real per capita GDP growth. To us, this is a clear symptom of a chronic deficiency in aggregate demand, and that it is essentially unreasonable to argue that the economy was operating at full employment prior to 2014/2015. This means that the Kalman filter is generating incorrect and unreliable estimates of the output gap, which means in turn that the filter’s estimation of R-star is almost assuredly wrong. Chart 6The US Economy Was Definitely Not At Full Employment In 2010 Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis How Can Investors Tell What The Neutral Rate Is? An Inferential Approach Table 2 presents the sensitivity of the original Q1 1961 LW estimate of R-star to a series of counterfactual scenarios for inflation, real GDP growth, nominal interest rates, and import and oil prices since mid-2009. While these scenarios do not in any way improve the validity of the LW R-star estimate, they do help clarify the theoretical basis of the model and they help reveal how investors may infer whether the neutral rate of interest is higher or lower than prevailing market rates, and whether it is rising or falling. Table 2 highlights that today’s estimate of R-star using the original LW approach is mostly sensitive to our counterfactual scenarios for growth and interest rates, but not inflation or oil prices. Shifting down import price growth also has a meaningful effect on R-star, but since core import price growth has been particularly weak over the past several years (Chart 7), it seems unreasonable to suggest that they have been abnormally high and thus “explain” a low R-star estimate today. Table 2Sensitivity Of Current LW R-star Estimate To Counterfactual Scenarios (2009 - Present) Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Chart 7Core Import Price Growth Has Been Weak On Average During This Expansion Core Import Price Growth Has Been Weak On Average During This Expansion Core Import Price Growth Has Been Weak On Average During This Expansion Table 2 essentially highlights that the entire question of the neutral rate of interest over the past decade, and the core contradiction that led to the re-emergence of the secular stagnation thesis, can effectively be boiled down to the following simple question: “Why hasn’t US economic growth been stronger this cycle, given that interest rates have been so low?” Based on the (hopefully uncontroversial) view that interest rates influence economic activity and that economic activity influences inflation, we propose the following checklist for investors to ask themselves in order to not only determine the answer to this important question, but to help identify whether R-star in any given country is likely higher or lower than existing policy rates at any given point in time. Are interest rates above or below the prevailing level of economic growth? Are interest rates rising or falling, and how intensely? Are there identifiable non-monetary shocks (positive or negative) that appear to be influencing economic activity? Is private sector credit growth keeping pace with economic growth? Are debt service burdens in the economy high or low? The first question reflects the most basic view of R-star, which is that the real neutral rate of interest should be equal to, or at least closely related to, the potential growth rate of the economy, ceteris paribus. Questions 2 through 5 attempt to determine whether ceteris paribus holds. In terms of how the answers to these questions relate to identifying the neutral rate, consider two economies, “Economy A” and “Economy B” (Chart 8). Economy A has broadly stable or slightly rising interest rates that are well below prevailing rates of economic growth (questions 1 & 2), no obvious beneficial shocks to domestic demand from fiscal policy or other factors (question 3), and strong private sector credit growth that is perhaps above or strongly above the current pace of GDP growth (question 4). Chart 8''Economy A'', Versus ''Economy B'' Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Inferentially, it would seem that interest rates in this hypothetical economy are below R-star today. Question 5 is in our list because the more that active private sector leveraging occurs (thus pushing up debt burdens), the more that we would expect R-star in the future to fall. This is because debt payments as a share of income cannot rise forever, and we would expect that the capacity of economy A’s central bank to raise interest rates in the future are negatively related to economy A’s private sector debt service burden today. Now, imagine another economy (“Economy B”) with interest rates well below average rates of economic growth, an interest rate trend that is flat-to-down, no identifiable non-monetary policy shocks that are restricting aggregate demand, persistently sluggish credit growth, and high private sector debt service burdens in the past. If economy B is growing (even sluggishly) and not in the middle of a recession, it would seem that prevailing interest rates are below R-star, but not significantly so. In this scenario it would seem reasonable to conclude that R-star in economy B has fallen non-trivially below its potential growth rate, and that interest rate increases are likely to move monetary policy into restrictive territory earlier than otherwise would be the case. Is The United States “Economy B”? From the perspective of some investors, our description of economy B above perfectly captures the experience of the US over the past decade: an extremely low Fed funds rate, sluggish to weak growth and inflation, all the result of a huge build-up in leverage and debt service burdens during the last economic cycle. We do not doubt that R-star fell in the US for some period of time during the global financial crisis and in the early phase of the economic recovery. But we doubt that it is as low today as the secular stagnation narrative would imply, in large part because it ignores several important aspects concerning questions 2 through 5 noted above. Chart 9Fiscal Austerity Has Been A Serious Non- Monetary Shock To Aggregate Demand Fiscal Austerity Has Been A Serious Non-Monetary Shock To Aggregate Demand Fiscal Austerity Has Been A Serious Non-Monetary Shock To Aggregate Demand Non-monetary shocks to the US and global economies: Over the past 12 years, there have been at least five deeply impactful non-monetary shocks to both the US and global economies that have contributed to the disconnect between growth and interest rates: 1) a prolonged period of US household deleveraging from 2008-2014, 2) the euro area sovereign debt crisis, 3) fiscal austerity in the US, UK, and euro area from 2010 – 2012/2014 (Chart 9), 4) the US dollar / oil price shock of 2014, and 5) the recent trade war between the US and China. Several of these shocks have been policy-driven, and in the case of austerity the negative consequences of that policy has led to a lasting change in thinking among fiscal authorities (outside of Japan) that is unlikely to reverse in the near-future. Private sector credit growth: Chart 10 highlights the extent of household deleveraging noted above by showing the growth in total household liabilities over the past decade alongside income growth. Panel 2 shows the leveraging trend of firms, as represented by the nonfinancial corporate sector debt-to-GDP ratio. Chart 10 underscores two points: the first is that while US household sector credit contracted for several years following the global financial crisis, it is now growing again and has largely closed the gap with income growth. The second point is that the nonfinancial corporate sector has clearly leveraged itself over the course of the expansion, arguing that interest rates have not in any way been restrictive for businesses. While it is true that firms have largely leveraged themselves to buy back stock instead of significantly increasing capital expenditures, in our view this reflects the fact that US consumer demand was impaired for several years due to deleveraging. We doubt that firms would have altered their capital structures to this degree if they did not view interest rates as extremely low. Chart 10Recent Trends In US Private Sector Leverage Do Not Suggest R-star Is Very Low Recent Trends In US Private Sector Leverage Do Not Suggest R-star Is Very Low Recent Trends In US Private Sector Leverage Do Not Suggest R-star Is Very Low Debt service burdens: Chart 11 highlights that US household debt service burdens were at very elevated levels prior to the financial crisis, suggesting that the neutral rate did fall for some time following the recession. But today, the debt burden facing households is the lowest it has been in the past 40 years due to both rate reductions and deleveraging, arguing against the view that household debt levels will structurally weigh on interest rates in the years to come. Chart 12 shows that the picture is different for nonfinancial corporations, as the substantial leveraging noted above has indeed raised debt service burdens for firms. However, the nonfinancial corporate sector debt service ratio remains 400 basis points below early-2000 levels when excess corporate sector liabilities had a clear impact on the economy, suggesting that the Fed’s capacity to raise interest rates still exists following the onset of economic recovery if corporate sector credit growth does not rise sharply relative to GDP over the coming 6-12 months. Chart 11The Debt Burden Facing US Households Is At A Record Low The Debt Burden Facing US Households Is At A Record Low The Debt Burden Facing US Households Is At A Record Low Chart 12Businesses Have Levered Up Their Balance Sheets, But There Is Still Room For Rates To Rise Businesses Have Levered Up Their Balance Sheets, But There Is Still Room For Rates To Rise Businesses Have Levered Up Their Balance Sheets, But There Is Still Room For Rates To Rise The intensity of recent interest rate changes: Finally, many investors have pointed to sluggish housing activity over the past three years as evidence of a low neutral rate. However, Chart 13 highlights that the rise in the 30-year US mortgage rate from late-2016 to late-2018 was one of the largest two-year changes in US history, and Chart 14 shows that the growth in household mortgage credit did not fall below its trend during this period until Q4 2018, when the US stock market fell 20% from its high in response to the economic consequences of the US/China trade war. Chart 14 also shows that mortgage credit growth responded sharply to a recent reduction in interest rates. All in all, Charts 13 & 14 cast doubt on the notion that the level of mortgage rates over the past three years reached restrictive territory. Chart 13Mortgage Rates Rose Very Significantly From Late-2016 To Late-2018 Mortgage Rates Rose Very Significantly From Late-2016 To Late-2018 Mortgage Rates Rose Very Significantly From Late-2016 To Late-2018 Chart 14A Record Rise In Mortgage Rates Did Not Crack The Housing Market A Record Rise In Mortgage Rates Did Not Crack The Housing Market A Record Rise In Mortgage Rates Did Not Crack The Housing Market Investment Conclusions In the face of a global pandemic and an attendant global recession this year, the idea of eventual Fed rate hikes and the notion that the US economy will be able to tolerate them likely seems preposterous to many investors. We agree that over the coming 6-12 months US Treasury yields are unlikely to rise; even at current levels of the 10-year Treasury yield, we are reluctant to call a trough. Chart 15US 10-Year Treasurys Are Mostly Priced For A Repeat Of The Past Decade US 10-Year Treasurys Are Mostly Priced For A Repeat Of The Past Decade US 10-Year Treasurys Are Mostly Priced For A Repeat Of The Past Decade However, Chart 15 highlights that over a long-term time horizon, the bond market is now essentially priced for a repeat of the ten-year path of the Fed funds rate following the global financial crisis. While some investors will view this as a reasonable expectation in the face of what they see as a persistent and unexplainable gap between growth and interest rates over the past decade, we think this gap is explainable and we highly doubt that a pandemic with minimal mortality risk to the working age population and the young will cause the US economy to be afflicted with active consumer deleveraging lasting 4 to 6-years, substantial and wide-ranging fiscal austerity, persistently rising trade tariffs, and sharply lower oil prices. So while we agree that the US economy will be substantially cyclically affected by COVID-19, US Treasury yields reflect a “low rates forever” view with much higher certainty than is analytically warranted and thus appear to be anchored by a false narrative. As such, we caution fixed-income investors against dogmatic medium-to-longer term views about bond yields, as their potential to rise following the upcoming recession may be larger than many investors currently believe.   Jonathan LaBerge, CFA, Vice President Special Reports jonathanl@bcaresearch.com   Footnotes 1    "IMF Fourteenth Annual Research Conference in Honor of Stanley Fischer," Washington DC, November 8, 2013. 2    "Measuring the Natural Rate of Interest," Federal Reserve Bank of New York. Global Investment Strategy View Matrix Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis Revisiting The Neutral Rate Of Interest: A Contrarian View In A Time Of Crisis

Highlights Since 2004, Sweden’s private sector leverage trend can be explained using a simple Taylor rule approach. The approach clearly highlights three distinct monetary policy phases, and underscores the singular role of inflation (not systemic risk from rising indebtedness) as a driving factor for Riksbank policy. Since 2015, the Riksbank has maintained interest rates well below what a Taylor rule approach would suggest, owing to the desire to raise inflation expectations and Sweden’s high trade exposure to the euro area. This highlights strong similarities between the experience of Sweden and Canada: both countries are in the orbit of a major neighboring central bank, which has created serious distortions in both economies. Given the extent of the spread of the SARS-CoV-2 virus, especially in Europe, our assessment of the Riksbank’s reaction function suggests the odds appear to be high that the repo rate will move back into negative territory at some point this year (despite their reluctance to do so). Over the near-term, Swedish policy easing suggests that investors should avoid the krona versus both the US dollar and euro. Over a medium-term time horizon, one implication of a return to negative interest rates is that Swedish house price appreciation is likely to trend higher once the economic impact of the COVID-19 pandemic ends, potentially to the benefit of Swedish consumer durable and apparel stocks. Finally, over the long-term, Sweden is very likely to face a period of domestic economic stagnation stemming from the extraordinary rise in private sector debt that has built up over the past two decades. The co-ordinated global response to the pandemic suggests that this is not the end of Sweden’s debt supercycle, but timing the transition from reflation to stagnation will be of crucial importance for investors exposed to the domestic Swedish economy over the coming few years. Feature One of the worrying legacies of the global financial crisis has been a substantial buildup in private sector debt in many economies around the world. This has most famously occurred in China, but private indebtedness is also very high in many developed economies. Among advanced countries, Sweden stands out as being particularly exposed to elevated private sector debt. Chart I-1 highlights that Sweden’s private sector debt-to-GDP ratio has ballooned to a massive 250% of GDP over the past 15 years, from a starting point of roughly average indebtedness. Chart I-1Sweden's Extremely Indebted Private Sector Sweden's Extremely Indebted Private Sector Sweden's Extremely Indebted Private Sector In this report we explore why Sweden has seen an explosion in private sector debt-to-GDP, and highlight that Sweden’s experience can be compared closely with that of Canada – both countries are in the orbit of a major neighboring central bank, which has created distortions in each economy. We also summarize what this implies for Riksbank policy, and what investment recommendations can be drawn from our analysis. We conclude that while the Riksbank is clearly reluctant to cut the repo rate after having just existed its negative interest rate position last year, it appears likely that they will forced to do so unless the negative economic impact from the COVID-19 pandemic abates very soon. Over the short-term, this suggests that investors should avoid the Swedish krona, versus either the US dollar or the euro. Why has Sweden seen such an explosion in private-sector debt? Over the medium-term, easy Riksbank policy and the probable absence of any additional macroprudential measures is likely to spur a renewed increase in Swedish house prices and household debt, which will likely benefit consumer durables and apparel stocks relative to the broad Swedish equity market.  But this will reinforce Sweden’s existing credit bubble, and similar to Canada will set the stage for domestic economic stagnation over the very long-term. Riksbank Policy and Sweden’s Private Sector Debt: A Tale Of Three Phases Much of the investor attention on Sweden's extremely high private sector debt load has occurred following the global financial crisis. But Chart I-1 clearly highlights that the process of private sector leveraging began in 2004, arguing that the Riksbank’s easy monetary policy stance following the global financial crisis is not the only cause of Sweden’s extremely elevated private debt-to-GDP ratio. In a previous Special Report for our Global Investment Strategy service,1 we investigated a similar experience in Canada and used a simple Taylor rule approach to show that the Bank of Canada’s decision to maintain interest rates below equilibrium levels for nearly two decades has contributed to a substantial buildup in private sector leverage. A similar approach for Sweden highlights similar conclusions, albeit with some complications: Chart I-2 shows our Taylor rule estimate for Sweden alongside the policy rate, and shows the deviation from the rule in the second panel. Chart I-2Since 2000, Sweden Has Had Three Distinct Monetary Policy Phases Since 2000, Sweden Has Had Three Distinct Monetary Policy Phases Since 2000, Sweden Has Had Three Distinct Monetary Policy Phases Compared with Canada’s experience, which has maintained too-low interest rates consistently for the past 20 years, Chart I-2 shows that the stance of Sweden’s monetary policy since 2000 falls into three distinct phases: Persistently easy policy from 2000 to 2008 A period of less easy and then relatively tight policy from 2009 to early-2014 A period of extremely easy policy from 2015 until today. The first phase noted above closely resembles the experience of Canada: policymakers in both countries simply kept interest rates too low during the last global economic expansion. In the second phase, the stance of monetary policy in Sweden became progressively less easy: the Taylor rule collapsed in 2009/2010, and trended lower again during the euro area sovereign debt crisis as well as its aftermath. In fact, Chart I-2 suggests that Sweden’s monetary policy stance was outrightly tight from 2012-2014, and in early-2014 the Taylor rule recommended negative policy rates while the actual policy rate was above 1%. In the third phase, the Riksbank appears to have overcompensated for the second phase of relatively less easy and eventually tight monetary policy. The Riksbank pushed policy rates into negative territory in late-2014, as had been recommended by the Taylor rule a year before, at a time when the rule was rising sharply. Roughly 2/3rds of the rise in the rule from early-2014 to late-2018 occurred due to the significant rise in Swedish inflation, with the rest due to a rise in Sweden’s output gap – which turned positive in late-2016 according to the OECD (Chart I-3). It is this third phase, featuring a massive and glaring gap between Swedish policy rates and a monetary policy rule that correctly recommended easy policy from 2010 – 2014, that has attracted global investor attention over the past few years. But Chart I-4 presents Sweden’s Taylor rule gap alongside its private sector debt-to-GDP ratio, and highlights that over 80% of the rise in the latter since 2000 actually occurred in the first phase described above – a period of persistently easy monetary policy as defined by our Taylor rule approach. The behavior of Sweden’s private sector debt-to-GDP ratio in the second and third phases also seems to validate our approach, as gearing essentially stopped during the second phase and restarted in the third phase. Chart I-3Since 2014, Sweden’s Rising Taylor Rule Has Been Driven Mostly By Inflation Since 2014, Sweden's Rising Taylor Rule Has Been Driven Mostly By Inflation Since 2014, Sweden's Rising Taylor Rule Has Been Driven Mostly By Inflation Chart I-4Sweden’s Monetary Policy Phases Explain Its Private Sector Leveraging Sweden's Monetary Policy Phases Explain Its Private Sector Leveraging Sweden's Monetary Policy Phases Explain Its Private Sector Leveraging The Riksbank: “Talk To Us About Inflation, Not Debt” Chart I-5During Phase 2, Households Clearly Took Advantage Of Low Mortgage Rates During Phase 2, Households Clearly Took Advantage Of Low Mortgage Rates During Phase 2, Households Clearly Took Advantage Of Low Mortgage Rates It is crucial to understand the motivations of Sweden’s central bank during each of these phases in order to be able to forecast the likelihood of a return to negative interest rates this year, as well as the Riksbank’s likely policy response once the COVID-19 pandemic subsides. In the first monetary policy phase that we have described, Sweden was not the only country to maintain persistently easy monetary policy. Given the relative scarcity of private sector deleveraging events in the post-war era, most policy makers, academic economists, and market participants were regrettably unconcerned about rising private sector indebtedness during this period, and only came to understand the consequences during the crisis and its aftermath. Most advanced economies leveraged during the first of Sweden’s monetary policy phases, and Sweden really only stands out as a major outlier from 2007 – 2009 when nearly 60% of the country’s total 2000-2019 private sector leveraging occurred (most of which, in turn, occurred before the collapse of Lehman Brothers in September 2008). In essence, by the time that Swedish policymakers were given a vivid and painful demonstration of the dangers of elevated private sector debt, it was too late to prevent most of the increase in debt-to-GDP that is facing the country today. In the second phase of Sweden's modern monetary policy, our Taylor rule framework highlights that the Riksbank largely acted as appropriate. One complication, however, is the difference in the leverage trend between Sweden's nonfinancial corporate and household sectors. Chart I-5 clearly highlights that Sweden's household sector took advantage of low interest rates during the country’s second monetary policy phase. Household sector leveraging began to rise again starting in late-2011, whereas it was completely absent for the corporate sector during the period. A crucial reason why the Riksbank ignored this renewed household sector leveraging is also part of the reason that it has maintained extremely low policy rates in the third phase noted above. The Riksbank’s monetary policy strategy, which is published in every monetary policy report, includes the following: “According to the Sveriges Riksbank Act, the Riksbank’s tasks also include promoting a safe and efficient payment system. Risks linked to developments in the financial markets are taken into account in the monetary decisions. With regard to preventing an unbalanced development of asset prices and indebtedness however, well-functioning regulation and effective supervision play a central role. Monetary policy only acts as a compliment to these.” In other words, the Riksbank has been very clear that preventing excessive leveraging is not its responsibility, and that the job ultimately falls to the Swedish government. But if the Taylor rule was recommending meaningfully higher interest rates during phase 3, then why did the Riksbank continue to lower interest rates into negative territory until last year? In our view, their behavior can be explained by the confluence of three factors: 1. Sweden’s deflation scare in 2014: Sweden’s underlying inflation rate had been trending lower for four years by the time that it dipped briefly into negative territory in March 2014. By this point, the Riksbank appears to have become increasingly concerned about inflation expectations rather than the trend in actual inflation. Chart I-6 presents Sweden’s underlying inflation rate and an adaptive-expectations based estimate of inflation expectations alongside the repo rate, and shows that inflection points in the repo rate match inflection points in expectations. Specifically, the repo rate continued to fall until inflation expectations stabilized in early-2016, and the Riksbank did not raise the repo rate until expectations crossed above 1.5%, a level that was reasonably close to the central bank’s 2% target. Chart I-6During Phase 3, The Riksbank Focused On Low Inflation Expectations During Phase 3, The Riksbank Focused On Low Inflation Expectations During Phase 3, The Riksbank Focused On Low Inflation Expectations 2. Sweden’s high trade sensitivity: Chart I-7 highlights that Sweden’s economy, like Canada and other Scandinavian countries, is highly exposed to exports to top trading partners. The euro area accounts for a large portion of Sweden’s exports, and Chart I-8 highlights that nominal euro area imports from Sweden remained very weak from 2012-2016. In addition, Sweden’s import sensitivity is also very high, with total imports of goods and services accounting for over 40% of Sweden’s GDP. By our calculations, roughly 2/3rds of Swedish imports are for domestic consumption,2 and Chart I-9 highlights how closely (inversely) correlated imported consumer and capital goods prices are to Sweden’s trade-weighted currency index. By pushing the repo rate into negative territory, the Riksbank reinforced rising inflation expectations by supporting exports and importing inflation from its trading partners via a weaker krona. Chart I-7Sweden, Like Other Small DM Countries, Are Highly Exposed To Trade Sweden, Like Other Small DM Countries, Are Highly Exposed To Trade Sweden, Like Other Small DM Countries, Are Highly Exposed To Trade Chart I-8Euro Area Demand For Swedish Goods Remained Weak For Several Years Euro Area Demand For Swedish Goods Remained Weak For Several Years Euro Area Demand For Swedish Goods Remained Weak For Several Years Chart I-9To 'Import' Inflation, The Riksbank Had To Weaken The Krona To 'Import' Inflation, The Riksbank Had To Weaken The Krona To 'Import' Inflation, The Riksbank Had To Weaken The Krona 3. The euro area’s persistently weak inflation and extremely easy monetary policy: While this is related to Sweden's overall trade sensitivity, the fact that the euro area had to combat persistently weak inflation with negative interest rates and asset purchases from late-2014 to late-2018 has had a particularly strong impact on Riksbank policy given the latter’s goal of boosting Swedish inflation via higher import prices. Chart I-10 highlights the strong link between the SEK-EUR exchange rate and the real interest rate differential between the two countries, and in particular shows that the Riksbank had to lower the differential into negative territory in order to bring the krona below “normal” levels (defined here as the average of the past global economic expansion). When faced with a real euro area policy rate of roughly -1.5% during the period (Chart I-11), the only way to achieve a negative real rate differential was to maintain the repo rate at an extremely low level as Swedish inflation rose. Chart I-10To Weaken The ##br##Krona... To Weaken The Krona... To Weaken The Krona... Chart I-11…Deeply Negative Real Policy Rates Were Required ...Deeply Negative Real Policy Rates Were Required ...Deeply Negative Real Policy Rates Were Required Where Next For The Repo Rate? In February 2019 the Riksbank was forecasting that the repo rate would return into positive territory by the end of this year, and would rise as high as 80 basis points by mid-2022. They downgraded this assessment in April, and again in October, highlighting that they expected a 0% repo rate for essentially the entire three-year forecast period. In other words, the Riksbank had been moving in a dovish direction even before the COVID-19 pandemic began. Prior to the outbreak, we would have been inclined to argue that the Riksbank’s forecast of a 0% repo rate beyond 2020 was suspect, given the budding recovery in global growth. Chart I-12 highlights that the global PMI had been improving for several months prior to the outbreak, and the Swedish PMI and consumer confidence index had recently rebounded sharply. A negative repo rate was essential to “import” inflation. But, given the extent of the spread of the SARS-CoV-2 virus, especially in Europe, and our description of the Riksbank mandate and reaction function, the odds appear to be high that the repo rate will move back into negative territory at some point this year. Besides the very negative direct impact to global trade from the pandemic, Chart I-13 highlights that Swedish inflation is now falling, and that our measure of inflation expectations has now peaked. Chart I-12Swedish Economic Momentum Was Building Prior To The Pandemic... Swedish Economic Momentum Was Building Prior To The Pandemic... Swedish Economic Momentum Was Building Prior To The Pandemic... Char I-13...But Inflation Is Falling And The Unemployment Rate Is Rising ...But Inflation Is Falling And The Unemployment Rate Is Rising ...But Inflation Is Falling And The Unemployment Rate Is Rising In addition, the Swedish unemployment rate has been trending higher since early-2018 (Chart I-13, second panel), in response to several factors: a shock to household wealth in late-2015/early-2016 due to sharply falling equity prices, a meaningful decline in house prices driven by newly introduced macroprudential policies, and a sharp albeit seemingly one-off decline in the contribution to Swedish economic growth from government expenditure (Chart I-14). These trends would have likely reversed at some point this year given the building economic momentum that was evident in January and early-February, but it is now clear that the pandemic will more than offset the budding improvement in economic activity. Chart I-14Swedish Policymakers Will Have To Reverse The Factors That Caused The Pre-Pandemic Slowdow Swedish Policymakers Will Have To Reverse The Factors That Caused The Pre-Pandemic Slowdow Swedish Policymakers Will Have To Reverse The Factors That Caused The Pre-Pandemic Slowdow Over the past week the Riksbank has announced two policies: it will provide cheap loans to the country’s banks (500 billion SEK) to bolster credit supply to Swedish small & medium-sized enterprises, and it will increase its asset purchase program by 300 billion SEK. The Riksbank is clearly reluctant to cut the repo rate after having just existed its negative interest rate position last year, and has argued that strong liquidity support and stepped up asset purchases are more likely to be effective measures in the current environment. However, Charts I-10 & I-11 underscored the link between real interest rate differentials and the currency, and the Riksbank will risk having the krona appreciate versus the euro and other currencies if inflation continues to fall and the policy rate is kept unchanged. Chart I-15 shows that market participants have already begun to price in cuts to the repo rate, and our sense is that the Riksbank will be forced to act in a way that is consistent with the market’s view. Chart I-15The Market Expects The Riksbank To Return To Negative Interest Rates. We Agree. The Market Expects The Riksbank To Return To Negative Interest Rates. We Agree. The Market Expects The Riksbank To Return To Negative Interest Rates. We Agree. Investment Conclusions Over a cyclical (i.e. 6-12 month) time horizon, the Swedish krona is the asset with the clearest link to our discussion of Riksbank policy, and investors should recognize that the krona call is now a binary one based on the evolution of the COVID-19 pandemic. It is one of the cheapest currencies in the G10 space, but foreign exchange markets have recently ignored fundamentals such as interest rate differentials and valuation. This is particularly true in the face of a spike in US dollar cross-currency basis swaps, which have started to send the dollar higher even against the safe haven currencies. In such an a environment, selling pressure could continue to push SEK lower, especially if the Riksbank is pushed to reduce the repo rate sooner rather than later. The SEK is one of the most procyclical currencies in the FX space, suggesting that investors should stand aside until markets stabilize (Chart I-16). Right now, the Swedish krona is the clearest play on Riksbank policy. As for the EUR/SEK cross, any renewed ECB stimulus suggests that Sweden will act accordingly to prevent the SEK from appreciating too far, too fast. EUR/SEK will top out after global growth is in an eventual upswing, and the Riskbank has eased policy further. Over the medium-term time horizon, one implication of a return to negative interest rates is that Swedish house price appreciation is likely to trend higher once the economic impact of the COVID-19 pandemic ends. House prices will likely decelerate in the near term given the shock to household wealth from falling equity prices, but we showed in Chart I-5 that Sweden’s household sector ultimately took advantage of low interest rates during Sweden’s second monetary policy phase. We expect a similar dynamic to unfold beyond the coming 6-9 months, and Chart I-17 highlights that overweighting Swedish consumer durable and apparel stocks within the overall Swedish equity market is likely the best way to eventually play a resumption of household leveraging and rising house prices. Chart I-16Avoid Krona Exposure ##br##For Now Avoid Krona Exposure For Now Avoid Krona Exposure For Now Chart I-17Swedish Consumer Durables & Apparel Stocks Linked To Domestic, Not Global, Demand Swedish Consumer Durables & Apparel Stocks Linked To Domestic, Not Global, Demand Swedish Consumer Durables & Apparel Stocks Linked To Domestic, Not Global, Demand With the exception of a selloff in 2013, the relative performance of the industry group has closely correlated with house price appreciation, and is now deeply oversold. The companies included the industry group earn a significant portion of their revenue from global sales, but the close correlation of relative performance with Swedish house prices and limited correlation with the global PMI suggests that domestic economic performance matter in driving returns for these stocks (Chart I-17, bottom panel). We are not yet prepared to recommend a long relative position favoring this industry group, but we are likely to view signs of policy traction and a relative performance breakout as a good entry point. Finally, the key long-term implication of our research is that Sweden will at some point likely face a period of stagnation stemming from the extraordinary rise in private sector debt that has built up over the past two decades. While regulators had begun to combat excessive debt with macroprudential measures, further measures to restrict household sector debt are extremely unlikely to occur until after another substantial reacceleration in Swedish house prices and another nontrivial rise in household sector leverage. This will be cyclically positive for Sweden coming out of the pandemic, but will ultimately make Sweden’s underlying debt problem meaningfully worse. Macroprudential control of rising nonfinancial corporate debt has not and is not likely to occur, and no regulatory control measure will be able to significantly ease the existing debt burden facing the private sector. Chart I-18 highlights that while Sweden’s private sector debt service ratio (DSR) is not the highest in the world, is it extremely elevated compared to other important DM countries such as the US, UK, Japan, and core euro area. Several other countries with higher private sector DSRs, such as Canada and Hong Kong, are also at serious risk of long-term stagnation. Chart I-18Swedish Domestic Economic Stagnation Is A 'When', Not An 'If' Swedish Domestic Economic Stagnation Is A 'When', Not An 'If' Swedish Domestic Economic Stagnation Is A 'When', Not An 'If' We have not yet identified a specific list of assets that will be negatively impacted by Swedish domestic economic stagnation over the longer term. Our European Investment Strategy service recently argued that Swedish stocks are attractive over the very long term versus Swedish bonds, based on valuations and the fact that the Swedish equity market as a whole is heavily driven by the global business cycle. We plan on revisiting the question of which equity sectors are most vulnerable to domestic stagnation in a future report, as the onset of stagnation draws nearer. As we noted in our report on Canada,3 it is difficult to identify precisely when Sweden’s high debt load will meaningfully and sustainably impact Swedish economic activity and related equity sectors. The acute shock to global activity from the COVID-19 pandemic is an obvious potential trigger, but the fact that policymakers around the world are responding forcefully to the pandemic suggests that this is not the end of Sweden’s debt supercycle. In this regard, the prospect of globally co-ordinated fiscal spending is especially significant. Our best guess is that Sweden’s true reckoning will come once US and global activity contracts for conventional reasons, instigated by tight monetary policy to control rising and above-target inflation. This may mean that Sweden will avoid a balance sheet recession for some time, but investors exposed to domestically-linked Swedish financial assets should take heed that the eventual consequences of such an event are likely to grow in magnitude the longer it takes to arrive. In short, beyond the acute nearer-term impact of the pandemic, Sweden is likely to experience short-term gain for long-term pain. The short- to medium-term focus of investors should be on the former, but with full recognition that the latter will eventually occur. Timing the transition between these two states will be of crucial importance for investors exposed to the domestic Swedish economy over the coming few years. Stay tuned!   Jonathan LaBerge, CFA, Vice President Special Reports jonathanl@bcaresearch.com Footnotes 1 Please see Global Investment Strategy Special Report "Canada: A (Probably) Happy Moment In An Otherwise Sad Story," dated July 14, 2017, available at uses.bcaresearch.com 2 We assume that all services imports are consumed domestically. Among goods exports, we assume domestic consumption of all imports of food & live animals, beverages & tobacco, mineral fuels, lubricants, and related materials, miscellaneous manufactured articles, road vehicles, and other goods. 3 Please see Global Investment Strategy Special Report "Canada: A (Probably) Happy Moment In An Otherwise Sad Story," dated July 14, 2017, available at uses.bcaresearch.com

Highlights Malaysian businesses and households have been deleveraging and the economy risks entering a debt deflation spiral. This macro-backdrop is bond bullish. EM fixed income-dedicated investors should keep an overweight position in both local currency and US dollar government bonds. In Peru, the central bank does not want its currency to depreciate rapidly; it will therefore defend the sol at the cost of slower economic growth. The outperformance of the Peruvian sol heralds an overweight stance in domestic and US dollar government bonds versus EM peers. Malaysia: In Deleveraging Mode Malaysian businesses and households have been deleveraging. The top panel of Chart I-1 illustrates that commercial banks’ domestic claims on the private sector – both companies and households – relative to nominal GDP have been flat to down in recent years. This measure is produced by the central bank and includes both bank loans as well as securities held by banks (Chart I-1, bottom panel). It does not include borrowing from non-banks or external borrowing. Other measures of indebtedness from the Bank of International Settlements (BIS) – which includes non-bank credit as well as foreign currency borrowing – portend similar dynamics: Household and corporate debt seem to have topped out as a share of GDP (Chart I-2). Chart I-1Malaysian Banks' Claims On The Private Sector Have Rolled Over Malaysian Banks' Claims On The Private Sector Have Rolled Over Malaysian Banks' Claims On The Private Sector Have Rolled Over Chart I-2Malaysia's Business And Household Total Leverage Has Peaked Malaysia's Business And Household Total Leverage Has Peaked Malaysia's Business And Household Total Leverage Has Peaked   Chart I-3Malaysia: The GDP Deflator Is About To Turn Negative Malaysia: The GDP Deflator Is About To Turn Negative Malaysia: The GDP Deflator Is About To Turn Negative The message is that after years of an unrelenting credit boom, households’ and companies’ appetite for new borrowing has diminished, and at the same time, creditors have become less willing to finance them.  At 136% of GDP, the combined total of household and company debt is non-trivial. If deleveraging among debtors intensifies, the economy risks entering a debt deflation spiral. To prevent such an ominous outcome, aggressive central bank rate cuts, sizable fiscal stimulus, some currency devaluation or a combination of all of the above is required. Not only is real growth very sluggish in Malaysia, but deflationary pressures are intensifying. Chart I-3 shows the GDP deflator is flirting with contraction. Moreover, headline and core consumer price inflation are both weak, while trimmed-mean inflation is at 1.1% (Chart I-4). Last year's spike in consumer inflation was due to low base effects from the abolishment of the country’s goods and services tax back in June 2018. Going forward, these base effects will dissipate, making deflation in consumer prices a likely threat. If prices or wages begin deflating, the highly-indebted Malaysian economy will fall into debt deflation. The latter is a phenomenon that occurs when falling level of prices and wages cause the real value of debt to rise. In such a case, demand for credit will plummet and banks could become unwilling to lend. A vicious cycle of further falling prices, income and credit retrenchment could grip the economy. Household and corporate debt seem to have topped out as a share of GDP. Nominal GDP growth has already dropped slightly below average lending rates (Chart I-5). When such a phenomenon occurs amid elevated debt levels, it can produce a lethal cocktail – namely, the debt-servicing ability of borrowers deteriorates, causing both demand for credit to evaporate and non-performing loans (NPLs) to rise. Chart I-4Malaysia: Consumer Price Inflation Is Very Low Malaysia: Consumer Price Inflation Is Very Low Malaysia: Consumer Price Inflation Is Very Low Chart I-5Malaysia: Nominal GDP Growth Dipped Below Lending Rates Malaysia: Nominal GDP Growth Dipped Below Lending Rates Malaysia: Nominal GDP Growth Dipped Below Lending Rates   Critically, falling inflation has caused real borrowing costs to rise. Lending rates in real terms are elevated, from a historical perspective (Chart I-6, top panel).1 Not surprisingly, loan growth has been decelerating sharply, posting a 13-year low (Chart I-6, bottom panel). Even though government expenditure growth has been accelerating over the past year or so and the central bank has cut interest rates twice in the past 8 months, economic conditions remain extremely feeble: Consumer spending has been teetering. Chart I-7 shows that retail sales are dwindling in nominal terms and have plummeted in volume terms. Chart I-6Malaysia: Real Lending Rates Have Risen & Credit Has Slowed Malaysia: Real Lending Rates Have Risen & Credit Has Slowed Malaysia: Real Lending Rates Have Risen & Credit Has Slowed Chart I-7Malaysia: Consumer Spending Is Teetering Malaysia: Consumer Spending Is Teetering Malaysia: Consumer Spending Is Teetering   Malaysian exports – which account for a 67% share of the economy – are still contracting 2.5% from a year ago, adding an additional unwelcome layer of deflation to the Malaysian economy. After years of travails, the property sector is not yet out of the woods. Residential property unit sales remain sluggish (Chart I-8, top panel). In turn, the number of unsold residential properties remains elevated and residential construction approvals are rolling over at lower levels (Chart I-8, second & third panels). As a result, residential property prices are beginning to deflate across various segments in nominal terms (Chart I-8, bottom panel). Listed companies’ earnings-per-share (EPS) in local currency terms are contracting (Chart I-9, top panel). Chart I-8Malaysia's Residential Property Market Is Struggling Malaysia's Residential Property Market Is Struggling Malaysia's Residential Property Market Is Struggling Chart I-9Malaysia: Capital Spending Is Contracting Malaysia: Capital Spending Is Contracting Malaysia: Capital Spending Is Contracting Chart I-10Malaysia: Weak Employment Outlook Malaysia: Weak Employment Outlook Malaysia: Weak Employment Outlook All of these ominous trends have induced Malaysian businesses to cut capital spending. The bottom three panels of Chart I-9 illustrate that real gross capital goods formation, capital goods imports and commercial vehicles units sales are all contracting. Equally important, the business sector slowdown is weighing on the employment outlook (Chart I-10). This will trigger a negative feedback loop of falling household income and spending. Bottom Line: Only by bringing borrowing costs down considerably for households and businesses and introducing large fiscal stimulus measures, can the Malaysian authorities prevent the economy from slipping into a vicious debt deflation spiral. On the fiscal front, the Malaysian government is committed to reducing its overall fiscal deficit from 3.4% to 3.2% of GDP this year, further consolidating it to 2.8% of GDP by 2021. Importantly, the government is also adamant about lowering its total public debt-to-GDP ratio from 77% to below 50% in the medium term by ridding itself of the outstanding legacy liabilities and guarantees incurred by the previous government. This leaves monetary policy and some currency depreciation as the likely levers to reflate the economy. Investment Recommendations We continue to recommend EM fixed -income dedicated investors keep an overweight position in local currency bonds within an EM local currency bonds portfolio. Malaysia’s macro-backdrop is bond bullish, and the central bank will cut its policy rate further. Consumer spending has been teetering. Consistent with further rate cut expectations, we also recommend continuing to receive 2-year swap rates. We initiated this trade on October 31, 2019, and it has so far produced a profit of 29 basis points. Furthermore, fiscal discipline and the government’s resolve to reduce public debt and government liabilities as a share of GDP will help Malaysian sovereign credit – US dollar-denominated government bonds – outperform their EM peers. Chart I-11The Malaysian Ringgit Is Cheap The Malaysian Ringgit Is Cheap The Malaysian Ringgit Is Cheap We recommend keeping a neutral allocation to Malaysian equities within an EM equity dedicated portfolio. In terms of the outlook for the currency, ongoing deflationary pressures are bearish for the MYR in the short-term. The basis is that the Malaysian economy needs a cheaper ringgit in order to help reflate the economy and boost exports. However, the Malaysian currency will sell off less than other EM currencies: First, foreign ownership of local bonds has declined from 36% in 2016-17 to 23% today. Likewise, foreign equity portfolios own about 31% of the stock market, which is less than in many other EMs. This has occurred because foreigners have been major net sellers of Malaysian equities. Overall, low foreign ownership of Malaysian financial assets reduces the risk of sudden portfolio outflows in case EM investors pull out en masse. Second, the current account balance is in surplus and will provide support for the Malaysian ringgit. Malaysia has become less reliant on commodities exports and more of a semiconductor exporter. We are less negative on the latter sector than on resources prices. Third, the currency is cheap, according to the real effective exchange rate, making further downside limited (Chart I-11). Finally, the ongoing purge in the Malaysian economy – deleveraging and deflation – is ultimately long-term bullish for the currency. Deflation brings down the cost structure of the economy and precludes the need for chronic currency depreciation in order to keep the economy competitive. All things considered, the risk-reward profile for shorting the MYR is no longer appealing. We are therefore closing this trade as of today. It has produced a 4% loss since its initiation on July 20, 2016.   Ayman Kawtharani Editor/Strategist ayman@bcaresearch.com Peru: A Pending Policy Dilemma Investors in Peruvian financial markets are presently facing three challenging macro issues: Will the currency appreciate or depreciate? If it depreciates, will the central bank cut or hike interest rates? If policy rates drop or rise, will bank stocks rally or sell off? Chart II-1Peru: Slow Money Growth Heralds Lower Inflation Peru: Slow Money Growth Heralds Lower Inflation Peru: Slow Money Growth Heralds Lower Inflation Looking forward, the central bank (also known as the BCRP) is facing a dilemma. On one hand, inflation is low and will likely drop toward the lower end of the central bank’s target band, as portrayed by narrow money (M1) growth (Chart II-1). Weak domestic demand and low and falling inflation – combined – justify additional rate cuts. On the other hand, the Peruvian currency – like most EM currencies – will likely depreciate versus the US dollar in the coming months, if our baseline view – that foreign capital will flow out of EM and industrial metals prices will drop further for a few months – transpires. In such a case, will the BCRP cut rates – i.e., will the monetary authorities choose to target the exchange rate, or inflation? If the Peruvian central bank follows its own historical footsteps, it will not cut rates, despite economic weakness and falling inflation. On the contrary, the BCRP will likely prioritize defending the nuevo sol by selling foreign currency reserves, as it has done in the past. This in turn will shrink banking system local currency liquidity and lift interbank rates (Chart II-2). Higher interbank rates will hurt the real economy as well as bank share prices. Chart II-2Peru: Selling BCRP FX Reserves Will Shrink Banking System Liquidity Peruvian Local Rates Have Risen Peru: Selling BCRP FX Reserves Will Shrink Banking System Liquidity Peruvian Local Rates Have Risen Peru: Selling BCRP FX Reserves Will Shrink Banking System Liquidity Is Peru more leveraged to precious or industrial metals? Precious and industrial metals account for 17% and 40% of Peruvian exports, respectively. Hence, falling industrial metals prices will be sufficient to exert meaningful depreciation on the sol, despite high precious metals prices. Foreign investors own about 50% of both Peruvian stocks and local currency bonds. Even if a fraction of these foreign holdings flees, the exchange rate will come under significant downward pressure.  Granted that Peru’s central bank does not want its currency to depreciate rapidly, it will defend the currency at the cost of the economy. All in all, the Impossible Trinity thesis is alive and well in Peru: In an economy with an open capital account, the central bank cannot target both interest rates and the exchange rate simultaneously. If the BCRP intends to achieve exchange rate stability, it needs to tolerate interest rate fluctuations. Specifically, interbank rates and other market-determined interest rates could diverge from policy rates. From a real economy perspective, it is optimal to target interest rates and allow the exchange rate to fluctuate. However, the Peruvian economy is still dollarized, albeit much less than before. Dollarization has been a motive to sustain exchange rate stability. If the Peruvian central bank follows its own historical footsteps, it will not cut rates, despite economic weakness and falling inflation. On the whole, Peru’s monetary authorities remain very mindful of exchange rate volatility. Odds are that they will sacrifice growth to avoid sharp currency fluctuations. This has ramifications for financial markets. The Peruvian sol will depreciate much less than other EM and Latin American currencies. This is why it is not in our basket of currency shorts. The central bank will not cut rates in the near term, even though the economy is weak and inflation is low. This is negative for the cyclical economic outlook. Growth will stumble further and non-performing loans (NPLs) in the banking system will rise. NPL growth (inverted) correlates with bank share prices (Chart II-3). Notably, the business cycle is already weak, as illustrated in Chart II-4. Higher interest rates and lower industrial metals prices will weigh further on the economy. Chart II-3Peru: Rising NPLs Will Depress Banks Share Prices Peru: Rising NPLs Will Depress Banks Share Prices Peru: Rising NPLs Will Depress Banks Share Prices Chart II-4Peru: The Economy Is Weak Peru: The Economy Is Weak Peru: The Economy Is Weak   Remarkably, local currency private sector loan growth has moderated, despite the 140 basis points decline in interbank rates over the past 12 months (Chart II-5). This indicates that either interest rates are too high, or banks are reluctant to originate more loans – or a combination of both. Whatever the reason, bank loan growth will decelerate further if interest rates do not drop. Investment Recommendations The Peruvian stock market has underperformed the aggregate EM index over the past five months (Chart II-6, top panel). This underperformance has not only been due to this bourse’s large weight in mining stocks but also because of banks’ underperformance (Chart II-6, bottom panel). Chart II-5Peru: Higher Rates Will Hinder Credit Growth Peru: Higher Rates Will Hinder Credit Growth Peru: Higher Rates Will Hinder Credit Growth Chart II-6Peruvian Equities Have Been Underperforming Peruvian Equities Have Been Underperforming Peruvian Equities Have Been Underperforming   Remarkably, bank shares have languished in absolute terms, even though their funding costs – interbank rates – have dropped significantly (Chart II-7). This is a definitive departure from their past relationship. Chart II-7Peruvian Bank Stocks Stagnated Despite Falling Interest Rates Peruvian Bank Stocks Stagnated Despite Falling Interest Rates Peruvian Bank Stocks Stagnated Despite Falling Interest Rates As interbank rates rise marginally, bank share prices will be at risk of selling off. This in tandem with lower industrial metals prices warrants a cautious stance on this bourse’s absolute performance. Relative to the EM benchmark, we remain neutral on Peruvian equities. The Peruvian sol will depreciate less than many other EM currencies, which will help the stock market’s relative performance versus the EM benchmark. Currency outperformance heralds an overweight stance in domestic bonds within the EM local currency bond portfolio. Dedicated EM credit portfolios should overweight Peruvian sovereign and corporate credit as well. The key attraction is that Peru’s debt levels are low, which will make its credit market a low-beta defensive one in the event of a sell off.   Arthur Budaghyan Chief Emerging Markets Strategist arthurb@bcaresearch.com Juan Egaña Research Associate juane@bcaresearch.com Footnotes 1 Deflated by the average of (1) the GDP deflator, (2) core consumer price inflation, and (3) 25% trimmed-mean consumer price inflation.   Equities Recommendations Currencies, Credit And Fixed-Income Recommendations

Highlights Global growth is poised to accelerate this year, although the spread of the coronavirus could dampen spending in the very short term. History suggests that the likelihood of a recession rises when unemployment falls to very low levels. Three channels have been proposed to explain why that is: 1) Low unemployment can prompt households and businesses to overextend themselves, making the economy more fragile; 2) Faster wage growth stemming from a tight labor market can compress profit margins, leading to less capital spending and hiring; 3) Shrinking spare capacity can fuel inflation, forcing central banks to raise rates. The first channel is highly relevant for some smaller, developed economies where housing bubbles have formed and household debt has reached very high levels. However, it is not an immediate concern in the US, Japan, and most of the euro area. We would downplay the importance of the second channel, as faster wage growth is also likely to raise aggregate demand and incentivize firms to increase capital spending on labor-saving technologies. The third channel poses the greatest long-term risk, but is unlikely to be market-relevant this year. Investors should remain bullish on global equities over the next 12-to-18 months. A more prudent stance will be warranted starting in the second half of 2021. Global Equities: Sticking With Bullish Global equities are vulnerable to a short-term correction after having gained 16% since their August lows. Nevertheless, we continue to maintain a positive outlook on stocks for the next 12 months due to our expectation that global growth will gather steam over the course of the year. The latest data on global manufacturing activity has generally been supportive of our constructive thesis. The New York Fed Manufacturing PMI beat expectations, while the Philly Fed PMI jumped nearly 15 points to the highest level in eight months. The business outlook (six months ahead) component of the Philly Fed index rose to its best level since May 2018. European manufacturing should also improve this year. Growth expectations for Germany in the ZEW index surged in January, rising to the highest level since July 2015 (Chart 1). The Sentix and IFO indices have also moved higher. Encouragingly, euro area car registrations rose by 22% year-over-year in December. In the UK, business confidence in the CBI survey of manufacturers surged from -44 in Q3 of 2019 to +23 in Q4, the largest increase in the 62-year history of the survey. Fiscal stimulus and diminished risk of a disorderly Brexit should also bolster growth this year. Chart 1Some Green Shoots Emerging In The Euro Area Some Green Shoots Emerging In The Euro Area Some Green Shoots Emerging In The Euro Area Chart 2EM Asia Is Rebounding EM Asia Is Rebounding EM Asia Is Rebounding The manufacturing and trade data in Asia have been improving. Following last week’s better Chinese trade data, Korean exports recovered on a rate-of-change basis for a fourth month in a row. Japanese exports to China increased for the first time since last February. In Taiwan, industrial production increased by more than expected in December, as did export orders. Our EM Asia Economic Diffusion Index has risen to the highest level since October 2018 (Chart 2). Coronavirus: Nothing To Sneeze At? The outbreak of the coronavirus represents a potential short-term threat to the budding global economic recovery. Conceptually, outbreaks can affect the economy in two ways. One, they can reduce demand by curtailing spending on travel, entertainment, restaurants, or anything that requires close proximity to others. Two, they can reduce supply by causing people to avoid going to work. In practice, the first effect usually dominates the second. As a result, such outbreaks tend to have a deflationary impact. The Brookings Institution estimates that the 2003 SARS epidemic shaved about one percentage point from Chinese growth that year.1 The fact that this outbreak is happening during the Chinese New Year celebrations, when over 400 million people will be on the move, has the potential to exacerbate the transmission of the virus, and in the process, amplify the economic damage. That said, while it is from the same class of zoonotic viruses, early indications suggest that this particular strain is less lethal than SARS. In addition, the Chinese authorities have moved faster to address the risks than they did during the SARS outbreak. The government has effectively quarantined Wuhan, a city of 11 million people, where the virus appears to have originated. They have also sequenced the virus and shared the information with the global medical community. This has allowed the US Centers for Disease Control (CDC) to develop a test for the virus, which is likely to become available over the coming weeks. The Dark Side Of Low Unemployment Provided the coronavirus outbreak is contained, stronger global growth should continue to soak up lingering labor market slack. This raises the question of whether, at some point, declining unemployment could become counterproductive. The outbreak of the coronavirus represents a potential short-term threat to the budding global economic recovery. The unemployment rate in the OECD currently stands at 5.1%, below the low of 5.5% set in 2007 (Chart 3). In the US, the unemployment rate has dropped to a 50-year low. Chart 3Unemployment Rates Are Below Their Pre-Crisis Lows In Most Economies Who’s Afraid Of Low Unemployment? Who’s Afraid Of Low Unemployment? No one would deny that the decline in unemployment since the financial crisis has been a welcome development. However, it does carry one major risk: Historically, the likelihood of a recession has risen when unemployment has fallen to very low levels (Chart 4). Chart 4Recessions Become More Likely When The Labor Market Begins To Overheat Who’s Afraid Of Low Unemployment? Who’s Afraid Of Low Unemployment? Three channels have been proposed to explain this positive correlation: 1) Low unemployment can prompt households and businesses to overextend themselves, making the economy more fragile; 2) Faster wage growth stemming from a tight labor market can compress profit margins, leading to less capital spending and hiring; 3) Shrinking spare capacity can fuel inflation.  This can force central banks to raise rates, choking off growth. Let’s examine each in turn. Unemployment And Irrational Exuberance Chart 5Growing Housing Imbalances In Some Economies Growing Housing Imbalances In Some Economies Growing Housing Imbalances In Some Economies A strong economy promotes risk-taking. While some risk-taking is essential for capitalism, an excessive amount can lead to the buildup of imbalances, thereby setting the stage for an eventual downturn. In Australia, New Zealand, Canada, and the Scandinavian economies, the combination of low interest rates and strong economic growth has stoked debt-fueled housing bubbles (Chart 5, panel 3). As we discussed last week, higher interest rates in those economies could sow the seeds for economic distress.2 In most other countries, financial imbalances are not severe enough to trigger recessions. Chart 6 shows that the private-sector financial balance – the difference between what the private sector earns and spends – still stands at a healthy surplus of 3.4% of GDP in advanced economies. In 2007, the private-sector financial balance fell to 0.4% in advanced economies, reaching a deficit of 2% in the US. The private-sector balance also deteriorated sharply in the lead-up to the 2001 recession (Chart 7). Chart 6The Private Sector Spends Less Than It Earns In Most Economies Who’s Afraid Of Low Unemployment? Who’s Afraid Of Low Unemployment? Chart 7The Private-Sector Surplus Is Larger Than It Was Before The End Of Previous Expansions The Private-Sector Surplus Is Larger Than It Was Before The End Of Previous Expansions The Private-Sector Surplus Is Larger Than It Was Before The End Of Previous Expansions   In the US, the personal savings rate has risen to nearly 8%, much higher than one would expect based on the level of household net worth (Chart 8). Despite growing at around 2.5% in 2018/19, real personal consumption has increased at a slower pace than predicted by the level of consumer confidence. This suggests that households have maintained a fairly prudent disposition. Consistent with this, the ratio of household debt-to-disposable income has declined by 32 percentage points since 2008. Chart 8Households Are Saving More Than One Would Expect Households Are Saving More Than One Would Expect Households Are Saving More Than One Would Expect Granted, some credit categories have seen large increases (Chart 9). Student debt has risen to 9% of disposable income. Auto loans have moved back to their pre-recession highs. We would not worry too much about the former, as the vast majority of student debt is guaranteed by the government. Auto loans are more of a concern. However, it is important to keep in mind that the auto loan market is less than one-sixth as large as the mortgage market. Moreover, after loosening lending standards for vehicle loans between 2011 and 2016, banks have since tightened them. This adjustment appears to be largely complete. Lending standards did not tighten any further in the latest Senior Loan Officer Survey, while demand for auto loans rose at the fastest pace in two years. The share of auto loans falling into delinquency has been trending lower, which suggests that delinquency rates are peaking (Chart 10). Chart 9US Household Debt Levels Have Fallen, Despite Increases in Student And Auto Loans US Household Debt Levels Have Fallen, Despite Increases in Student And Auto Loans US Household Debt Levels Have Fallen, Despite Increases in Student And Auto Loans Chart 10Auto Loans: Monitoring Trends In Credit Standards And Delinquency Rates Auto Loans: Monitoring Trends In Credit Standards And Delinquency Rates Auto Loans: Monitoring Trends In Credit Standards And Delinquency Rates Lastly, we would point out that despite all the hoopla over the state of the auto market, auto loan asset-backed securities have performed well (Chart 11). While default rates have risen, lenders have generally set interest rates high enough to absorb incoming losses. Chart 11Securitized Auto Loans Have Performed Well Securitized Auto Loans Have Performed Well Securitized Auto Loans Have Performed Well Will Falling Profit Margins Derail The Expansion? Profit margins usually peak a few years before the onset of a recessions (Chart 12, top panel). This has led some to speculate that falling margins could usher in a recession by curbing companies’ willingness to hire workers and invest in new capacity. Chart 12A Peak In Profit Margins: An Ominous Sign? A Peak In Profit Margins: An Ominous Sign? A Peak In Profit Margins: An Ominous Sign? While it is an interesting theory, it does not stand up to closer scrutiny. Surveys of business sentiment clearly show that capital spending intentions are positively correlated with plans to raise wages (Chart 13, left panel). Far from cutting capital expenditures in response to rising wages, firms are more likely to boost capex if they are also planning to increase labor compensation.  Chart 13AFaster Wage Growth, Increased Hiring, And More Capex Go Hand In Hand (I) Faster Wage Growth, Increased Hiring, And More Capex Go Hand In Hand (I) Faster Wage Growth, Increased Hiring, And More Capex Go Hand In Hand (I) Chart 13BFaster Wage Growth, Increased Hiring, And More Capex Go Hand In Hand (II) Faster Wage Growth, Increased Hiring, And More Capex Go Hand In Hand (II) Faster Wage Growth, Increased Hiring, And More Capex Go Hand In Hand (II) One reason for this is that rising wages make automation more attractive. By definition, automation requires more capital spending. However, that is not the entire story because firms also tend to hire more workers during periods when wage growth is rising (Chart 13, right panel). This implies that a third factor – strong economic growth – is responsible for both accelerating wages and rising hiring intentions. The fact that real business sales are strongly correlated with both employment growth and nonresidential investment is evidence for this claim (Chart 12, bottom panel). Falling Margins: A Symptom Of A Problem The discussion above suggests that faster wage growth is unlikely to dissuade firms from either hiring more workers or boosting capital spending. Indeed, the opposite is probably true: Since workers normally spend more of every dollar of income than firms do, an increase in the share of national income flowing to workers will lift aggregate demand. So why do profit margins usually peak before recessions? The answer is that declining labor market slack tends to push up unit labor costs, forcing central banks to hike interest rates in an effort to stave off rising inflation. Thus, falling margins are just a symptom of an underlying problem: economic overheating. Don’t blame lower margins for recessions. Blame central banks. Inflation Is Not A Threat... Yet For now, unit labor cost inflation remains reasonably well contained in the major economies (Chart 14). However, there is little evidence to suggest that the historic relationship between labor market slack and wage growth has broken down (Chart 15). Barring a major surge in productivity growth, inflation is likely to accelerate eventually as companies try to pass on higher labor costs to their customers. Chart 14AUnit Labor Costs Are Well Behaved For Now (I) Unit Labor Costs Are Well Behaved For Now (I) Unit Labor Costs Are Well Behaved For Now (I) Chart 14BUnit Labor Costs Are Well Behaved For Now (II) Unit Labor Costs Are Well Behaved For Now (II) Unit Labor Costs Are Well Behaved For Now (II)       Chart 15Correlation Between Labor Market Slack And Wage Growth Remains Intact Correlation Between Labor Market Slack And Wage Growth Remains Intact Correlation Between Labor Market Slack And Wage Growth Remains Intact We do not know exactly when such a price-wage spiral will emerge. Inflation is a notoriously lagging indicator (Chart 16). Our best guess is that inflation could become a serious risk for investors in late 2021 or 2022. Thus, investors should remain overweight global equities for the next 12-to-18 months, but be prepared to turn more cautious in the second half of 2021.  Chart 16Inflation Is A Lagging Indicator Who’s Afraid Of Low Unemployment? Who’s Afraid Of Low Unemployment?   Peter Berezin Chief Global Strategist peterb@bcaresearch.com   Footnotes 1   Jong-Wha Lee and Warwick J. McKibbin, “Globalization and Disease: The Case of SARS,” Brookings Institution, dated February 2004. 2  Please see Global Investment Strategy Weekly Report, “Bond Yields: How High Is Too High?” dated January 17, 2020.   Global Investment Strategy View Matrix Who’s Afraid Of Low Unemployment? Who’s Afraid Of Low Unemployment? MacroQuant Model And Current Subjective Scores Who’s Afraid Of Low Unemployment? Who’s Afraid Of Low Unemployment? Strategic Recommendations Closed Trades

Highlights Investors should remain overweight global stocks relative to bonds over the next 12 months and begin shifting equity exposure towards non-US markets. Bond yields will rise next year as global growth picks up, while the dollar will sell off. The extent to which bond yields increase over the long term depends on whether inflation eventually stages a comeback. Today’s high debt levels could turn out to be deflationary if they curtail spending by overstretched households, firms, and governments. However, high debt levels could also prompt central banks to engineer higher inflation in order to reduce the real burden of debt obligations. Which of these two effects will win out depends on whether central banks are able to gain traction over the economy. This ultimately boils down to whether the neutral rate of interest is positive or negative in nominal terms. While there is little that policymakers can do to alter certain drivers of the neutral rate such as the trend rate of economic growth, they do have control over other drivers such as the stance of fiscal policy. Ironically, a structural shift towards easier fiscal policy could lead to a decline in government debt-to-GDP ratios if higher inflation, together with central bankers' reluctance to raise nominal rates, pushes real rates down far enough. This suggests that the endgame for today’s high debt levels is likely to be overheated economies and rising inflation.   Stay Bullish On Stocks But Shift Towards Non-US Equities We returned to a cyclically bullish stance on global equities following the stock market selloff late last year, having temporarily moved to the sidelines in June 2018. We have remained overweight global equities throughout 2019. Two weeks ago, we increased our pro-cyclical bias by upgrading non-US stocks within our recommended equity allocation at the expense of their US peers. Our decision to upgrade non-US equities stems from the conviction that global growth has turned the corner. Manufacturing has been at the heart of the global slowdown. As we have often pointed out, manufacturing cycles tend to last about three years – 18 months of weaker growth followed by 18 months of stronger growth (Chart 1). The current slowdown began in the first half of 2018, and right on cue, the recent data has begun to improve. The global manufacturing PMI has moved off its lows, with significant gains seen in the new orders-to-inventories component. Global growth expectations in the ZEW survey have rebounded. US durable goods orders surprised on the upside in October. The regional Fed manufacturing surveys have also brightened, suggesting upside for the ISM next week (Chart 2). Chart 1A Fairly Regular Three-Year Manufacturing Cycle A Fairly Regular Three-Year Manufacturing Cycle A Fairly Regular Three-Year Manufacturing Cycle Chart 2Some Manufacturing Green Shoots Some Manufacturing Green Shoots Some Manufacturing Green Shoots     Unlike in 2016, China has not allowed a major reacceleration in credit growth this year. Instead, fiscal policy has been loosened significantly. The official general government deficit has increased from around 3% of GDP in mid-2018 to 6.5% of GDP at present. The augmented budget deficit – which includes spending through local government financing vehicles and other off-balance sheet expenditures – is on track to reach nearly 13% of GDP in 2019. This is a bigger deficit than during the depths of the Great Recession (Chart 3). As a result of all this fiscal easing, the combined Chinese credit/fiscal impulse has continued to move up. It leads global growth by about nine months (Chart 4). Chart 3China Has Been Stimulating, Fiscally China Has Been Stimulating, Fiscally China Has Been Stimulating, Fiscally Chart 4Chinese Stimulus Should Boost Global Growth Chinese Stimulus Should Boost Global Growth Chinese Stimulus Should Boost Global Growth The dollar tends to weaken when global growth strengthens (Chart 5). The combination of stronger global growth and a softer dollar will disproportionately benefit cyclical equity sectors. Financials will also gain thanks to steeper yield curves (Chart 6). The sector weights of non-US stock markets tend to be more tilted towards deep cyclicals and financials. As a consequence, non-US stocks typically outperform when global growth picks up (Chart 7). Chart 5The Dollar Is A Countercyclical Currency The Dollar Is A Countercyclical Currency The Dollar Is A Countercyclical Currency Chart 6Steeper Yield Curves Will Benefit Financials Steeper Yield Curves Will Benefit Financials Steeper Yield Curves Will Benefit Financials In addition, valuations favor stocks outside the US. Non-US equities currently trade at 13.8-times forward earnings, compared to 18.1-times for the US. The valuation gap is even greater if one looks at price-to-book, price-to-sales, and other measures (Chart 8). Chart 7Non-US Equities Usually Outperform When Global Growth Improves Non-US Equities Usually Outperform When Global Growth Improves Non-US Equities Usually Outperform When Global Growth Improves Chart 8US Stocks Are Relatively More Expensive US Stocks Are Relatively More Expensive US Stocks Are Relatively More Expensive Trade War Remains A Key Risk The US-China trade war remains a key risk to our bullish equity view. President Trump continues to send conflicting signals about the status of the talks. He complained last week that Beijing is not “stepping up” in finalizing a phase 1 agreement, adding that China wants a deal “much more than I do.” This Wednesday he struck a more optimistic tone, saying that negotiators were in the “final throes” of deal. However, he made this statement on the same day that he decided to sign the Hong Kong Human Rights and Democracy Act into law, a decision that was bound to antagonize China. According to our BCA geopolitical team, Trump had little choice but to sign the bill. The Senate approved it unanimously, while the House voted for it 417-1. Failure to sign it would have resulted in an embarrassing veto by the Senate. The key point is that the new law does not force Trump to take any immediate actions against China. This suggests that the trade talks will continue. In fact, from China's point of view, Congress’ desire to pass a Hong Kong bill may provide a timely reminder that getting a deal done with Trump now may be preferable to waiting until after the election and potentially facing someone like Elizabeth Warren who is likely to make human rights a key element of any deal to roll back tariffs. Waiting For Inflation If global growth accelerates next year, history suggests that bond yields will rise (Chart 9). Looking further out, the extent to which bond yields will continue to increase depends on whether inflation ultimately stages a comeback. Right now, most of our forward-looking inflationary indicators remain well contained (Chart 10). However, this could change if falling unemployment eventually triggers a price-wage spiral. Chart 9Stronger Economic Growth Will Put Upward Pressure On Government Bond Yields Stronger Economic Growth Will Put Upward Pressure On Government Bond Yields Stronger Economic Growth Will Put Upward Pressure On Government Bond Yields Chart 10An Inflation Breakout Is Not Imminent An Inflation Breakout Is Not Imminent An Inflation Breakout Is Not Imminent     Many investors are skeptical that such a price-wage spiral could ever emerge. They argue that automation, globalization, weak trade unions, and demographic changes make an inflationary outburst rather implausible. We have addressed these arguments in the past1 and will not delve into them in this report. Instead, we will focus on one argument that also gets a fair bit of attention, which is that high debt levels will prove to be deflationary. Are High Debt Levels Inflationary Or Deflationary? Total debt levels in developed economies are no lower today than they were during the Great Recession. While private debt has fallen, public debt has risen by roughly the same magnitude, leaving the overall debt-to-GDP ratio unchanged (Chart 11). Meanwhile, debt levels in emerging markets have risen substantially. A common rebuttal to any suggestion that inflation might rise over the medium-to-longer term is that high debt levels around the world will cause households, firms, and governments to pare back spending. While this may be true, it could also be argued that high debt levels could prompt central banks to engineer higher inflation in order to reduce the real burden of debt obligations. So which effect will win out? Given the choice, it is likely that most policymakers would opt for higher inflation. This is partly because high unemployment and fiscal austerity are politically toxic. It is also because falling prices make it very difficult to reduce real debt burdens. The experience of the Great Depression bears this out: Private debt declined by 25% in absolute terms between 1929 and 1933. However, due to the collapse in nominal GDP, the ratio of debt-to-GDP actually increased more in the first half of the 1930s than during the Roaring Twenties (Chart 12). Chart 11Global Debt Levels Remain High Global Debt Levels Remain High Global Debt Levels Remain High Chart 12The Experience Of The Great Depression Shows Deleveraging Is Impossible Without Growth The Debt Supercycle Endgame: Deflation Or Inflation? The Debt Supercycle Endgame: Deflation Or Inflation?   Means, Motive And Opportunity Chart 13A Kinked Relationship: It Takes Time For Inflation To Break Out The Debt Supercycle Endgame: Deflation Or Inflation? The Debt Supercycle Endgame: Deflation Or Inflation? There is a big difference between wanting to engineer higher inflation and being able to do so. The distinction between success and failure ultimately boils down to a seemingly technical question: Is the neutral rate of interest – the interest rate consistent with full employment and stable inflation – positive or negative in nominal terms? When the neutral rate is above zero, central banks can gain traction over the economy. Even if the neutral rate is only slightly positive, a zero rate would be enough to keep monetary policy in expansionary territory. When monetary policy is accommodative, the unemployment rate will tend to drop. Eventually the “kink” in the Phillips curve will be reached, resulting in higher inflation (Chart 13). In contrast, when the neutral rate is firmly below zero, monetary policy loses traction over the economy. Since there is a limit to how deeply negative policy rates can go before people decide to hold cash, the central bank could find itself out of ammunition. This could set off a vicious circle where high unemployment causes inflation to drift lower, leading to an increase in real rates. Rising real rates will then further curb spending, causing inflation to fall even more. Drivers Of The Neutral Rate Two of the more important determinants of the neutral rate of interest are the growth rate of the economy and the national savings rate. If either the savings rate rises or economic growth slows, the stock of fixed capital will tend to pile up in relation to GDP, leading to a higher capital-to-output ratio.2 As Chart 14 shows, this has already happened in Europe and Japan. An increase in the capital-to-GDP ratio will drag down the rate of return on capital. A lower interest rate will be necessary to ensure that the capital stock is fully utilized. Chart 14Capital Stock-To-Output Ratios Have Risen The Debt Supercycle Endgame: Deflation Or Inflation? The Debt Supercycle Endgame: Deflation Or Inflation? Realistically, there is not much that policymakers can do to raise trend GDP growth. While looser immigration policy would allow for a faster expansion of labor force growth, this is politically contentious. Increasing productivity growth is also easier said than done. Fiscal Policy And The Neutral Rate In contrast, policymakers already have a ready-made mechanism for lowering the savings rate: fiscal policy. The fiscal balance is a component of national savings. If the government runs a larger budget deficit in order to finance tax cuts or higher transfer payments to households, national savings will decline and aggregate demand will rise. Is the endgame for today’s high debt levels deflation or inflation? The answer is inflation. Since one can think of the neutral rate as the interest rate that brings aggregate demand in line with the economy’s supply-side potential, anything that raises demand will also lift the neutral rate. Once the neutral rate has risen above the zero bound, monetary policy will gain traction again. This implies that central banks should never run out of ammunition in countries whose governments can issue debt in their own currencies. While higher inflation stemming from fiscal stimulus will erode the real value of private sector debt obligations, won’t the impact on total debt be offset by the increase in public debt? Not necessarily. True, larger budget deficits will raise the stock of government debt. However, nominal GDP will also rise on account of higher inflation. Standard debt sustainability equations state that the government debt-to-GDP ratio could actually fall if higher inflation pushes real policy rates down far enough. As discussed in Box 1, such an outcome is quite likely when inflation accelerates in response to an overheated economy, but the central bank nevertheless refrains from raising nominal rates. The Final Verdict We are finally ready to answer the question posed in the title of this report: Is the endgame for today’s high debt levels deflation or inflation? The answer is inflation. People with a 30-year fixed rate mortgage will always favor inflation over deflation. And there are more voters who owe mortgage debt than own mortgage debt. Chart 15Germany's Competitive Advantage Over The Rest Of The Euro Area Is Deteriorating Germany's Competitive Advantage Over The Rest Of The Euro Area Is Deteriorating Germany's Competitive Advantage Over The Rest Of The Euro Area Is Deteriorating Politics is moving in a more populist direction. Whether it is left-wing populism of the Elizabeth Warren/Jeremy Corbyn variety or right-wing populism of the Donald Trump/Matteo Salvini variety, the result is usually bigger budget deficits and higher inflation. Even in those countries where populism has been slow to take hold, there may be pragmatic reasons for loosening fiscal policy. For example, Germany’s trade surplus with the rest of the euro area has fallen in half since 2007, largely because German unit labor costs have increased more than elsewhere (Chart 15). As Germany loses its ability to ship excess production to the rest of the world, it may end up having to rely more on easier fiscal policy to bolster demand. Of course, the path to higher inflation is paved with interest rates that stay lower for much longer than the economy needs to reach full employment. This means we are entering a period where first the US economy, and then many other economies, will start to overheat, and yet central banks will still refrain from tightening monetary policy until inflation rises well above their comfort zones. Such an environment will be positive for stocks for as long as it lasts, even if it eventually produces a mighty hangover.   Peter Berezin Chief Global Strategist peterb@bcaresearch.com   Box 1 When Does A Large Budget Deficit Lead To A Lower Government Debt-to-GDP Ratio?   The Debt Supercycle Endgame: Deflation Or Inflation? The Debt Supercycle Endgame: Deflation Or Inflation? Footnotes 1    Please see Global Investment Strategy Special Report, “1970s-Style Inflation: Could It Happen Again? (Part 2),” dated August 24, 2018. 2   This point can be seen through the lens of the widely used Solow growth model. In steady state, the desired level of investment in the model is given by the formula: I=(a/r)(n+g+d)Y where a denotes the output elasticity of capital, r is the real rate of interest, n is labor force growth, g is productivity growth, d is the depreciation rate, and Y is GDP. Savings is assumed to be a constant fraction of income, S=sY. Equating savings with investment yields: r=(a/s)(n+g+d). A decrease in the growth rate of the economy (n+g) shifts the investment schedule downward, leading to a lower equilibrium rate of interest. This initially makes investing in fixed capital more attractive than buying bonds. Over time, however, the marginal return on capital will fall as the capital stock expands in relation to GDP.    Strategy & Market Trends MacroQuant Model And Current Subjective Scores The Debt Supercycle Endgame: Deflation Or Inflation? The Debt Supercycle Endgame: Deflation Or Inflation? Strategic Recommendations Closed Trades