+1 Added to my documents.
Please be aware your selection is temporary depending on your cookies policy.
Remove this selection here

Thematics Views - June 2020



Controlling the yield curve during the recovery phase

State of the US consumer: weakened but resilient


June 2020


Juin 2020


Auteurs 1

The ECB would face operational diffculty and political obstacles

Controlling the yield curve during the recovery phase

As part of their toolkit to support the economic recovery during the Covid-19 crisis, central banks could implement yield-curve control. Although appealing, the implementation and exit risks of such a policy counterbalance the benefits, particularly in the Eurozone. Moreover, the impact on financial markets could be significant since chained risk-free assets could temporarily leave risky assets unsettled.

With central bank (CB) key rates close to zero, yield-curve control (YCC) policies are back in the spotlight. Under such a policy, the CB seeks to control a particular point or segment of the yield curve (often the 10-year maturity). To do so, the CB implicitly commits itself to buy (or sell) the amounts of securities needed to achieve its objective. If the CB is credible, thanks to market arbitrage, bond yields tend to move close to their target. In practice, YCC is often similar to a form of QE but with very different modalities, in that security purchases adjust to bond yields and not the other way around. The CB loses the control of its balance sheet. The yield peg could even lead the CB to shrink its balance sheet. This could be the case if private-investor demand exceeds financing needs (new issuance and refinancing).

YCC is unlikely for the ECB

YCC was implemented in the US in the post-war period (until 1947), more recently in Australia and since 2016, the BoJ has explicitly targeted its 10-year yield at zero. This policy, combined with negative key rates, has been successful in Japan because it has allowed the BoJ to reduce its bond purchases and contain its balance sheet. YCC is conceivable in “true” monetary unions like US, Japan or the UK where there is only one (central) government debt instrument.

On the other hand, it is not feasible in the Eurozone for at least two reasons. First, the ECB would have to manage several yield curves simultaneously. On top of the operational difficulty, there would be a political obstacle: the ECB has no legitimacy to estimate the credit risk embedded in each sovereign debt. Secondly, the ECB would have a legal problem because it would have to commit to buying assets without any rule of proportionality, or even to monetise public debt, which is prohibited by the Lisbon Treaty. That said, if the Fed and/ or the BoE opt for YCC, that would likely have repercussions on German yields, too, as investors would arbitrate the spreads between nations. Lastly, the exit of YCC bears significant risks as the bond market could suddenly adjust and potentially offset the benefits.




note thematic

When jailed risk free rates unleash risk assets

For the Fed, a temporary control of the short-end of the curve aligned with the “dots”, combined with QE on the mid-tolong- end, looks like an efficient strategy both in terms of cost/benefit and exit risk. A YCC + QE combination means that the adjustment of the “risk-free asset” to the nominal economic cycle will not happen, at least temporarily, unless there is a shift in the term premium. These yield cap policies should therefore affect the comovement of financial assets (volatility and correlation) during the recovery phase away from their historical norm and boost unanchored risk-assets returns.

In effect, if Treasury bill and bond yields remain low and stable, and the curve stays flat while investors’ expectation for growth improves, “uncontrolled assets” such as equities should benefit from artificial high valuations. These assets will become the market proxy for changes in nominal growth expectations while the correlation of their returns with bonds will get closer to zero. Yet, equities would also suffer from a more acute version of the volatility regime we have seen over the last 3 years with long periods of very low volatility and sudden spikes to extreme levels. Moreover, risk free rates should also influence IG credit offering higher expected returns, more than HY credit on a volatility-adjusted basis. Lastly, as the yield curve stays flat regardless of the economic recovery, EM HCD would find valuation support and inflows from investors hunting for yield.

However, the uncertainty around the “natural” risk free rate should also lead higher risk premia, lower crossasset correlations and potentially lower diversification benefit for balanced strategies. Therefore, YCC + QE could put the US economy back on track quicker, but with the risk derailing financial markets at least temporarily.

Finalised on 22/05/2020

Auteurs page 2

As total labour income falls sharply, personal consumption is expected to collapse in Q2, in a size well beyond the depths seen in the GFC

State of the US consumer: weakened but resilient

Massive government support and healthy consumer balance sheets are providing the means for the consumer to weather a relatively short but very deep recession. We are closely watching the magnitude and duration of unemployment to gauge the consumer’s debt-repayment capacity. The government backstop continues to provide stability for the financial markets and we believe it is a fertile environment for active management.

As the US plunges into its worst recession in the post-WW2 era, we examine four key questions: (1) How has the crisis affected the US labour market so far? (2) What condition is the US consumer in? (3) What are the primary debt vulnerabilities? (4) What are the prospects for defaults in mortgages, credit cards, and auto loans? How has the crisis has affected the US labour market so far?

The Covid-19 crisis has led to unprecedented rise in jobless claims and payrolls cuts, erasing almost completely the jobs created in the longest economic expansion on record. Looking at the available data, job losses are concentrated so far in sectors on the “front line” of the crisis, i.e. services. The most affected both in both absolute and relative terms are leisure and hospitality, followed by trade, transportation and utilities. In March-April more than 21 million US jobs were lost. They were concentrated in the private service sector (84% of the total jobs lost in the economy), with almost 18 million layoffs, i.e., 17% of pre-Covid-19 available jobs. The service sector accounts for approximately 71% of the U.S. labour market. So far, layoffs have affected people and sectors with mostly the lowest salaries, and this has created a compositional shift, which has boosted temporarily average hourly earnings both in aggregate and within each category. Yet, we expect this phenomenon to be only temporary. As total hours worked in the economy decline due to a combination of lower hours and lower employment, the total labour income proxy we track to understand future trends in consumption is expected to collapse in Q2, to an extent well beyond the depths seen during the GFC.

How the US consumer will resurface from this shock relates critically to persistent weakness in the labour market. In April records, a rather large proportion of people were reported to be on temporary lay-off rather than unemployed. In order to assess how fast the labour market will be able to recover, it will be crucial to monitor how quickly continuing claims will be falling in reopening states. In addition, while all eyes will be on the unemployment rate, it is worth mentioning that this indicator cannot fully capture the weakness in the labour market, and in our opinion, should be monitored closely together with the labour force participation rate. The LFPR in April dropped by 2.5pp to 60.2% (-0.7pp in March), and will be a key variable to watch in discerning the persistence of the shock, i.e., whether he decline is a temporary response to the restriction of people’s mobility or a signal of much greater weakness in the labour market).

Tableau 1 page 2

We expect the US economy to contract between -4.5% and -6.5% in 2020, and the unemployment rate to rise up to 21% between May and June. We then assume a gradual decline throughout the year but maintaining the average unemployment rate in a range of 10 to 12% on average in 2020.

US consumer fundamentals have been strong, especially the household balance sheet

Consumers started from a position of strength

US consumer fundamentals have been strong, especially household balance sheets. According to the Fed, household debt as a percentage of disposable personal income fell sharply from a peak of 133.5% in Q4 2007 to 97.1% in Q4 2019. (Chart 3). Household debt-service payments, or the ability to pay interest on debt owed as a percentage of disposable income, stood at 9.7%, the lowest since the inception of the series in 1980. Finally, the personal savings rate stood at 9% in February, the highest level since 2012, and the early indication from the pandemic is that consumers have sharply increased their savings, with the latest savings rate skyrocketing to 13.1% in March. This is the highest level since 1981.

Graph 1 page 3

The Covid-19-related lockdown of the US economy and health concerns have weighed heavily on the US consumer. The Conference Board (CB) Consumer Confidence Index declined from 118.8 in March to 86.9 in April, its lowest level since May 2014. In particular, consumer sentiment reflects labour market weakness. The labour differential, i.e., the difference between jobs plentiful minus jobs hard to get, fell from +29.5 in March to -13.6 in April (Chart 4). Thirty-three million people lost jobs between March 20 and May 1; 61% of consumers surveyed were anxious about job security, and 78% said their household income had been negatively affected by the pandemic recession, according to an Experian poll.


However, consumers remain optimistic that conditions will improve within six months, with those expecting a net rise increasing from +2.3 in March to +14.3 in April, according to the CB. The May 12 National Federation of Independent Business survey found that 78% of unemployed workers expect to be rehired before year-end. The government’s fiscal stimulus packages and more generous unemployment insurance benefits may be serving to counter some consumer trepidation.

Graph 2 page 3

We see three areas where the consumer may be more vulnerable: credit card debt, auto loans and student loans

Key debt vulnerabilities

Unsurprisingly, many Americans are becoming concerned about their ability to pay their debts, such as student loans, mortgage payments and credit card bills. The rate of new household delinquencies, defined as missing a payment for 30 days or more, was 4.61% in Q1 2020. Our analysis from the New York Federal Reserve Bank’s Household Debt and Credit Report revealed that total household debt rose from 3.5% year-over-year in Q1 2019 to 4.6% in Q1 2020, the fastest growth in nearly three years. However, household debt to GDP rests considerably lower today at 75.4, than the highs of 98.6 during the global financial crisis. We expect loan growth to slow dramatically.

We see three areas where the consumer may be more vulnerable: credit card debt, auto loans and student loans. While residential mortgages do not appear to be stressed, based on our analysis, the severity of this recession is increasing risk in this area (Chart 5). The CARES Act allows borrowers of federally insured mortgage loans facing financial difficulties to seek up to one year of forbearance. Through April, holders of more than 7% of all mortgages chose to postpone payments for 90 days, according to Black Knight, a mortgage research firm. That said, there is evidence that consumers continue to make payments even on loans in forbearance. Student loans have the highest delinquencies, but the US government has automatically suspended principal and interest payments on federal student loans through September 30, 2020, equating to 85% of all student loans.

Overall credit card, auto and residential mortgage delinquencies remain stable according to the Fed’s quarterly report on household debt.

  • Credit card delinquencies rose modestly from 6.71% in Q1 2019 to 6.84% in Q1 2020, a gain of 1.77% since the trough in Q2 2016. The rate of serious delinquencies (missing a payment exceeding three months), nudged up from 5.04% to 5.31% in Q1 2020. The card balances that went delinquent remained unchanged from one year ago at 4.6% in Q1 2020.
  • Auto loan delinquencies improved from 7.07% in Q1 2019 to 6.89% in Q1 2020. Shorter-term auto loans and those for used cars tend to be made to more conservative buyers who repay more reliably.
  • Residential mortgage delinquencies slipped from 3.5% in Q1 2019 to 3.48% in Q1 2020. Rising house prices (and ensuing lower LTV), steady income growth and an overall robust economy in 2019 helped lower the delinquency rate in Q4.


Assessing consumer debt risk

We are using a framework that provides an economic-adjusted measure of asset performance by comparing the percentage of loans becoming newly delinquent per 100,000 initial jobless claims. When the ratio is high, it indicates that job losses are causing a higher-than-average increase in delinquencies, signalling weak credit performance given the state of the economy. This ratio is a good tool to assess potential vulnerabilities to debt payment, adjusted for the state of the economy.

According to most economists, consumer credit performance is strongly correlated with initial jobless claims or other labour market metrics like the unemployment rate. In our analysis, we found that credit cards and auto loans were at risk of delinquencies, breaching the weak economic-adjusted performance territory. The heightened level of 4Q19 auto delinquencies is puzzling, since a robust economy and low unemployment over the last few years should have resulted in a declining delinquency trend. Subpar income growth and an increase of auto loans for subprime borrowers with low credit scores could explain the rise in delinquencies.

On the other hand, first-quarter results indicate that solid economic-adjusted performance of residential mortgages improved. The rate of new delinquencies  was flat, while initial unemployment claims rose modestly in Q4 2019. This suggests there may not be measurable stress in the mortgage market.

Graph page 4

The magnitude of defaults will depend on the length and severity of the Pandemic Recession

The prospects of default in mortgages, credit cards and auto loans

US household debt increased by $155bn in Q1 2020 to $14.3 trillion, led by mortgages ($9.7 trillion), student loans ($1.54 trillion), auto debt ($1.35 trillion) and credit card loans ($0.89 trillion).

To help us determine default risk, we regressed annual changes in the unemployment rate vs. the 90-day delinquency rate since 1999. We utilized linear trend lines to reveal the anticipated relation between rises in loan defaults and unemployment. Our analysis revealed that with a 1% rise in the unemployment rate, defaults might rise 0.1%, 0.9% and 0.9% for auto, credit card and mortgage loans, respectively. Using Bloomberg’s weighted average forecast for unemployment of 9.7% by end-2020, we estimate the 90-day delinquency rates of 3.0%, 10.6% and 6.5%, respectively, vs. 2.4%, 5.3% and 1.1% at Q4 2019, and the 2009 highs of 3.5%, 11% and 8.4% in a typical recession cycle.


Why this time may be different

Our potential default analysis assumes conditions of a typical recession; however, this is far from “normal”.

  • The nonfarm payroll data revealed that temporary layoffs tallied 18.1 mil, or almost 90% of the job losses. The CARES Act makes it easier to claim unemployment benefits by lifting the requirement to actively look for work. This could mean a quicker reversal in the labour market as the economy reopens.
  • The $600/week increase in unemployment benefits on top of state benefits has led to an unusually high number of people experiencing income improvement. State & federal benefits average $978/week, vs. pre-crisis benefits of $378/week. Pre-crisis, more than half of workers brought home less than $957/week, according to Countable.com. The New York Times reported recently that “workers in more than half of states will receive, on average, more in unemployment benefits than their normal salaries”. This will support consumer’s ability to pay bills. Another positive is  that the CARES Act has increased the generous unemployment benefits for an additional 13 weeks, for a total of 39 weeks of benefits.
  • Consumers plan to spend and save the direct payment checks and unemployment benefits more carefully. According to the Magnify Money LendingTree consumer survey of April 2020, 42.6% plan to pay bills, 28.5% to pay their rent/mortgage and 26% to save. This behaviour goes a long way in mitigating broad-based defaults.



The Covid-19 crisis has generated a wave of layoffs, which may well push the unemployment rate to a peak above 20%. Yet, as many layoffs are registered as temporary, a faster than usual recovery could follow as states reopen and economic activity resumes. Although we are not expecting the labour market to heal fully, we do expect the unemployment rate to retrace lower in H2, alongside economic growth. In this scenario, we should monitor two key aspects: (a) the rate of change in jobless claims in reopening states, to measure how fast unemployed people return to back work; and (b) the labour force participation rate, to evaluate the persistence of the damage created to the labour force. A healthy labour market, although not the level of full employment seen in early 2020, will be key in assessing the resilience of the US consumer.

Despite the severity of the Covid-19 Pandemic Recession, we believe delinquencies may fall short of the rise seen in a typical downturn, especially considering the previous robust economy and the lowest unemployment rate in 50 years. The potential temporary nature of unemployment, the unprecedented levels and duration of government benefits, and consumer frugality should help to minimize the prospects of large-scale defaults.

Ultimately, the magnitude of defaults will depend on the length and severity of the Pandemic Recession. A sharp recession but a quicker reduction in the unemployment rate will lessen the default prospects. As long as the peak in unemployment is transitory, we believe the consumer balance sheet will remain healthy enough for the US consumer to weather a relatively short but very deep recession.

BLANCHET Pierre , Head of Investment Intelligence
BOROWSKI Didier , Head of Global Views
UPADHYAYA Paresh , Director of Currency Strategy, US Portfolio Manager, US
USARDI Annalisa , Senior Economist
Send by e-mail
Thematics Views - June 2020
Was this article helpful?YES
Thank you for your participation.
0 user(s) have answered Yes.
Related articles