Who Gets the Key to the Backstop?
Central bank lending facilities are also systems for deciding which parts of the financial system the central bank can reach. This chapter examines counterparty eligibility, with special attention to banks, non-bank financial institutions, and indirect access through dealers. The practical judgment is not whether access should be broad in the abstract, but whether a particular route reaches the liquidity problem while keeping credit, market, and incentive risks within a defensible boundary.
Why banks are the default counterparties
Banks and other depository institutions occupy the broadest position in the cross-jurisdiction evidence. The BIS Compendium reports that banks are eligible in virtually all of the jurisdictions covered, not just for standing lending facilities but also for discretionary facilities and open market operations. That pattern follows from what banks do inside the monetary system.
Banks are central to financial intermediation: they transform deposits and other short-term funding into loans and securities, and they connect the policy rate to the terms faced by households and businesses. They also perform maturity transformation, funding assets that may be difficult to sell quickly with liabilities that can move or be withdrawn sooner. When reserve demand changes or markets become unsettled, a bank's liquidity problem can therefore affect payment flows, credit supply, and monetary-policy transmission at the same time.
A central bank also has an established supervisory relationship with banks. That does not eliminate risk, and it does not make every bank automatically sound. It does mean that the central bank and other supervisors usually have a recognized institutional perimeter within which to assess condition, collateral, and controls. Eligibility is consequently not a reward for being a bank. It is a decision about which institutions are sufficiently central to the policy and payment system, and sufficiently knowable within the regulatory framework, to be connected to a liquidity channel.
The word “eligible” needs careful handling. It identifies a class of institutions that may use at least one tool under its rules; it does not guarantee approval of every request, acceptance of every asset, or funding at any desired time. A bank may be eligible for a standing facility but still need legal agreements, pledged collateral, and an acceptable borrowing request. The counterparty rule opens a door. Other parts of the facility determine whether a particular transaction can pass through it.
The reason banks are broadly eligible does not mean the same rule should automatically apply to every financial institution. A securities dealer, clearing platform, insurance company, pension fund, or money-market fund may have a different balance sheet, funding pattern, regulatory status, and relationship to the market the central bank is trying to stabilize. Counterparty design is therefore a map of the financial system's functions, not merely an institutional list.
How NBFI access differs across the evidence
Non-bank financial institutions, or NBFIs, include several distinct categories. The BIS table separates securities brokers and dealers, financial market infrastructures, money-market funds, insurance companies, pension funds, and other entities. Their access varies by category, by the type of liquidity tool, and by whether the jurisdiction is an advanced or emerging market.
The comparison needs a consistent denominator. The table reports the share of jurisdictions in which a counterparty type has access to at least one central-bank liquidity tool. It does not report the amount of credit extended, the probability that a request will be approved, or the share of a financial sector's assets. For example, among the 12 advanced economies in the table's standing-lending-facility sample, securities brokers and dealers have access in 58 percent of jurisdictions and financial market infrastructures in 67 percent. Among the 13 emerging-market jurisdictions in that same standing-facility comparison, the corresponding shares are 31 percent and 15 percent. These are differences in the presence of an access route, not measures of how heavily a facility is used.
The pattern is narrower for other NBFIs. In the same standing-facility comparison, money-market funds have access in 17 percent of the advanced-economy jurisdictions and 15 percent of the emerging-market jurisdictions. Insurance companies are at 17 percent and 8 percent, while pension funds are at 8 percent in both groups. Those figures should not be read as a ranking of institution quality. They show that access is much less common outside banks, and that the difference depends on the type of institution and the facility being considered.
Why are securities firms and financial market infrastructures more likely than funds or insurers to appear in access rules? Brokers and dealers are active intermediaries in money and securities markets, so their funding can transmit stress quickly through market-making and secured financing. Financial market infrastructures, such as clearing and payment platforms, are important because their failure or delay can interrupt settlement even when the underlying financial institutions remain open. Funds, insurers, and pension funds have different liabilities and asset pools. Their liquidity needs can be large, but their access raises distinct questions about credit assessment, asset valuation, regulation, and the central bank's authority.
The tool's function also changes the answer. A payment-system facility may prioritize an infrastructure that must settle transactions. A standing facility used for ordinary reserve management may be limited to institutions with regular operational relationships. A discretionary stability facility may be opened for a market segment whose liquidity demand becomes systemically important during a specific disruption. The same NBFI can therefore be outside one channel and inside another without the rules being inconsistent.
The BIS gives examples of both broad and restricted approaches. At some central banks, insurance companies, money-market funds, and investment or pension funds can access open market operations and standing facilities if they meet regulatory criteria and are important for money-market liquidity. Elsewhere, regulated NBFIs have gained access mainly to discretionary facilities during financial stress. These examples do not establish one preferred model. They show that eligibility follows a jurisdiction's market structure, legal framework, and assessment of where liquidity support can prevent wider damage.
Direct access versus an intermediary route
The most consequential choice is not always whether an NBFI can benefit from central-bank liquidity, but whether it can borrow directly. Direct access puts the central bank closer to the institution experiencing the funding strain. That can reduce a delay and avoid forcing the institution to sell assets immediately. It also puts the central bank closer to the institution's credit risk, balance-sheet complexity, and potentially less standardized collateral.
Indirect access inserts a bank or dealer between the central bank and the market participant. The intermediary borrows or receives funds under the facility, then uses them to purchase assets or provide financing to the strained segment. This can preserve a familiar central-bank counterparty relationship and make the risk assessment more manageable. But the route may be slower, more expensive, or less certain if the intermediary is unwilling or unable to expand its balance sheet when markets are under pressure.
The Federal Reserve's money-market mutual-fund liquidity facility illustrates the indirect route. During episodes including the subprime crisis and the Covid-19 pandemic, eligible banks and dealers received non-recourse loans to purchase high-quality assets from money-market mutual funds. The banks and dealers pledged those assets as collateral. The facility could therefore address redemption pressure at funds without making every fund a direct borrower from the Federal Reserve.
That arrangement changes the path of support, not the underlying reason for it. If investors redeem fund shares quickly, a fund may need cash and could otherwise sell assets into a falling or illiquid market. A bank or dealer purchasing high-quality assets can provide a buyer and reduce pressure for fire sales. The central bank, however, now has exposure to the intermediary and to the collateral it accepts. “Indirect” does not mean “riskless”; it means the counterparty and the transmission path are different.
Direct access has the opposite profile. If a central bank can assess an NBFI, control its collateral, and impose appropriate terms, lending directly can reach the source of the liquidity need without relying on an intermediary's balance-sheet capacity. That may be valuable when a market-wide shock affects many participants at once. Yet many NBFIs are more lightly regulated than banks, and their funding structures may make credit, leverage, and maturity risks harder to evaluate. The central bank may also take market risk if it accepts assets whose value changes quickly or whose markets have become illiquid.
This is the core trade-off. Broader access can reduce the damage caused by a temporary shortage of liquidity. It can also create an expectation that the central bank will finance institutions or strategies that would otherwise face market discipline. If participants believe central-bank funding will always be available, they may have less incentive to hold liquid assets, limit leverage, or control the gap between asset and liability maturities. The risk is not necessarily immediate misuse; it is a change in behavior before the next stress episode.
The BIS identifies proportionate regulation of maturity mismatch and leverage as one possible way to address this concern. The logic is conditional: if an institution is given a direct liquidity route because it is important to a market, the rules around that route may need to address the risks that made the institution's liquidity demand difficult to absorb. Eligibility, supervision, collateral, and pricing then work as a package. Expanding one without considering the others can shift risk rather than solve it.
The visual should not be mistaken for a universal sequence. Some tools admit several counterparty types, some NBFI access is indirect, and the function of the facility can change the route. Its point is to show the balancing problem: reach can stabilize the market, while the chosen route determines which institution bears the first layer of credit, market, and incentive risk.
Judge access by purpose, not by a universal rule
A sensible access decision begins with the facility's purpose. If the goal is monetary-policy transmission, the central bank may prioritize institutions that trade or transmit short-term rates and reserves. If the goal is payment-system support, a financial market infrastructure may matter because timing failures can interrupt settlement. If the goal is financial stability, the relevant question may be which institutions hold assets or liabilities at the center of the stressed market and whether direct support can prevent a destructive sale.
That purpose-based approach avoids two opposite errors. The first is assuming that banks should always be the only counterparties because they are familiar. That can leave a major market segment outside the liquidity channel when non-banks are the source of demand or the place where fire-sale pressure appears. The second is assuming that every systemically important NBFI should receive the same direct access as a bank. Importance does not by itself solve questions about supervision, collateral, legal authority, or the central bank's ability to control losses.
Access should also be read alongside the facility's safeguards. The Federal Reserve's risk-management description emphasizes ongoing condition monitoring, identification of institutions that pose higher risk, communication with supervisors, and controls that become more stringent as risk rises. Those layers matter because collateral does not replace borrower analysis. A collateralized loan still depends on the asset's value, the lender's legal claim, the borrower's condition, and the process used to mobilize and monitor the collateral.
The most defensible conclusion is therefore neither “broader access is better” nor “restricted access is safer.” Broader access may be more stabilizing when liquidity stress sits outside the banking perimeter, but it can increase exposure and complicate supervision. Restricted or indirect access may preserve stronger controls, but it can leave support too far from the market where stress is forming. The right comparison is conditional: for this function, in this market structure, which counterparty route reaches the problem quickly enough, and what controls make the resulting risk acceptable?
Once that question is answered, eligibility is only the beginning of the transaction. An institution may be allowed to enter a facility, but the central bank still needs collateral it can value, control, and turn into a defined amount of lendable credit. That is where the access decision meets the balance sheet.