On the response of the financial authorities to Covid-19. Jon Danielsson. Modelsandrisk.org

On the response of the financial authorities to Covid-19

December 2, 2020
The financial authorities have just bailed the financial system out for the second time in a decade. While the authorities are proud of having prevented a financial crisis, are we really better off? No, one cannot judge the policy intervention a success if it is only due to the promise of the intervention that the problem arose in the first place.

I just finished a piece on The Covid-19 bailouts and the future of the capitalist banking system, promising at the end to come up with some solutions. There are alternatives: the official one discussed below, and others listed at the end of this piece.

Central bank liquidity interventions saved us from a Covid-19 financial crisis, so, judged by that alone, the authorities did a great job. They certainly are proud, as you can read in the financial stability board's (FSB) Holistic overview.

What happened in March and April was a bailout, not of any individual financial institution, but of the financial system. Government money and guarantees ensured the markets remained liquid, and privately owned financial institutions benefited mightily. In my book a bailout.

In the words of Neel Kashkari, the president of the Minneapolis Fed

I don’t know what the best policy solution is, but I know we can’t just keep doing what we’ve been doing, ... As soon as there’s a risk that hits, everybody flees and the Federal Reserve has to step in and bail out that market, and that’s crazy. And we need to take a hard look at that.”

I am not against bailouts, per se, and don't object, in principle, to the financial authorities calming the financial markets when they are heading for a disastrous financial crisis. As I will argue later, a key purpose of central banks is to help the financial system operate at higher levels of risk than it otherwise could do.

But it matters a great deal how it is done.

To start with, what objectives should the financial authorities have?

A common view is that of Olivier Blanchard, then the Chief Economist at IMF, back in 2009:

"Crises feed uncertainty. And uncertainty affects behaviour, which feeds the crisis.[...] So what are policymakers to do? First and foremost, reduce uncertainty."

Precisely what the financial authorities did in 2020. It all sounds sensible, pouring oil on troubles waters. But, while seemingly laudable, the calming raises questions.

There is a crucial distinction between stabilising the financial system and protecting banks from the consequences of their actions, a point made by Bagehot back in 1873. In his principles of lending a last resort, as a response to a crisis:

  1. The central bank should lend freely;
  2. At a high rate of interest;
  3. On good banking securities.

In other words, the central bank has to respond and calm the markets, but in the process should not give the private sector a free ride. If a bailout is needed, the financial institutions should pay for it. And that is the crucial difference between the 2020 bailouts and Bagehot.

The bailouts Bagehot discussed were designed to help financial institutions monetise their temporarily illiquid assets, and the Bank of England knew exactly who got help and at what terms, allowing the bank to price the bailouts.

The most powerful interventions in 2020 were the Federal Reserve's US dollar swap lines. Foreign creditors have access to dollar liquidity via their central banks, just like they already get in their own currencies.

It's a broad intervention, benefiting the entire market, including those individual institutions who did not partake. It is impossible to apply Bagehot's rules to FX swap lines since we cannot easily identify the beneficiaries, (the direct beneficiaries typically pay a small fee) and hence price them correctly.

However, the reason the swap lines were needed is because they exist.

If the central banks promise to provide liquidity when is market turmoil, of course the private banks will take full advantage. The effectiveness of the interventions does not prove they were needed, except, if one only does static analysis.

And, the ultimate reason for disregarding Bagehot's advice is that the cost of not doing anything is higher than the cost of intervention. We don't have to price the bailouts, they already pay for themselves. Correct, but only if one looks at the 2020 turmoil in isolation, disregarding how we got there, and what it heralds.

So what is the problem with the liquidity bailouts in 2020?

Back to the the holistic FSB document. It is a very good overview of how the crisis played out, and also makes several interesting points:

To begin with, it argues that the policy response was correct:

"The policy response was speedy, sizeable and sweeping. The unprecedented policy actions by central banks alleviated market stress through different channels: asset purchases; liquidity operations, including for US dollars; and backstop facilities designed to provide targeted liquidity to specific financial entities (e.g. MMFs and primary dealers). Regulatory and supervisory measures as well as fiscal policies complemented these central bank interventions. Securities regulators also took measures to support market functioning. The policy measures succeeded in alleviating market strains to date, with announcement effects appearing to be particularly important in restoring confidence and shaping the expectations of market participants. Absent central bank intervention, it is highly likely that the stress in the financial system would have worsened significantly."

But it it wasn't really the banks that needed help but the nonbanks:

" Some parts of the system, particularly banks and financial market infrastructures, were able to absorb rather than amplify the macroeconomic shock, supported by the post-crisis reforms"

So, the banks are resilient but not the non-bank financial intermediation (NBFI) sector:

The March turmoil has underscored the need to strengthen resilience in the NBFI sector

So to summarize:

  1. There was excessive market turmoil;
  2. The part of the system regulated after the last crisis in 2008 did well;
  3. Unfortunately, we as authorities had to provide liquidity to calm the markets, especially the NBFIs, but we really don't want to;
  4. So afterwards we will regulate the parts of the system that needed liquidity, the NBFIs, to prevent a recurrence.

That analysis raises four issues: Two philosophical and two practical.

The economic profession went off the rails in the 1950s and 1960s, after John Maynard Keynes passed and before the pathbreaking work of Robert Lucas. The prevailing view in these lost years was that one could model and control the economy in a static framework. Achieve policy objectives — high growth, low unemployment, low inflation — by just tweaking the parameters of a static model. The financial authorities and the governments loved it. This way of doing economics made them feel powerful, even omnipotent.

Except, just like the other idiocies of the economic policy of the era, such as industrial policy, it didn't work. The Lucas critique showed us why. Expectations matter and economic agents react to policies in a way that undermines the policy objective. All we got was the word of the 1970s stagflation — inflation and stagnation.

Industrial policy is now coming back, and so is static economic policymaking via the doctrine of macroprudential regulations.

What undermines the static approach is crystallised in the difference between 2008 and 2020. In 2008, it was mostly banks that were affected, so over the subsequent decade, we sharply increased the level of regulations on banks. And they worked so well that when the next liquidity crisis came along, the banks did OK, but now the risk had spilled over to the shadow banking sector.

You see, the economic agents that comprise the financial system do not take regulations lying down, they react to them, changing the financial system in the process. Once the regulations take effect, they apply to a system that no longer exists.

The Lucas critique.

In the ensuing cat and mouse game, the private sector financial institutions have the advantage.

The second philosophical problem is a fallacy of composition. Even if we manage to make every financial institution prudent, that does not mean the system is safe. The reason is shock absorption. If some shock comes along, a prudent financial institution has to react, selling its risky assets to stay within its prudential constraints. But, because every financial institution is prudent, there is nobody to buy, so we end up with a vicious feedback spreading to ever more assets and asset classes, even causing a financial crisis.

And now I get to one of the practical problems, procyclicality. The other is moral hazard, but as I discussed that in the last piece, I won't repeat it here.

The most potent force of financial stability is diversity. The more financial institutions see the world differently and react differently, the more stable the system is. If I react to a shock by buying and my buddy Bob by selling, our reaction in aggregate is random noise.

Why Baron Rothschild was so prescient when he wrote two and half centuries ago that the best time to buy property was in the middle of a Civil War. We need the Baron Rothschilds, today the Soros' and Buffetts', someone who can buy in crises unhindered by constraints, someone who sees an opportunity in turmoil. For such individuals and entities to exist, they cannot be regulated, and have to be able to invest in the way they see best, unencumbered by anybody else's views.

Procyclicality is driven by financial institutions having the same beliefs and action. If they see the world in the same way and are compelled to react to it in the same way, financial institutions behave as a crowd — they are procyclical.

The most powerful force of homogeneous beliefs and actions is those financial regulations that force banks to model risk with similar methodologies and mandate particular reactions to that risk. The main driver of those regulations is the Basel Committee, the progenitor of the Basel Accords.

The FSB makes two points. First, that the official response to the Covid-19 liquidity turmoil was correct. And the second, that those financial institutions that are not banks — shadow banks — should be more regulated.

Here it gets interesting. The holistic document does not make the argument that the nonbanks should be regulated like banks, and if one says that to the policymakers, they are likely to react in horror, "nooo, that is not the intention".

So, while the nonbanks will not be brought under the same regulatory umbrella as the banks, we will borrow from the bank regulations, the idea being that since the banks came out of the Covid-19 shock relatively unscathed, the relevant bank regulations worked well and should be applied to the rest of the system.

I think both points are incorrect. Minsky points the way "stability is destabiliziing".

The financial institutions will be strengthened in the belief that the authorities are ready to step in to bail them out in times of market turmoil, and the beliefs and actions of financial institutions will be harmonised, reducing diversity and increasing procyclicality — moral hazard. Both will increase systemic risk and destabilise the financial system.

When the next crisis comes along, the forces of monoculture and regulations will be even stronger, calling into question the need for the private sector financial system, the topic of my next piece.


The Covid-19 bailouts and the future of the capitalist banking system
The socialist response to the Covid-19 financial turmoil

Models and risk
Bloggs and appendices on artificial intelligence, financial crises, systemic risk, financial risk, models, regulations, financial policy, cryptocurrencies and related topics
© All rights reserved, Jon Danielsson,