FTX Comment Letter: Basel Committee on Banking Supervision
Basel Committee on Banking Supervision
Bank for International Settlements
Centralbahnplatz 2
CH-4002 Basel
Switzerland
Re: Response to the Consultative Document on the Second consultation on the prudential treatment of cryptoasset exposures
Ladies and Gentlemen:
FTX US appreciates the opportunity to respond to the second consultative document published on 30 June 2022 by the Basel Committee on Banking Supervision (“BCBS”) regarding the “Prudential treatment of cryptoasset exposures” (the “Second Consultation”).1 FTX US appreciates the BCBS’s active engagement in the ongoing discussions on designing a prudential framework for cryptoassets. Additionally, FTX US would like to thank the BCBS and its members for its dedication to an in-depth and comprehensive process of proposing this prudential framework and for engaging the public. FTX US looks forward to continuing its engagement with the BCBS on this topic.
The banking industry is an important part of the cryptoasset ecosystem as well as a critical partner for responsible innovation in the digital asset space. However, the standards as proposed in the Second Consultation would make it unnecessarily difficult for banks to participate in cryptoasset markets in notable ways. As such, we recommend several changes to the standards that would facilitate bank participation while also protecting consumers and the marketplace moving forward.
Background on FTX US
FTX US includes a spot cryptocurrency exchange, FTX US Derivatives (“FUSD”) (https://derivs.ftx.us/), a nonfungible token marketplace, a broker dealer (FTX Capital Markets) regulated by the Financial Industry Regulatory Authority (“FINRA”), and a South Dakota Trust Company. FTX US is housed under a separate corporate entity from FTX International and is headquartered in Miami, FL.2 Like other crypto-platforms in the U.S., the spot cryptocurrency exchange is primarily regulated through state money-transmitter laws and is also licensed by the U.S. Treasury Department’s Financial Crimes Enforcement Network (“FinCEN”) as a money services business. FUSD offers futures and options contracts on digital assets (or commodities) to both U.S. and non-U.S. persons. FUSD operates with three primary licenses from the CFTC: a Designated Contract Market (“DCM”) license, a Swap Execution Facility (“SEF”) license, and a Derivatives Clearing Organization (“DCO”) license. Prior to its acquisition by FTX, this business was the first crypto-native platform issued a DCO license by the CFTC in 2017, which was a milestone for the agency and the digital-asset industry. That license was later amended in 2019 to permit the clearing of fully collateralized futures contracts on all commodity classes.
Comments
In the Second Consultation, the BCBS stated that it welcomes comments on “all aspects” of its proposals, which includes draft standards text to be included in the Basel Framework. The BCBS also highlighted certain areas of its proposals for which it would particularly welcome feedback. These areas include:
- The structure and specification of the BCBS’s contemplated “redemption risk test” and “basis risk test” for determining whether a cryptoasset should be included in Group 1.
- Whether or not to set out an option that would recognize banks’ exposure to stablecoins issued by regulated entities as generally lower risk than those issued by unregulated entities.
- What modifications to the Group 1 classification conditions would be required to permit the inclusion in Group 1 of cryptoassets that use permissionless blockchains, the risk that modifications would raise, and ways to mitigate such risks.
In addition to addressing these issues, the discussion below addresses several other areas highlighted by market participants in feedback provided on the BCBS’s first consultation document regarding the prudential treatment of cryptoasset exposures (the “First Consultation”), published in July 2021, that FTX US believes the BCBS should reconsider.
We believe that many elements of the Basel Framework draft standards text contemplated by the BCBS do not accurately represent the actual risks posed by cryptoassets, nor does it adequately account for the diversity of cryptoasset types and underlying markets. If adopted as proposed, the standards could have consequences for the development of the sector moving forward. In particular, policies endorsed by the Second Consultation would drive banks away from participating in cryptoasset markets in important ways, thus increasing these markets’ overall risks. Rather than allowing for institutions to use market-based approaches to risk management such as hedging strategies embraced by the BCBS in other contexts, the draft Basel Framework text would, if implemented, generally set forth a highly prescriptive and punitive approach towards the prudential treatment of cryptoassets.
A. The Group 1 classification should be simplified and amended to create a clearer path for stablecoins based on permissionless blockchains to qualify.
The Second Consultation indicates that, under the BCBS’s draft standards text, “it is highly unlikely that any cryptoassets based on permissionless blockchains will be able to meet the classification conditions to be included in [the] Group 1 [cryptoassets category].”3 In other words, it seems that the BCBS may be of the view that a stablecoin based on a popular permissionless blockchain such as Ethereum or Solana would under the proposed standards text likely not be treated as a Group 1 cryptoasset, meaning that a bank could need to maintain capital at least equivalent to the dollar value of stablecoins based on permissionless blockchains (“Permissionless Blockchain Stablecoins”) that the bank holds.4 In July 2022, data indicate roughly 85 percent of crypto transactions had on one side a Permissionless Blockchain Stablecoin.5
Accordingly, without appropriate modifications, the BCBS’s proposals could potentially incentivize banks to participate in less liquid cryptoasset markets relative to non-banks, which unlike banks could easily hold Permissionless Blockchain Stablecoins to be used to buy other cryptoassets for customers. More broadly, banks could find it unnecessarily costly to participate in budding web3 ecosystems built upon maturing permissionless blockchains. Cryptoassets broadly, and stablecoins in particular, should not be excluded from the Group 1 category simply for being based on permissionless blockchains. To avoid this outcome – and thus ensure that regulated banks are able to bring increased stability into the cryptoasset markets – Basel should make adjustments to each of its four classification conditions for Group 1 cryptoassets.
Classification condition 1 should be modified to eliminate the basis risk test and allow for regulatory approval as a substitute.
The redemption risk test set forth in classification condition 1 contemplates that, in order for a cryptoasset with a stabilization mechanism to not be treated as a Group 2 cryptoasset, the value of its reserve assets “must at all times, including during periods of extreme stress, equal or exceed the aggregate peg value of all cryptoassets.”6 It goes on to propose that the cryptoasset would require a level of overcollateralization that ensures that the value of reserve assets does not breach this threshold even during extreme scenarios.7 The redemption risk test also, among other things, necessitates that a bank ensure the governance arrangements relating to reserve assets management for a cryptoasset arrangement be comprehensive and transparent, including that “reserve assets are managed and invested with an explicit legally enforceable objective of ensuring that all cryptoassets can be redeemed promptly at the peg value,” as well as that the arrangement maintains a robust operational risk and resilience framework to ensure the availability and safe custody of the reserve assets.8
In other words, classification condition 1 sets forth via the redemption risk test very robust requirements to dramatically reduce capital risk associated with bank exposure to a cryptoasset that meets this test. Accordingly, if the redemption risk test is met, FTX US fails to see the need for the BCBS’s contemplated basis risk test. As the BCBS notes, the redemption risk test sufficiently controls for capital risk by “ensur[ing] that the reserve assets are sufficient to enable the cryptoassets to be redeemable at all times . . . for the peg value.”9 Accordingly, meeting the basis risk test may be unrealistic, given that other assets designed to maintain the values of reference assets would likely not fully pass the test (e.g., commodity ETFs).10
Even if the market value of a cryptoasset designed to maintain the value of a reference asset deviates from that reference asset by 20 basis points more than 10 times over the course of 12 months – the BCBS’s contemplated threshold for failing the basis risk test – as long as the redemption risk test is met, there is no reason for a Group 1 cryptoasset to then be treated the same as a much more speculative cryptoasset just because it breached this arbitrary threshold. That such an outcome is possible suggests that the BCBS’s draft standards text still includes undesirable “cliff effects.” Moreover, price deviations on crypto trading platforms can be caused by temporary supply and demand dynamics, but such dynamics do not impede the ability of an institution to redeem a stablecoin with an issuer for its underlying value. Indeed, leading stablecoin arrangements redeem dollar-pegged stablecoins 1-to-1 with a dollar regardless of the market price of a stablecoin.11
Recommendation 1: Eliminate the basis risk test.12 The redemption risk test is sufficient for the purposes of classification condition 1.
We also believe the BCBS could provide additional clarity regarding the application of the redemption risk test to real market scenarios. The redemption risk test references the necessity to maintain redeemability during “periods of extreme stress,” but does not provide further clarity or examples regarding the definition of periods of extreme stress.13 The BCBS should provide greater clarity regarding what it means by “periods of extreme stress” should it choose to include this condition as part of the redemption risk test.
We also recommend augmenting an element of classification condition 1 introduced by the BCBS that would require that, in order for a stablecoin to be classified as Group 1b, “the issuer must be supervised and regulated by a supervisor that applies prudential capital and liquidity requirements.”14 To begin with, we suggest that the BCBS clarify that the issuer may be a limited purpose bank, such as certain state trust companies in the United States or national trust companies chartered by the U.S. Office of the Comptroller of the Currency. Should this condition be met, there is no need for a bank to apply a redemption risk test or a basis risk test to the stablecoin, as a developed regulatory framework controlling for these risks applies.15 However, if a stablecoin issuer does not meet the condition that it is supervised and regulated by a regulator that applies prudential and liquidity requirements, they should nevertheless still be able to qualify under the redemption risk test.
Recommendation 2: A stablecoin should be considered to meet classification condition 1 for the purposes of determining whether it is a Group 1 cryptoasset if its issuer is supervised and regulated by a regulator that applies prudential and liquidity requirements.16
Classification condition 2 should be modified to reflect permissionless blockchain considerations.
Even if the redemption risk test is met by a cryptoasset with a stabilization mechanism, the ability of a bank to redeem a cryptoasset could be impacted by the potential manifestation of operational or other risks that potentially impede the ability of the bank to redeem a cryptoasset for the value of its underlying asset. Accordingly, we are supportive of the aims of classification condition 2 to ensure that a stablecoin, in order to have favorable capital treatment, be issued through an arrangement with sufficient transparency and legal clarity. We are also supportive of certain requirements regarding the stablecoin’s redeemability, including that in order for it to be treated as a Group 1 cryptoasset, the full value of the cryptoasset be redeemable within five calendar days. It is reasonable for a bank to be expected to confirm these features of a particular stablecoin arrangement.
However, some elements of classification condition 2, as contemplated in the Second Consultation, risk impeding banks' use of Permissionless Blockchain Stablecoins under certain circumstances.17 One potential issue is the currently contemplated approach towards assessing whether applicable legal frameworks ensure settlement finality for a particular cryptoasset. This proposed condition for Group 1 cryptoassets should better take into account that there may be a lack of regulatory clarity regarding when settlement finality occurs when a transaction is recorded on a permissionless blockchain. Indeed, ongoing policy initiatives aim to accommodate the lack of clarity with regards to this issue,18 which the International Monetary Fund’s primer for supervisors on blockchain consensus mechanisms (“IMF Supervisory Blockchain Primer”) notes can make fitting certain blockchains into existing broader settlement regulation difficult.19
Moreover, many new Proof-of-Stake blockchains upon which stablecoins are run achieve settlement finality near-instantaneously (e.g., within 400 milliseconds).20 In fact, the IMF Supervisory Blockchain Primer has noted that regulatory sandboxes that accommodate cryptoassets built on these blockchains could be particularly important, as the security of these networks inherently improves with scale.21 Accordingly, it would be best for the BCBS to consider adjustments to the proposed requirement regarding documentation of settlement finality for cryptoassets based on a permissionless blockchain, while maintaining the spirit of this element of the proposal.
Recommendation 3: Classification condition 2 should be reworded to make clear that approval by a relevant regulator of a cryptoasset arrangement to offer a particular cryptoasset is sufficient for the purposes of determining whether this condition is met. If a cryptoasset arrangement does not meet this condition, the BCBS should nevertheless consider introducing further alternative approaches to ensuring that issues regarding settlement finality can be adequately addressed for stablecoins based on a permissionless blockchain.
Classification conditions 3 and 4 are redundant and should be removed.
We appreciate the spirit of the Basel Committee’s “classification condition 3” and “classification condition 4” for determining whether cryptoassets should be categorized in Group 1. However, when considering the other elements of the BCBS’s proposed framework, these conditions appear redundant and add confusion. Classification condition 3 introduces requirements that the network on which a cryptoasset operates be designed and operated to mitigate and manage any material risks.22 Presumably, a cryptoasset meeting classification conditions 1 and 2 edited per our suggestions would indicate that material risks surrounding the underlying stablecoin arrangements’ design and operation are sufficiently mitigated. Indeed, classification condition 2 edited per our suggestions already requires banks or regulators to validate cryptoasset arrangements to meet conditions designed to mitigate risk, including those related to issuance, validation, redemption, and transfer of the asset. Separately, our proposed changes to the BCBS’s contemplated infrastructure risk add-on would address risks that classification condition 3 as proposed evidently aims to mitigate.
With respect to classification condition 4, as written, it is potentially unworkable given existing regulatory frameworks for cryptoassets and the novelties of the networks upon which cryptoassets are built. For example, classification condition 4 states that in order for a stablecoin to be classified as a Group 1 asset, wallet providers should be regulated and supervised, or subject to “appropriate risk management standards,”23 even though this may not be required in certain jurisdictions and there is no indication as to who would determine whether risk management standards in place are “appropriate.” Similarly, what it would mean for a node validator to be subject to appropriate risk management standards – as classification condition 4 as written also contemplates – is also unclear and is an unrealistic requirement given existing regulatory frameworks.
Recommendation 4: Classification condition 3 should be removed. The modified classification condition 2 and modified infrastructure risk add-on proposed below adequately address these risks.
Recommendation 5: Classification condition 4 should be removed as it is inconsistent with existing regulatory frameworks.
FTX US believes that the five recommendations set forth above could be implemented without a significant increase in risk to the banking system. The recommendations simply ensure the Basel Framework does not penalize exposure to certain well-regulated stablecoin arrangements built on permissionless blockchains.
B. The infrastructure risk add-on should be removed or substituted with a bank-conducted risk assessment regarding the risks applicable to a particular blockchain.
The framework contemplates an infrastructure risk add-on which is designed to control for any “unforeseen risks” posed by distributed ledger technology (“DLT”). The infrastructure risk add-on presumes cryptoassets to be inherently risky without specifying the characteristics that create such risks and allowing cryptoasset firms to mitigate accordingly. While we disagree with the current approach to the infrastructure risk add-on, which treats all DLT as the same for the purposes of calculating the infrastructure risk add-on for Group 1 cryptoassets,24 in the event that the BCBS proceeds with the approach, classification condition 3 becomes redundant as the add-on is designed to capture additional risk. In place of the infrastructure risk add-on, we believe that prudential regulators should facilitate the mitigation of risks associated with DLT networks upon which cryptoassets are built by allowing banks to identify unique operational risks posed by particular blockchains through a transparent approach that results in a determination of whether cryptoassets built on that blockchain face a systematic, inadequately-controlled additional risk. Such a review could consider factors including the history of transactions on that blockchain and how it has performed during periods of market stress.
Recommendation 6: The infrastructure risk add-on should ideally be removed. If that is not possible, it should be replaced with a framework that allows banks to tailor the add-on based on the bank’s determination of the infrastructure risks associated with a particular form of DLT.
C. Group 2 cryptoassets should be broken down into sub-categories and subject to more appropriate risk weights.
The Second Consultation creates a new Group 2a cryptoasset sub-category defined as a cryptoasset meeting both of the following conditions:25
- Either (i) a spot Group 2 cryptoasset for which a derivative or ETF/ETN is traded on a regulated exchange that solely references the cryptoasset; or (ii) a cash-settled derivative or ETF/ETN that: (a) references a reference rate published by a regulated exchange, (b) has been explicitly approved by markets regulators for trading or is cleared by a qualifying central counterparty, or (c) references a regulated derivative or ETF/ETN meeting the conditions set forth in (b).
- The underlying Group 2 cryptoasset is sufficiently liquid, meaning that it has both (i) a market capitalization of at least USD 10 billion; and (ii) a 10% trimmed mean of daily trading volume against major fiat currencies of at least USD 50 million.
Cryptoassets within the Group 2a cryptoasset sub-category would be subject to modified versions of the market risk requirements, permitting a limited degree of hedging when determining a bank’s net exposure for purposes of determining capital exposure. Group 2 cryptoasset exposures would generally still, however, be subject to a 100% capital charge after the hedging is taken into account,26 and depending on the circumstances could also be subject to other add-on capital requirements such as for counterparty credit risk27 and an even higher capital charge.28
FTX US appreciates that the BCBS took into account feedback from industry stakeholders that the originally contemplated Group 2 cryptoasset category was too broad. However, the two sub-categories proposed in the Second Consultation still fail to reflect the diversity of cryptoassets in the marketplace and still subject the vast majority of cryptoassets to unreasonably punitive capital treatment.
The proposed risk weight of 1250% for Group 2 cryptoassets is arbitrary and unreasonable for many cryptoassets. The table below contrasts the contemplated Group 2 cryptoasset risk weight against the Basel III risk weights for other asset classes, including those that are often associated with an elevated degree of risk. The table demonstrates that banking book claims on high-risk, illiquid sovereign debt is risk weighted under the simple risk weight approach at 100%, compared to 1250% contemplated for Group 2 cryptoassets. Further, Basel III risk weights equity holdings which are not publicly traded at 400%. Unlike many Group 2 cryptoassets, unlisted equity holdings do not benefit from the price discovery, market transparency, and liquidity granted by certain cryptoasset markets. Notably, the table below also indicates that some cryptoasset markets are even larger and more liquid than markets for listed equities, which are subject to a 300% risk weight.
In our view, given the risk weights in place for other asset classes and the characteristics for these markets, it is unreasonable to assign a 1250% weight to all Group 2 cryptoassets currently included in the BCBS’s contemplated Group 2b, and the thresholds for Group 2a are too restrictive.
Recommendation 7: We support as an alternative to two static Group 2 sub-categories a quantitative approach that accounts for volatility similar to that proposed in response to the First Consultation by other industry commenters.40 Under this approach, risk weights would be determined for all Group 2 cryptoassets based on the individual cryptoasset’s performance against a holistic, multi-factor risk assessment. Factors that would be appropriate to include are those that determine the risk of a particular cryptoasset such as liquidity, volatility, use, and relevant design features. However, if the BCBS does not find such an alternative agreeable, we encourage the BCBS to at least incorporate two additional Group 2 sub-categories, as outlined below, as well as a less restrictive threshold for Group 2a. The proposed sub-categories would be split by the market cap and trading volume of the cryptoassets, as well as whether an accompanying derivative/exchange-traded product (“ETP”) market exists. The treatment of trading book Group 2a cryptoasset exposures would be based on the BCBS market risk framework.
The criteria for inclusion in our proposed Group 2a is stringent enough that currently only the most liquid and well-developed cryptoassets in Group 2 will fall into the sub-category. Accordingly, we suggest the creation of two additional Group 2 sub-categories, with each Group 2 sub-category having risk weights tailored based on cryptoasset market size and liquidity. Furthermore, sub-categories 2a through 2c should each permit hedging, as this incentivizes the utilization of hedging strategies for a broader range of cryptoassets, contributing to the evolution of safer and more stable markets overall.
Because cryptoassets such as Bitcoin and Ether appear fundamentally less risky than unlisted equity, the Basel Committee should at the least consider a 300% risk weight – comparable to listed equity – for this sub-category. For cryptoassets that are similar in market size and liquidity but lack a derivative or exchange traded product, a slightly higher risk weight of 350% would be appropriate.
For those cryptoassets with a less established market position and lower levels of liquidity, we recognize that higher risk weights may be appropriate. We propose a sub-category 2c with an 400% risk weight – equivalent to unlisted equities – for assets which meet a threshold of $2 billion market cap and 10% trimmed mean of daily trading volume of at least $25 million. These thresholds demonstrate market interest and some levels of liquidity, but elevated levels of risk relative to Group 2b. For any asset which does not meet the standards of market legitimacy and liquidity posed by the thresholds in the first three proposed categories, we support a conservative 800% risk weight.
We believe that by expanding the sub-categories in Group 2 the framework will more appropriately treat cryptoassets based on their actual stability and riskiness. These lower risk weights would lessen the cliff effects currently present in the Basel Committee’s proposed standard where a stablecoin falls from Group 1 to Group 2.
Importantly, across all of Group 2 sub-categories, it is important for the BCBS to provide additional clarity regarding the volume prong of its approach to determining in what subcategory a particular cryptoasset belongs. The BCBS currently contemplates only calculating trading volume “with major fiat currencies”46 for purposes of determining the appropriate Group 2 sub-category. However, as mentioned above, most cryptoasset trading involves a stablecoin on one side of the transaction. Further, a considerable portion of cryptoasset trading volume occurs against leading tokens including Bitcoin and Ether, tokens which have a daily trading volume higher than that of many fiat currencies. Accordingly, by only measuring trading volume with fiat currency trading volume, the BCBS would be artificially and dramatically reducing the number of cryptoassets eligible for less punitive Group 2 capital treatment.
Recommendation 8: The Basel Framework’s measurement of daily trading volume should take into account trading volume against fiat currency as well as both: (i) fiat-pegged stablecoins backed by fiat currency-denominated assets and with appropriate disclosure regimes in place; and (ii) Group 2a cryptoassets including Bitcoin and Ether.47 This would result in a more accurate understanding of the liquidity of cryptoasset markets, allowing for more cryptoassets that align with the intent of the exemption to be eligible.
D. Treatment of hedging strategies for cryptoassets should be less restrictive.
The BCBS’s approach also fails to adequately accommodate the risk mitigation opportunities presented by hedging in the context of digital assets and digital asset derivatives. The BCBS has long recognized the benefits of hedging for exposures to assets, but as written, the BCBS’s approach would not adequately recognize hedging of digital asset exposures. Accordingly, what constitutes “exposure” to Group 2 cryptoassets should be modified in the proposed framework to appropriately recognize hedging and netting benefits across a broader range of Group 2 cryptoassets. Unfortunately, the proposed framework discourages appropriate risk mitigation in its application of a punitive risk weight to the greater of the absolute value of long and short positions.
Recommendation 9: We recommend that for cryptoassets that meet the criteria we set forth above for Group 2b and Group 2c cryptoassets, the BCBS allows for netting of long and short derivatives exposures that reference the same asset, as well as hedging strategies that banks use when hedging against the risks of other exposures. Indeed, the BCBS should generally treat hedged digital asset exposures in accordance with how it treats other hedged exposures with similar risk characteristics. This updated treatment would more appropriately reflect the risk presented by these exposures and would encourage market participants to engage in risk-mitigating behavior.
E. The unnecessarily restrictive aggregate Group 2 cryptoasset exposure limit should be raised.
The BCBS contemplates an aggregate exposure limit of 1% of Tier 1 capital for all Group 2 cryptoassets. This limit is a departure from the BCBS’s approach to risk management in other contexts, as the BCBS allows a global systemically important bank (“G-SIB”) to maintain exposures to another G-SIB of up to 15% of Tier 1 capital.48 If the Basel Committee thinks that it is appropriate for a major global bank to maintain interbank liabilities with another major global bank equivalent to 15% of its Tier 1 capital, then it seems as if the 1% aggregate threshold currently contemplated for all Group 2 cryptoassets is unreasonable and arbitrary.
Recommendation 10: We ask that any such threshold be grounded in analysis performed by prudential regulators, rather than handed down by the BCBS, and believe that an exposure limit of 15% of Tier 1 capital for all Group 2 cryptoassets is likely more appropriate. This exposure limit should take into account offsetting of positions.49 Indeed, any exposure limit applied must be adjusted to account for the fact that all cryptoassets are being grouped together for the limit whereas other assets are generally grouped by counterparty for purposes of the exposure limit.
F. Off-balance sheet cryptoasset exposures should not be treated the same as on-balance sheet exposures.
The Second Consultation clarifies the definition of “exposure” to include off-balance sheet amounts, including cryptoassets held by a bank through non-fiduciary custodial services.50 This sweeping applicability would impose an unreasonably punitive capital treatment on banks simply for offering cryptoasset non-fiduciary custodial services, an area where banks are highly regulated and assume no balance sheet exposure. This in turn would disincentivize banks from bringing rules and stability to crypto markets, thus threatening the health and vibrancy of these markets by dramatically undermining the competitiveness of well-regulated banks with a long history of offering non-fiduciary custodial services. It is unclear what policy goal is served by this anomalous treatment of custody of cryptoassets and why that treatment is so different from all other assets held in custody by banks.
Recommendation 11: We recommend that the Basel Committee’s proposed standards text for prudential treatment of cryptoassets not be applicable to off-balance sheet non-fiduciary custody of cryptoassets.
Conclusion
FTX US appreciates the BCBS’s active engagement in the ongoing discussions on designing a prudential framework for cryptoassets, in both the First and Second Consultations. The BCBS posed a number of thoughtful questions in its consultations, which we hope will help contribute towards a productive policy framework.
As it stands, the Second Consultation will deter banks from participating in cryptoasset activity. It is excessively conservative, fails to account for many unique characteristics of blockchain technology, and does not sufficiently leverage the expertise and discretion of local prudential regulators. If banks do not participate in the crypto ecosystem, this will hinder technological innovation, exacerbate consumer risks, and elevate risks in the financial sector as a whole.
If implemented, FTX US believes the recommendations discussed above will provide banks with a robust, yet not overly prescriptive and punitive, framework for the prudential treatment of cryptoasset exposures, while satisfactorily managing risk.
We look forward to continuing our engagement with the BCBS on this topic and are happy to answer any questions the BCBS may have regarding the contents of this letter.
Sincerely,
Samuel Bankman-Fried
Chief Executive Officer, FTX US
1 BCBS, Prudential treatment of cryptoasset exposures (June 2022), https://www.bis.org/bcbs/publ/d533.pdf [hereinafter “BCBS Second Consultation”].
2 The FTX group of companies was established by three Americans, Samuel Bankman-Fried, Gary Wang and Nishad Singh, with international operations commencing in May 2019 and the U.S. exchange commencing in 2020. The business was established in order to build a digital-asset trading platform and exchange with a better user experience, customer protection, equitable access, and innovative products, and to provide a trading platform robust enough for professional trading firms and intuitive enough for first-time users. FTX’s international exchange, which is not available to U.S. users, is subject to prudential supervision and licensing in the European Union (EU), Japan, Australia, and Dubai. FTX has aimed to combine the best practices of the traditional financial system with the best from the digital-asset ecosystem.
3 BCBS Second Consultation at 4.
4 If a cryptoasset is not determined to be a Group 2 cryptoasset, then it would be subject to a highly punitive risk-weight. See infra notes 25-28 and accompanying text.
5 The Block, Share of Trade Volume by Pair Denomination (accessed Sept. 29, 2022), https://www.theblock.co/data/crypto-markets/spot/share-of-trade-volume-by-pair-denomination.
6 BCBS Second Consultation, SCO60.13(1). Peg value refers to “the value of the reference asset(s) to which one unit of the cryptoasset is designed to be redeemable.” Id. at 3.
7 Id., SCO60.13(1).
8 Id., SCO60.13(2).
9 Id. at 3.
10 An analysis from a group of trade associations which measured the delta between daily returns of ETFs and daily returns of the underlying index found that the United States Oil Fund exceeded a 10 bps breach 220 times in 2020 and 115 times in 2021. See Global Financial Markets Association, Financial Services Forum, Futures Industry Association, Institute of International Finance, the International Swaps and Derivatives Association, and Chamber of Digital Commerce, Comments in Response to the Consultative Document on the Prudential
Treatment of Cryptoasset Exposures 16 (Sept. 20, 2021), https://www.bis.org/bcbs/publ/comments/d519/gfma_fsf_fia_iif_isda_chamber.pdf [hereinafter “Joint Trades Letter”].
11 See, e.g., Circle, Crypto that’s held to a higher standard (accessed Sept. 27, 2022), https://www.circle.com/en/usdc (stating that “every digital dollar of USDC can always be exchanged 1:1 for cash”).
12 We recommend deleting sections SCO60.14 through SCO60.16.
13 BCBS Second Consultation, SCO60.12.
14 BCBS Second Consultation, SCO60.17.
15 The New York Department of Financial Services (“NYDFS”) has in place requirements for issuers of stablecoins. Among other requirements, NYDFS requires stablecoin issuers to adopt redemption policies approved by NYDFS conferring on any lawful holder the “right to redeem units of the stablecoin in a timely fashion at par” and to maintain full backing of reserve assets for the stablecoin. These requirements control risks that classification condition 1 aims to address. NYDFS, Guidance on the Issuance of U.S. Dollar-Backed Stablecoins (June 8, 2022), https://www.dfs.ny.gov/industry_guidance/industry_letters/il20220608_issuance_stablecoins.
16 Should BCBS choose to maintain a basis risk test, we also believe that an issuer being supervised and regulated by a regulator that applies prudential liquidity requirements should be an alternative to both tests.
17 BCBS states that with a permissionless blockchain, the “validation of transactions can be done by any participating agent, or distributed among several agents or intermediaries, which could be unknown to the users.” BCBS Second Consultation, SCO60.130.
18 See, e.g., European Union, Regulation on a pilot regime for market infrastructures based on distributed ledger technology (May 30, 2022), https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32022R0858.
19 IMF, Blockchain Consensus Mechanisms: A Primer for Supervisors (Jan. 2022), https://www.imf.org/-/media/Files/Publications/FTN063/2022/English/FTNEA2022003.ashx.
20 Centre Consortium, Centre Consortium Announces Solana as Official Chain for USDC (Oct. 21, 2020), https://medium.com/centre-blog/centre-consortium-announces-solana-as-official-chain-for-usdc-6e12be8f764.
21 See supra note 19 at 11 (“Weaknesses within PoS consensus mechanisms might impact stability and integrity, such as centralization in smaller networks and the inefficiency of staking. In the long term, these flaws can lead to centralization-related issues, which can be particularly problematic in smaller networks or networks in their infancy-and can therefore impact market integrity. Authorities could consider elements like sandboxing to help protect nascent networks”).
22 BCBS Second Consultation, SCO60.21.
23 BCBS Second Consultation, SCO60.23.
24 BCBS Second Consultation, SCO60.57.
25 BCBS Second Consultation, SCO60.60.
26 Id. at 5 (“[T]hese modified approaches retain a conservative 100% capital charge (consistent with the 1250% risk weight), but permit banks to recognise, to a limited extent, the hedging of long and short exposures”).
27 BCBS Second Consultation, SCO60.103-104.
28 A 1250% risk weighting would result in at least a 100% capital charge (given an 8% minimum capital ratio), but could be higher given other factors that increase the minimum capital ratio, such as certain jurisdiction-specific buffers in the U.S. and the U.K.
29 BCBS, Calculation of RWA for credit risk (Dec. 15, 2019), https://www.bis.org/basel_framework/standard/CRE.htm?export=pdf. This column presents risk weights under the simple risk weight method.
30 S&P Global Ratings, Sovereign Risk Indicators 2022 Estimates (July 11, 2022), https://disclosure.spglobal.com/sri/.
31 Jevans Nyabiage, Angola looks at refinancing debt as it faces higher repayments on Chinese loan, South China Morning Post (Mar. 11, 2022), https://www.scmp.com/news/china/diplomacy/article/3170020/angola-looks-refinancing-debt-it-faces-higher-repayments.
32 Kristjan Mee, The essential guide to investing in emerging market debt, Schroders (Feb. 2, 2021), https://prod.schroders.com/en/sysglobalassets/digital/insights/2021/02-february/the-essential-guide-to-investing-in-emd/600267_sc_idd_essential-guide-to-investing_02.02.21.pdf.
33 S&P Dow Jones Indices, S&P MidCap 400 (accessed Sep. 30, 2022), https://www.spglobal.com/spdji/en/indices/equity/sp-400/#data.
34 Nasdaq, Comparing Stocks and ETFs and Futures (July 22, 2021), https://www.nasdaq.com/articles/comparing-stocks-and-etfs-and-futures-2021-07-22.
35 Allison Herren Lee, Going Dark: The Growth of Private Markets and the Impact on Investors and the Economy, SEC (Oct. 12, 2021), https://www.sec.gov/news/speech/lee-sec-speaks-2021-10-12.
36 Laurent Belsie, The Liquidity Cost of Private Equity Investments, NBER (Oct. 10, 2016), https://www.nber.org/digest/oct16/liquidity-cost-private-equity-investments.
37 BCBS Second Consultation, SCO60.88.
38 Cryptocurrency Prices by Market Cap, CoinGecko (accessed Sept. 29, 2022), https://www.coingecko.com/
39 See supra note 38.
40 See, e.g., Coinbase, Comments in Response to the Consultative Document on thePrudential Treatment of Cryptoasset Exposures 24 (Sept. 10, 2021), https://www.bis.org/bcbs/publ/comments/d519/coinbase.pdf.
41 This is a modified version of a condition set forth in BCBS’s proposed definition of Group 2a cryptoassets. We believe that the term “exchange traded product,” defined to include exchange-traded funds and exchange-traded notes, should be used rather than “exchange-traded fund (ETF)/exchange-traded note (ETN).”
42 We recommended an updated approach to calculating daily trading volume, described below in recommendation 8, that more appropriately reflects the characteristics of crypto trading markets.
43 The 300% figure is equivalent to the risk-weight assigned to mid-cap publicly traded securities for which markets are oftentimes much less deep and liquid than cryptoasset markets, and for which derivatives solely referencing the asset do not exist.
44 The risk weight for Group 2b reflects that markets for cryptoassets in this category are generally more liquid and less volatile than markets for unlisted equities.
45 Notably, Group 2c captures cryptoassets with relatively lower market capitalization and thus not captured in either of the previous two thresholds but with regulated exchange traded products referencing the underlying cryptoasset. See, e.g., VanEck, VanEck Algorand ETN Product Description (Aug. 31, 2022), https://www.vaneck.com/at/en/library/fact-sheets/vgnd-fact-sheet.pdf/. Further, the $25 million daily trading volume threshold is roughly equivalent to the average daily trading volume for Russell 2000 equities. See supra note 33 and accompanying text.
46 BCBS Second Consultation, SCO60.60(2)(b).
47 If BCBS accepts our recommendations regarding Group 1 classification conditions, then volume thresholds for Group 2 subcategories could be determined using combined trading volume against Group 1 stablecoins, Group 2a cryptoassets, and fiat currency.
48 See BCBS, The treatment of large exposures in the Basel capital standards - Executive Summary (Apr. 30, 2018), https://www.bis.org/fsi/fsisummaries/largeexpos.htm.
49 Based on capitalization ratios for Global Systemically Important Banks (GSIBs) data from 2017, increasing the total exposure threshold for Group 2 cryptoassets to 15% of common equity tier 1 capital would only lead to a maximum exposure level of 2% of risk-weighted assets for U.S. banks. FDIC, Global Capital Index: Capitalization Ratios for Global Systemically Important Banks (GSIBs) (Dec. 31, 2017), https://www.fdic.gov/about/learn/board/hoenig/capitalizationratio4q17.pdf.
