Who Holds the Promise—and Who Acts on It?
For a deposit insurance system, institutional design is a question of decision rights: who can reimburse depositors, who can fund a resolution, who holds the information, and who is accountable when those choices conflict. This chapter compares the main mandates and legal forms so readers can judge them by the capabilities they create—not by whether an agency is called independent, integrated, public, or private.
Four mandates, four distributions of responsibility
A mandate is the set of responsibilities and powers assigned to the deposit insurance function. It determines what the insurer can do when a member bank fails and how its work connects to the resolution authority, supervisor, central bank, and fiscal authority. The labels are useful because they reveal different distributions of responsibility, but they are not four universal organizational templates. Actual laws can combine functions or assign them in different ways.
The narrowest arrangement is a paybox. The deposit insurer reimburses insured deposits after a bank is closed or another legal trigger occurs. Its task is focused: determine which claims qualify, obtain the money, and pay depositors. A narrow mandate can be easier to explain and can reduce overlap with supervision or resolution. But it also leaves the insurer outside the decision to preserve a bank’s deposit-taking business. If a transfer to another bank would protect access more smoothly than liquidation, a paybox may not have the legal authority to help finance that option.
A paybox-plus arrangement gives the insurer an additional role while keeping it legally separate from the resolution authority. The deposit insurance fund may contribute to a resolution transaction, such as transferring insured deposits and assets to an acquiring bank or bridge bank, provided the legal safeguards are met. The insurer remains responsible for protecting depositors and its fund; it does not become a general rescuer of shareholders, managers, or every asset of a failing bank. The benefit is an additional way to preserve continuity. The cost is that two institutions must coordinate closely when the resolution authority is deciding what to do and the deposit insurer is being asked to use its resources.
A loss-minimizer arrangement places responsibility for both resolution and the deposit insurance fund within the resolution authority. This can reduce the distance between the institution choosing a resolution strategy and the institution managing the money available for depositor protection. It may also concentrate expertise and make an emergency decision chain shorter. The same concentration can make conflicts and accountability more important: the authority must decide how to preserve a bank’s critical functions while also protecting the fund and carrying out the public objectives of deposit insurance.
A risk-minimizer arrangement places the deposit insurance function within a broader agency that may include supervision, early intervention, resolution, and deposit insurance. Its potential advantage is a wide view of risk across the bank’s life cycle. An authority that sees supervisory information, intervention decisions, resolution planning, and insurance exposure may identify connections that separate agencies would have to negotiate. Its challenge is that a broader agency must manage competing objectives and preserve the operational independence of the deposit insurance function inside a larger institution.
These mandates should be compared on the same questions. What can the insurer do before and during failure? Who decides whether a resolution tool is used? Who controls the fund? Which institution has the expertise to value assets and execute a transfer? How does confidential supervisory information reach the people who need it? What happens when the institution selling a failed bank, the supervisor of the acquiring bank, the central bank with a claim on the failed institution, and the deposit insurer have different priorities?
The international standard does not choose a universal winner among these arrangements. IADI recognizes different mandates and institutional settings, while the IMF staff paper gives a more specific recommendation: a narrow paybox should consider moving toward paybox-plus where country circumstances allow it, because funding a resolution can preserve depositor access and use the fund more efficiently. That difference is important. The Core Principles are outcome-focused and institutionally adaptable; the IMF view offers policy advice about implementation trade-offs, not a binding rule that every country must adopt.
The diagram shows relationships, not a requirement that every box be a separate agency. In a loss-minimizer or risk-minimizer system, several boxes may sit inside one institution. Even then, the functions remain conceptually distinct. Combining them does not remove the need to define who is acting as supervisor, resolution authority, deposit insurer, or liquidity provider in a particular decision.
Integration can create capacity, and separation can create clarity
The next choice is where the deposit insurance function should live. A stand-alone deposit insurance agency can give senior management one clear focus, a distinct budget, and a visible line of accountability. It may be easier for the public to understand who is responsible for reimbursement. It can also make it clearer that the deposit insurance fund is separate from the balance sheets and policy objectives of the central bank, supervisor, or resolution authority.
But separation has a price. A new agency must recruit staff, build information technology, establish legal and communications capabilities, manage the fund, and develop expertise in deposit records and bank failures. In a small banking system or a country with limited specialist capacity, duplicating those functions may divert contributions away from building the fund. A stand-alone authority may also have to negotiate for information and staff support at exactly the moment when a bank’s condition is deteriorating.
Housing the function inside an existing authority can provide economies of scale. A central bank or another established public authority may already have information technology, human resources, legal, communications, and investment capabilities. A resolution authority may already hold the information and expertise needed to assess a failing bank. Integration can make routine cooperation easier because staff work within the same operational environment, and it may help a small system redeploy scarce specialists when a failure approaches.
Integration does not automatically solve coordination. If the deposit insurance function is placed inside the same institution as the supervisor, internal departments can still have different responsibilities, incentives, and information boundaries. The institution may need internal directives specifying which supervisory information must reach the deposit insurance team and when. If a fund is administered by a central bank, the law should also make clear that the fund’s liabilities are not automatically liabilities of the central bank. Otherwise, a structure intended to use capacity efficiently may blur financial responsibility.
The choice can change as the banking system develops. A country with few failures, a small banking sector, and limited capacity may gain more from a hosted function than from a costly new agency. A larger or more complex system with established resolution and deposit insurance expertise may be able to support a stand-alone authority or a more clearly separated resolution-and-insurance structure. Neither conclusion follows from the agency label alone. The relevant comparison is the capability gained, the duplication created, and the risks that must be governed.
The legal form creates another important distinction. A public DIS is established by law and led by a public authority. A private DIS is established and administered by the banking industry, usually with limited public oversight. A hybrid DIS is government-regulated but privately administered under a private-law body operating within public parameters. These forms can coexist with different mandates, and the international principles do not simply declare that one form is invalid. The question is whether the chosen form can perform the public function with the required information, independence, funding, and accountability.
The IMF staff recommendation favors a public authority, especially when the mandate extends beyond a narrow paybox. The reason is practical rather than symbolic. A paybox-plus insurer may need confidential information about supervisory ratings, problem banks, resolution plans, and individual depositor records. Bank supervisors and resolution authorities may be reluctant to share sensitive information with a private or hybrid body whose governance includes representatives of the institutions being assessed. Industry representatives may also have an incentive to resist higher contributions, even when the fund needs to grow.
A public form does not remove conflicts. A public resolution authority may want to preserve a bank for financial-stability reasons; the supervisor may worry about the condition of the bank that would acquire it; the central bank may be a creditor; the Treasury may be concerned about fiscal exposure; and the deposit insurer may need to limit the loss to its fund. Public governance must make these conflicts visible and provide rules for resolving them. A private form can face the same conflicts with an additional problem: the institutions that pay premiums may also influence the body responsible for deciding how much protection and funding the system needs.
For that reason, the governing body needs more than industry knowledge. It needs operational autonomy, technical expertise, internal controls, transparency, legal protection for good-faith decisions, and accountability to a higher authority. Active bankers may still provide useful specialist advice in a separate advisory setting, but allowing them to control decisions about confidential information, premiums, or the use of the fund can create real or perceived conflicts. The protection of a public function cannot depend on the goodwill of the institutions that bear its cost.
The information boundary is a design boundary
The most important institutional question may be what crosses the boundary between agencies. A deposit insurer needs timely information about the condition of member banks, the structure of their deposits, and the plans for recovery or resolution. Without that information, it cannot assess the risk to its fund, prepare reimbursement, or judge whether it has the capacity to support a resolution. Yet the information may be confidential and market-sensitive, so the law must authorize sharing while imposing clear confidentiality obligations.
This is why institutional proximity is useful but not sufficient. A hosted deposit insurance team may receive information more easily, but only if internal rules require the relevant departments to share it. A separate agency may have a clearer accountability line, but only if legislation, regulations, or formal agreements give it access to supervisory and resolution information. The practical test is not whether the institutions attend the same meetings. It is whether the deposit insurer can obtain accurate information early enough to act.
The same boundary applies to fund control. Resolution powers should be assigned clearly so that the resolution authority can choose and execute a resolution strategy without competing authorities each trying to direct the failure. At the same time, a separate deposit insurer should not be treated as an unlimited account that another authority can draw from without safeguards. The legal design must define when the fund can be used, what information the insurer receives, what conditions must be satisfied, and how disagreements are resolved quickly enough to avoid either unchecked spending or a dangerous stalemate.
That balance is difficult because speed and accountability pull in different directions. A resolution weekend does not allow every institutional disagreement to become a prolonged negotiation. But urgency is not a reason to eliminate controls over public or member-funded resources. A workable arrangement makes ordinary decision paths clear in advance, gives the resolution authority unambiguous resolution powers, gives the deposit insurance function the information and protections needed to guard its fund, and provides review after the action. The exact mechanism depends on national law; the underlying requirement is that neither authority should discover its role for the first time during a failure.
The practical choice is therefore not “integration or independence” in the abstract. It is a sequence of more concrete questions. Does integration give the deposit insurance function access to the people and information it needs without burying its mandate? Does separation provide genuine autonomy without creating an agency too small to operate? Can a private or hybrid body receive confidential information and obtain a credible public backstop? Can a resolution authority act quickly while the fund’s owner remains accountable for the conditions under which money is used?
A mandate is credible when the answers line up. Paybox may be appropriate where the system is deliberately limited and reimbursement capability is strong. Paybox-plus can add continuity if the resolution authority and deposit insurer have compatible powers, information, and safeguards. A loss-minimizer or risk-minimizer arrangement can concentrate expertise, but must manage conflicts among supervision, resolution, liquidity, and depositor protection. A stand-alone, hosted, public, private, or hybrid form can work only when it supplies the required capacity and preserves the public purpose.
The architecture is now ready for a harder test. It is one thing to decide who should hold the mandate; it is another to follow a bank as it weakens and ask whether information, authority, governance, and resources arrive in time. That test will show whether the institutional design is a working safety net or merely a plausible diagram on paper.