Public Comments on the Proposed Guidelines for Evaluating Requests for Accounts and Services

I’m grateful to the Federal Reserve Board for allowing me to comment on its proposed guidelines to evaluate requests for accounts and services at Federal Reserve Banks. Among other things, the Board asks prospective commentators whether its proposed guidelines “are sufficiently clear and appropriate to achieve their intended purpose,” whether other criteria “may be relevant to evaluate accounts and services requests,” whether the proposed guidelines will “support responsible financial innovation,” and whether it “should consider other steps or actions to facilitate the review of requests for accounts and services in a consistent and equitable manner.”

Before taking up these questions, I wish to commend the Federal Reserve Board for recognizing the need for “a more transparent and consistent approach” to evaluating requests for master accounts from holders of special‐​purpose depository institution charters, and for proposing such an approach.

I agree wholeheartedly with the Board’s aim of having “the Reserve Banks apply a consistent set of guidelines when reviewing such access requests…to facilitate equitable treatment across institutions,” and I believe that the proposed guidelines represent an important step toward this end. However, I also believe that the guidelines can be made more conducive than they are at present to encouraging beneficial financial innovations, and that this can be done without increasing the risk of financial instability and without complicating the conduct of monetary policy.

As Cleveland Federal Reserve Bank President Loretta Mester observed last November, the digitalization of finance “holds the promise of increasing the efficiency, productivity, and inclusiveness of the financial sector, thereby increasing the economic welfare of households and businesses.”¹ Digital financial service providers other than traditional banks, or fintechs, have often led the way in such digitalization, helping thereby to achieve substantial worldwide gains in financial inclusion.² Fintechs are also likely to be the main source of future digital‐​finance innovations, and especially so if they can compete effectively, on truly a level playing field, with ordinary banks.

The Board has elsewhere made clear its own commitment “to supporting responsible innovation, both by the firms we regulate directly, and in the financial market broadly.”³ Many of the innovative services fintechs offer, or may offer in the future, depend on their having access to the Federal Reserve’s wholesale settlement services. At present, many fintechs secure such access indirectly, using traditional banks as their agents. This arrangement obliges them to rely on rival payment‐​service providers in order to be able to compete with them, while also placing them at a cost disadvantage relative to those rivals. Furthermore, because it means having an additional legal entity involved in the settlement process, it leads to longer settlement delays, tying‐​up capital in unsettled transactions.

To the extent that fintechs are able to gain the same direct access to the Fed’s wholesale purchases that ordinary banks enjoy, they can operate more efficiently, and compete more effectively with those banks in providing alternative retail products and services. Many foreign central banks have already taken steps to grant fintechs such access. Although the risks involved in taking this step justify proceeding cautiously, as the Bank of England observed in announcing its own decision to give “non‐​bank payment service providers” direct access to its wholesale settlement services, doing so can ultimately enhance financial stability by

• creating more diverse payment arrangements with fewer single points of failure;
• identifying and developing new risk‐​reducing technologies; and
• expanding the range of transactions that can take place electronically and be settled in central bank money.⁴

At present, fintechs that are not and do not wish to become full‐​service banks are only able to apply for Reserve Bank master accounts after securing special purpose or “novel” depository‐​institution (“bank”) charters.⁵ Such charters are offered by the OCC as well as by some state banking authorities. The Board does not propose to change this eligibility condition, and it isn’t clear whether the Federal Reserve Act allows for any change. The limited availability of special‐​purpose charters, and the far from lenient conditions fintechs must meet to obtain them, already make it very difficult for most fintechs to qualify for Fed master accounts. For that reason, if the Board wishes to encourage financial innovation, it is essential that the guidelines for granting such accounts to eligible fintechs not be unnecessarily burdensome.

In the remainder of this letter, I will argue that some aspects of the Board’s proposed guidelines make them needlessly burdensome to many actual and potential special‐​purpose master account applicants. I will also suggest ways in which I believe this unnecessary burden might safely be lightened.

Under the current master account review procedures, the time Reserve Banks may take to reach their decisions is left entirely to their discretion. Consequently, an applicant cannot rule out being kept waiting, not just for many months, but indefinitely, for a decision. Because the cost of enduring a long delay might itself discourage potential applicants, leaving the length of the review process entirely to Reserve Bank’s discretion can prove contrary to the Board’s goal of encouraging financial innovation. Unfortunately, the proposed guidelines do not address this undesirable feature of the status quo. Including among the guidelines a prospective timeline for Reserve Banks’ decision process, however speculative, would be a considerable improvement.⁶

More importantly, the proposed guidelines consist of many necessary conditions for having master account applications approved, with no intimation of sufficient ones. “These guidelines,” the Board’s proposal states, “broadly outline considerations for evaluating access requests but are not intended to provide assurance that any specific institution will be granted an account and services.” Although it is understandable that the Board may not wish to offer any absolute guarantee of success, it might nonetheless indicate particular conditions which, if met, make it especially likely that a request will be granted. A complementary revision would require Reserve Banks to justify their denial of applicants meeting these conditions, making such denials the exception rather than the rule.

With rare exceptions, the Fed routinely grants master accounts to banks and credit unions possessing ordinary bank charters and insured by the FDIC. The rationale for having firms holding special‐​purpose charters meet special requirements to obtain such accounts consists of the fact that those special‐​purpose charters typically subject their holders to less‐​stringent regulation than holders of ordinary bank charters, in part by making it unnecessary for them to insure their deposits. That exemption allows some special‐​purpose banks to operate with capital levels below the FDIC’s requirements, though not necessarily below those of traditional or full‐​service state banks.⁷

According to spokesmen for traditional banks, were the Fed to award master accounts to fintechs and other firms with special‐​purpose charters without compelling them to be insured or otherwise subjecting them to bank‐​like regulatory requirements, it would be giving them an unfair advantage. In particular, the bankers, besides opposing special charters themselves, would have the Fed insist that holders of those charters maintain approximately the same capital cushions insured banks are required to have. In short, banking industry representatives assert that unless fintechs and traditional banks are subject to similar regulations, granting the former master accounts is inconsistent with having them compete on a “level playing field.”

But this argument is valid only assuming that all potential master account applicants, which is to say all firms that manage to acquire any sort of “bank” charter, expose the Fed and the financial system as a whole to the same risks. This would be so if they all engaged, or intended to engage, in similar activities. But that is seldom the case for fintechs. Most fintechs seek “bank” charters, not so they can replicate ordinary banks’ bundle of services, but because they wish to supply a subset of such services, and having Fed master accounts will assist them in doing so.

As Cornell Law Professor Dan Awrey argues in a recent working paper, by “unbundling” the various services conventional banks offer, more specialized fintechs can perform some of those services without taking the same risks conventional banks take.⁸ For this reason, in seeking to limit “the risks that may arise when an institution gains access to accounts and services,” including “risks to the Reserve Banks, to the payment system, to the financial system, and to the effective implementation of monetary policy,” the Board should ask, not whether the holder of a special‐​purpose charter has as much capital as ordinary banks must have, or whether it is otherwise held to the same regulatory standards, such as having to be insured by the FDIC, but whether the standards it does meet are adequate to sufficiently limit the risks associated with its particular business plans. In other words, as Loretta Mester remarked last fall, “Existing regulatory and supervisory structures…need to adapt to keep up with the new ways that financial services are being delivered and the new players delivering them.” The Reserve Banks’ master account approval guidelines clearly qualify as “existing regulatory structures.”

BPI (the Bank Policy Institute), which represents the nation’s largest commercial banks, has been the most vocal opponent of both special‐​purpose “bank” charters and the granting of Fed master accounts to firms that secure such charters. According to it, and to most authorities, “Banking at its core is the business of funding illiquid assets—loans—with highly liquid liabilities—deposits.”⁹ In other words, traditional banks borrow short and lend long, and may find themselves insolvent or illiquid because of this. That is why traditional banks are subject to very strict capital and liquidity requirements. It is also why full‐​service national banks must be insured by the FDIC.

Most fintechs, in contrast, do not combine deposit taking and lending; and many don’t engage in any sort of maturity transformation. Some, including peer‐​to‐​peer and other “marketplace” lenders, extend credit, but do not fund their lending with, or even take, deposits.¹⁰ Strictly “custodial” fintechs also don’t take deposits.¹¹ Others, including money transmitters, payment processors and cryptocurrency exchanges, accept dollar deposits from their clients to be remitted to others, but do not invest them in risky assets for the brief intervals in which they possess them. It is not generally the case, therefore, as BPI elsewhere claims, that in acquiring special‐​purpose charters for the sake of qualifying for Fed master accounts “fintechs are trying to act like [traditional] banks while avoiding the supervisory and regulatory framework that applies to actual banks.”¹²

In light of the above considerations, I hope the Board will consider a suggestion I’ve sketched elsewhere for improving its master account review guidelines.¹³ The proposal is very simple: let the guidelines include a “fast approval lane” for “narrow” special‐​purpose “banks” that don’t engage in any maturity transformation, meaning those that either don’t take U.S. dollar deposits at all or agree to fully back any dollar deposits they take with 100‐​percent master account balances. Allowing fast‐​lane treatment for such “narrow” master account seekers seems fully consistent with Reserve Banks right, as indicated in the proposed guidelines, to

The proposed “fast lane” option resembles one of the Bank of England’s conditions for granting settlement accounts to non‐​bank Payment Service Providers: although hundreds of fintechs are potentially eligible for Bank of England settlement accounts, the Bank only grants such accounts to those that “do not undertake maturity transformation activities.”¹⁴ However, my proposal would not altogether rule‐​out master accounts for maturity‐​transforming applicants with special purpose charters. It only means that they would have to undergo a more involved and time‐​consuming evaluation process.¹⁵

I am also not proposing that “narrow” special‐​purpose banks should not be subject to any further requirements to qualify for master accounts: the Board need not be satisfied with the standards set either by the OCC or by any state banking authority. However, these additional requirements should also reflect the significantly reduced risks such narrow institutions pose, and should be specified in such a way as will allow prospective narrow‐​bank applicants that abide by them to be reasonably certain that they stand a good chance of having their applications approved in a timely manner.

Two examples of the sort of fintech that would benefit from the proposed “fast lane” treatment are Kraken Financial and Avanti Bank. Kraken Financial, a division of Kraken, a cryptocurrency and bitcoin exchange, received one of Wyoming’s Special Purpose Depository Institution (SPDI) charters last September; Avanti Bank, which seeks to offer cryptocurrency custodial services, secured its own Wyoming SPDI charter soon after. Both subsequently applied to the Federal Reserve Bank of Kansas for master accounts. Their applications are still under consideration.

Unlike ordinary banks, Wyoming’s SPDIs are not required to have deposit insurance.¹⁶ This exemption means that, although they are subject to minimum capital requirements,¹⁷ can operate with less capital than the FDIC requires, while avoiding the (not inconsiderable) risk of being either turned down or kept in limbo by the FDIC even if they appear to meet its regulatory standards.¹⁸

But it doesn’t follow that either SPDI will pose a greater risk to the financial system than an ordinary bank, because Wyoming SPDIs are also “generally prohibited from making loans with customer deposits [and] must at all times maintain unencumbered level 1 high‐​quality liquid assets valued at 100% or more of their depository liabilities.” In response to criticism from BPI, pointing out that even these level‑1 assets may expose SPDIs to interest‐​rate risk, Wyoming’s Banking Division has since proposed that SPDIs be allowed to invest in liquid government and agency level‑1 securities only.¹⁹

As a practical matter, assuming their master account balances bear interest at the same rate as ordinary bank reserves, it should not matter much to Kraken or Avanti whether they can hold some “cash” in the form of level‑1 government and agency securities or must maintain 100‐​percent master account backing for their dollar deposits. The difference in returns is slight; and these firms’ clients are generally looking, not for any return on their briefly‐​held deposits, but for convenient cryptocurrency exchange and custodial services. Consequently, although Wyoming’s law may not require it, it is doubtful that either firm would mind fully backing its deposits with master account balances if doing so would improve its chances of receiving a master account, while simplifying and hastening the application assessment process.

The Board’s proposed guidelines include requirements that appear unnecessary for “narrow” applicants, yet would not automatically be met by them. As part of the proposed “fast lane” approval process for “narrow” master account applicants, the Board should consider relaxing these.

As the now substantial literature on narrow banking makes clear, in its 1oo‐​percent reserve form, with no maturity transformation or interest rate risk, narrow banking does away with any need for either deposit insurance or substantial capital requirements.²⁰ Narrow banking also makes other precautions meant to guard against a master account holder’s insolvency, several of which are included in the proposed guidelines, redundant. For example, the guidelines call for Reserve Banks to “clearly identify all risks that may arise related to the institution’s business” (my emphasis). They also require that applying firms have “adequate capital to continue as a going concern and to meet its current and projected operating expenses under a range of scenarios.” Such requirements make sense when an account applicant’s failure would pose “risks to the Reserve Banks, to the payment system, to the financial system, and to the effective implementation of monetary policy.” But they may not be needed for at least some “narrow” fintechs, which might fail without posing any risk that their dollar obligations will not be fully met. A cryptocurrency custodian, for example, might have its custodial accounts hacked, exposing its custodial clients to potential losses. Yet so long as its U.S. dollar deposits are fully backed by Fed account balances, its failure need not put any strain on the U.S. dollar‐​based payments and financial system.

One relatively straightforward way in which the Board might haveavoided subjecting master account applicants to unnecessary or redundant requirements would be to treat possession of any special‐​purpose charter as itself satisfying certain requirements, depending on which sort of charter. The Board and Reserve Banks could, in other words, subject bank charter types themselves to a sort of “approval” process, and then stipulate additional requirements for applicants equipped with each charter type, with relaxed requirements in each case for “narrow” applicants. Although this procedure would have the Fed ride “piggyback” on chartering agencies’ regulatory requirements, it would not have it defer to those agencies. Nor is it meant to preclude having the Reserve Banks exercise their own due diligence by separately determining whether applicants are conforming to their charters’ requirements.

This “piggyback”approach could greatly simplify the master‐​account application‐​approval process, for by doing away with all redundant requirements it would allow the Reserve Banks to devote their assessment efforts to the Board’s additional requirements only. Because the number of charter types is relatively small, it could ultimately save both the Fed itself and banks applying for accounts much time and trouble. Provided those extra requirements are such as will make the full regulatory burden any applying “bank” must satisfy (which may be less but not more than what it must actually bear) roughly the same regardless of what charter it holds, this procedure would also ensure more equitable treatment of similar special‐​purpose banks possessing different types of charters.

Paradoxically, the very safety of Fed master account balances that would prevent the failure of a narrow fintech from posing undue payment or financial system risk may make the granting of such accounts to fintechs dangerous for a very different reason: should their dollar deposits appear more attractive than deposits at commercial banks, fintech depositories might attract such a large flow of funds previously directed to commercial banks as would subject the latter to an disintermediation crisis. This could happen either because fintech balances pay more interest than ordinary bank deposits, or because holders of ordinary bank deposits, and of less than fully insured deposits especially, worry that their banks might fail.

Several passages in the proposed guidelines address this other danger. “The extent to which the institution’s use of a Reserve Bank account and services,” one says, “might restrict funds from being available to support the liquidity needs of other institutions.” Another says that Reserve Banks should consider the extent to which, especially during times of financial or economic stress, access to an account and services by an institution itself (or a group of like institutions) could affect deposit balances across U.S. financial institutions more broadly and whether any resulting movements in deposit balances could have a deleterious effect on U.S. financial stability.” Still another adds that the risk of a disintermediation crisis will be especially great “if a non‐​commercial bank master account holder is not subject to capital requirements similar to a federally‐​insured institution.”

Having warned of this disintermediation risk myself (though with reference to proposals for personal Fed accounts or Fed “digital currency”), I understand that the Fed’s master account guidelines must take it into account.²¹ However, it is a danger few fintechs pose. Many receive dollars only for the sake of completing their clients’ payments requests, rather than to be kept on deposit. The U.S. dollar deposits such fintechs have on hand at any time are minimal deposits needed to perform their payment‐​related functions. To the extent that fintechs receive and hold deposits only for such purposes, they don’t pose any substantial disintermediation risk.

“Pass‐​Through Investment Entities” (“PTIEs”) like The Narrow Bank (TNB) are quite distinct from fintechs in this regard. Their sole raison d’être is to allow investors to take advantage of the fact that the interest rate paid on Fed master accounts sometimes exceeds the yield on other safe, short‐​term securities. PTIEs exist for the sole purpose of giving their clients (in TNB’s case, mutual funds) direct access to that interest return, by accepting their deposits and investing them in Fed master account balances. Because they are a special case, PTIEs may well merit special scrutiny, or extra conditions, for master account approval. But it would be wrong to impose the same conditions on all fintech master account applicants.

Because it may not always be clear whether a master account applicant use its master account to offer pass‐​through investment services, the Reserve Bank’s guidelines could stipulate that, to qualify for fast‐​lane treatment, master accounts applicants must agree to use their dollar assets for payment purposes only, and not as substitutes for commercial bank deposits. The Bank of England includes such a stipulation among its own conditions for granting a settlement account to a non‐​bank “Payment Service Provider” or PSP:

Besides its other advantages, the “fast lane” treatment I propose for “narrow” fintech master account applicants would be a natural complement to the OCC’s actual and proposed special‐​purpose charters for “narrow” national banks.

Although several fintechs are operating with or have secured special purpose charters from state banking authorities, for many, such state charters are far from ideal, because their business models call for them to do business across the country, while state charters make this possible only if they secure separate money‐​transmission licenses from each of the states they wish to do business in. In contrast, a national charter granted by the OCC preempts state money transmission laws, automatically allowing its holder to serve clients in all 50 states. One fintech offering custodial services has secured an OCC national trust bank charter, while two others have had their applications for such charters conditionally approved, but these cases are controversial.²³ Otherwise, until quite recently, most other fintechs could only secure national charters by becoming full‐​service banks, and meeting all the regulatory burdens this entailed.

It was owing to its desire to have limited‐​purpose fintechs take advantage of its national charters that the OCC began accepting applications for what it calls “special purpose narrow bank” (SPNB) charters in 2018.²⁴ According to its guidelines for evaluating these applications, its SPNB charters are only available to fintechs that “have nontraditional or limited business models, do not take deposits, and rely on funding sources different from those relied on by insured banks.”²⁵ While fintechs offering payment services can apply for the OCC’s SPNB charter, its absolute prohibition of deposit taking makes it most suitable for fintech lenders. Last June then‐​Comptroller Brian Brooks announced that the OCC planned to create yet another special‐​purpose charter designed for fintechs offering payments services, which receive and hold clients’ funds in anticipation of transferring them to others.

These OCC efforts have met with considerable resistance. The SPNB charter has for some time now been mired in litigation launched by New York’s Department of Financial Services,²⁶ while the payments charter, which state authorities have also criticized,²⁷ remains a proposal only, which may or may not be pursued by the current Comptroller. The OCC may nevertheless prevail, in which case it will be highly desirable for the Fed’s master account assessment guidelines to accommodate fintechs that secure OCC special charters by acknowledging their “narrow” status, and automatically shunting their holders into its “fast lane” approval process.

By suggesting ways to improve upon the Board’s proposed guidelines, I hope I haven’t given the impression that I consider any of them “easy” to implement. On the contrary: I understand that they all involve details and difficulties that will require some effort to work‐​out. Still, I hope that the Board of Governors won’t be deterred by the challenges involved from considering my advice, and I thank it once again for inviting me to offer it.