VoxEU Column COVID-19 Monetary Policy

Three strikes against the Fed

The US Federal Reserve – the world’s most important central bank – is not in a good place. This column outlines three flaws in the operating practices of the Fed – (i) its refusal to adopt negative policy rates, (ii) the build-up of significant credit risks through non-transparent (quasi-)fiscal actions, and (iii) stress testing analysis which fails to account for the severity of the COVID-19 crisis. It proposes a number of ways forward, including a symmetric policy rate around zero, a temporary ban on dividend payments, new equity issuance, and conducting a comprehensive stress test of the financial system.

The Fed’s current operating practices are afflicted with three serious flaws. Like most other central banks, the Fed refuses to set seriously negative policy rates; Like many other central banks, including the ECB, the Fed acts as an unaccountable fiscal principal rather than as a transparent and accountable fiscal agent of the federal government; and, unlike most other advanced economy central banks, it has emasculated the stress tests it imposes on systemically important banks to ensure their capital adequacy.

Going deeply negative

Regarding negative policy rates, the Fed does not even go down to the effective lower bound (ELB), which equals the zero interest rate on currency minus the carry cost of currency (storage, insurance, etc.). The example of the ECB and other European central banks suggests that, even without reforms, the lower bound on the Fed’s target federal funds rate could be set at -75 basis points rather than at its current level of 0.00. Indeed, the rates on required reserves and on excess reserves, both currently 0.10%, could be lowered to -75 basis points, providing a modest but non-trivial financial stimulus.

It would be better, however, to get rid of the ELB altogether. The reason is straightforward: a policy rate at the ELB is useless as an instrument for providing a stimulus to aggregate demand. Abolishing the ELB could be achieved either by taxing currency (as proposed by Gesell 1916), or by introducing a variable exchange rate between currency and deposits with the central bank (as proposed by Eisler 1932), or by abolishing currency altogether and offering every member of the public access to a central bank digital currency. This could take the form of a checkable interest-bearing account managed by commercial banks, savings banks, the post office, and other convenient financial institutions, integrated with Apple Pay and similar mobile payments systems, and guaranteed by the central bank. If there are social concerns about the inability of part of the population to manage without cash (including the elderly and those who don’t have any experience with bank accounts and electronic/digital payments technologies), the temporary retention of low-denomination currency notes could address this issue. It would lower the ELB without abolishing it (see Buiter 2009, Lilley and Rogoff 2020).

The only economic argument against going seriously negative with the policy rate is the ‘reversal interest rate’ proposition of Brunnermeier and Koby (2018), that a sufficiently low level of interest rates can distort the incentives faced by banks and result in a lower level of bank lending. This does not by itself pin down the level of the ‘reversal interest rate’ (which could be materially negative), nor would it allow for intermediation between savers and investors through capital markets rather than through banks, which avoids the ‘reversal interest rate’ obstacle.

Quasi-fiscal and outright fiscal actions of the Fed and the legitimacy of its operational independence

With regards to the non-transparent and unaccountable (quasi-)fiscal actions of the central bank, I will focus on the size and composition of the Fed’s balance sheet. I recognise that the setting of the policy rate(s), forward guidance, and yield curve control have unavoidable fiscal consequences – redistribution between borrowers and lenders and profits for the Fed and thus for its beneficial owner, the federal treasury. It is key to recognise that, whatever the formal (often bizarre) ownership structure of a central bank, the national treasury is its beneficial owner, ultimately entitled to its profits and responsible for any losses.

The Fed also pays annual remittances to the US Treasury. If there is any systematic and transparent dialogue between the Fed and the US Treasury about these remittances, it remains well hidden. The amounts of money involved are non-trivial.1

Since the COVID-19 pandemic struck, the Fed has taken on significant credit risk by lending to and purchasing risky debt instruments from private financial and non-financial corporations, state and local governments, and households. Only a small fraction, generally not more than 10% of the Fed’s maximum possible exposure to these high-risk activities, is covered by US Treasury guarantees or other means of indemnification. Consider the following examples, all taken from the Fed’s own press releases and associated background papers:2

1. Municipal Liquidity Facility (MLF): The MLF will provide a liquidity backstop to issuers of Eligible Notes through an SPV (special purpose vehicle). The Treasury will make an initial equity investment of $35 billion in the SPV, which will have the ability to purchase up to $500 billion of Eligible Notes.

2. Main Street Lending Program (MLP): The Treasury provides $75 billion equity to the Main Street SPV, which will purchase up to $600 billion of participations in eligible loans.

3. Primary and Secondary Market Corporate Credit Facilities (PMCCF and SMCCF): The combined size of the CCFs is up to $750 billion. The Treasury will make a $75 billion equity investment in the SPV for both CCFs.

4. The Term Asset-Backed Securities Loan 2020 Facility (TALF 2020): The Treasury will make a $10 billion equity investment in the TALF SPV. The initial size of the facility is up to $100 billion.

The US dollar liquidity swap lines with other central banks are another example of the US Treasury allowing (or encouraging) the Fed to take on material credit risk and other price risk without offering the Fed full, or indeed any, indemnification.

The guiding principles for the Fed’s (quasi-)fiscal operations should be simple: the US Treasury guarantees the entire balance sheet of the Fed. Actions undertaken by the Fed that involve taking on material credit risk, price risk, and other risk must be joint decisions of the Fed and the US Treasury. Because some of the joint actions by the Fed and the US Treasury may have commercially sensitive and market-sensitive dimensions, temporary discretion, confidentiality, and secrecy may be warranted. Full openness and transparency should, however, be restored as soon as possible. Unless these principles are adhered to, many of the Fed’s balance sheet actions cause it to lose its legitimacy and threaten its operational independence even in those limited areas where it makes sense: the setting of the policy rate(s) and the provision of funding liquidity and market liquidity as lender of last resort and market maker of last resort.

Note that the ECB is in even worse shape regarding the legitimacy of its fiscal and quasi-fiscal operations compared to the Fed. The ECB is owned by the 19 National Central Banks (NCBs) of the euro area according to the capital key. Each NCB is beneficially owned by its national treasury/ministry of finance. There is no euro area fiscal authority, however, that guarantees the assets on the ECB’s (and the Eurosystem’s) balance sheet or engages with the ECB’s Governing Council in meaningful, open, and transparent discussions about the appropriate management of the ECB’s and Eurosystem’s balance sheet. The result is, since the Global Crisis, a growing illegitimacy of the ECB/Eurosystem as it has taken large amounts of both sovereign and private credit risk onto its balance sheet on terms that involve a meaningful fiscal transfer to the private and public counterparties.

Stress testing

The Fed Board released, on 25 June 2020, the results of the full Dodd Frank Act Stress Test (DFAST) 2020, including the performance of 33 individual banks, designed in 2020 Q1, before the coronavirus.3

It also released the results of an additional sensitivity analysis that did try to take into account the economic consequences of the COVID-19 pandemic, by testing the resilience of 34 large banks under three recession and recovery scenarios: V-shaped, U-shaped, and W-shaped.4 Only aggregate results for loan losses and capital ratios of the 34 banks included in the sensitivity analysis were provided, however. The sensitivity analysis did not allow for the potential effects of government stimulus payments and expanded unemployment insurance. On the other hand, the recession and recovery alphabet soup that was considered leaves out the more pessimistic, and in my view realistic, L-shaped scenario, as well as other scenarios (V-shaped, U-shaped, and W-shaped), where the recovery does not reach the pre-COVID-19 path of potential output for many years, if ever. Nor does it consider scenarios where the post-recovery growth rate of potential (and actual) output is persistently (or even permanently) below the pre-COVID-19 growth rate. Lower global and US potential output growth may well be the result of increased economic nationalism and growing anti-globalisation sentiment and policy measures. It can also be driven by a cautionary and uncertainty-motivated shift from ‘just-in-time’ economics (think supply chains) to ‘just-in-case’ economics where more redundancy is built in, rationally, but at the expense of growth.

Not surprisingly, the full stress test downside scenario based on pre-COVID-19 information is comparable to the least dire, V-shaped recession and recovery scenario in the sensitivity analysis. It is extraordinary that the “Board will use the results of this test to set the new stress capital buffer requirement for these firms, which will take effect, as planned, in the fourth quarter.”5This full stress test should have been discarded as irrelevant when COVID-19 hit and should have been replaced by a full stress test based on one or more downside scenarios that incorporate the COVID-19 pandemic and its economic and financial consequences. If necessary, the stress testing process could have been delayed. It is better to be a bit late but relevant than right on time but irrelevant.

In the sensitivity analysis, aggregate loan losses for the 34 banks involved ranged from $560 billion to $700 billion and aggregate capital ratios declined from 12% in the fourth quarter of 2019 to between 9.5% and 7.7% under the hypothetical downside scenarios. Risk-weighted assets of the 33 banks participating in the stress test were $10,354 billion in 2019 Q4.

Based on the stress test and the sensitivity analyses the Fed suspended share repurchases by the banks for the third quarter. There are two caps on dividend payments. Dividend payments in the third quarter are capped by the lesser of the amount paid in the second quarter and the average net income over the prior four quarters – the first three of which were pre-COVID.

I consider the decision by the Fed to permit dividend payments in 2020 Q3 to be unwise. I do not share the sanguine attitude of the Fed about the capital adequacy pre-COVID-19 of the 33 largest banks. Corporate lending and investments in high-yield corporate debt, including asset-backed securities (ABS) like covenant-lite collateralised loan obligations (CLO), implied a potential vulnerability that the Fed chose to ignore. The widespread impairment of the creditworthiness of non-financial corporate and household borrowers since the COVID-19 pandemic started must have made for losses, including honest mark-to-market losses, that could be well in excess of the $700 billion that is the maximum aggregate loan loss the Fed considers. At the very least, there should have been a complete ban on dividend payments (as on share purchases) for 2020 Q3. A cautious Fed would indeed have insisted on additional equity issuance in 2020 Q3 by the 34 banks involved in the sensitivity analysis.


The Fed – the world’s most important central bank – is not in a good place. Like all other central banks, it accepts the Effective Lower Bound on the policy rate as an absolute constraint rather than as something that, from a technical perspective, can be eliminated easily. For legitimacy reasons, eliminating the ELB and creating full symmetry of the policy rate around zero should be a decision that is approved by the Treasury and possibly by both Houses of Congress.

Like most advanced economy central banks since the Global Crisis, there has been massive quantitative and qualitative easing by the Fed. Since the start of the COVID-19 pandemic, the size of the Fed’s balance sheet has increased massively, and the composition of its balance sheet has shifted towards less liquid, higher-risk assets on an unprecedented scale. For those (risky) Fed programs for which both a maximum scale and a Treasury guarantee/equity injection can be established easily, we find up to $1,950 billion of potential Fed exposure and a mere $195 billion of Treasury equity to back it up. If any of these Fed balance sheet risks materialise, this is likely to involve a hit to the Fed’s capital and an associated redistribution from the tax payer (or the beneficiary of public spending) to the defaulting bank, non-financial corporation or household. Such (quasi-)fiscal redistribution is not part of the Fed’s mandate. Its legitimacy and ultimately its operational independence in setting the policy rate and acting as lender of last resort and market maker of last resort is undermined and may be taken away from it.

Stress testing of the largest bank in the COVID-19 era has been a debacle this year. It is time to undertake immediately a new comprehensive DFAST 2020 stress test, simultaneously with the Comprehensive Capital Analysis and Review (CCAR), which is a complementary exercise to DFAST.

As part of CCAR, the Fed “evaluates institutions' capital adequacy, internal capital adequacy assessment processes, and their individual plans to make capital distributions, such as dividend payments or stock repurchases.”6 The manifest overlap with DFAST is a fact of life and not necessarily a negative feature, because in a world shaken by the pervasive uncertainty created by the COVID-19 epidemic and the political and economic responses to it, redundancy and duplication can, up to a point, be a blessing rather than a curse.

In the meantime – until the results of the new DFAST and CCAR stress tests are known, the uncertainty principle calls for extreme caution in the regulator’s approach to the capital adequacy of systemically important banks (and indeed of banks in general). I would favour not just a ban dividend payments in 2020 Q3, but the imposition of a requirement for additional equity issuance to boost the capital adequacy of the US banking system. New equity issuance sufficient to ensure that, under the worst downside scenario, capital ratios do not decline below 10% in 2020 Q4, would be a modest step in the right direction.

There is hope for the Fed, but it must change its ways.


Brunnermeier, M K and Y Koby (2018), “The reversal interest rate”, NBER Working Paper, No. 25406.

Buiter, W H (2009), "Negative Nominal Interest Rates; Three Ways to Overcome the Zero Lower Bound", North American Journal of Economics and Finance 20: 213-238. NBER working paper version.

Eisler, R (1932), Stable Money: the remedy for the economic world crisis: a programme of Financial reconstruction for the international conference 1933 (with a preface by Vincent C. Vickers), London: The Search Publishing Co.

Gesell, S (1916), Die Natuerliche Wirtschaftsordnung, Rudolf Zitzman Verlag, available in English as The Natural Economic Order, London: Peter Owen Ltd, 1958.

Lilley, A and K Rogoff (2020), “Negative interest rate policy in the post-Covid-19 world”,, 17 April.


1 See a recent press release from the Fed here.

2 See here

3 See the press release here.

4 In the report from June 2020 the Fed states that “The sensitivity analysis includes the 33 firms required to participate in DFAST 2020”. It is unclear where the 34th bank comes from that is used in the sensitivity analysis but is not part of the DFAST 2020.

5 See the press release here.

6 See here

2,310 Reads