Ever since the Global Financial Crisis (GFC) tackling the ‘too-big-to-fail’ (TBTF) problem without having to rely on taxpayer-funded bailouts has been a regulatory priority. Consequently, complex resolution frameworks have been installed in all major financial jurisdictions. The experience over the last decade has shown, however, that even a sound resolution regime cannot prevent bailouts when faced with firm failure of potentially systemic proportions.
Author: Professor Michael Schillig
Published: 17 May 2022
Ever since the Global Financial Crisis (GFC) tackling the ‘too-big-to-fail’ (TBTF) problem without having to rely on taxpayer-funded bailouts has been a regulatory priority. Consequently, complex resolution frameworks have been installed in all major financial jurisdictions. The experience over the last decade has shown, however, that even a sound resolution regime cannot prevent bailouts when faced with firm failure of potentially systemic proportions. This has become even more pertinent in the face of the COVID-19 crisis. Although governments and central banks have launched vast stimulus and rescue programs, further down the line financial institutions will likely be confronted with increased loan losses potentially putting them at risk. In a market economy, the state cannot support insolvent firms forever; eventually failing banks will have to be resolved with as little public financial support as possible, if only to fortify public finances against the next exogenous shock.
The aftermath of the GFC also saw the launch of the first successful cryptocurrency.[1] Bitcoin’s genesis block contained the following message: ‘The Times 03/Jan/2009 Chancellor on brink of second bailout for banks.’ What Satoshi Nakamoto intended to convey, other than proof of the date, remains speculation. Still, as a decentralized peer-to-peer payment network that operates without intermediaries (in the traditional sense), Bitcoin provides an alternative value transfer system that has no need for bailouts. Its underlying blockchain technology has inspired an entire new ecosystem of platforms and protocols with modified and improved technical features, ever widening the range of possible applications. Following initial skepsis and hostility, the financial industry has been very receptive and has recognised the technology’s potential across a wide range of sectors.[2]
Certain core characteristics of blockchain-based value transfer systems, stemming from their reliance on peer-to-peer networks with decentralised consensus mechanisms and ‘smart contract’ capability, render the technology particularly interesting from a bank resolution perspective. Conceptually, a network can be established so that no single participating node is in control and responsible for the maintenance and operation of the system. A blockchain based system can operate globally across jurisdictional boundaries. The failure of a single node or even a limited number of nodes has no effect on the continued operation of the network overall, provided that uncorrupted nodes maintain a complete copy of the blockchain. The stored transaction and account data is authenticated, transparent and auditable, with the possibility to limit transparency on a need-to-know basis. Decentralization makes it difficult to sabotage smart contract execution and limits the possibility for parties to act opportunistically for the benefit of the few at the expense of the many.
From the perspective of TBTF resolution, these key attributes have an immediate appeal. We can imagine the global financial system as a complex network of different types of bank and non-bank financial institutions, interconnected across jurisdictions through interbank lending, repo financing, stock lending, derivative transactions, inter-institutional guarantees, brokerage service contracts, and the issuing and holding of cash equivalents and long-term debt instruments. Within this system, shocks may be transmitted directly through contractual exposures and interconnections, and/or indirectly through downward asset price spirals and wholesale or retail runs on cash and cash equivalents. The opacity of the various transmission channels makes it difficult if not impossible to correctly appreciate the effect that TBTF firm failure may have on the system.
Now imagine a blockchain-based system in which the failure of a node, or even of several nodes, has virtually no effect on the continuous operation of the system as a whole; where the functionalities carried out by the failing nodes are seamlessly continued by their peers, and clients would suffer no detriment because network access would still be available through non-failing peer nodes. Moreover, the blockchain database contains every transaction carried out within the system as well as its status. With this enhanced transparency and a much-reduced potential effect of failure on the network, the strong bailout incentives of regulators and policymakers would be much curtailed.
For the future, it is possible to imagine different scenarios in which blockchain technology has been adopted to a greater or lesser degree.[3] Over the next three to five years more and more financial products may be issued on blockchains. The key advantage here would be enhanced transparency which could facilitate the effective application of resolution tools and powers. Ten years from now we may live in a ‘Smart Securities World’ where blockchain has become the internet of value and most financial transactions take place on-chain. Smart contract technology may have evolved so that state-contingencies can be robustly embedded in smart securities and oracles[4] can be relied on to bring off-chain data on-chain. Parts of the resolution process could then be automated, significantly reducing the scope for regulators to opportunistically exercise their discretion at the resolution stage.
For example, it has been suggested that the private sector contribution to a failing firm’s recapitalization should be separated from the subsequent restructuring or liquidation of the firm’s business. It should occur upon a clearly defined trigger event and with predictable consequences for the liabilities subject to bail-in, with no need for a perfect adjustment to the firm’s recapitalization needs in resolution.[5] In a ‘Smart Securities World,’ this elegant solution could be implemented using blockchain. An institution’s capital structure debt could be issued in tokenized form by deploying a smart contract that records in-contract the token balances of the investors holding capital structure debt instruments. The write down and/or conversion upon a certain trigger event could be encoded as part of the smart contract logic. It is of a state-contingent nature: If the trigger conditions are satisfied, then a write down or conversion occurs, the former by reducing investors’ token balances, and the latter by ‘burning’ the debt tokens and replacing them with equity tokens at an appropriate rate of conversion.[6]
As for the bail-in trigger, a capital-structure-debt-smart-contract could encode three alternative trigger events: lack of good collateral, failure of a private sector recapitalization attempt, and actual default on capital structure debt. Provided that capital structure debt, collateral and equity are tokenized and managed through linked smart contracts, the capital-structure-debt-smart-contract would have access to the necessary data on-chain, no oracles required. As for the amount of bail-in, we can draw on Bebchuk’s approach for addressing the valuation uncertainty in corporate restructurings.[7] Provided the ranking and nominal value of pre-resolution entitlements is known, the value to which each participant is entitled in the resolved entity can be precisely expressed as a function of the resolved firm’s value. A set of rights that precisely matches participants’ post-resolution entitlements can be designed and distributed accordingly. There would be no need for extensive and unreliable valuations ex ante with possibly expensive and litigation-prone corrections ex post; no need either for an estimation of losses and potentially market confidence destroying repeated rounds of bail-in. Investors could appreciate the treatment of their entitlements in the event of bail-in and could price their investments more accurately.
Of course, the amount of bail-inable debt may be insufficient to completely absorb the institution’s losses so that run-able liabilities (cash equivalents, derivatives subject to close-out etc) would no longer be immune. Drawing on Acharya et al.,[8] in a ‘Smart Securities World,’ ‘systemically important assets and liabilities’ could be ‘automatically stabilized’ through ‘distributed financial market infrastructures’ (dFMIs) based on smart contracts running on a blockchain-based network. A dFMI can be envisaged as a decentralised network of connected nodes operating on a peer-to-peer basis subject to a certain governance regime and regulatory oversight. For example, derivative contracts or repos could be initialized on a smart contract-capable platform. Counterparties would interact directly with each other so that the risk inherent in a trade would directly affect only the parties to that transaction. [9]
Market participants could pool assets (initial margin; repo collateral) in a default-fund-smart-contract account. The account’s smart contract code could embed rules for the release of funds, including automatic rules that would make funds available to affected counterparties in the case of a member’s default. The event of default – failure to post requested margin or to make a payment – would be on-chain data directly accessible by the default-fund-smart-contract. Upon default of a member, the default-fund-smart-contract would make an automatic liquidity payment based on the initial margin or repo collateral held in escrow, to be liquidated later so as to prevent asset price spirals. These liquidity payments would prevent runs, whilst non-defaulting members would still be faced with the remaining credit risk of their counterparties and collateral, thereby enhancing market discipline.
This setup would be superior to the mandatory clearing of eligible (standardized) derivatives through central counterparties (CCPs). There would be no market distorting TBTF institution charged with maintaining financial stability in the public interest. Instead, a government-sponsored dFMI would directly provide this stabilizing function whilst at the same time keeping moral hazard in check.[10] Moreover, smart contract automation could address one of the key challenges that significantly obstructed the resolution process during the Lehman bankruptcy: the non-termination and non-payment of amounts due by out-of-the money counterparties.[11] Importantly, in this ‘Smart Securities World,’ TBTF would not be eradicated completely. Only parts of the resolution process can be automated in this way and many complex decisions still require expert human judgment on an ad hoc basis. Still, a predictable part-automation of private sector contributions and the automatic stabilization of ‘systemically important assets and liabilities’ could significantly enhance the credibility of the resolution process.
To banish TBTF for good, a still more radical overhaul of the entire financial system will be required. In the longer term – 20 to 50 years from now – we can imagine a comprehensively redesigned blockchain-powered financial system in which runs on short-term debt and consequently TBTF have been eliminated:
Based on the notion that the government and commercial banks are engaged in a public/private partnership, or finance franchise,[12] when issuing and allocating the money supply, we can imagine a system with only one category of money: a fiat cryptocurrency as a form of (synthetic) CBDC[13] as ‘legal tender.’ It is issued only by the member banks of the system and only in recorded form.[14] Like physical currency today, it would bear no interest and constitute non-defaultable sovereign fiat currency. Economic agents would hold fiat-crypto accounts at their respective member banks. Privately owned member banks would be chartered by the government and authorized to issue fiat-crypto in exchange for financial assets in the form of loans and bonds. Enrolling the private sector in the issuance and circulation of the money supply is necessary to improve asset allocations, provided member banks and their shareholders benefit from good portfolio performance and are first in line to absorb losses.
Member banks would maintain a government-sponsored permissioned blockchain, subject to a consensus mechanism to address the double spending problem. Like Ethereum, the blockchain network could be account-based. Customer accounts would store fiat-crypto balances of every customer within the system; and every block would contain a summary of the current state of every customer account. Fiat-crypto units would not be claims on an issuer, but cryptographic tokens created in accordance with the system’s underlying protocol by the member banks. In addition to operating full nodes, maintaining the network, and providing a network interface for their customers, member banks would make investment decisions. In exchange for (tokenized) loans and bonds, member banks would create fiat-crypto tokens in accordance with the network protocol and subject to embedded portfolio constraints and capital requirements. Conversely, fiat-crypto could be destroyed (‘burned’) by letting financial assets mature or by selling them. Controlled by wallet software provided by their member banks, customer accounts would allow the ‘owner’ to send and receive fiat-crypto directly and peer-to-peer. Customers would essentially maintain accounts with the network, operating like a single bank. Insolvency of a member bank, in principle, has no effect on its customers’ accounts. These simply remain available on the network, maintained by the remaining member banks who will be able to continue lending and create new fiat-crypto.
Importantly, only member banks can issue fiat-crypto and fiat-crypto equivalents. Financial firms (other than member banks) would be prevented from funding their activities (including structured finance, derivatives, proprietary trading etc.) with run-able short term debt instruments. They would have to rely on longer-term debt and equity instead. No doubt, such a rule would be costly for the firms affected by it. However, for society and financial system stability overall the transition might be beneficial:[15]
In this system, there would be no room for runs and no need for deposit insurance. Any member bank can drop out at any point with no effect on their customers’ liquidity which is entirely maintained by the network itself. No single member bank would be TBTF, although, like a central counterparty today, the network itself would be. But it would be backed directly by the government which could be compensated ex ante for any bailout intervention through risk-calibrated franchise fees charged to the member banks for the privilege of participating in the money supply.
Perhaps these are ‘brave new worlds.’ However, their fundamental technical building blocks are up and running today. They may lack scalability, reliability, and cyber security, but technological ingenuity is relentless. Whether and to what extent any of these futures will come to pass will depend as much on political and economic leadership as on the pace of technological innovation.
[1] S Nakamoto, ‘Bitcoin – A Peer-to-Peer Electronic Cash System’ (2008) 1 (available at https://bitcoin.org/bitcoin.pdf).
[2] JP Morgan, JP Morgan Perspectives: Blockchain, digital currency and cryptocurrency: Moving into the mainstream? (21 February 2020).
[3] M Schillig ‘The Too-Big-to-Fail Problem and the Blockchain Solution’ (2022) 19.1 Berkeley Business Law Journal 126-174.
[4] In the blockchain context, an ‘oracle’ is a system that can provide blockchain-external data (e.g. share prices, CDS spreads and other financial market information) and make it available to the network; A Antonopulos and G Wood, Mastering Ethereum (O’Reilly, 2019) 253.
[5] T Tröger, Too Complex to Work: A Critical Assessment of the Bail-in Tool under the European Bank Recovery and Resolution Regime, (2017) SAFE Working Paper No. 179, 33.
[6] For details: Schillig (n 3).
[7] L Bebchuk, ‘A New Approach to Corporate Reorganizations’ (1988) 101 Harvard Law Review 775.
[8] V Acharya, B Adler and M Richardson, ‘A Proposal to Resolve the Distress of Large and Complex Financial Institutions’ 2 (available at http://www.lse.ac.uk/fmg/events/conferences/past-conferences/2011/DBWorkshop_14Mar2011/13_Acharya_et_al.pdf); V Acharya and T Öncü, ‘A Proposal for the Resolution of Systemically Important Assets and Liabilities: The Case of the Repo Market’ (2013) 9 International Journal of Central Banking 291.
[9] S Feenan, D Heller, A Lipton, M Morini, R Ram, R Sams, T Swanson, S Yong and D Barrero Zalles, ‘Decentralized Financial Market Infrastructures’ (2020) 15-18 (available at https://static1.squarespace.com/static/57af6f83893fc027c794e637/t/5e5fdb96d00cf530ae6ed3ec/1583340439946/2020+Feenan+Decentralized+Financial+Market+Infrastructures.pdf) 18-19.
[10] Acharya and Öncü (n 8) 334-336.
[11] M Fleming and A Sarkar, ‘The Failure Resolution of Lehman Brothers’ (2014) FRBNY Economic Policy Review 175, 182-187.
[12] R Hockett and S Omarova, ‘The Finance Franchise’ (2017) 102 Cornell Law Review 1143.
[13] T Adrian, ‘Stablecoins, Central Bank Digital Currencies, and Cross-Border Payments: A New Look at the International Monetary System’, Remarks at the IMF-Swiss National Bank Conference, Zurich, May 2019 ( available at: https://www.imf.org/en/News/Articles/2019/05/13/sp051419-stablecoins-central-bank-digital-currencies-and-cross-border-payments).
[14] This draws on M Ricks, The Money Problem: Rethinking Financial Regulation (University of Chicago Press 2017).
[15] ibid 17-18, 25.
