Jon Danielsson, Robert Macrae, Dimitri Vayanos, Jean-Pierre Zigrand 26 March 2020
Coronavirus is having a strong impact on the world economy.
What about the financial system? Is a systemic financial crisis likely, perhaps along the lines of the crisis 12 years ago in 2008, or even the Great Depression of the 1930s?
There are certainly similarities: widespread bankruptcies, liquidity shortages, large losses and some financial institutions may fail.
Still, that in itself does not mean a systemic crisis is inevitable or even likely.
By using the lens of endogenous risk, we can compare and contrast the coronavirus crisis with 2008, and in the process, gain some intelligence on the likelihood of a systemic crisis as a consequence of COVID-19, as well as the appropriate policy response (Baldwin 2020, Baldwin and Weder di Mauro 2020).
Endogenous and exogenous systemic risk
We first proposed the classification of financial risk into an endogenous and exogenous in 2002 (Danielsson and Shin 2002).
Exogenous risk arrives to the financial system like an asteroid might hit earth – it comes as a surprise, there is nothing we do to precipitate its arrival, and it can cause enormous damage. To the financial system, the coronavirus shock is purely exogenous.
However, the distinguishing feature of all serious financial crises is that they gather momentum from the endogenous responses of the market participants themselves. Rather like a tropical storm over a warm sea, they gain more energy as they develop. As financial conditions worsen, the willingness of market participants to bear risk seemingly evaporates and liquidly evaporates. They curtail their exposures and generally attempt to take on a more prudent, conservative stance (Danielsson et al. 2009).
The damage in every systemic crisis in history is caused by endogenous risk amplification, even if the trigger can be an exogenous shock as it was in the 1914 systemic crisis.
Fortunately, systemic crises are not frequent. The typical OECD country can expect to experience such a crisis at most one year out of 43. But when they happen, the damage is enormous. As a rule of thumb, a systemic crisis will cost more than a tenth of GDP, so for the US and Europe that would be in the trillions of dollars.
2008 was a typical endogenous risk crisis
The crisis event in 2008, at its core, was caused by the interaction of market participants, who came to doubt assumptions that had previously usually been made almost without question. The result was synchronised decisions to sell similar assets and to avoid the same exposures, causing an acute lack of liquidity.
The crisis preyed on the weaknesses of the financial system: ill-advised and poorly risk-assessed structured credit products, extreme maturity mismatches, illusionary bank capital, regulatory fragmentation, extensive regulatory arbitrage and large off-balance sheet liabilities. The process was primarily endogenous.
Seen through the lens of endogenous risk, 2008 was typical – not too different from the crises of 1914, 1866 and even 1766. The market participants before the crisis all had infinite liquidity as an unrealistic central assumption, and when the market questioned that assumption, liquidity evaporated and a crisis ensued.
What about coronavirus?
The coronavirus shock is, of course, purely exogenous to the global financial system.
The question is whether the financial system absorbs this shock, as it usually does for exogenous shocks, or whether the shock interacts adversely with the latent feedback loops in the financial system and exposes existing vulnerabilities that culminate in a systemic financial crisis.
While not out of the realm of possibility, we think it unlikely for four main reasons.
- The very fact that the coronavirus shock was exogenous and not itself the consequence of inherent system weaknesses is encouraging as it requires the coincidental presence of some additional vulnerability.
- The parts of the financial system most prone to causing systemic crises – the banks – are in better shape now than they were in 2008. Memories are fresh. Caution is higher and the regulators more powerful and better informed.
- The highest-risk lending is now increasingly done by non-bank institutions, sovereign wealth funds, hedge funds, bond markets, and the like. While unfortunate for those investors, large COVID-induced losses for end investors will not trigger the same feedback loops as would the same losses taken by a bank because most employ much less leverage.
- The high levels of capital and more sensible maturity mismatches mean banks are more able to absorb defaults and give the authorities more room to use forbearance to forestall widespread bankruptcies.
Central bank reactions to crisis
The central bankers reacted appropriately in 2008, sharply increasing liquidity. In that, they had learned both from the disastrous consequences of their predecessors in the Great Depression, who refused to provide liquidity, and the Japanese policy mistakes after its 1990s crisis.
However, 2008 was at its core a liquidity crisis, and providing liquidity then was the appropriate response. It is a policy remedy that directly targets the immediate vulnerability.
Since 2008, the central banks have continued to provide massive amounts of liquidity, keeping interest rates low in the interests of stimulating the economy.
The perverse consequence of that is historical levels of leverage. The private sector is highly indebted, increasing its vulnerability to a demand shock like Corona as we warned here on Vox in 2016 (Csullag et al. 2016).
Providing liquidity to banks, as the central banks are doing now, is perhaps a sensible precaution but in itself seems unlikely to be effective because it is a policy remedy directed at a different type of crisis. The vulnerability today lies not primarily within the financial sector, but in the broad swathe of over-indebted companies that will see revenues collapse below any levels, they may have anticipated. Ironically, these firms are, in part, so highly indebted because of the post-2008 low interest rates.
As the primary channel for a systemic crisis is be though cascading commercial-sector bankruptcies, policymakers need to contain and mitigate such an eventuality, which is not the main domain of monetary policy.
Comparison between the coronavirus crisis and the global systemic crisis of 2008 is inevitable, but seen through the lens of exogenous and endogenous risk they are quite different.
2008 was a global systemic financial crisis fuelled by the endogenous interactions of market participants. The forces of the crisis fed on deep weaknesses in the financial system that had built up out of sight.
COVID-19 is an exogenous shock to the economy, and the question is whether there are sufficient latent weaknesses for it to prey on. We think this unlikely. Instead, the locus of the problem lies outside the financial industry, in a real economy in which shops, services and business are being closed by state fiat, and the income of employees involved is collapsing.
In turn, that means the appropriate policy response cannot be limited to reducing interest rates or purchases of corporate or sovereign bonds on the open market, but should also encompass forbearance and targeted help and similar policies.
With the most vulnerable part of the financial system – the banks – in much better shape now than in 2008, we feel that this is not the real problem and that any solution focused primarily on the financial system would fail.
Richard B (2020), “Keeping the lights on: Economic medicine for a medical shock”, VoxEU.org, 13 March.
Baldwin, R and B Weder di Mauro (2020), Mitigating the COVID economic crisis: Act fast and do whatever it takes, a VoxEU.org eBook, CEPR Press.
Csullag, B, J Danielsson and R Macrae (2016), “Why it doesn’t make sense to hold bonds”, VoxEU.org, 27 June.
Danielsson, J and H S Shin (2002), “Endogenous risk”, in Modern Risk Management — A History, Risk Books.
Danielsson, D, H S Shin and J-P Zigrand (2009), “Modelling financial turmoil through endogenous risk”, VoxEU.org, 11 March.