Stablecoins: Growth Potential and Impact on Banking

Board of Governors of the Federal Reserve System International Finance Discussion Papers ISSN 1073-2500 (Print) ISSN 2767-4509 (Online) Number 1334 January 2022 Stablecoins: Growth Potential and Impact on Banking Gordon Y. Liao and John Caramichael Please cite this paper as: Liao, Gordon Y. and John Caramichael (2022). “Stablecoins: Growth Potential and Impact on Banking,” International Finance Discussion Papers 1334. Washington: Board of Governors of the Federal Reserve System, https://doi.org/10.17016/IFDP.2022.1334. NOTE: International Finance Discussion Papers (IFDPs) are preliminary materials circulated to stimulate discussion and critical comment. The analysis and conclusions set forth are those of the authors and do not indicate concurrence by other members of the research staff or the Board of Governors. References in publications to the International Finance Discussion Papers Series (other than acknowledgement) should be cleared with the author(s) to protect the tentative character of these papers. Recent IFDPs are available on the Web at www.federalreserve.gov/pubs/ifdp/. This paper can be downloaded without charge from the Social Science Research Network electronic library at www.ssrn.com.

Stablecoins: Growth Potential and Impact on Banking GordonLiao* JohnCaramichael† January 2022 Abstract Stablecoins have experienced tremendous growth in the past year, serving as a possible breakthrough innovation in the future of payments. In this paper, we discuss the current use cases and growth opportunities of stablecoins, and we analyze the potential for stablecoins to broadly impact the banking system. The impact of stablecoin adoption on traditional banking and credit provision can vary depending on the sources of inflow and the composition of stablecoin reserves. Among the various scenarios, a two-tiered banking system can both support stablecoin issuance and maintain traditional forms of credit creation. In contrast, a narrow bank approach for digital currencies can lead to disintermediation of traditional banking, but may provide the most stable peg to fiat currencies. Additionally, dollar-pegged stablecoins backed by adequately safe and liquid collateral can potentially serve as a digital safe haven currency during periods of crypto market distress. Keywords: Stablecoins, Digital currencies, Credit intermediation, Banking, Systemic risk, Fintech, Financial innovation, Payment system JEL Classifications: E40, E50, F33, G10, G20, O30 *Liao:Whenthispaperwaswritten,LiaowasaSeniorEconomistattheBoardofGovernorsoftheFederal ReserveSystem.Email:gliao@post.harvard.edu †Caramichael:BoardofGovernorsoftheFederalReserveSystem.Email:john.caramichael@frb.gov. WethankShaghilAhmed,BrettBerger,JorgeHerrada,AlexLee,PeterLone,YiSun,FrankWarnock,Mark VanDerWeide,PaulWong,JeffZhang,membersoftheFederalReserveBoardTechnologyLabintheDivisionofReserveBankOperationsandPaymentSystems,andparticipantsattheFederalReserveInternational FinanceWorkshopforhelpfulfeedback.Theviewsexpressedinthispaperaresolelytheresponsibilityofthe authorsandshouldnotbeinterpretedasreflectingtheviewsoftheBoardofGovernorsoftheFederalReserve SystemorofanyotherpersonassociatedwiththeFederalReserveSystem. Firstdraft:May2021. Finaldraft: September2021.

Stablecoins are digital currencies that peg their value to an external reference, typically the U.S. dollar (USD). Stablecoins play a key role in digital markets, and their growth couldspurinnovationsinthebroadereconomy. Inthepastyear,USD-peggedstablecoins circulatingonpublicblockchainshaveseenexplosivegrowth,withacombinedcirculating supply of nearly $130 billion as of September 2021 – a more than 500% increase from one yearago. As stablecoins gain increasing attention in public discourse, a host of issues have been raised, including the stability of their pegs, consumer protection, know-your-customer andanti-moneylaunderingcompliance,andthescalabilityandefficiencyofsettlements.1 In this note, we focus our discussion on the potential impact of stablecoins on the banking system and credit intermediation.2 While a range of stablecoin-related issues may beresolvedwithappropriateinstitutionalsafeguards,regulations,andtechnicaladvancements, sustained growth in stablecoins in circulation would ultimately impact the traditionalbankingsysteminsignificantwaysthatareimportanttounderstand. In this note, we first discuss the basics of stablecoins, their current use cases, and their growthpotential. Second,westudyhistoricalbehaviorsofstablecoinsduringpastepisodes ofcryptoandbroadfinancialmarketdistress. Wefindthatdollar-peggedstablecoinshave exhibited safe asset qualities in that their prices in the secondary market temporarily rise abovethepegduringtimesofextrememarketdistress,incentivizingtheissuanceofmore stablecoins. Wealsohighlighttheriskofa”run”oncertainstablecoinsthatarebackedby non-cash-equivalentriskyassets. Finally, we outline possible scenarios for bank reserves, credit intermediation, and central bank balance sheets should stablecoins gain broader traction. Our research suggests thebroadadoptionofasset-backedstablecoinscanpotentiallybesupportedwithinatwotiered, fractional reserve banking system without a negative impact on credit intermediation. In such a framework, stablecoin reserves are held as commercial bank deposits, and commercial banks engage in fractional reserve lending and maturity transformation astheynormallywouldwithtraditionalbankdeposits.3 Wealsofindthatthereplacement of physical cash (banknotes) with stablecoins could result in more credit intermediation. Incontrast,anarrowbankingframework,inwhichstablecoinissuersarerequiredtoback their stablecoins with central bank reserves, minimizes the risk of ”runs” on stablecoins butcanpotentiallyreducecreditintermediation. 1Amongthevariousissuesassociatedwithstablecoinadoptionandregulations,thestabilityand“runrisk” areofprimaryconcern. SeeGortonandZhang(2021)foradiscussionofregulatorysafeguardssurrounding stablecoins. 2This paper does not consider all potential impacts of stablecoins on the banking system. For example, severalkeyareasremainunexplored,suchaschangestoleverageratios;liquiditycoverageandtherunrate ofdifferentformsofbankdeposits;netstablefundingratios;thedistributionofdepositsandreservesacross banks; the challenges of know-your-customer and anti-money laundering policies; and the transmission of monetarypolicy. 3Thisnecessarilyassumesthatstablecoindepositsaretreatedsimilarlyastransactionaldepositsforliquiditymanagement,depositoryinsurance,andregulatorypurposes. 1

I. The basics of stablecoins Stablecoinsaredigitalcurrenciesrecordedondistributedledgertechnologies(DLTs),usually blockchains, that are pegged to a reference value.4 The majority of outstanding stablecoins are pegged to the U.S. dollar, but stablecoins can also be pegged to other fiat currencies, baskets of currencies, other cryptocurrencies, or commodities such as gold. Stablecoins serve as a store of value and a medium of exchange on DLTs, which enable stablecoinstobeexchangedorintegratedwithotherdigitalassets. Stablecoinsdifferfromtraditionaldigitalrecordsofmoney,suchasbankdepositaccounts, in two primary ways. First, stablecoins are cryptographically secured. This allows users to settle transactions near-instantaneously without double-spending or an intermediary that facilitates settlements. On public blockchains, this also allows for 24-hours-a-day/7days-a-week/365-days-a-yeartransactions.5 Second,stablecoinsaretypicallybuiltonDLT standards that are programmable and allow for the composability of services.6 In this context,“composability”meansstablecoinscanfunctionasself-containedbuildingblocks that interoperate with smart contracts (self-executing programmable contracts) to create payment and other financial services.7 These two key features underpin the current use casesofstablecoinsandsupportinnovationinboththefinancialandnon-financialsectors. The use of stablecoins recorded on public blockchains such as Ethereum, Binance Smart Chain,orPolygonhassurgedsince2020. AsoftheendofSeptember2021,thecirculating supply of the largest USD-pegged public stablecoins was almost $130 billion. In Figure 1, we show that the growth in the circulating supply of public stablecoins was especially strong in early 2021, averaging around 30% month-on-month for the first five months of theyear. Currenttypesofstablecoins The stablecoin is a nascent, broadly defined technology that can potentially take many forms. Thistechnologyiscurrentlyimplementedinspecificformsthatwedescribebelow andsummarizeinTable1. However,notethatstablecointechnologiesareintheirinfancy withahighpotentialforinnovation. Thecurrentimplementationsofstablecoinsdiscussed below,aswellastheircurrentstatusintheregulatorylandscape,donotreflectallpotential deploymentsofstablecointechnologies. 4Adistributedledgertechnology(DLT)isadecentralizeddatabasedistributedacrossmultiplenodes(devices).DLTsarecryptographicallysecuredanduseaconsensusmechanismtosynchronizethedatabaseacross their nodes instead of relying on a centralized administrator. A blockchain is a form of DLT where lists of records,orblocks,arechainedinsequence. 5FordiscussionsonDLTsinpayments,clearing,andsettlements,seeMillsetal.(2016). 6Composability is a systems design principle emphasizing interoperability of individual components in formingamorecomplexsystem. 7SeeLeeetal.(2021)foradiscussionof“Whatisprogrammablemoney?”andSzabo(1994)foradiscussion ofsmartcontracts. 2

Figure1: CirculatingsupplyofUSD-peggedpublicstablecoins ($) Billions 140 140 120 Other 120 Dai 100 Binance USD 100 USD Coin 80 Tether 80 60 60 40 40 20 20 0 0 2019 2020 2021 CirculatingsupplyofthetenlargestUSD-peggedpublicstablecoinsbymarketcapitalization. Data extends from January 2019 through September 2021. Other category consists of Fei, TerraUSD, TrueUSD,PaxosDollar,NeutrinoUSD,andHUSD.Thelegendcorrespondstothepositionofeach stablecoininthefigure. Source: Authorcalculationsbasedonpublicblockchains. Publicreserve-backedstablecoins Mostexistingstablecoinscirculateonpublicblockchains,suchasEthereum,BinanceSmart Chain, or Polygon. Of these public stablecoins, most are backed by cash-equivalent reservessuchasbankdeposits,Treasurybills,andcommercialpaper. Thesereserve-backed stablecoins are also referred to as custodial stablecoins, as they are issued by intermediaries who serve as custodians of cash-equivalent assets and offer 1-for-1 redemption of theirstablecoinliabilitiesforU.S.dollarsorotherfiatcurrencies. The full backing and soundness of some public reserve-backed stablecoins have been called into question. In particular, Tether, the largest stablecoin by circulating supply, agreed to pay $41 million to settle a dispute with the U.S. Commodity Futures Trading Commission, which alleged that Tether misrepresented the sufficiency of its dollar reserves.8 Other widely used reserve-backed USD-pegged public stablecoins with varying levelsoffinancialauditsincludeUSDCoin,BinanceUSD,TrueUSD,andPaxosDollar. 8See Prentice (2021). Tether has also been investigated by the New York Attorney General’s office, and theU.S.DepartmentofJusticeisreportedlyinvestigatingwhetherTethercommittedbankfraud(Schoenberg, Robinson,&Faux,2021). 3

Table1: Currenttypesofstablecoins Type Description Examples Publicreserve-backed Backedbycash-equivalent Tether,USDCoin(USDC), reserves(deposits,T-bills, BinanceUSD(BUSD),Paxos commercialpaper),issuedby Dollar(USDP). centralizedfirms. Publicalgorithmic Backedbyovercollateralized Dai,TerraUSD,Fei,IRON cryptocurrencyand/orsmart (failed),Basis(failed). contractsthatautomatically defendthepegbybuyingor sellingthestablecoin. Institutionalorprivate Issuedbyfinancialand JPMCoin* non-financialinstitutionsfor internalaccounttransactions, liquiditymanagement,and transactionsbetweenuser accountswithinthesame privatenetwork. *Tokenizeddepositsissuedonpermissionedblockchain. Publicalgorithmicstablecoins The remaining fraction of existing public stablecoins use other mechanisms to stabilize their price instead of relying on the soundness of underlying reserves. These stablecoins are often called algorithmic stablecoins. While reserve-backed stablecoins are issued as a liability on the balance sheet of a legally incorporated firm, algorithmic stablecoins are maintainedbysystemsofsmartcontractsthatoperateexclusivelyonapublicblockchain. The ability to control these smart contracts is often conferred by the possession of a governance token, a specialized token primarily used for voting on changes to protocol or governance parameters. These governance tokens can also potentially serve as direct or indirectclaimaonfuturecashflowsfromtheusageofastablecoin’sprotocols. The public algorithmic stablecoin sector is highly innovative and difficult to categorize. However,onecangenerallythinkofthedesignofthesestablecoinsasbasedontwomechanisms: (1) the collateralized mechanism and (2) the algorithmic peg mechanism. Collateralized public stablecoins, such as Dai, are minted when a user deposits a volatile cryptocurrency,suchasEthereum,intoDai’ssmartcontractprotocols.9 Theuserthenreceives a loan of Dai (which is pegged to the dollar) against their crypto collateral, at a greater than100%collateralizationratio. IfthevalueoftheEthereumdepositfallsbelowacertain threshold,theloanisautomaticallyliquidated. 9Inpractice,Dai’scollateralalsoincludespublicreserve-backedstablecoinssuchasUSDCoin. Inthefuture,theprotocolmayfurtherdiversifyitscollateraltoperformliquiditytransformation. Recently,adigital currency-focusedsubsidiaryofSocie´te´ Ge´ne´ralsubmittedanapplicationtoreceive$20millioninDaiinexchangeforatokenizedAAA-ratedeuro-denominatedbond. 4

Incontrast,thealgorithmicpegmechanismusesautomatedsmartcontractstodefendthe pegbybuyingandsellingthestablecoinagainstanassociatedgovernancetoken.10 However,thesepegsmayexperienceinstabilityordesignflawsthatleadtode-pegging,asexemplifiedbythetemporarycollapseofFei,apublicalgorithmicpegstablecointhatbriefly de-peggedafteritslaunchinApril2021. Additionally, some algorithmic stablecoins use a blend of the collateralized and algorithmic peg mechanisms. For example, the failed IRON public algorithmic stablecoin drew elements from both mechanisms, as its peg was partially backed by USD Coin, a public reserve-backedstablecoin,andTITAN,thegovernancetokenfortheIRONFinanceprotocol. Institutionalorprivatestablecoins In addition to reserve-backed stablecoins that circulate on public blockchains, traditional financialinstitutionshavealsodevelopedreserve-backedstablecoins,alsoknownas”tokenizeddeposits”.11 Theseinstitutionalstablecoinsareimplementedonpermissioned(private)DLTs,andtheyareusedbyfinancialinstitutionsandtheirclientsforefficientwholesale transactions. The most well-known institutional stablecoin is JPM Coin.12 JPMorgan anditsclientscanuseJPMCoinfortransactionssuchasintradayreposettlementsandto manageinternalliquidity.13 These private, reserved-backed stablecoins are functionally and economically comparable to products offered by some money transmitters. For example, Paypal and Venmo (a Paypal subsidiary) allow users to make near-instant transfers and payments within their network,andbalancesheldatthesefirmsarebackedsimilarlytoareserve-backedstablecoin. The key difference is the use of centralized databases rather than a permissioned DLT. Otherpotentialtypesofstablecoins Asnotedpreviously,thestablecoinisanincipienttechnology,anditispossibletoimagine manywaysstablecoinscouldbeimplementedthroughouttheglobalfinancialsystem. For example,paymentscompaniescoulduseaninternal,permissionedDLTtosettlepayments efficiently, which would be conceptually equivalent to a stablecoin. One implementation 10Thisisroughlyanalagoustohowacentralbankmightdefendacurrencypegbybuyingandsellingits currencyagainstforeigncurrencyreserves.Thekeydifferenceisthatinsteadofanothercryptocurrencyasits ”foreigncurrencyreserves”,thealgorithmicpegmechanismusesthegovernancetoken. 11While these stablecoins are often described as ”tokenized deposits”, they share many similarities. The maindifferenceappearstobetheprivateandclosednatureofitsnetwork(JPMorgan,2020). 12Inarecentearningscall,JPMorgan’sCFOJenniferPiepszakstatedthatJPMCoinisnotastablecoin,but ratheraformof”tokenizingdepositstomakepaymentseasierforclients”(4Q20FinancialResults: Earnings CallTranscript,2021). 13JPMorgan(2019)providesexampleusageofJPMCoin. SeeCorrea, Du,andLiao(2020)andCopeland, Duffie,andYang(2021)forin-depthdiscussionsofinternalliquidityconstraintsandintra-dayliquidityneeds inthebankingsector. 5

of this is Visa’s B2B Connect system, a DLT-based payment system for wholesale interbank transactions. We may also see exchanges and clearinghouses rely on stablecoins or stablecoin-likeproductsfortransactingintokenizedfinancialmarkets. In the following section, we discuss the current use cases that are driving the growth of existing stablecoins, as well as potential innovations that could drive further growth and morediverseimplementationsinthefuture. II. Use cases and growth potential of stablecoins Robustusecasesaredrivingthecurrentgrowthinvariousformsofstablecoins. Wesummarize these use cases in Table 2. The most important current use case of stablecoins is theirroleintransactingincryptocurrencyonpublicblockchains. Investorsoftenpreferto usepublicstablecoinsinsteadoffiatbalancestotradecryptocurrency,becausethisallows fornear-instantaneous24/7/365tradingwithoutrelyingonnon-DLTpaymentsystemsor custodialholdingsoffiatcurrencybalances.14 Table2: Currentstablecoinusecases Usecase Details Digitalmarkets Stablecoinsareusedtotradedigitalassetsandserveasan onrampfromfiatcurrencytodigitalassetsrecordedon blockchains. Payments Stablecoinsareusedtofacilitatefastpeer-to-peerand cross-borderpayments. Theyalsoholdthepotentialfornew paymentinnovations,suchasprogrammablemoney(see below). Internaltransfers Institutionalstablecoinsfacilitatetransfersoffundswithina andliquidity firmandallowefficientmovementofinternalcashacross management subsidiariestomanageliquidityriskandregulatory requirements. DeFi Theprogrammabilityandcomposabilityofstablecoins currentlysupportsdecentralized,blockchain-based cryptocurrencymarketsandservices,knownas decentralizedfinanceorDeFi. Protocolsallowformarket making,collateralizedlending,derivatives,asset management,andotherservices. Besides their use in crypto trading, both public and institutional stablecoins are currently usedfortheirnear-instant,24/7,non-intermediatedpaymentswithpotentiallylowfees.15 This is especially relevant for cross-border transfers, which ordinarily can take multiple 14Manyexchangesdonotallowuserstoconverttheircryptoholdingsintoafiatcurrencybalance,sothe useofstablecoinsontheseexchangesisparticularlyimportant. 15Inthiscontext,anon-intermediatedtransactiondoesnotrelyonacentralizedintermediarytovalidatethe transactionandpreventdouble-spending. 6

daysanddemandhighfees. Firmsarealsousinginstitutionalstablecoinstonear-instantly movecashacrosstheirsubsidiariestomanageinternalliquidity,andtofacilitatewholesale transactionsinexistingfinancialmarkets,suchasintradayrepotransactions. Andfinally, becausepublicstablecoinsareprogrammableandcomposable,theyareusedheavilyindecentralized,publicblockchain-basedmarketsandservices,knownasdecentralizedfinance or DeFi.16 Systems of DeFi protocols allow users to use stablecoins to directly and transparently participate in a variety of cryptocurrency-related markets and services, such as market-making,collateralizedlending,derivatives,andassetmanagement,withouttraditionalintermediaries. AsofSeptember2021,about$60billionindigitalassetswerestaked (locked)inDeFiprotocols.17 Futuregrowthpotential The defining features of stablecoins, their cryptographic security and programmability, support the robust use cases that are currently driving the usage of existing public and institutional stablecoins. However, these features have the potential to drive innovation beyond current uses cases, which are mostly confined to cryptocurrency markets, certain peer-to-peerpayments,andinstitutionalliquiditymanagementbyverylargebanks. Looking forward, stablecoin technologies may see diverse implementations and drive innovationinseveralgrowthareas: moreinclusivepaymentandfinancialsystems, tokenizedfinancial markets, and the facilitation of microtransactions for technological advancements suchasWeb3.18 Moreinclusivepaymentandfinancialsystems Stablecoinshavethepotentialtospurgrowthandinnovationinpaymentsystems,allowing for faster, cheaper payments. Because stablecoins can be used to transfer funds nearinstantaneously peer-to-peer between digital wallets for potentially low fees, stablecoins may lower payment barriers and exert pressure on existing payment systems to provide better services.19 This is especially important for cross-border transfers, which can take several days to clear and carry high fees. These fees and delays are a burden on low and middle-incomecountries,whichreceivefinancialsupportfromremittances.20 Stablecoins may also support a more inclusive financial system through the growth of DeFi, which likely requires stablecoins as a necessary building block. It should be noted thatDeFifacesseriouschallenges,includingacomplexuserexperience,alackofconsumer 16ForanoverviewofdevelopmentsinDeFi,seeDeFiBeyondtheHype(2021). 17Source:TheBlock. 18Additionally,WongandManiff(2020)outlinefurtherusecasesofadigitalcurrencyissuedbyacentral bank. 19AsGovernorChristopherWallerrecentlynoted, ”Onecaneasilyimaginethatcompetitionfromstablecoinscouldpressurebankstoreducetheirmarkupforpaymentservices”(Waller,2021). 20TheWorldBankestimatedthatin2020,low-andmiddle-incomecountriesreceivedabout$540billionin remittances,withtransactionfeesaveraging6.5%–alossofabout$35billioninfinancialsupport(Ratha,Kim, Plaza,&Seshan,2021). 7

protection,frequenthacking,protocoldysfunctions,andmarketmanipulations. Additionally,virtuallyallDeFiprotocolsonlysupportthetradingorlendingofcryptocurrenciesor non-fungible tokens (NFTs). Should DeFi protocols mature beyond the current state and become integrated with the broader financial market to support real-world economic activities, DeFi could encourage a more inclusive financial system that allows investors to directly participate in markets without intermediation. This growth in DeFi would likely drivegrowthintheusageofstablecoins. Tokenizedfinancialmarkets Additionally,stablecoinsmayplayakeyroleintokenizingfinancialmarkets. Thiswould entail converting securities into digital tokens on DLTs and trading and servicing them with stablecoins. For delivery-versus-payment (DvP) transactions, such as security purchases, a tokenized market would allow for real-time settlement at very low costs. This couldincreaseliquidity,transactionspeeds,andtransparencywhilereducingcounterparty risk,tradingcosts,andotherbarrierstomarketparticipation. Thismightespeciallybenefit certainassetclasses,suchasrealestate,byallowingforfractionalownershipoftokenized assetsandmoretransparentpricediscovery. Forpayment-versus-payment(PvP)transactions,suchasacross-currencyswap,tokenizationwouldalsoallowfornear-instantaneous executioninsteadofthemarket’scurrentconventionalT+2framework,inwhichaswap’s paymentsaresettledtwobusinessdaysaftertheswapisstruck. Moreover,forbothkinds of transactions, tokenized financial markets would benefit from the programmability of DLTs, which could automate security servicing and regulatory requirements such as required holding periods. If financial markets were to become partially or completely tokenized,thiswouldlikelydrivefurthergrowthinstablecoinusage. Next-generationinnovations Finally,stablecoinsholdthepotentialtosupportnext-generationinnovations. OneexampleofsuchaninnovationisWeb3,apossiblemoveawayfromcentralizedwebplatforms and data centers towards decentralized networks.21 Under this paradigm, Internet services and social media platforms would shift their revenue from advertisements to microtransactions, facilitated by the advent of efficient, integrated online payment systems. One could imagine, for example, a search engine or video streaming platform supported by near-instant micropayments of stablecoin instead of advertising revenue and the sale of user data. If this shift in web services were to take hold, it would likely drive further growthinstablecoins. Inconclusion,thecurrentusageofstablecoinsisprimarilydrivenbycryptocurrencytrading, limited peer-to-peer payments, and DeFi. Looking forward, stablecoins may see further growth through their facilitation of more inclusive payments and financial systems, 21ForageneraldiscussionofWeb3andthenextgenerationofpayments,seeDixon(2018)andDixonand Haun(2020). 8

the tokenization of financial markets, and possible next-generation innovations such as Web3. III. Peg stability The stability of a stablecoin’s peg to its reference value is a central issue. It is not the focus of our paper, but we briefly discuss this important issue here.22 In this section, we willfirstoutlinethesourcesofpeginstabilityforcurrentpublicreserve-backedstablecoins and discuss how those sources may be addressed. We will then review how stablecoins could serve as a potential safe asset in digital markets, and provide evidence that current publicreserve-backedstablecoinsmayalreadyservethatroleincryptocurrencymarkets. Presently, peg instability for public reserve-backed stablecoins comes in two forms: investor redemption risk from the issuer and secondary market price dislocations. The formerrelatestothesafetyandsoundnessofastablecoin’sreserves. Ifstablecoinholderslose confidence in the soundness of a stablecoin’s backing, a run dynamic could ensue. A run on a stablecoin poses a risk of spillovers to other asset classes, as stablecoin reserves are soldofforunloadedtomeettheredemptiondemand.23 Additionally,arunonastablecoin coulddisruptthemarketsandservicesthatrelyonthestablecoinviainteroperablesmart contracts, causing further distress. We think this type of instability is addressable with properinstitutionaland/orregulatoryguardrailssuchastransparentfinancialauditsand adequate requirements on the liquidity and quality of stablecoin reserves. The concerns surrounding redemption risk and the extent to which they can be addressed have been notedrecentlyinQuarles(2021). Thesecondformofpeginstabilityforpublicreserve-backedstablecoinsarisesfromsupply anddemandimbalancesinthesecondarymarket. Asthesestablecoinsaretradedonboth centralizedanddecentralizedexchanges, theyarevulnerabletodemandshocksthatmay temporarilydislocatetheirpeguntilthestablecoinissueradjuststhesupply. Inparticular, because public stablecoins serve as a store of value on public blockchain-based markets, thesestablecoinsexperiencehighdemandduringcryptomarketdistressasinvestorsrush to liquidate their speculative positions into stablecoins. During these episodes, the price ofmajorpublicreserve-backedstablecoinstendstotemporarilyappreciateuntiltheissuer adjuststhesupply. Toprovideanexample,Figure2displaysthecryptomarketcrasheson March 12, 2020 and May 19, 2021. The first episode occurred during a period of general market turmoil surrounding concerns with the spread of Covid-19. The second episode occurred in a crypto market downturn associated with heavy deleveraging. In both periods,asthepriceofthespeculativecryptocurrenciesBitcoinandEthereumcrashed30to50 percent, the prices of major public reserve-backed stablecoins largely spiked upwards.24 22Fordiscussionofthestabilityofstablecoins,seeLyonsandViswanath-Natraj(2020).Additionally,Gorton andZhang(2021)outlinespossibleregulationsthatcouldmitigateconcernsaroundstablecoinstability. 23As a concrete example, a “run” on Tether could conceivably force the issuer to sell off its purportedly sizableportfolioofcommercialpaper,whichcouldcausedistressintheshort-termfundingmarket. 24OneexceptionisthatBinanceUSDtemporarilyde-peggedonthedownside. 9

For these episodes of extreme crypto market distress, stablecoins served as a digital safe asset, appreciating while more speculative crypto assets were temporarily in freefall, until the stablecoin issuers were able to increase their supply and purchase reserves and/or the stablecoins experienced downward price pressure from arbitrageurs.25 The behaviors of these public stablecoins are unique and differentiated from prime money market funds,whichexperiencedlargeoutflowsthatpromptedsellingofcommercialpaperholdings during the height of the 2008 Global Finance Crisis and the 2020 Covid-19 market turmoil.26 Figure2: Publicstablecoinsappreciateduringcryptomarketdistress Note: Hourlypricesofstablecoins,BitcoinandEthereum. TimeisinGMT.BitcoinandEthereum prices are in U.S. dollars, indexed to March 11, 2020 and May 18, 2021. Source: CryptoCompare API. These episodes demonstrate the potential for stablecoins to serve as a digital safe haven during market distress. While discussions about the financial stability risk from public reserve-backed stablecoins have largely focused on redemption risks that are unique to the form of reserves of individual stablecoins, our analysis suggests that counter-cyclical 25Griffin and Shams (2020) find an increase in Tether purchases and issuance following large declines in cryptopricesthroughanalysisofblockchaindata. 26Baba,McCauley,andRamaswamy(2009);Eren,Schrimpf,Sushko,etal.(2020) 10

demand for stablecoins in the secondary market can ameliorate risks of redemption runs duringtimesofbroadermarketdownturns. Withappropriatesafeguardsandregulations, stablecoins have the potential to provide a level of stability that is on par with traditional formsofsafevalue. IV. The potential impact of stablecoins on credit intermediation Ifstablecoinsweretoseebroadadoptionthroughoutthefinancialsystem,theycouldhave asignificantimpactonthebalancesheetsoffinancialinstitutions. Regulators,marketparticipants,andacademicsareparticularlyfocusedonthepotentialforstablecoinstodisrupt bank-ledcreditintermediation.27 Inthissection,weanalyzeseveralplausiblescenariosin whichreserve-backedstablecoinsseewidespreadadoptioninthefinancialsystem. Wefocus on reserve-backed stablecoins, rather than algorithmic stablecoins, as reserve-backed stablecoins are currently the largest and the most closely tied to the existing banking system. Usingthesescenarios, wehighlighthowtheimpactofstablecoinadoptiononcredit provisiondependscriticallyontwofactors: thesourcesofinflowintostablecoinsandthe compositionofastablecoin’sreserves.28 We summarize our results in Table 3. We find that in most scenarios we consider, credit provision would likely not be negatively affected. In fact, the replacement of physical currency(banknotes)bystablecoinscouldpotentiallyallowformorebank-ledcreditprovision. Anotableexceptionthatcanleadtosizablecreditdisintermediationisthescenario in which stablecoins are required to be fully backed by central bank reserves, which we callthenarrowbankframework. Inthisframework, redemptionrunriskisminimizedat theexpenseoflargercreditdisintermediation. Sourcesofinflows If stablecoins were to see widespread adoption, major inflows could come from three sources: physical currency (banknotes), commercial bank deposits, and cash-equivalent securities (or money market funds). These sources of inflows are summarized as rows in Table 3. First, as a form of digital currency, stablecoins stand to replace some portion of banknotesincirculation, especiallyastheeconomybecomesmoredigital. Insomeofour scenarios,asuserssubstituteawayfromphysicalcashintoreserve-backedstablecoins,we see an increase in credit provision. This is because banknotes, which are a direct liability ofthecentralbank, arereplacedbyreserve-backedstablecoins, whichcanbeinstruments ofcreditcreationvialoansorsecuritypurchases,dependingonthereserveframework. 27Forexample,arecentBankofEnglanddiscussionpaperpositedascenarioinwhichoutflowsfromcommercialbankdepositsintostablecoinsledtohigherinterestrates(Newformsofdigitalmoney,2021). 28Otherstudieshavealsoanalyzedbalancesheetimpactsfromtheintroductionofdigitalcurrencieseither issuedbythecentralbank(Centralbankdigitalcurrencies,2018)ortheprivatesector(Malloy&Lowe,2021). Relativetothesestudies,weanalyzeagreatersetofpossiblescenarioswithmorefocusonthegeneralequilibriumoutcomeandemphasisontheimpactoncreditintermediation. 11

Table 3: Impact on credit intermediation by stablecoin reserve framework and source of inflow Second, stablecoins could see inflows from commercial bank deposits should households andfirmsprefertoholdstablecoinsinsteadofatraditionalbalanceatacommercialbank. This source of inflow is of great interest to policymakers, as there is a common concern that a significant substitution away from deposits could disrupt credit provision by commercialbanks. Weshowthattheimpactofdepositsubstitutiononcreditprovisioncanbe positive, negative, or neutral, depending on the reserve framework. Finally, stablecoins could see inflows from cash-equivalent securities (or equivalently, money market funds). This would likely have no impact on credit provision, as it would entail recycling funds backintothebankingsystem,whichwediscussinalatersection. Compositionofreserves Theimpactofwidespreadreserve-backedstablecoinadoptiononcreditprovisionalsodependsonthecompositionofstablecoinreserves. Wepresentthreeplausiblestablecoinreserveframeworks: narrowbank, two-tieredintermediation, andsecurityholdings. These frameworksaresummarizedascolumnsinTable3. Under the narrow bank framework, stablecoins would be required to be backed by commercial bank deposits that are fully backed by central bank reserves. Equivalently, it is possible commercial banks could issue stablecoins (or tokenized deposits) that are fully 12

backed by central bank reserves. The narrow bank approach is roughly equivalent to a form of retail central bank digital currency where the digital currency is a liability of the centralbankbutaccessedbyhouseholdsandfirmsthroughanintermediarysuchasacommercialbankorfintechcompany. ThisframeworkhasbeenadoptedbythePeople’sBank ofChinainitsstate-backeddigitalcurrencyknownasDigitalCurrencyandElectronicPayments,thedigitalyuan,ore-CNY.Therequirementforstablecoinstomaintainreservesat the central bank has also been mentioned as a possibility in the proposed STABLE Act in theUnitedStates.29 While a narrow bank framework would guarantee the stability of a stablecoin’s peg as it iseffectivelyapass-throughcentralbankdigitalcurrency(CBDC),thisreserveframework posesthelargestriskofcreditdisintermediation. Periodsoffinancialstressorpaniccould leadtolargemigrationsofregularcommercialbankdepositsintonarrowbankstablecoins, which could disrupt credit provision. Though this credit disruption effect could be mitigated by limits on stablecoin holdings and differential reserve interest rates, the overall structure of the narrow bank approach to stablecoin reserves is potentially destabilizing for the banking system. Additionally, the narrow bank approach could lead to an expansion of the central bank’s balance sheet in order to accommodate the demand for reserve balancesfromstablecoinissuers. Theseconcernsaboutnarrowbankstablecoinsmirrortheconcernsaboutnarrowbanking more generally, which have been noted by the Federal Reserve. In a recently proposed regulation that would impact narrow banks (officially, pass-through investment entities or PTIEs), the Federal Reserve stated that it was ”concerned that [narrow banks] could disrupt financial intermediation in ways that are hard to anticipate, and could also have a negative effect on financial stability” (Regulation D: Reserve Requirements of Depository Institutions, 2019). Additionally, the Federal Reserve outlined serious concerns about the demandforreservebalances,stating,”Thedemandforreservebalancesby[narrowbanks] could become quite large. In order to maintain the desired stance of monetary policy, the FederalReservewouldlikelyneedtoaccommodatethisdemandbyexpandingitsbalance sheetandthesupplyofreserves”. In contrast to the narrow bank framework, under the two-tiered intermediation framework, stablecoins would be backed by commercial bank deposits that are used for fractionalreservebanking. Equivalently,itispossiblethatcommercialbanksissuestablecoins or provide tokenized deposits that are used for fractional reserve banking. To be clear, thisdoesnotmeanthatthestablecoinsarenotfullybacked. Rather,thestablecoinissuers rely on commercial bank deposits as assets, and the commercial banks practice fractional reserve banking withthe stablecoins and/or stablecoin deposits, meaning the stablecoins are ultimately backed by a mix of loans, assets, and central bank reserves. It would ef- 29STABLEActof2020(H.R.8827),forinstance,setsfortharequirementforcentralbankreservebacking ofstablecoins,“AnyissuerofstablecoinsshalldepositreserveswiththeapplicableFederalreservebankina segregatedaccountinanamountequaltothenominalredemptionvalueofalloutstandingstablecoinsissued bytheissuer,andsuchreservesshallserveascollateralforsuchstablecoins.” 13

fectively relabel some portion of regular deposits as stablecoin deposits. Importantly, for bankintermediationtoremainthesame,thetreatmentofstablecoindepositshastobethe same as non-stablecoin deposits in terms of the required reserve ratio, liquidity coverage andotherregulatoryandself-imposedrisklimits.30 Finally,stablecoinissuerscouldholdcash-equivalentsecuritiessuchasTreasurybillsand high-quality commercial paper instead of depositing their funds at commercial banks. These securities could be purchased directly or indirectly through money market funds. This is the main framework adopted by current issuers of public reserve-backed stablecoins,suchasTether,whichFederalReserveChairJeromePowellrecentlynotedare”like moneymarketfunds”(OversightoftheTreasuryDepartment’sandFederalReserve’sPandemic Response,2021). Scenarioconstruction In our scenarios, we consider the impact if one or several fiat-reserve backed stablecoins weretogainbroadadoptionwithinastylizedversionofthebankingsystem. Thebaseline balance sheet of this banking system is displayed in Table 4. Specifically, we consider a scenarioinwhichhouseholdsandfirmssubstitute$10awayfrombanknotes,commercial bank deposits, or securities, and we then conduct an accounting exercise to determine how the stablecoin’s adoption impacts the balance sheets of the central bank, commercial banks, and households and firms. We analyze how this impact differs depending on the stablecoin’sreserveframeworkanditssourceofinflows. Table4: Baselinebalancesheet Centralbank Commercialbanks Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Securities 18 Reserves 8 Reserves 8 Deposits 80 Deposits 80 Debt& 200 equity Physical 10 Loans& 92 Debt& 20 Physical 20 cash securities equity cash Securities& 100 other It is important to note that in constructing these scenarios, we are making several key assumptions. Thefirstisthatweareagnosticonthespecificformofthestablecointhatis adopted. Ourscenariosarenotintendedtoanalyze,forexample,thespecificimpactofthe widespreadadoptionofexistingstablecoinssuchasTether. Wedonotdistinguishwhether the adopted stablecoin is an institutional tokenized deposit, or a stablecoin circulating on apublicblockchain,orsomeotherform. Second,weareonlypresentingillustrativeedge 30Itisconceivablethatdepositsassociatedwithstablecoinissuancearecategorizedaseithertransactional orbrokereddeposits. Theformertypehasalowerassumed“runrate”inassessmentsofliquiditycoverage. Toachievefullequivalencetoretaildeposits,stablecoinswouldalsorequireFDICinsurance. 14

casesthatarenotexhaustive. Inreality, stablecoinscanseeinflowsfrommultiplesources and hold a variety of assets as reserves. Third, these scenarios do not capture secondary knock-on effects or feedback loops, and they do not address heterogeneous within-sector impacts. Finally, we assume that traditional deposits at commercial banks have a 10% requiredreserveratio. To illustrate the complex flows between the various parts of the banking system that underpinouredgecasescenarios,wevisualizeinFigure3asubsetofthestablecoininflows and reserve allocations we have discussed. Specifically, we use a diagram to show the flows of commercial bank deposits (Inflow A) and banknotes (Inflow B) into stablecoins, as well as the allocation of those funds into reserves in the form of commercial bank deposits(ReserveflowA)andsecurities(ReserveflowB).31 Figure3: Illustrationofstablecoininflowsandreserves Central bank Cash in circulation Cash is a liability of the central bank Inflow B Firms and households switch from cash to stablecoins Commercial bank reserves held at central bank Commercial banks Stablecoin issuers Inflow A Firms and households switch from commercial bank deposits to stablecoins Commercial banks lend to firms and households Reserve flow A Stablecoin issuers hold reserves as deposits at commercial banks Reserve flow B Stablecoin issuers buy securities Security sellers (firms, households, banks) recycle Firms and households proceeds back into the banking system as deposits Cash-equivalent securities In Figure 3, we see how stablecoin inflows and reserve flows are interconnected. In the diagram, firms and households substitute away from deposits (Inflow A) and banknotes (Inflow B) into stablecoins. The stablecoin issuer deposits some of these funds back into the commercial banking system to hold reserves as commercial bank deposits (Reserve flowA),andalsousesthefundstopurchasesecuritiesforreserves(ReserveflowB).These security purchases also recycle funds back into the banking system, because the sellers of thesecuritiesultimatelytaketheproceedsofthesecuritysalesanddepositthembackinto thebankingsystem. AsillustratedinFigure3,theseflowsimpactthecentralbank,which 31In Figure 3, we separate stablecoin issuers from commercial banks, but it is plausible that commercial banksdirectlyissuestablecoins. 15

maintains cash and central bank reserves as liabilities, as well as firms and households, whichreceiveloansfromcommercialbanks. Whilethisdiagramdoesnotcapturethefull set of flows between these entities, it is emblematic of how the widespread adoption of stablecoinscouldreshufflecomplexfinancialrelationshipswithinthebankingsystem. Scenarioanalysis Narrowbankframework Asdiscussedearlier,thenarrowbankframeworkposesthelargestrisktocreditprovision, dependingonthesourceofinflow. Inournarrowbankscenarios, depictedinTable5, we find that physical cash inflows into narrow bank stablecoins would have a neutral effect oncreditprovision,whilecommercialbankdepositswoulddisruptcreditprovision. In Panel A, the cash inflows scenario, we see stablecoins replacing cash on the household and firm balance sheet. This influx of cash results in a pass-through increase in the commercialbankbalancesheetandthecommercialbank’sreserves. Thecentralbank’sbalance sheetisreshuffled,withreserveliabilitiesreplacingcashliabilities. Theneteffectisthatthe commercial bank balance sheet expands, but there is no change in credit provision. This scenario assumes that banks are not balance-sheet size constrained. That is, narrow bank depositsandassociatedreserveholdingsareexemptfromleverageratiocalculation. This type of leverage ratio exemption for central bank reserve holdings has been previously appliedbyregulatorsindifferentjurisdictions.32 Panel B presents the narrow bank scenario with deposits migrating into stablecoins. As stablecoin deposits are fully reserved on commercial banks’ balance sheets, banks must reduceassetholdingstoaccommodatethedeclineinnon-stablecoindepositfunding. The central bank balance sheet then expands to accommodate the increased demand for reservebalanceswithoutanoffsettingdeclineincashliabilities. Inthisscenario,weassume the central bank will accommodate the increased demand for reserves by purchasing securities. ThisassumptionofcentralbankaccommodationisinformedbypreviousFederal Reserve proposed rulings on narrow banks as discussed above relating to Regulation D: ReserveRequirementsofDepositoryInstitutions(2019). However,shouldthecentralbankfix the size of its balance sheet, we present two alternative scenarios in Table A1 in the appendix. In the first alternative scenario, the commercial banks significantly contract their balance sheets to compensate for the lack of deposit funding. In the second scenario, the commercialbankscompensateforthelostdepositfundingbyissuingdebtsecurities. The resultisanevenlargerreductioninbank-ledcreditcreation.33 32Forinstance,theFederalReserveandtheEuropeanCentralBankbothexemptedcentralbankreservesin thecalculationofsupplementaryleverageratioin2020duetotheinfluxofdepositsandexpansioninbank balancesheets. 33As illustrations, these scenarios might not capture the full spectrum of scenarios and secondary effects stemmingfromstablecoingrowth.Forinstance,anexpansionofthecentralbankbalancesheetrequiresasset purchasesthatmightspursecurityissuancebyhouseholdsorcommercialbanks. Thiscouldleadtoalower costoffinancingandcreditexpansion. Thecentralbankcouldalsosourcethesecuritypurchasesfromthe assetholdingsofcommercialbanksleavingthebanks’loanportfoliosandhouseholddebtunchanged. 16

Table5: Changesfrombaselinefornarrowbankstablecoins PanelA:Physicalcashinflows Centralbank Commercialbanks Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Reserves +10 Reserves +10 Stablecoin +10 Stablecoins +10 deposits Physical -10 Physical -10 cash cash Net 0 +10 +10 0 PanelB:Commercialbankdepositinflows Centralbank Commercialbanks Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Securities +9 Reserves +9 Reserves +9 Stablecoin +10 Stablecoins +10 Debt -9 deposits (loans)* Loans -9 Retail -10 Deposits -10 deposits Securities† -9 Net +9 +9 0 0 -9 -9 * The households and firms sector could possibly experience credit contraction as commercial banks’ loan booksarereduced. †Householdswouldhavetosellassetstomeetrepaymentofloanobligations.Theseassetsalesareillustrated assecuritysales, matchingcentralbanksecuritypurchases. Thoughinreality, householdassetscouldtake otherforms(e.g. realestate)thataresecuritizedasmortgages. Adeclineinthehouseholdsector’ssecurities holdingsissimilartoareductioninrealassetsunderthisexample. Wedonotvisualizethescenarioinwhichnarrowbankstablecoinsseelargeinflowsfrom security holdings. In this scenario, the impact on credit provision would likely be neutral. Under the same assumption as above in which the central bank accommodates the increaseddemandforreservesbypurchasingsecurities(fromhouseholds),thenetimpact on credit provision should be minimal. Instead of holding securities directly, a migration tostablecoinswouldseehouseholdsowningstablecoinsbackedbycentralbankreserves, which are in turn backed by securities. This scenario also makes the assumption that the addednarrowbankreservesareexemptedfromleverageratiosasdiscussedearlier. Two-tieredintermediationframework For the two-tiered intermediation framework, presented in Table 6, we find that large inflowsintostablecoinswouldhaveaneutraltopositiveimpactoncreditprovision. PanelA showsthecaseinwhichcashisexchangedforstablecoins. Ascommercialbanksengagein fractional-reserve banking with stablecoin deposits, their balance sheet expands with expansionsincreditandsecurityholdingsaccountingformostoftheexpansion. Thecentral bank shrinks its balance sheet on the net, as reserves increase slightly while cash liabilities decrease significantly. Households accumulate more assets, funded by the expansion 17

in bank loans. The effect on credit provision is positive. Panel B shows the two-tiered intermediation scenario with deposit substitution. The overall balance sheets and asset holdingsofcommercialbanksandthecentralbankareunchanged. Theonlyshiftisinthe composition of commercial bank liabilities, as regular deposits are shifted into stablecoin deposits. Asnotedearlier,thisscenarioassumesthetreatmentofstablecoindepositsisthe sameasnon-stablecoindepositsintermsoftherequiredreserveratio, liquiditycoverage, andotherregulatoryandself-imposedrisklimits. Table6: Changesfrombaselinefortwo-tieredintermediationstablecoins PanelA:Physicalcashinflows Centralbank Commercialbanks Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Securities -9 Reserves +1 Reserves +1 Stablecoin +10 Stablecoins +10 Debt +9 deposits (loans) Physical -10 Loans +9 Physical -10 cash cash Securities +9 &other* Net -9 -9 +10 +10 +9 +9 PanelB:Commercialbankdepositinflows Centralbank Commercialbanks Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Stablecoin +10 Stablecoins +10 deposits Retail -10 Deposits -10 deposits Net 0 0 *Households/firmsusetheaddedbankloanfundingtopurchasemoreassets,possiblyintheformofsecuritiesfromthecentralbank. Alternatively,households/firmscanincreaserealassetholdings(e.g. housesand factories). Securityholdingsframework The impact of widespread adoption of security-backed stablecoins, presented in Table 7, is the most difficult to anticipate. Many scenarios are possible. In Panel A, we present a scenarioinwhichsecurity-backedstablecoinsseeinflowsfromcommercialbankdeposits. We assume the stablecoin issuer is sourcing securities from the commercial banks, not the households and firms sector. In this scenario, as households exchange deposits for stablecoins,commercialbanksmakeupthelostdepositfundingbyconductingtheirown security issuance.34 Additionally, commercial banks can reduce their security portfolio to accommodate the loss in deposit funding. The size of banks’ loan portfolios can possibly 34Issuing of debt securities by commercial banks might affect the banks’ regulatory metrics such as Net StableFundingRatio.Weassumeherethattheseeffectsaresecondorder. 18

remainunchangedifbanksadjusttheassetsideofthebalancesheetprimarilybychanging securityholdings. Inthisscenario,thecentralbankbalancesheetalsoshrinksslightlydue tolossinbankingreserves. Panel B of Table 7 presents a scenario in which households exchange holdings of cashequivalentsecuritiesforstablecoins. Thiswouldleadtoeffectivetokenizationofcash-like securitieswithoutadirectimpactoncreditprovisionbythebankingsystem. Wealsoconsider an alternate scenario (not shown) in which security-backed stablecoins experience depositinflowsfromhouseholdsandfirmssectorthatsimultaneouslysellssecurityholdingstothecommercialbanks. Thesecurityselleristhehouseholdsandfirmssectorinstead ofcommercialbanksasdepictedinTable7PanelA.Thenetimpactoncreditprovisionis neutral, as the commercial bank deposit balances held by the households and firms that purchasestablecoinsareultimatelyrecycledbackintothebankingsystembytransferring themtootherhouseholdsandfirmsthatsellsecuritiestothestablecoinissuer. ThisreshufflingofsecurityholdingsisillustratedinFigure3byInflowAandReserveflowB.Theend resultisabalancesheetshiftthatisthesameasTable7PanelB. Finally, wedonotdepictthescenariowheresecurity-backedstablecoinsseeinflowsfrom physical cash. However, this could have a neutral or positive impact on credit creation. If the stablecoin issuers use the banknotes to purchase existing securities, and those banknotesareultimatelynotdepositedintothebankingsystem,thiswouldhavenoimpacton creditprovisionasitwouldconstituteadirectexchangeofbanknotesforsecurities. However,ifthebanknotesfrompurchasesofexistingsecuritiesaredepositedintothebanking system, or if the banknotes are used to fund the issuance of new securities, this could increasecreditprovisionbyincreasingloansandsecuritypurchasesbycommercialbanksor byloweringtheequilibriumcostofissuingsecurities. Altogether,thelikelyimpactwould beamodestincreaseincreditprovision. V. Conclusion Stablecoins have grown tremendously over the past year as digital assets gain broader adoptionandtheusecasesofprogrammabledigitalcurrenciesareclarified. Thisrapidascensionhasraisedconcernsthattheremightbenegativeimpactsonbankingactivitiesand thetraditionalfinancialsystem. Inthisnote,wediscussthecurrentusecasesandpotential growthofstablecoins,analyzehistoricalepisodesofpeginstability,andillustratedifferent scenariosofstablecoins’impactonthebankingsystem. Asnotedintheintroduction,this paperdoesnotconsiderallthepotentialimpactsofstablecoinsonfinancialstability,monetary policy, consumer protection, and other important unexplored issues. We focus on thebalancesheeteffectsandcreditintermediationunderasetofplausibleassumptions. We examine reserve-backed stablecoins and find the impact of stablecoins’ adoption on traditional banking and credit provision can vary depending on the source of inflow and thecompositionofstablecoinreserves. Amongthevariousscenarios,atwo-tieredbanking 19

system can support both stablecoin issuance and maintain traditional forms of credit creation. Incontrast,anarrow-bankstablecoinframeworkcanbringthemoststabilitybutat thepotentialcostofcreditdisintermediation. Finally, dollar-peggedstablecoinscanserve as a safe haven relative to other crypto-assets during times of market distress if they are perceivedtobesufficientlycollateralized. 20

Table7: Changesfrombaselineforsecurity-backedstablecoins PanelA:Depositsubstitution Centralbank Commercialbanks Stablecoinissuers Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Assets Liabilities Securities -1 Reserves -1 Reserves -1 Security +5 Securities +10 Stablecoins +10 Stablecoins +10 issuance Securities -4 Retail -10 Deposits -10 deposits Net -1 -1 -5 -5 +10 +10 0 PanelB:Householdsecuritysubstitution Centralbank Commercialbanks Stablecoinissuers Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Assets Liabilities Securities +10 Stablecoins +10 Stablecoins +10 Securities -10 Net +10 +10 0 21

References 4Q20financialresults: Earningscalltranscript. (2021). JPMorganChase&Co. Baba, N., McCauley, R. N., & Ramaswamy, S. (2009). US dollar money market funds and non-USbanks. BISQuarterlyReview,March. Centralbankdigitalcurrencies. (2018). BankforInternationalSettlements. Copeland, A. M., Duffie, D., & Yang, Y. D. (2021). Reserves were not so ample after all. FRBofNewYorkStaffReport(974). Correa,R.,Du,W.,&Liao,G.Y. (2020). USbanksandgloballiquidity(Tech.Rep.). National BureauofEconomicResearch. Defi beyond the hype. (2021). Wharton Blockchain and Digital Asset Project and the World EconomicForum. Dixon,C. (2018). Introducinga16zcrypto. AndreessenHorowitz. Dixon,C.,&Haun,K. (2020). CryptofundII. AndreessenHorowitz. Eren,E.,Schrimpf,A.,Sushko,V.,etal. (2020). USdollarfundingmarketsduringthecovid-19 crisis-themoneymarketfundturmoil. BankforInternationalSettlements. Gorton,G.B.,&Zhang,J. (2021). Tamingwildcatstablecoins. SSRN. Griffin,J.M.,&Shams,A. (2020). Isbitcoinreallyuntethered? TheJournalofFinance. JPMorgan. (2019). Corporate investment bank. Retrieved from https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/events/2019/2019-investor-day/2019 ib nvestor ay a56d0e8.pdf c i d b JPMorgan. (2020). Blockchain, digital currency and cryptocurrency: Moving into the mainstream? Lee,A.,etal. (2021). Whatisprogrammablemoney? FEDSNotes. Lyons,R.K.,&Viswanath-Natraj,G.(2020).Whatkeepsstablecoinsstable? NBERWorking Papers. Malloy, M.,&Lowe, D. (2021). Globalstablecoins: MonetarypolicyimplementationconsiderationsfromtheUSperspective. FEDSWorkingPapers. Mills, D. C., Wang, K., Malone, B., Ravi, A., Marquardt, J., Badev, A. I., ... Kargenian, V. (2016). Distributed ledger technology in payments, clearing, and settlement. FEDS WorkingPapers. Newformsofdigitalmoney. (2021). BankofEngland. OversightoftheTreasuryDepartment’sandFederalReserve’sPandemicResponse. (2021). House FinancialServicesCommittee,116thCongress. Prentice, C. (2021). Crypto firms Tether, Bitfinex to pay $42.5 mln to settle U.S. CFTC charges. Reuters. Quarles,R. (2021). Parachutepantsandcentralbankmoney. Speechatthe113thAnnualUtah BankersAssociationConvention,SunValley,Idaho. Ratha, D., Kim, E., Plaza, S., & Seshan, G. (2021). Migration and development brief 34. KNOMAD-WorldBank. RegulationD:ReserveRequirementsofDepositoryInstitutions. (2019). BoardofGovernorsof theFederalReserveSystem. 22

Schoenberg, T., Robinson, M.,&Faux, Z. (2021). Tetherexecutivessaidtofacecriminalprobe intobankfraud. Bloomberg. Szabo,N. (1994). Smartcontracts. Waller, C. J. (2021). CBDC: A solution in search of a problem? Speech at the American EnterpriseInstitute. Wong,P.,&Maniff,J. (2020). Comparingmeansofpayment: Whatroleforacentralbank digitalcurrency? FEDSNotes. 23

TableA1: Alternatenarrowbankscenarios PanelA:Commercialbankbalancesheetshrinks Centralbank Commercialbanks Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Loans -45 Stablecoin +5 Stablecoins +5 Debt -45 deposits (loans) Retail -50 Deposits -50 deposits Net -45 -45 -45 -45 PanelB:Commercialbanksissuedebtsecurities Centralbank Commercialbanks Households/Firms Assets Liabilities Assets Liabilities Assets Liabilities Stablecoin +5 Stablecoins +5 deposits Retail -50 Deposits -50 deposits Debt +45 Securities +45 securities Net 0 0 24