Finance: Research, Policy and Anecdotes

I have completed another , this time on the effect of the pandemic and lockdown policies on banks’ health, lending growth and loan pricing in the US.  Jointly with Jan Keil, we exploit geographic variation in the exposure of US banks to COVID-19 and lockdown policies using branch network and branch-level deposit data.  We find first that unemployment rates (the most accurate and most rapidly available indicator of economic activity) co-varies significantly across counties in the US in the first three quarters of 2020 with COVID outbreaks and lockdown policies. Second, focusing on banks, we find that  banks geographically more exposed to the pandemic and lockdown policies show (i) an increase in loan loss provisions and non-performing loans, (ii) an increase in lending to small businesses, but not in other lending categories, and (iii) an increase in interest spreads and decrease in loan maturities. These results are consistent with previous findings that show a general increase in lending after the onset of the crisis (related mostly to the drawdown of credit lines by large firms), while the bank-specific increase in small business lending might be explained with demand for government support programmes by small businesses in the areas most affected.  In a nutshell, these findings show that banks have already seen the negative impact of the pandemic and have reacted to higher lending risk with an adjustment in loan conditionality, but have also responded to higher loan demand and government support programmes.

Last week saw the 7th edition of the Emerging Scholars in Banking and Finance conference, in virtual format given circumstances.  I do not have time or space to mention all the interesting papers, so herewith just some highlights:


In Monetary Policy Corporate Debt Maturity, Andrea Fabiani and co-authors show that a loosening of the US monetary policy rate lengthens corporate debt maturity, an effect entirely driven by the adjustments of very large firms. They explain this empirical finding with a theoretical model that combines standard financial frictions of moral hazard with short-termist, yield-oriented investors who rebalance their portfolios towards longer-term securities  when the policy rate and thus short-term rate descends. This effect seems to be driven by corporate bond funds that increase their holdings of corporate bonds in reaction to looser monetary policy and rebalance their holdings towards longer-term debt securities. It is then larger firms that can take advantage of this higher demand by issuing more longer-term bonds. As Andrea discusses in this blog entry, these findings can have quite important policy implications in times of increasing corporate debt and show another channel through which loose monetary policy and the search-for-yield channel affect firm financing patterns.



Leslie Shen presented a very interesting paper (Global Banking and Firm Financing: A Double Adverse Selection Channel of International Transmission) that explores theoretically and empirically global monetary policy transmission through cross-border lending to firms, both on the extensive (global vs. local banks) and the intensive (interest rate) margins. The key mechanism (for which she provides empirical evidence) is that cross-border lending relies more on information on the global dimension of borrowers’ returns, while domestic lending relies more on information on the local dimension of borrowers’ returns. This has important repercussion for the transmission of the monetary policy.  Take as example an increase in US interest rates: this results in more firms at the margin in the euro area switching away from US banks, while it raises (lowers) interest rates for firms that keep borrowing from US (Euro area) banks. This is a very nice paper that uses micro-data to test theories of international finance and links nicely to both the banking and the international macro literatures.


Carola Mueller and co-authors focus on the effectiveness of stress tests in Europe in The disciplining effect of supervisory scrutiny in EU wide stress tests.  Stress tests have been an important tool to test the resilience of banks against tail risk and guide supervisors in their dialogue with banks and setting pillar 2 capital requirements.  Using proprietary data on the EU-wide Stress Test conducted in 2016 by the European Banking Authority and the European Central Bank, the authors provide evidence that supervisory scrutiny that comes with stress testing can have a disciplining effect on bank risk, i.e., banks that received more supervisory scrutiny (as measured by the intensity of communication between banks and supervisors during the stress test exercise)  reduce risk weighted density more than banks that were under less intense scrutiny. On the other hand, there is no significant evidence that higher capital requirements resulting from the stress tests and higher publication requirements resulting in higher market discipline might have resulted in lower risk taking. Overall, this is compelling evidence in support of supervisory discipline. This paper provides complementary evidence to research from the US on the effect of stress tests on banks’ risk-taking (part of a special JFI issue I discussed earlier), but provides more detail on the actual channels through which stress tests work.

The last months have seen an explosion of research papers on the pandemic (to which I have done my small contribution, with another one coming up shortly).  Herewith three recent ones I liked especially, two of them by Thiemo Fetzer (who has achieved certain fame among political economist by linking austerity policies in the UK to the Brexit vote):  In Subsidizing the Spread of COVID-19: Evidence from the UK’s Eat-Out-to-Help-Out Scheme, he shows that government subsidies for the hospitality sector in England  (50% of food and non-alcoholic beverages on Monday through Wednesday in August, up to 10 pounds per client and order – and yes, yours truly also benefitted from it) are responsible for an increase in new COVID-19 infection clusters.   He does so by linking the number of participating restaurants in an area to mobility and restaurant booking data and COVID infection data, comparing Monday-Wednesday in August to the same weekdays in July and September. To control for any simultaneity bias, he uses rainfall data (during lunch and dinner hours) to show that rainfall dampened restaurant bookings and COVID infections during Monday-Wednesday in August but not the months before and after and not Thursday-Sunday in August (weekdays with no subsidies). His estimates suggest that the subsidy scheme may have been responsible for around 8 to 17% of all new detected COVID19 clusters emerging during August and into early September in the UK.  If this does not sound too dangerous, remember that due to the exponential growth of infection, this economic support programme might very well have contributed to the second wave (on 10 August, a week after the programme started there were 826 new cases in the UK, while there were 2,948 on 7 September, a week after the programme ended). While some might argue that this shows the trade-off between economy and public health, that would certainly be penny-wide and pound-fool, given the second lock-down we are currently in, plus the enormous economic (not to speak of human) costs of COVID patients.


Another depressing paper by Thiemo and Thomas Graeber considers the importance of a functioning track-and-trace system. In Does Contract Tracing Work? Quasi-experimental Evidence from an Excel Error in England, they exploit a unique quasi-experiment in England that generated exogenous variation in the intensity of contact tracing: On October 3, 2020, the government announced that due to a “technical error,” 15,841 COVID-19 cases that should have been reported between September 25 and October 2 had not entered the official case statistics and had therefore not been referred to the central contact tracing system (around 15-20% of all cases during this time period and randomly distributed across England)).  As Thiemo and Thomas show, this mistake had deadly consequences: In areas with higher exposure to the contact tracing mistake, they find a notable subsequent increase in COVID-19 infections and, with the usual delay, in COVID-19- related deaths, for a total of between 1,500 and 2,000 additional COVID-19-related deaths. One important channel was a sharp decline in the efficiency of the track-and-trace system in contacting positive cases, once they were overwhelmed with the additional 16,000 cases. Far from the world-beating track-and-trace system that Boris Johnson promised, England rather got a world-killing system.  


A final paper (Revenge of the Experts: Will COVID-19 Renew or Diminish Public Trust in Science?), gauges the effect that pandemics have on the public’s trust in scientists.  Using data from the 2018 Wellcome Global Monitor (WGM), which includes responses to from over 75,000 individuals in 138 countries, Barry Eichengreen, Cevat Giray Aksoy and Orkun Saka link individual responses to questions about trust in science and scientists to the incidence of epidemics since 1970. Focusing on individuals that were between 18 and 15 during an epidemic in their country (“impressionable late-adolescent and early-adult years”), they find that epidemics do not significantly affect trust in science, but reduce trust in scientists, with the consequence that these individuals are subsequently less likely to vaccinate their children. While the authors are careful to not use their findings to predict a similar reaction to the COVID-19 pandemic, their findings should clearly ring alarms across the globe.

The last two weeks have seen a sad political spectacle in the US, where an incumbent president who fairly and squarely lost his bid for re-election, is refusing to accept the results.  One might see this as just the last political chapter of a deranged Donald Trump who has never allowed to get reality in his way. Rudy Giuliani’s press conferences in front of a garden centre and in an overheated room (with his hair dye dripping down his face) will serve as material for Saturday night shows for years to come.  It is clear that the legal path through courts is close to failure.  But there is a much more sinister side to Trump’s attempt to overthrow the democratic verdict.  His attempts to openly pressure Republican legislators and state officials into setting aside legally cast votes and overturn popular majorities in their respective states is as close to a coup that the US has come over the past decades. And as much as one appreciates the few Republicans that are doing the obvious – recognising and congratulating Joe Biden as president-elect – the more worrisome is the fact that most Republican leaders are either silent or openly support Donald Trump in his attempt to overthrow the electoral verdict. It has become clear that the Republican party has taken yet another step towards turning itself into an authoritarian caudillo-type support group for one man.


It has also become clear over the past two weeks that these attempts to overthrow the electoral decision will almost certainly fail.  So, chapter closed, heal the nation and move forward?  As much as one can wish for that, I do not think that this will happen.  To the contrary, even if he fails to overthrow the results of the election, Donald Trump and his henchmen will have cast sufficient doubt on the legitimacy of Joe Biden’s presidency  - at least in the mind of a large minority of the US population. And that is where the danger for the republic stands.


I am aware that it is easy to reject history lessons by Germans for today’s political situation in the US, but there is a dangerous parallel to post-WW I Germany history.  It was the German military leadership who requested on 29 September 1918 that a new democratically legitimised government negotiate a cease fire with the Allied forces, given that German troops were facing defeat. At the same time, the centre-left parties, unions and Jewish groups were blamed for this defeat by having undermined the home support for the successful German troops. This stab-in-the-back myth has been debunked extensively, but was successfully used by conservative groups and parties and – most prominently – by Hitler’s Nazi party to delegitimise centre and centre-left parties that participated in most of the short-lived governments  of the Weimar Republic and ultimately the Weimar Republic itself. It served as justification for more than 300 murders by the terrorist group Organisation Consul.


If one considers the infamous (hair dye) conference, there are scary parallels to the stab-in-the back myth – in the case of the US elections, it seemed to have been Venezuelan and Cuban communists, George Soros (the inevitable anti-Semitic link) and “lock-her-up” Hillary Clinton (how could she not be involved?) who are responsible for the election outcome. And Donald Trump’s landslide win (according to his lawyers) resembles the undefeated German troops in 1918.


“History doesn’t repeat itself but it often rhymes,” Mark Twain is reputed to have said. The risk is not that Donald Trump and his henchmen will be able to overthrow the results of a democratic election, but that they manage to undermine the legitimacy of democratic elections themselves.

It was not a landslide, but it was also not tight – the drawn-out US elections were only drawn-out because of the peculiarities of the US election system, organised on the county and state-level, underfunded and technologically challenged. The elections were a clear vote against Trump (with the popular vote difference most likely amounting to four or five percentage points once all votes have been counted). The obvious question for many outside observers (and many Americans) is why the vote difference was not even larger – Donald Trump received more votes than he did in 2016. Why would voters support an authoritarian crook? Lots has been written about this; one answer lies in one of my summer readings:  The Righteous Mind, by Jonathan Haidt, a social and cultural psychologist who argues that moral judgments and political opinions arise not from logical reasoning, but from gut feelings and “groupish righteousness”, in turn driven by culture; liberals, conservatives and libertarians have different intuitions about right and wrong as they prioritize different values. As powerful as this explanation is, however, over the eight years since this book has been written, another important development has taken place: alternative facts.  Watching CNN and Fox one might get the impression that they report about different countries. What in the UK are the open lies and mis-representations by the Brexit press, are right-wing radio and TV stations in the US – the result: conspiracy theories about child-trafficking rings, Bill Gates, COVID-19 etc.


Will this  be the end of Trump and Trumpism? Well, a lot will depend on what happens next to private citizen Donald Trump. He is facing a fair number of lawsuits that he was able to delay by being a sitting president. The outstanding tax charge will most likely come up shortly as will the billion or more of debt repayments. So, in one thing he was right: the presidency might have cost him dearly.  There are talks about him running again in 2024 – he would be 78, as old as Joe Biden will be in a few weeks. A lot will depend on what will become public in the next few years on his shady business dealings over the past decades, his tax evasion and his corruption. There has been only one previous president (Grover Cleveland), running for a second non-consecutive term after having lost reelection.  While one might think that today’s politics is too fast-moving (there has not been a losing presidential candidate presenting himself for a second time since Richard Nixon in 1968) for Donald Trump to return as candidate, it seems that he has managed to build a personality cult that resembles more that of a religious sect than a political party, so one cannot exclude this possibility. And being charged (and possibly convicted) in several cases might actually turn him into a martyr. The biggest post-election soul searching is certainly for the Republican party, even though they seem in a pole position to maintain control of the Senate and have gained seats in the House. Which way should they go – follow the populist authoritarian personality cult of Trump or return to a somewhat more sober version of conservatism a la Mitt Romney and Paul Ryan (anyone remember the former Speaker)?


As shocked as many of us were on the morning of 9 November 2016, as relieved are we in the afternoon of 7 November 2020 – I will certainly remember where I was when the election was called for Joe Biden.  My expectations are not high, but as many I hope for a stop of presidential hatred, foul language, and racism and for more respect for institutions and norms (it is rather bizarre that Republican Trump enablers focus so much on an originalist interpretation of the constitution but have supported Trump in the destruction of long-standing norms and institutional arrangements). Just a return to normalcy would be welcome, both domestically within the US and on the international level. One can hope for more competence in the US government – as Trump kicked out the B-team running government during the first two years, we are now at the C-team.  This utter incompetence is most prominently illustrated by the Trump campaign’s invitation to a press conference at the Four Seasons….  Total Landscaping company in Philadelphia, located between a crematorium and a sex shop, rather than at the Four Seasons hotel as one might have expected. Finally, one can hope for less corruption and theft in the government, as soon as the crook and his henchmen have left the White House.


There are lots of implications of this elections beyond the U.S.  Closest to (my current) home, the EU-UK trade negotiations. For a long time, the British government has hoped for an easy and early trade deal with the US. While this was naïve from the start, the defeat for Donald Trump buries these plans definitely.  Joe Biden’s Irish roots and the importance of Ireland in US politics (to remind ourselves: the Good Friday Agreement was negotiated by former US senator George Mitchell) will make a US-UK trade agreement only possible if it does not undermine the open border between Northern Ireland and the Republic.  However, the fading dream of an easy and early US-UK trade deal makes a EU-UK trade deal politically even more important for Boris Johnson (obviously also for economic reasons, but we know how Johnson thinks about business). So, I think the odds of a last minute deal between the UK and the EU just went up.


But there will certainly be a reset in many relationships across the globe, starting from the pressing issue of the trade war with China (and others) and the more general competition with China to the conflict with Iran (which I have a small personal stake in, as a former PhD student is living and working there). It is naïve to think that the US-EU relationship will go back to what it used to be, but at a minimum one can expect them to be on a more rational and civilised basis; similarly, the conflict with China will continue, but maybe with less megaphone and more direct talks.