Europe

Highlights The six-month increase in European bank credit flows amounts to an underwhelming $70 billion, compared to a record high $660 billion in the US and $550 billion in China. Underweight European domestic cyclicals versus their peers in the US and China. Specifically, underweight euro area banks versus US banks. Overweight equities on a long-term (2 years plus) horizon. The mid-single digit return that equities are offering makes them attractive versus ultra-low yielding bonds. But remain neutral equities on a 1-year horizon, until it becomes clear that we can prevent a second wave of the pandemic. Fractal trade: long bitcoin cash, short ethereum. Feature Chart I-1Bank Credit 6-Month Flow Up $70 Bn ##br##In The Euro Area… Chart I-2…But Up $700 Bn ##br##In The US Governments and central banks are dishing out an alphabet soup of stimulus. The question is: how much is reaching those that need it? Our preferred approach to assessing monetary stimulus is to focus on the evolution of bank credit flows and bond yields over a six-month period. Bank Credit Flows Have Surged In The US And China, Not In Europe On our preferred assessment, Europe’s monetary stimulus is underwhelming compared with that in the US and China. The six-month increase in US bank credit flows, at $660 billion, is the highest in a decade and not far from the highest ever. In China, the equivalent six-month increase is $550 billion. But in the euro area, the six-month increase in bank credit flows amounts to an underwhelming $70 billion (Charts I-1 - Chart I-4). Chart I-3Bank Credit 6-Month Flow Up $550 Bn In China… Chart I-4...And Up ##br##Globally Admittedly, US firms are drawing on pre-arranged bank credit lines rather than taking out new loans. Furthermore, the link between bank credit flows and final demand might be compromised during the current economic shutdown. For example, if firms are borrowing to pay workers who are not producing any output, then the transmission of a credit flow acceleration to a GDP acceleration would be weakened. Europe’s monetary stimulus is underwhelming compared with that in the US and China. Nevertheless, some bank credit flows will still reach the real economy. And the US and China are creating more bank credit flows than Europe. Focus On The Deceleration Of The Bond Yield Turning to the bond yield, it is important to focus not on its level, and not on its decline. Instead, it is important to focus on its deceleration. The focus on the deceleration of the bond yield sounds counterintuitive, but it results from a fundamental accounting identity. The next two paragraphs may seem somewhat technical but read them carefully, as they are important for understanding the transmission of stimulus. GDP is a flow. It measures the flow of goods and services produced in a quarter. Hence, GDP receives a contribution from the flow of credit. The flow of credit, in turn, is established by the level of bond yields. When we talk about stimulating the economy, we mean boosting the GDP growth rate from, say, -1 percent to +1 percent, which is an acceleration of GDP. This acceleration in the GDP flow must come from an acceleration in the flow of credit. This acceleration in the flow of credit, in turn, must come from a deceleration of bond yields. In other words, the bond yield decline in the most recent period must be greater than the decline in the previous period. Banks tend to perform better after bond yields have decelerated. The good news is that in the US and China, bond yields have decelerated; the bad news is that in Europe, they have not. Over the past six months, the 10-year bond yield has decelerated by 40 bps in the US and by 65 bps in China. Yet in France, despite the coronavirus crisis, the 10-year bond yield has accelerated by 60 bps (Charts I-5 - Chart I-8).1 Chart I-5The Bond Yield Has Accelerated ##br##In The Euro Area... Chart I-6...Decelerated ##br##In The US... Chart I-7...Decelerated In China... Chart I-8...And Decelerated Globally European bond yields are struggling to decelerate because of their proximity to the lower bound to bond yields, at around -1 percent. The inability to decelerate the bond yield constrains the monetary stimulus that Europe can apply compared to the US and China, whose bond yields are much further from the lower bound constraint. Compared to Europe, the US and China have much stronger decelerations in their bond yields and much stronger accelerations in their bank credit flows. This suggests underweighting European domestic cyclicals versus their peers in the US and China. Specifically, banks tend to perform better after bond yields have decelerated; and they tend to perform worse after bond yields have accelerated. On this basis, underweight euro area banks versus US banks (Chart I-9). Chart I-9Banks Perform Better After Bond Yields Have Decelerated, Worse After Bond Yields Have Accelerated Long-Term Asset Allocation Is Straightforward, Shorter-Term Is Not The level of the bond yield, or of so-called ‘financial conditions’, does not drive the short-term oscillations in credit flows. To repeat, it is the acceleration and deceleration of the bond yield that matters. Yet when it comes to the long-term valuation of assets, the level of the bond yield does matter, and when the bond yield is ultra-low it matters enormously. An ultra-low bond yield justifies a much lower prospective return on competing long-duration assets, like equities. The reason is that when bond yields approach their lower bound, bond prices can no longer rise, they can only fall. This higher riskiness of bonds justifies an abnormally low (or zero) ‘risk premium’ on equities. In this world of ultra-low numbers – for both bond yields and equity risk premiums – the low to mid-single digit long-term return that equities are offering makes them attractive versus bonds (Chart I-10). Chart I-10Equities Are Offering Mid-Single Digit Long-Term Returns But this long-term valuation argument only works for those with long-term investment horizons. What does long-term mean? There is no clear dividing line, but we would define long-term as two years at the very minimum. For a one-year investment horizon, the much more important question is: what will happen to 12-month forward earnings (profits)?  In the stock market recessions of 2008-09 and 2015-16, the stock market reached its low just before forward earnings reached their low. Assuming the same holds true in 2020-21, we must establish whether forward earnings are close to their low or not. In 2008-09, world forward earnings collapsed by 45 percent. In the current recession, which is putatively worse, world forward earnings are down by less than 20 percent to date. To have already reached the cycle low in forward earnings with only half the decline of 2008, the current recession needs to be much shorter than the 2008-09 episode (Chart I-11 and Chart I-12). Chart I-11In The Global Financial Crisis, Forward Earnings Collapsed By 45 Percent Chart I-12In The Current Crisis, Forward Earnings Are Down 20 Percent. Is That Enough? Whether this turns out to be the case or not hinges on the pandemic and our response to it. A controlled easing of lockdowns will boost growth as more of the economy comes back to life. But too rapid an easing of lockdowns will unleash a second wave of the pandemic, requiring a second wave of economic shutdowns, a double dip recession and a new low in the stock market. Hence, if you have a long-term (2-year plus) investment horizon, the choice between equities and bonds is very straightforward: overweight equities. On this long-term horizon, German and Swedish equities are especially attractive versus negative-yielding bonds. On a 1-year investment horizon, the key question is: can we avoid a second wave of the pandemic? But if you have a 1-year investment horizon, the choice is less straightforward, because it hinges on whether we can avoid a second wave of the pandemic or not. Until it becomes clear that governments will not reopen economies too quickly, remain neutral equities on the 1-year horizon. Fractal Trading System* This week’s recommended trade is a pair-trade within the cryptocurrency asset-class. Long bitcoin cash / short ethereum. Set the profit target at 21 percent with a symmetrical stop-loss. The 12-month rolling win ratio now stands at 61 percent. Chart I-13Bitcoin Cash Vs. Ethereum When the fractal dimension approaches the lower limit after an investment has been in an established trend it is a potential trigger for a liquidity-triggered trend reversal. Therefore, open a countertrend position. The profit target is a one-third reversal of the preceding 13-week move. Apply a symmetrical stop-loss. Close the position at the profit target or stop-loss. Otherwise close the position after 13 weeks. * For more details please see the European Investment Strategy Special Report “Fractals, Liquidity & A Trading Model,” dated  December 11, 2014, available at eis.bcaresearch.com.   Dhaval Joshi Chief European Investment Strategist dhaval@bcaresearch.com Footnotes 1 In the US, the 10-year bond yield has declined by 120 bps in the past six months compared with 80 bps in the preceding six months, which equals a deceleration of 40 bps; in China, the 10-year bond yield has declined by 73 bps in the past six months compared with 18 bps in the preceding six months, which equals a deceleration of 65 bps; but in France, the 10-year bond yield has increased by 12 bps in the past six months compared with a 48 bps decline in the preceding six months, which equals an acceleration of 60 bps. Fractal Trading System   Cyclical Recommendations Structural Recommendations Closed Fractal Trades Trades Closed Trades Asset Performance Currency & Bond Equity Sector Country Equity Indicators Bond Yields Chart II-1Indicators To Watch - Bond Yields Chart II-2Indicators To Watch - Bond Yields Chart II-3Indicators To Watch - Bond Yields Chart II-4Indicators To Watch - Bond Yields Interest Rate Chart II-5Indicators To Watch - Interest Rate Expectations Chart II-6Indicators To Watch - Interest Rate Expectations Chart II-7Indicators To Watch - Interest Rate Expectations Chart II-8Indicators To Watch - Interest Rate Expectations

Special Report

Dear Client, This week, in lieu of our regular report, we are sending you a Special Report from my colleague Jonathan LaBerge. Jonathan discusses the global fiscal response to the COVID-19 pandemic, and provides some perspective on whether the response will be enough to prevent an "L-shaped" economic outcome. I hope you find the report insightful. Last Friday, I participated in a webcast with my fellow BCA Strategists Caroline Miller and Arthur Budaghyan where we discussed the outlook for developed and emerging market equities over the immediate (0-3 month) and cyclical (12 month) horizon. You can access a replay of the webcast here. Please note we will be sending out our regular weekly report next week. Best regards, Peter Berezin, Chief Global Strategist Highlights In this Special Report we explore in detail the fiscal response amongst advanced economies, with the goal of judging whether the response is large enough to prevent an “L-shaped” recession. The crisis remains in its early days and new information about the size and character of the response, as well as the magnitude of the economic shock, continues to emerge on a near-daily basis. As such, our conclusions may change over the coming weeks in line with incoming data. Even when narrowly-defined, the announced (or likely) fiscal response of the US, China, and Germany is quite large and appears to be adequate to prevent the direct and indirect effects of the lockdowns from causing an “L-shaped” event. This is not the case, however, in other euro area economies (France, Italy, and Spain), or in emerging markets. Our analysis also suggests that the global fiscal response will need to increase if the global economy faces a W-shaped shock caused by another round of aggressive containment measures later this year. This underscores the importance of ensuring that the “Great Lockdown” succeeds at reducing the spread of the disease to a point that does not necessitate widespread renewed restrictions on economic activity. Feature The global economic expansion that began in 2009 has come to an abrupt end due to the COVID-19 pandemic. Aggressive containment measures necessary to control the spread of the disease and prevent the collapse in health care systems around the world have caused a large and sudden stop in global economic activity, which has prompted unprecedented responses from governments around the world. In this Special Report we explore in detail the fiscal response amongst advanced economies, with the goal of judging whether the response is large enough to prevent an “L-shaped” recession (characterized by a very prolonged return to trend growth). The crisis remains in its early days and new information about the size and character of the response, as well as the magnitude of the economic shock, continues to emerge on a near-daily basis. As such, our conclusions may change over the coming weeks in line with incoming data. But for now, we (tentatively) conclude that the fiscal response appears to be adequate to prevent the direct and indirect effects of the lockdowns from causing an “L-shaped” event. However, there are two important caveats. First, while Germany has provided among the strongest fiscal responses globally, measures in France, Italy, and Spain are still lacking and must be stepped up. Second, the announced fiscal measures will not be sufficient if the global economy faces a W-shaped shock caused by another round of aggressive containment measures later this year – more will have to be done. For policymakers, this underscores the importance of ensuring that the “Great Lockdown” succeeds at reducing the spread of the disease to a point that does not necessitate widespread renewed restrictions on economic activity. In this regard, the gradual re-opening of several US states by early-May, while positive for economic activity in the short-run, is a non-trivial risk to the US and global economic outlooks over the coming 6-12 months. This risk must be closely watched by investors. The Global Fiscal Response: Comparing Across Countries And Across Measures The flurry of policy announcements from national governments over the past six weeks has led to a great degree of confusion about the size and disposition of the global COVID-19 fiscal response. Our analysis is based heavily on the IMF’s tracking of these measures, albeit with a few adjustments. We also rely on analysis from Bruegel, a prominent European macroeconomic think-tank, as well as our own Geopolitical Strategy team and a variety of news reports. Chart 1 presents the IMF’s estimate of the total fiscal response to the crisis across major countries, as of April 23, broken down into “above-the-line” and “below-the-line” measures. Above-the-line measures are those that directly impact government budget balances (direct fiscal spending and revenue measures, usually tax deferrals), whereas below-the-line measures typically involve balance sheet measures to backstop businesses through capital injections and loan guarantees. Chart 1 makes it clear that the fiscal response of advanced economies is enormous when including both above- and below-the-line measures. By this metric, the response of most developed economies is on the order of 10% of GDP, and well above 30% in the case of Italy and Germany. However, using the sum of above- and below-the-line measures to gauge the fiscal response of any country may not be the ideal approach, given that below-the-line measures are contingent either on the triggering of certain conditions or on the provision of credit to households and firms from the financial system. Below-the-line measures also likely increase the liability position of the private sector, thus raising the odds of negative second-round effects. Chart 1The Global Fiscal Response Is Huge When Including All Measures Instead, Chart 2 compares the countries shown in Chart 1 based only on the IMF’s estimate of above-the-line measures, and with a 4% downward adjustment to Japan’s reported spending to account for previously announced measures.1 The chart shows that countries fall into roughly three categories in terms of the magnitude of their above-the-line response: in excess of 4% of GDP (Australia, the US, Japan, Canada, and Germany), 2-3% (the UK, Brazil, and China), and sub-2% (all other countries shown in the chart, including Spain, Italy, and France). Chart 2The Picture Changes When Excluding Below-The-Line Measures Analysis by Bruegel, provides somewhat different estimates of the global COVID-19 fiscal response for select European countries as well as the US (Table 1). Bruegel breaks down discretionary fiscal measures that have been announced into three categories: those involving an immediate fiscal impulse (new spending and foregone revenues), those related to deferred payments, and other liquidity provisions and guarantees. Bruegel distinguishes between the first and second categories because of their differing impact on government budget balances. Deferrals improve the liquidity positions of individuals and companies but do not cancel their obligations, meaning that they result only in a temporary deterioration in budget balances. Table 1 highlights that Bruegel’s estimates of the sum of above- and below-the-line measures are similar to the IMF’s estimates for the US, the UK, and Spain, but are smaller for Italy and larger for France and Germany (particularly the latter). These differences underscore the extreme uncertainty facing investors, who have to contend not only with varying estimates of the magnitude of government policies but also a torrent of news concerning the evolution of the pandemic itself. Table 1The Type Of Fiscal Response Varies Significantly Across Countries Chart 3 presents our best current estimate of the above-the-line fiscal response of several countries (the measure we deem to be most likely to result in an immediate fiscal impulse), by excluding loans, guarantees, and non-specified revenue deferrals to the best of our ability.2Chart 3 is based on a combination of data from the IMF, Bruegel analysis, and BCA estimates and news analysis. Chart 3When Narrowly Defined, Several Countries Are Responding Forcefully, But Many Countries Are Not Overall, investors can draw the following conclusions from Charts 1 – 3 and Table 1: When measured as the total of above- and below-the-line measures, nearly all large developed market countries have responded with sizeable measures. Emerging market economies are the clear laggards. Excluding below-the-line measures and using our approach, Australia, the US, China, Germany, Japan, and Canada appear to be spending the most relative to the size of their economies. While Japan’s “headline” fiscal number was inflated by including previously-announced spending, it is still decently-sized after adjustment. Outside of Germany, the rest of Europe appears to be providing a middling or poor above-the-line fiscal response. The UK appears to be providing between 4-5% of GDP as a fiscal impulse, whereas the fiscal response in Italy, Spain, and France looks more like that of emerging markets than of advanced economies. Measuring The Stimulus Against The Shock Despite the substantial amount of new information over the past six weeks concerning the evolution of the pandemic and the attendant policy response, it remains extremely difficult to judge what the balance between shock and stimulus will be and what that means for the profile of growth. Nonetheless, below we present a framework that investors can use to approach the question, and that can be updated as new information emerges concerning the impact of the shutdowns and the extent of the response. Our approach involves analyzing four specific questions: What is the size of the initial shock? What are the likely second-round effects on growth? What is the likely multiplier on fiscal spending? Will the composition of fiscal spending alter its effectiveness? The Size Of The Initial Shock Chart 4 presents the OECD’s estimates of the initial impact of partial or complete shutdowns on economic activity in several countries. The OECD first used a sectoral approach to estimating the impact on activity while lockdowns are in effect, assuming a 100% shutdown for manufacturing of transportation equipment and other personal services, a 50% decline in activity for construction and professional services, and a 75% decline for retail trade, wholesale trade, hotels, restaurants, and air travel. Chart 4 illustrates the total impact of this approach for key developed and emerging economies. Chart 4Annual GDP Will Be 1.5%-2.5% Lower For Each Month Lockdowns Are In Effect The OECD’s approach provides a credible estimate of the impact of aggressive containment policies, and implies that annual real GDP is likely to be 1.5-2.5% lower for major countries for each month that lockdown policies are in effect. This implies that output in major economies is likely to fall 3.5% - 6% for the year from the initial shock alone, assuming an aggressive 10-week lockdown followed by a complete return to normal. Estimating Potential Second Round Effects Chart 5 presents projections from the Bank for International Settlements on the spillover and spillback potential of a 5% initial shock to the level of global GDP from the COVID-19 pandemic (equivalent to a 20% impact on an annualized basis). The chart shows that the cumulative impact of the initial shock rises to 7-8% by the end of this year for the US, euro area, and emerging markets, and 6% for other advanced economies. These estimates account for both domestic second round effects of the initial shock, as well as the reverberating impact of the shock on global trade. Chart 5 also shows the devastating effect that a second wave of COVID-19 emerging in the second half of the year would have after including spillover and spillback effects, assuming that only partial lockdowns would be required. In this scenario, the level of GDP would be 10-12% lower at the end of the year depending on the region, suggesting that investors should be more concerned about the possibility of additional lockdown events than they should be about the after-effects of the first wave of infections (more on this below). Chart 5Additional Lockdown Events Are A Greater Risk Than First Wave After-Effects Will Fiscal Multipliers Be High Or Low? When examining the academic literature on fiscal multipliers, the first impression is that multipliers are likely to be extremely large in the current environment. Tables 2 and 3 present a range of academic multiplier estimates aggregated by the IMF, categorized by the stage of the business cycle and whether the zero lower bound is in effect. The tables tell a clear story: multipliers are typically meaningfully larger during recessions than during expansions, and extremely large when the zero lower bound (ZLB) is in effect. However, there are at least two reasons to expect that the fiscal multiplier during this crisis will not be as large as Tables 2 and 3 suggest. First, it is obviously the case that the multiplier will be low while full or even partial lockdowns are in effect, as consumers will not have the ability to fully act in response to stimulative measures. This will be partially offset by a burst of spending once lockdowns are removed, but the empirical multiplier estimates during recessions shown in Table 2 have not been measured during a period when constraints to spending have been in effect, and we suspect that this will have at least somewhat of a dampening effect on the efficacy of fiscal spending relative to previous recessions (even once regulations concerning store closures are removed). Table 2Fiscal Multipliers Are Much Larger During Recessions Than Expansions Table 3Models Suggest The Multiplier Is Quite High At The Zero Lower Bound Second, Table 3 likely overestimates the multiplier at the ZLB. These estimates have been based on models rather than empirical analysis, and appear to be in reference to the prevention of large subsequent declines in output following an initial shock. The modeled finding of a large multiplier at the ZLB occurs because increased deficit spending will not lead to higher policy rates in a scenario where the neutral rate has fallen below zero. But it seems difficult to believe that the fiscal multiplier during ZLB episodes, defined as the impact of fiscal spending on the path of output relative to the initial shock (not relative to a counterfactual additional shock), is larger than the highest empirical estimates of the multiplier during recessions. The only circumstance in which we can envision this being the case is an environment where long-term bond yields are capped and remain at zero, alongside short-term interest rates, as the economy improves. The IMF has provided a simple rule of thumb approach to estimating the fiscal multiplier for a given country. The IMF’s approach involves first estimating the multiplier under normal circumstances based on a series of key structural characteristics that have been shown to influence the economy’s response to fiscal shocks. Then, the “normal” multiplier is adjusted higher or lower depending on the stage of the business cycle, and whether monetary policy is constrained by the ZLB. For the US, the IMF’s approach suggests that a multiplier range of 1.1 – 1.6 is reasonable, assuming the highest cyclical adjustment but no ZLB adjustment (see Box 1 for a description of the calculation). Given the unprecedented nature of this crisis, we are inclined to use the low end of this range (1.1) as a conservative assumption when judging whether fiscal responses to the crisis are sufficient. For investors, this means that governments should be aiming, at a minimum, for fiscal packages that are roughly 90% of the size of the expected shock of their economies, using our US fiscal multiplier assumption as a guide. Box 1The 'Bucket' Approach To Estimating Fiscal Multipliers The Composition Of The Response: Helping Or Hurting? The last of our four questions deals with the issue of composition and whether the form of a country’s fiscal response is likely to alter its effectiveness. We implicitly addressed the first element of composition, whether measures are above-the-line or below-the-line, by comparing Charts 1 - 3 on pages 3-5. Our view is that above-the-line measures are far more important than below-the-line measures, as the former provides direct income and liquidity support. Below-the-line measures are also important, as they are likely to help reduce business failure and household bankruptcies. The fiscal multiplier on these measures has to be above zero, but it is likely to be much lower than that of an above-the-line response. Chart 6Much Of The US Fiscal Response Is Going To Households And Small Businesses The second element of composition concerns the appropriate distribution of aid among households, businesses, and local governments. On this particular question, it remains extremely challenging to analyze the issue on a global basis, owing to a frequent lack of an explicit breakdown of fiscal measures by recipient. For now, we limit our distributional analysis to the US, and hope to expand our approach to other countries in future research. Chart 6 presents a breakdown of the US fiscal response by recipient, which informs the following observations. Households: Chart 6 highlights that US households will receive approximately $600 billion as part of the CARES Act, roughly half of which will occur through direct payments (i.e. “stimulus checks”) and another 40% from expanded unemployment benefits. In cases where the federal household response has been criticized by members of the public as inadequate, it has often been compared to income support programs of other countries. The Canada Emergency Response Benefit (“CERB”) is a good example of a program that seems, at first blush, to be superior: it provides $2,000 CAD in direct payments to individuals for a 4 week period, for up to 16 weeks (i.e. a maximum of $8,000 CAD), which seems better than a $1,200 USD stimulus check. However, Table 4 highlights that this comparison is mostly spurious. First, the CERB is not universal, in that it is only available to those who have stopped or will stop working due to COVID-19. At a projected cost of $35 billion CAD, the CERB program represents 1.5% of Canadian GDP. By comparison, $600 billion USD in overall household support represents 2.75% of US GDP; this number drops to 1.75% when only considering support to those who have lost their jobs, but this is still higher as a share of the economy than in Canada. Moreover, there is little question that Congress is prepared to pass more stimulus for additional weeks of required assistance.   Table 4US Household Relief Is Just As Generous As Seemingly Better Programs The discrepancy between the perception and reality of US household sector support appears to be rooted in the speed of payments. Speed is the one area where Canada’s household sector response appears to have legitimately outperformed the US; CERB payments are received by applicants within three business days for those registered for electronic payment, and in some cases they are received the following day. By contrast, it has taken some time for US States to start paying out the additional $600 USD per week in expanded unemployment benefits, but as of the middle of last week nearly all states had started making these payments.   Firms: On April 16 the Small Business Administration announced that the Paycheck Protection Program (“PPP”) had expended its initial budget of $350 billion. While additional funds of $320 billion have subsequently been approved (plus $60 billion in small business emergency loans and grants), the run on PPP funds was, to some investors, an implicit sign that the CARES Act was inadequately structured.   Table 5Help For Small Businesses Will Replace A Significant Amount Of Lost Income However, the fact that the initial funds ran out in mid-April simply reflects the reality that social distancing measures had been in place for 3-4 weeks by the time that the program began taking applications. Table 5 highlights that $350 billion was large enough to replace nearly 90% of lost small business income for one month, assuming that overall small business revenue has fallen by 50% and that small businesses account for 44% of total GDP. The table also shows that a combined total of $730 billion is enough to replace almost 80% of lost small business income for 10 weeks, given these assumptions. With loan forgiveness at least partially tied to small businesses retaining employees on payroll for an 8-week period, the PPP is also essentially an indirect form of household income support. State & Local Governments: The magnitude of support for state & local (S&L) governments appears to be the least-well designed element of the US fiscal response.  The CARES Act provides for $170 billion in support to S&L, which at first blush seems large as it is approximately 25% of S&L current receipts in Q4 2019 (i.e. it stands to cover a 25% loss in revenue for one quarter). However, this does not account for the significant reported increase in S&L costs to combat the pandemic, nor does it provide S&L governments with any revenue certainty beyond June 30 when most of the assistance from CARES must be spent.   Chart 7Persistent State & Local Austerity Must Be Avoided This Time Unlike households or firms, who also face significant uncertainty, nearly all US states are subject to balanced budget requirements, which prevent them from spending more than they collect in revenue. When faced even with projected revenue losses in the second half of this year and into 2021, states are likely to aggressively and immediately cut costs in order to avoid budgetary shortfalls. Chart 7 highlights that S&L austerity was a significant element of the persistent drag on real GDP growth from overall government expenditure and investment in the first 3-4 years of the post-GFC economic expansion. A repeat of this episode would significantly raise the odds of an “L-type” recession (and thus should certainly be avoided). This is why Congress is moving to pass larger state and local aid. Our Geopolitical Strategy team argues that neither President Trump nor Senate Majority Leader Mitch McConnell will prevent the additional financial assistance that US states will require, despite their rhetoric about states going bankrupt.3 A near-term, temporary standoff may occur, but Washington will almost certainly act to provide at least additional short-term funding if state employment starts to fall due to budget pressure. So while we recognize that the state & local component of the US fiscal response is currently lacking, it does not seem likely to represent a serious threat to an eventual economic recovery in the US. Putting It All Together: Will It Be Enough? Chart 8 reproduces Chart 3 with an assumed fiscal multiplier of 1.1, and with shaded regions denoting the likely initial and total impact on GDP from aggressive containment measures (based on the OECD and BIS’ estimates). Based on our analysis of the US fiscal response, we make no adjustments for the composition of the measures beyond defining the fiscal response on a narrow basis (i.e. excluding loans, guarantees, and non-specified revenue deferrals). The chart highlights that the narrowly-defined fiscal response of three key economies driving global demand, the US, China, and Germany, is either at the upper end or above the total impact range. Thus, for now, we tentatively conclude that the fiscal response that has or will happen appears to be adequate to prevent the direct and indirect effects of the lockdowns from causing an “L-shaped” event, especially since Chart 8 explicitly excludes below-the-line measures. Chart 8Several Important Countries Seem To Be Doing Enough, But More Is Needed In Europe Ex-Germany However, there are two important caveats to this conclusion. First, Chart 8 makes it clear that measures in France, Italy, and Spain are still lacking and must be stepped up. Italy and France have provided a substantial below-the-line response, but it is far from clear that a debt-based response or one that only temporarily improves access to cash for households and businesses will be enough to prevent a prolonged fallout from the sudden stop in economic activity and income. Second, our analysis suggests that the announced fiscal measures will not be sufficient if the global economy faces a W-shaped shock caused by another round of aggressive containment measures later this year or if these measures remain in place at half-strength for many months. This underscores how sensitive the adequacy of announced fiscal measures are to the amount of time economies remain under full or partial lockdown. As such, it is crucial for investors to have some sense of when advanced economies may be able to sustainably end aggressive containment measures. When Can The Lockdowns Sustainably End? Several countries and US states have already announced some reductions in their restrictions, but the question of how comprehensive these measures can be without risking a second period of prolonged stay-at-home orders looms large. Table 6 presents two different methods of estimating sustainable lockdown end dates for several advanced economies. First, we use the “70-day rule” that appears to have succeeded in ending the outbreak in Wuhan, calculated from the first day that either school or work closures took effect in each country.4 Second, using a linear trend from the peak 5-day moving average of confirmed cases and fatalities, we calculate when confirmed cases and fatalities may reach zero. Table 6By Re-Opening Soon, The US May Be Risking A Damaging Second Wave The table highlights that these methods generally prescribe a reopening date of May 31 or earlier, with a few exceptions. The UK’s confirmed case count and fatality trends are still too shallow to suggest an end of May re-opening, as is the case in Canada. In the case of Sweden, no projections can truly be made based on the 70-day rule because closures never formally occurred. But the most problematic point highlighted in Table 6 is that US newly confirmed cases are only currently projected to fall to zero as of February 2021. Chart 9 highlights that while new cases per capita in New York state are much higher than in the rest of the country, they are declining whereas they have yet to clearly peak elsewhere. Cross-country case comparisons can be problematic due to differences in testing, but with several US states having already begun the gradual re-opening process, this underscores that US policymakers may be allowing a dangerous rise in the odds of a secondary infection wave. Chart 9No Clear Downtrend Yet Outside Of New York State Investment Conclusions Our core conclusion that an “L-shaped” global recession is likely to be avoided is generally bullish for equities on a 12-month horizon. However, uncertainty remains extremely elevated, and the recent rise in stock prices in the US (and globally) has been at least partially based on the expectation that lockdowns will sustainably end soon, which at least in the case of the US appears to be a premature conclusion given the current lack of large-scale virus testing capacity. As such, we are less optimistic towards risky assets tactically, and would recommend a neutral stance over a 0-3 month horizon. As noted above, our cross-country comparison of narrowly-defined fiscal measures suggested that euro area countries (excluding Germany) will likely have to do more in order to prevent a long period of below-trend growth. In the case of highly-indebted countries like Italy, this raises the additional question of whether a significantly increased debt-to-GDP ratio stemming from an aggressive fiscal impulse will cause another euro area sovereign debt crisis similar to what occurred from 2010-2014. Chart 10Italy's Debt Sustainability Hurdle Is Lower Than It Used To Be Government debts are sustainable as long as interest rates remain below economic growth, and from this vantage point Italy should spend as much as needed in order to ensure that nominal growth remains above current long-term government bond yields. Chart 10 highlights that, despite a widening spread versus German bunds, Italian 10-year yields are much lower today than they were during the worst of the euro area crisis, meaning that the debt sustainability hurdle is technically lower. However, we have also noted in previous reports that high-debt countries often face multiple government debt equilibria; if global investors become fearful that that high-debt countries may not be able to repay their obligations without defaulting or devaluing, then a self-fulfilling prophecy will occur via sharply higher interest rates (Chart 11). Chart 11Multiple Equilibria In Debt Markets Are Possible Without A Lender Of Last Resort For now, we view the risk of a renewed Italian debt crisis from significantly increased spending related to COVID-19 as minimal, and it is certainly lower than the status quo as the latter risks causing a sharp gap between nominal growth and bond yields like what occurred from 2010 – 2014. First, Chart 12 highlights that Italy has succeeded in somewhat reducing its structural balance, which averaged -4% for many years prior to the euro area crisis. Assuming an adequate global response to the crisis and that economic recovery ensues, it is not clear why global bond investors would be concerned that Italian structural deficits would persistently widen. Second, the ECB is purchasing Italian government bonds as part of its new Pandemic Emergency Purchase Program, which will help cap the level of Italian yields. Third, Chart 13 shows what will occur to Italy’s government debt service ratio (general government net interest payments as a percent of GDP) in a scenario where Italy’s gross debt to GDP rises a full 20 percentage points and the ratio of net interest payments to debt remains unchanged. The chart shows that while debt service will rise, it will still be lower than at any point prior to 2015. Chart 12Italy's Structural Budget Balance Has Improved Chart 13Italy's Debt Service Ratio Won't Go Up Much, If Yields Are Unchanged   So not only should Italy spend significantly more to combat the severely damaging nature of the pandemic, we would expect that Italian spreads would fall, not rise, in such an outcome.   Jonathan LaBerge, CFA, Vice President Special Reports jonathanl@bcaresearch.com Footnotes 1 “Skeptical economists call Japan’s largest-ever stimulus package ‘puffed-up’, Keita Nakamura, The Japan Times, April 8, 2020. 2 Please note that Chart 3 differs somewhat from a chart that has been frequently shown by our Geopolitical Strategy service. Both charts are accurate; they simply employ different definitions of the fiscal response to the pandemic. 3 Indeed, McConnell has already walked back his comments that states should consider bankruptcy. President Trump is constrained by the election, as are Senate Republicans, and the House Democrats control the purse strings. Hence more state and local funding is forthcoming. At best for the Republicans, there may be provisions to ensure it goes to the COVID-19 crisis rather than states’ unfunded pension obligations. See Geopolitical Strategy, “Drowning In Oil (GeoRisk Update),” April 24, 2020, available at gps.bcaresearch.com. 4 School and work closure dates have been sources from the Oxford COVID-19 Government Response Tracker. Global Investment Strategy View Matrix Current MacroQuant Model Scores  

Highlights Inflation-Linked Bonds: The plunging price of oil has put renewed downward pressure on global bond yields via lower inflation expectations. With oil prices set to recover over the next 6-12 months as the global economy awakens from the COVID-19 slumber, depressed market-derived inflation expectations can move higher across the developed markets – most notably in the US, the UK, Australia and Canada. Favor inflation-linked government bonds versus nominals in those countries on a strategic (6-12 months) basis. UK Corporates: The Bank of England (BoE) is supporting the UK investment grade corporate bond market with an unprecedented level and pace of purchases, with credit spreads at attractive levels. Upgrade UK investment grade corporates to overweight on a tactical (0-6 months) and strategic (6-12 months) basis. Across sectors, favor debt from sectors such as non-bank Financials and Communications that are less exposed to pandemic-related uncertainty but still benefit from BoE buying. Feature Chart of the WeekThe Link Between Oil & Bond Yields Remains Strong The shocking, albeit brief, journey of the West Texas Intermediate (WTI) oil price benchmark below zero last week was another in a long line of stunning market moves seen during the COVID-19 pandemic. Those negative oil prices were technical in nature and lasted all of one day, but the ramifications for global bond markets of the falling cost of oil in 2020 have been more enduring. Government bond yields have largely followed the ebbs and flows in energy markets for most of the past decade, and this year has been no exception (Chart of the Week). That link from oil has been through the inflation expectations component of yields, which have been (and remain) highly correlated to oil prices in virtually every developed market country. This is likely due to the persistent low global inflation backdrop since the 2008 financial crisis, which has made cyclical swings in energy prices the marginal driver of both realized and expected inflation. Chart 2BCA's Commodity Strategists Expect Oil Prices To Recover Our colleagues at BCA Research Commodity & Energy Strategy now anticipate higher oil prices over the next 12-18 months.1 Global growth is expected to recover from the COVID-19 recession sooner (and faster) than global oil production, helping to improve the demand/supply balance in energy markets and boost oil prices (Chart 2). Our energy strategists expect the benchmark Brent oil price to rise to $42/bbl by the end of 2020 and $78/bbl by the end of 2021. Those are big moves compared to the current spot price around $20/bbl, and would impart significant upward pressure on inflation expectations if the history of the past decade is any guide. That kind of move in oil prices should also help lift overall nominal government bond yields. Although the real (inflation-adjusted) component of yields is likely to remain low as major central banks like the Fed and ECB will remain highly accommodative, even when growth and inflation begin to recover, given the severity of the COVID-19 global recession. With market-based inflation expectations now at such beaten-up levels, and with the disinflationary effect of falling energy prices set to fade, we see an opportunity to play for a cyclical rebound in inflation breakevens across the developed markets by favoring inflation-linked government bonds versus nominal yielding equivalents. A Simple Framework For Finding Value In Inflation Breakevens Given the remarkably tight correlation between oil prices and market-determined inflation expectations in so many countries, it should be fairly straightforward to model the latter using the former as the main input. We have developed a series of fair value regressions for breakevens in the major developed countries which do exactly that. In this simple approach, we attempt to model the 10-year breakeven from inflation-linked bonds for eight countries – the US, the UK, Germany, Japan, France, Italy, Canada and Australia - as a function of a short-run variable (oil prices) and a long-run variable (the trend in realized inflation). Specifically, we are using the annual percentage change in the Brent oil price benchmark in local currency terms (i.e. converted from US dollars at spot exchange rates) as the short-run variable and a five-year moving average of realized headline CPI inflation as the long-run variable. The latter is included to provide an “anchor” for breakevens based on the actual performance of inflation in each country. In other words, expectations about what inflation will look like in the future are informed by what it has done in the past – what economists refer to as “adaptive” expectations. The generic regression equation used for each country is: 10-year inflation breakeven = α + β1 * (annual % change of Brent oil price in local currency terms) + β2 * (60-month moving average of headline CPI inflation) In Table 1, we present the results of the regressions of each of the eight countries, which use weekly data dating back to the start of 2012 to capture the period when oil prices have most heavily influenced inflation expectations. The coefficients, R-squareds and standard errors of the regressions are all shown, as well as the most recent model residual (i.e. the deviation of 10-year inflation expectations from model-determined fair value). All the coefficients for each model are significant. The R-squareds of the models vary, with the models for France and Australia doing the best job of explaining changes in inflation expectations in those two countries. Table 1Details Of Our New 10-Year Inflation Breakeven Models For the UK and Japan, we added an additional “dummy” variable to control for the unique situations that we believe have influenced inflation breakevens in those countries. For the UK, the period since the June 2016 Brexit vote has seen the path of inflation expectations stay nearly 50bps higher than implied by moves in GBP-denominated oil prices and the trend in actual UK inflation. For Japan, the period since the Bank of Japan initiated its Yield Curve Control policy in September 2016 has seen breakevens stay nearly 60bps below fair value as derived from JPY-denominated oil prices and the trend in actual Japanese inflation. Bond investors with longer-term investment horizons looking to play for a global growth recovery from the COVID-19 recession over the next 12-18 months should position for some widening of breakevens by favoring inflation-linked bonds over nominal paying government debt. In Charts 3 to10 over the next four pages, we show the models for each country. 10-year inflation breakevens versus the independent variables in the models are shown in the top two panels, the model fair value is presented in the 3rd panel, and the deviation from fair value is in the bottom panel. In all cases, breakevens are below fair value, suggesting that inflation-linked bonds look relatively attractive versus nominal government bonds. Chart 3Our US 10-Year TIPS Breakevens Model Chart 4Our UK 10-Year Breakeven Inflation Model Chart 5Our France 10-Year Breakeven Inflation Model Chart 6Our Italy 10-Year Breakeven Inflation Model Chart 7Our Japan 10-Year Breakeven Inflation Model Chart 8Our Germany 10-Year Breakeven Inflation Model Chart 9Our Canada 10-Year Breakeven Inflation Model Chart 10Our Australia 10-Year Breakeven Inflation Model Chart 11Real Inflation-Linked Bond Yields Will Remain Subdued For Longer The largest deviations from fair value can be found in Canada (-70bps), Australia (-48bps), the UK (-29bps), and the US (-26bps). 10-year breakevens are also below fair value in the euro zone countries and Japan, but not by more than one standard deviation as is the case for the other four countries. Bond investors with longer-term investment horizons looking to play for a global growth recovery from the COVID-19 recession over the next 12-18 months should position for some widening of breakevens by favoring inflation-linked bonds over nominal paying government debt. Focus on the four markets with breakevens furthest from fair value, although from a market liquidity perspective it is easier to implement those positions in the US and UK, which represent a combined 69% of the Bloomberg Barclays Global Inflation-Linked bond index. A rise in inflation expectations should also, eventually, put some sustained upward pressure on nominal bond yields. We would rather play that initially by positioning for higher inflation breakevens, rather than having outright below-benchmark duration exposure, as developed market central banks will stay accommodative for longer given the severity of the COVID-19 recession - that will keep real bond yields lower for longer (Chart 11). Breakevens from inflation-linked bonds are now too low across the developed markets – most notably in the US, the UK, Australia and Canada. Bottom Line: The plunging price of oil has put renewed downward pressure on global bond yields via lower inflation expectations. With oil prices set to recover over the next 6-12 months as the global economy starts to awaken from the coronavirus induced slumber, breakevens from inflation-linked bonds are now too low across the developed markets – most notably in the US, the UK, Australia and Canada. Favor linkers over nominals in those countries. Where Is The Value In UK Corporate Bonds? Chart 12Upgrade UK IG Corporates To Overweight On BoE Buying The Bank of England (BoE) initiated 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 vote to exit the European Union. As we noted in recent joint report with our sister service, BCA Research US Bond Strategy,2 the CBPS helped tighten spreads by lowering downgrade and default risk premiums and also helped spur corporate bond issuance (Chart 12). Shortly after that report was published, the BoE announced that it would be purchasing a further £10 billion in investment grade nonfinancial corporate bonds in the coming months, doubling the scheme’s aggregate holdings to £20 billion. In addition, the bank would make these purchases at a significantly faster pace than in 2016, which implies a faster transmission towards tightening of spreads. Compared to other central bank peers, however, the BoE’s program still has room to expand, which makes UK investment grade credit attractive over tactical and strategic investment horizons. Using the market value of the Bloomberg Barclays UK corporate bond index (excluding financials) as a proxy for the total value of eligible bonds, the CBPS is on track to own roughly 9% of all eligible bonds by the time the £20 billion target is reached. The neighboring European Central Bank, on the other hand, already owns 23% of the stock of eligible euro area corporate bonds in its market, and that figure is only set to increase with policymakers set to do “whatever it takes” to backstop the investment grade market. Year-to-date, UK corporate bonds appear to have recovered somewhat from the panicked selloff earlier this quarter (Table 2), with the Bloomberg Barclays UK investment grade corporate bond index down only -0.3% in total return terms. In excess return terms relative to duration-matched UK corporate bonds, however, the index is down -5.2%, indicating that weakness has persisted in the pure credit component. Table 2UK Investment Grade Corporate Bond Returns At the broad sector level, Other Industrials appear to be the outlier, having delivered positive excess returns (+0.6%) and significant total returns (+16%). These returns are not nearly as attractive, however, on a risk-adjusted basis once you consider that this sector has an index duration more than three times that of the overall index.3 Outside of that sector, the best performers, in excess return terms, are predominantly the more “defensive” sectors—Utilities (-3.4%), Technology (-3.7%), Communications (-4.2%) and Consumer Non-Cyclical (-4.6%). Meanwhile, the sectors most exposed to vanishing consumer demand and weak global growth have performed the worst—Transportation (-9.5%), Capital Goods (-7%), Energy (-6.8%), and Basic Industry (-6.2%). Credit spreads in the UK indicate that the market has already begun to stabilize in response to the BoE’s new round of corporate bond purchases. Credit spreads in the UK indicate that the market has already begun to stabilize in response to the BoE’s new round of corporate bond purchases (Chart 13). The overall index spread, although still elevated at 228bps, has already tightened by 57bps from the peak in late March. The gap between the index spreads of Baa-rated and Aa-rated UK debt remained relatively stable through the wave of sell-offs, peaking at +53bps, below the 2019 high of +55bps, and settling now to +36bps. Outside the purview of the CBPS, however, the situation is a bit rockier, with the overall high-yield index spread +590bps above that of the investment grade index. Broadly speaking, there is a clear disparity between those credit tiers that have the support of the monetary authorities and those that do not. Investment grade spreads will continue to tighten as the BoE rapidly increases its holdings of investment grade corporate bonds. However, high-yield bonds remain exposed to downgrade/default risk and ongoing uncertainty stemming from the COVID-19 economic shock. To drill down into which credit tier spreads offer the most value within the UK investment grade space, we use the 12-month breakeven spread percentile rankings. This is one of the tools we use to assess value in global credit spreads, as measured by historical “spread cushions”. Specifically, we calculate how much spread widening is required over a one-year horizon to eliminate the yield advantage of owning corporate bonds versus duration-matched government debt. We then show those breakeven spreads as a percentile ranking versus its own history, to allow comparisons over periods with differing underlying spread volatility. Chart 14 shows the 12-month breakeven spread percentile rankings for all the credit tiers in the UK investment grade space. Aaa-rated debt appears most unattractive, with the spreads currently ranking below the historical median. Between the other three tiers, Aa-rated debt offers the most value, although all three are at historically attractive levels. Chart 13UK IG Has Held Up Well During The COVID-19 Shock Chart 14UK IG Breakeven Spreads Look Most Attractive For Aa-Rated Bonds On the sector-level, the disparity in spreads is most clearly visible in the sectors most exposed to the pandemic. In Charts 15 & 16, we show the history of option-adjusted spreads (OAS) for the major industrial sub-groupings of the Bloomberg Barclays UK investment grade corporate index. Spreads look widest relative to history for sectors such as Energy and Transportation, while spread widening has been contained in more insulated sectors such as Financials. Chart 15A Mixed Performance For UK IG By Sector In 2020 … Chart 16… But Spreads, In General, Remain Below Previous Cyclical Peaks Another way to assess value across UK investment grade corporates is our sector relative value framework. Borrowing from the methodology used for US corporate credit by our colleagues at BCA Research US Bond Strategy, the sector relative value framework determines “fair value” spreads for each of the major and minor industry level sub-indices of the overall UK investment grade universe. The methodology takes each sector's individual OAS and regresses it in a cross-sectional regression with all other sectors. The dependent variables in the model are each sector's duration, 12-month trailing spread volatility and credit rating - the primary risk factors for any corporate bond. Using the common coefficients from that regression, a risk-adjusted "fair value" spread is calculated. The difference between the actual OAS and fair value OAS is our valuation metric used to inform our sector allocation ranking. We see this as an opportune time to upgrade our recommended allocation for UK investment grade corporates to overweight. The latest output from the UK relative value spread model can be found in Table 3. We also show the duration-times-spread (DTS) for each sector in those tables, which we use as the primary way to measure the riskiness (volatility) of each sector. The scatterplot in Chart 17 shows the tradeoff between the valuation residual from our model and each sector's DTS. Table 3UK Investment Grade Corporate Sector Valuation & Recommended Allocation Chart 17UK Investment Grade Corporate Sectors: Valuation Versus Risk We can then apply individual sector weights based on the model output and our desired level of overall spread risk to come up with a recommended credit portfolio. The weights are determined at our discretion and are not the output from any quantitative portfolio optimization process. The only constraints are that all sector weights must add to 100% (i.e. the portfolio is fully invested with no use of leverage) and the overall level of spread risk (DTS) must equal our desired target. Amid a backdrop of global uncertainty, we reiterate one of our major themes this quarter—buy what the central banks are buying. Given that UK corporate spreads are attractive on a breakeven basis, and with the BoE purchasing corporate debt at an even faster pace than during the volatile period following the shock Brexit vote in 2016, we see this as an opportune time to upgrade our recommended allocation for UK investment grade corporates to overweight. This is both on a tactical (0-6 months) and strategic basis (6-12 months). In our model bond portfolio, we have added two percentage points to our recommended UK corporate bond allocation, funded by reducing further our existing underweight on Japanese government bonds. At the sector level, given this positive backdrop for credit performance, we do not see a need to favor lower risk sectors with a DTS score below that of the overall UK investment grade index. On that basis, we are looking to go overweight sectors with higher relatively higher DTS and positive risk-adjusted spread residuals from our relative value model (and vice versa). Those overweight candidates would ideally be located in the upper right quadrant of Chart 17. Based on the latest output from the relative value model, the strongest overweight candidates are the following UK investment grade sectors: selected Financials (Insurance, Subordinated Bank Debt, and Other Financials), Media Entertainment, Cable Satellite, Tobacco, Diversified Manufacturing, and Communications. The least attractive sectors within this framework are: Packaging, Lodging, REITs, Other Industrials, Metals, Natural Gas, Restaurants, Transportation Services, Financial Institutions, and Midstream Energy. Bottom Line: The BoE is supporting the UK investment grade corporate bond market with an unprecedented level and pace of purchases. Spreads have already begun to tighten in response but are still at attractive levels. Upgrade UK investment grade corporates to overweight on a tactical (0-6 months) and strategic (6-12 months) basis. Across credit tiers, favor Aa-rated debt. Across sectors, favor debt from sectors such as non-bank Financials and Communications that are less exposed to pandemic-related uncertainty but still benefit from the CBPS.   Robert Robis, CFA Chief Fixed Income Strategist rrobis@bcaresearch.com Shakti Sharma Research Associate shaktis@bcaresearch.com Footnotes 1 Please see BCA Commodity & Energy Strategy Weekly Report, "US Storage Tightens, Pushing WTI Lower", dated April 16, 2020, available at ces.bcaresearch.com. 2 Please see BCA US Bond Strategy Special Report, "Trading The US Corporate Bond Market In A Time Of Crisis", dated March 31 2020, available at usbs.bcaresearch.com. 3 Other Industrials has an index duration of 28.6 years, compared to 8.5 years for the overall UK investment grade corporate bond index. Recommendations The GFIS Recommended Portfolio Vs. The Custom Benchmark Index Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Highlights Social distancing makes it impossible to do jobs that require close personal interaction, yet these are the very job sectors that have kept jobs growth alive in recent decades. If social distancing persists, then AI will penetrate these job sectors too. Aggregate wage inflation is set to collapse – not just temporarily, but structurally. Structurally overweight US T-bonds versus the core European bonds in Germany, France, Netherlands, Switzerland and Sweden. Structurally overweight big technology, structurally underweight banks. Structurally overweight S&P 500 versus Euro Stoxx 50. Fractal trade: long Australian 30-year bond versus US 30-year T-bond. Feature Social distancing will feature large in our lives for the foreseeable future, and it carries a profound consequence. Social distancing really means physical distancing. And physical distancing diminishes the ways that we can interact with other humans – through the qualities of empathy, sympathy, the ability to recognise and respond to emotional cues, and to express ourselves through complex movements. You cannot hug someone on Facetime. Social distancing makes it impossible to do jobs that require close personal interaction. From an economic perspective, social distancing makes it impossible to do jobs that require close personal interaction. It follows that in the recent bloodbath of job losses, the biggest casualties have been in employment sectors that rely on this close personal interaction: food services and drinking places (waitresses, bartenders, and baristas), ambulatory healthcare services, hotels, and social assistance (Table I-1). Table I-1Social Distancing Is Destroying Jobs That Require Close Personal Interaction A profound consequence arises because these are the very sectors that have kept jobs growth alive in recent decades (Table I-2). Millions of new jobs that rely on close personal interaction have more than offset the structural job destruction in manufacturing and finance. As well as being export-proof, jobs that require this close personal interaction have been ‘artificial intelligence (AI) proof’. That is, until now. Table I-2Jobs That Require Close Personal Interaction Have Been The Engine Of Jobs Growth One UK doctor told the New York Times “we’re basically witnessing 10 years of change in one week”. Before the virus, online consultations made up only 1 percent of doctors’ appointments. But now, three in four UK patients are seeing their doctor remotely. Moravec’s Paradox + Social Distancing = A Very Tough Jobs Market Regular readers will know that one of our mega-themes is the far-reaching societal and economic implications of Moravec’s Paradox. Named after the professor of robotics, Hans Moravec, the paradox points out that: For AI the hard things are easy, but the easy things are hard. By the hard things, we mean things that require ‘narrow-frame pattern recognition’ within a defined body of knowledge. For example, playing chess, translating languages, diagnosing medical conditions, and analysing legal problems. We find these tasks hard, but AI finds them effortless. By the easy things, we mean our social skills: empathy, sympathy, the ability to recognise and respond to emotional cues, and to express ourselves through complex movements. To us, all these things are second nature, but AI finds them very hard to replicate. The reason, it turns out, is that the higher brain that enables us to learn and play chess and solve similar abstract problems evolved relatively recently. Whereas the ancient lower brain that enables complex movement and the associated giving and receiving of emotional signals took much longer to evolve. As AI is just reverse engineering the human brain, AI has found it easy to replicate the less-evolved higher brain functions, but very difficult to replicate the skills that emanate from the deeply evolved lower brain. Millions of new jobs that rely on close personal interaction have more than offset the structural job destruction in manufacturing and finance. The far-reaching societal and economic implication is that we have misunderstood and mispriced what is difficult and what is easy. By reverse engineering the brain, AI is correcting this mispricing. So far, AI has been most disruptive to high-paying jobs requiring abstract problem-solving skills, such as in finance. AI has been less disruptive to jobs requiring close personal interaction (Table I-3). But if social distancing persists, then AI will disrupt those jobs too, especially during a recession. Table I-3New Jobs That Require Close Personal Interaction Have Offset Lost Jobs In Manufacturing And Finance Labour Market Disruption Intensifies During A Recession To paraphrase Ernest Hemingway, industries adopt labour-saving technologies gradually then suddenly. And the suddenly tends to be during a recession. This is because once an industry has already shed many workers, it is easier to restructure the industry with a new labour-saving technology that reduces labour input permanently. At the start of the Great Depression a substantial part of the US automobile industry was still based on skilled craftsmanship. These smaller, less productive craft-production plants were the ones that shut down permanently, while plants that had adopted labour-saving mass production had the competitive advantage that enabled them to survive. The result was a major restructuring of the auto productive structure. Likewise, until the late 1990s, the ‘typing pool’ was a ubiquitous feature of the office environment. But once the 2000 downturn arrived, these typing jobs became extinct to be replaced by the wholesale roll-out of Microsoft Word. After the 2008-09 recession, UK economic power became focussed in a few large firms that could access the finance to ensure their survival. As small firms went by the wayside, job growth came disproportionately from self-employment and the ‘gig economy’. In this case, the labour market disruption hurt productivity as an army of freelancers ended up doing their own sales, marketing and accounts in which they had no specialism (Chart I-1 and Chart I-2). Chart I-1The 1990s UK Recovery Produced No Increase In Self-Employment... Chart I-2...But The 2010s UK Recovery Produced A Huge Increase In Self-Employment The point is that all recessions produce major structural changes in the labour market and the current recession will be no different. If social distancing persists, it will nullify the social skill advantage that humans have over AI. Therefore, one structural change will be that AI disrupts the more ‘human’ job sectors that have so far escaped its penetration. All recessions produce major structural changes in the labour market. To repeat, labour market disruption arrives suddenly. Within the space of a few weeks, most UK patients have switched to receiving their medical care online or by telephone. Admittedly, the patients are still ‘seeing’ a human doctor, but the question and answer consultations are a classic example of narrow-frame pattern recognition. Meaning that it would be a small step to upgrade the human doctor to the superior diagnosis from AI. And if AI can produce a superior diagnosis to your human doctor, why can’t AI also produce a a superior legal analysis to your human lawyer? The Investment Implications Even when the labour market seemed to be humming and unemployment rates were at multi-decade lows, aggregate wage inflation was anaemic (Chart I-3 and Chart I-4). A major reason was the hollowing out of high paying jobs and substitution with low paying jobs. Now that unemployment rates are surging, and AI is penetrating even more job sectors, aggregate wage inflation is set to collapse – not just temporarily, but structurally. Chart I-3Unemployment Rates Have Been At Multi-Decade Lows... Chart I-4...But Wage Inflation Has Been ##br##Anaemic This leads to the following investment implications: 1. All bond yields will gravitate to their lower bound, so any bond yield that can go lower will go lower. 2. It follows that bond investors should continue to overweight US T-bonds versus the core European bonds in Germany, France, Netherlands, Switzerland and Sweden (Chart I-5). Chart I-5Any Bond Yield That Can Go Lower Will Go Lower 3. Underweight banks structurally. Depressed and flattening yield curves combined with shrinking demand for private credit constitutes a strong headwind. Banks are now underperforming in both up markets and in down markets (Chart I-6). Chart I-6Banks Are Underperforming In Both Up Markets And Down Markets 4. Overweight technology structurally. As AI penetrates even more job sectors, the superstar companies of big tech will continue to thrive. The duopoly of Apple and Google are designing proximity-tracking apps for every smartphone in the world. Big tech is laying down the law to governments, and there is not even a hint of antitrust suits. Tech is now outperforming in both up markets and in down markets (Chart I-7). Chart I-7Tech Is Outperforming In Both Up Markets And Down Markets 5. Finally, if big tech outperforms banks, the sector composition of the S&P 500 versus the Euro Stoxx 50 makes it inevitable that the US equity market will structurally outperform the euro area equity market (Chart I-8). Chart I-8If Big Tech Outperforms Banks, The S&P 500 Must Outperform The Euro Stoxx 50 Fractal Trading System* The steep decline in the US 30-year T-bond yield means that it has crossed below the Australian 30-year bond yield for the first time in recent history. Resulting from this dynamic, this week’s recommended trade is long the Australian 30-year bond versus the US 30-year T-bond. Set the profit target at 9 percent with a symmetrical stop-loss. Chart I-930-Year Govt. Bonds: Australia Vs. US In other trades, long IBEX versus Euro Stoxx 600 hit its 3 percent stop-loss, while long nickel versus copper is half way to its 11 percent profit target. The rolling 12-month win ratio now stands at 63 percent. When the fractal dimension approaches the lower limit after an investment has been in an established trend it is a potential trigger for a liquidity-triggered trend reversal. Therefore, open a countertrend position. The profit target is a one-third reversal of the preceding 13-week move. Apply a symmetrical stop-loss. Close the position at the profit target or stop-loss. Otherwise close the position after 13 weeks. * For more details please see the European Investment Strategy Special Report “Fractals, Liquidity & A Trading Model,” dated  December 11, 2014, available at eis.bcaresearch.com. Dhaval Joshi Chief European Investment Strategist dhaval@bcaresearch.com Fractal Trading System   Cyclical Recommendations Structural Recommendations Closed Fractal Trades Trades Closed Trades Asset Performance Currency & Bond Equity Sector Country Equity Indicators Bond Yields Chart II-1Indicators To Watch - Bond Yields Chart II-2Indicators To Watch - Bond Yields Chart II-3Indicators To Watch - Bond Yields Chart II-4Indicators To Watch - Bond Yields   Interest Rate Chart II-5Indicators To Watch - Interest Rate Expectations Chart II-6Indicators To Watch - Interest Rate Expectations Chart II-7Indicators To Watch - Interest Rate Expectations Chart II-8Indicators To Watch - Interest Rate Expectations