Moral hazard, the fear of the markets, and how central banks responded to Covid-19
Mattia Bevilacqua, Lukas Brandl-Cheng, Jon Danielsson, Jean-Pierre Zigrand 28 January 2021
The direct economic consequences of Covid-19 have been significant (Baldwin and Weder di Mauro 2020) but the impact on the financial markets has been more nuanced, with considerable disruption but no crisis – unlike in 2008 – perhaps thanks to prompt central bank responses.
While central bank interventions did alleviate immediate market fears, it remains an open question whether that success came at the expense of higher longer-term moral hazard and the increased global importance of the US Federal Reserve.
Shedding more light on those issues motivates our companion work (Bevilacqua et al. 2021), where we document how the various categories of Fed policy responses at the height of the Covid-19 crisis affected market fears across a wide range of countries, maturities, levels of losses and assets.
Data and methodology
Our empirical investigation is based on a uniquely rich global options data set provided by the data vendor IHS Markit, containing maturities ranging from one week to 30 years, strike prices from drops up to 80% all the way to 200% price increases, across 242 indices and 3,334 individual stocks. We use the options data to obtain a time series of the risk-neutral distribution of market outcomes – the term structure of fear – the main ingredient in our empirical analysis. We show a snapshot of the data in Figure 1.
Figure 1 Probabilities (risk-neutral) of a 20% or higher drop in the S&P 500 over the next month, year and decade.
The difference between how the markets saw the 2008 and the Covid crises is remarkable. The increase in fear was much more fleeting in 2020, and then mostly manifesting in short-term risk.
In order to better understand the reactions to the Fed interventions in the Covid crisis, we classify them into five categories.
1. Wider economy support: Credit to households, businesses and the public sector
2. Interest rate changes and forward guidance
3. Liquidity support for domestic financial markets
4. Macroprudential regulations
5. US dollar swap lines and FIMA repo facility
We run a series of regressions, with either the one day change in the perceived quantile of expected losses or volatility (VIX) as dependent variables and policy announcements and controls as exogenous variables, and do this across a wider range of maturities and levels of losses, identifying the policy shocks by their immediate impact. Below we only report the quantiles changes, termed fear, but the companion paper does both.
We start by directly identifying how the Fed policies affect fear in the central US stock market index, the S&P 500. As it turned out, neither wider economy support nor interest rate policy announcements had any economically significant impact on market fear, regardless of maturity.
Figure 2 shows the impact of the other three categories on market fear (tail losses, i.e. quantile) associated with 10% risk-neutral probability across maturities from two weeks to ten years, for the three significant types of policy.
Figure 2 Impact of Fed Covid-19 policy interventions on S&P 500’s fear of losses (negative of the change of the 10% quantile)
The negative values tell us fears fell significantly in response to the Fed interventions. For instance, at the three-year horizon, an average normalised US dollar swap announcement lowers the 10% fear of losses by 11%
Policies directed at financial institutions overall had a strong impact on market risk across maturities, highlighting the perceived long-term consequences of improved short-term money market liquidity today.
The strongest policy, measured by how it affected market fear, are the FX swap lines, whose impact is at least twice as large as that of liquidity and macroprudential easing.
To augment the aggregate results from the S&P 500 index, we also repeat the analysis for individual stocks, showing below the results for the largest bank in the US, J.P. Morgan, and its largest retailer, Amazon. The size of the dots again reflects their statistical significance.
Figure 3 Impact of Fed policies on 10% quantile of Amazon and J.P. Morgan stock
Not surprisingly, the impact on J.P. Morgan is much stronger than on Amazon across all four policy categories, as the immediate benefits of Fed policies would likely accrue to financial markets and its institutions rather than to retailers.
With the exception of interest rate policies, interventions affect market fear across all maturities, and unlike for the broad stock market index S&P 500, there is little discernible difference between how the three policy categories affect fear.
The impacts of the macroprudential relaxations on Amazon and J.P. Morgan is of particular interest. It is not surprising that macroprudential relaxations for financial companies reduce fears of Amazon shareholders, say through less troublesome future bank financing. However, the effects on J.P. Morgan are less obvious. As these policies are explicitly designed to relax the risk constraints on regulated banks, increasing their riskiness, one could have expected the likelihood of tail losses – especially in the long term – to increase, not decrease.
However, the markets perceived the macroprudential relaxations as reducing long-term fear. One possible explanation is that the Fed is viewed as signalling that it will intervene again in the future, so the market has a lower willingness to pay for private disaster insurance provided via options. This then is an indication of the moral hazard caused by Fed crisis interventions.
We have further results on the global impact of the Fed’s Covid-19 policy interventions on stock markets’ fear around the world.
Figure 4 Impact of Fed policies on fear (10% quantile) in key global markets
While the global impact of the Fed’s FX swaps is similar to the domestic impact, we see significant heterogeneity. We have countries that always had access to them (such as the UK, Japan and Germany), those who never did and those who got them in the middle of the crisis (like Korea).
The dollar swap lines were particularly effective in reducing long-term fear in countries lucky enough to have confirmed access or have gained access to the facility. Meanwhile, we show in our main paper that the swap line announcements had no discernible impact on countries without access.
Central bank relevance
The Fed swap lines were implemented by domestic central banks acting as vehicles for the execution of US monetary policy. The swap lines are thus a double-edged sword. While they immediately calmed the markets, they also incentivised local markets to continue relying on a funding currency that the local central bank does not control, perhaps generating further risk and moral hazard, and reinforcing the pivotal role of the US central bank.
We suspect the Fed’s primary objective was to alleviate immediate market fear, pour oil on troubled waters, targeting short-term risk but not specifically the long-term.
However, some of the strongest impacts the Fed interventions had were a reduction in long-term market risk, years even decades into the future.
That may be suggestive of a sharp increase in moral hazard. If the market sees the central bank as willing and able to calm short-term fear in March 2020, it also expects the same in the future, which may lead today to less vigilance and further build-ups of risky positions, both in the US and globally.
A common narrative maintains that prompt central bank responses in March and April 2020 saved us, and if they had reacted as decisively back in 2008, that crisis could have been averted as well.
We partially agree, but only if viewed through the lens of static equilibria. These two crises were different as shown by Danielsson et al. (2020), and a more dynamic analysis looking at multiple horizons, reveals a more complex picture.
The US dollar swap lines had both a short- and long-term impact on global market fear (tail risk). We surmise that there are two reasons for this.
The first was to reassure US dollar borrowers that they would not be starved of the dollars needed to fulfil their obligations, important as the market saw dry-ups in dollar liquidity as a significant concern, as noted by Boissay et al. (2020).
But secondly, and even more importantly, they reaffirmed the US’s commitment to the global financial community, even when such commitments have at the time come under increasing strain elsewhere in US policymaking.
However, success comes at a cost.
Powerful central banks such as the ECB, the Bank of England, and the Bank of Japan act as the conduit of US dollar liquidity to their own domestic markets.
The US dollar swaps may further encourage global dollar borrowers to continue business as usual and even intensify their dependence on the dollar. The resulting moral hazard would then incentivise market participants to take even more liquidity risk in the future, in the strength and belief that the Fed is ready willing and able to step in and bail the markets out.
This is at the root of Minsky’s dictum that “stability is destabilising”.
By calming the markets now, they come to expect the same to happen in the future and thus are incentivised to take on more risk than they otherwise would do.
The end result may be increased systemic risk.
Authors’ note: We thank the Economic and Social Research Council (UK) (grant number ES/K002309/1) for supporting this work and special thanks to IHS Markit for providing the options data essential to this work.
Baldwin, R and B Weder Di Mauro (2020), Mitigating the COVID economic crisis: Act fast and do whatever it takes, CEPR Press.
Bevilacqua, M, L Brandl-Cheng, J Danielsson, L Ergun, A Uthemann and J-P Zigrand (2021), “The calming of short-term market fears and its long-term consequences: The Federal Reserve’s reaction to Covid-19“, SSRN.
Boissay F, N Patel and H S Shin (2020), “Trade credit, trade finance, and the Covid-19 Crisis”, BIS Bulletin 24.
Danielsson J, R Macrae, D Vayanos, D. and J-P Zigrand (2020a), “The coronavirus crisis is no 2008”, VoxEU.org, 26 March.