How a Bank Group Can Sidestep a Lending Constraint
This chapter follows one specific mechanism inside a banking group: a tighter loan-supply rule binds a regulated bank subsidiary, and the parent responds by using an affiliated NBFI as another lending vehicle. The point is not to decide whether the move is clever or improper. It is to identify the steps that make it possible, the resources the parent must bring to it, and the limits that remain. That mechanism gives readers a more precise way to distinguish a change in regulatory exposure from a complete escape from regulation.
When the rule binds the bank
A loan-supply rule is meant to affect how much credit a bank can extend, or the terms under which it can do so. It may restrict credit growth, impose loan-related requirements, limit certain forms of lending or otherwise make additional bank origination more costly. The immediate effect is not necessarily that a borrower becomes unable to obtain any financing. The first effect is that one legal entity—the regulated bank subsidiary—has less room to supply it.
The distinction is important because a banking group is not a single balance sheet in the operational sense, even when supervisors also examine it on a consolidated basis. The parent controls several entities with different licenses, funding sources and business models. A bank subsidiary may be the group’s natural vehicle for deposits and conventional lending, but an affiliated investment bank or broker-dealer may already arrange or hold corporate loans. If the rule attaches more directly to the bank subsidiary’s lending capacity, the parent can compare the cost of keeping the loan there with the cost of assigning some part of the relationship to the affiliate.
That comparison is shaped by the regulatory perimeter. Banks generally face extensive prudential oversight, including requirements related to capital, liquidity, activity and safety and soundness. Many NBFI affiliates face a different regime. They may be supervised for investor protection, market integrity or conduct without being subject to the same bank-like capital and liquidity requirements. In some jurisdictions, the affiliate is also touched by consolidated supervision, but the direct constraints on a specific loan product may still be weaker or different.
This is not a claim that the nonbank is outside all oversight. It is a claim about incidence: which entity bears the immediate force of the rule, which entity can still use its balance sheet, and which entity’s activities are counted in the relevant regulatory calculation. A measure that constrains a local loan-supply margin may leave a larger substitution opportunity than a measure calculated against the parent’s consolidated capital position. The distinction between those instruments will matter later; for now, it explains why a parent can have a choice at all.
The difference can be understood as two versions of the same decision. Under a local loan-supply constraint, the parent may see the bank subsidiary lose room in a particular lending channel while an affiliated NBFI remains able to participate. The group can then compare the cost of reducing the loan with the cost of using the affiliate. Under a capital-based constraint applied to consolidated accounts, moving the loan may not change the group’s relevant capital position. The alternative entity still exists, but the parent has less ability to use it to relieve the binding constraint. A legal chart of subsidiaries is therefore not enough; the analyst must identify the level at which the rule is calculated.
The same distinction prevents a common analytical mistake. Observing that a nonbank is subject to lighter direct rules does not establish that it can expand without limit. It may lack funding, lending capacity or the ability to absorb the exposure. It may also be affected indirectly by the parent’s consolidated supervision. The mechanism is best treated as a conditional margin: it is available when the affiliate can perform the lending function and when the relevant rule does not already reach the group as a whole. Those conditions are why the empirical result can be strong for some measures and absent or reversed for others.
The choice also depends on whether the affiliate is capable of doing the job. An NBFI that is already active in the relevant corporate-lending market can take on the lending role more readily than an entity with no such capacity. The parent’s existing group relationships and funding channels can reduce the organizational friction of the move. The legal form changes, but the group does not start from zero.
The process can be represented as a sequence rather than a slogan.
The sequence does not imply that every reallocation is a prearranged instruction from the parent. It shows the economic path that the evidence makes plausible. The regulator changes the conditions facing the bank subsidiary. The parent then has to decide whether an affiliated NBFI can carry the relationship. If it can, the borrower may receive credit from a different legal entity while the parent continues to coordinate the group’s resources and strategy.
That last point is why “sidestep” needs careful use. In ordinary language, it can sound like evasion or misconduct. Here it describes a shift toward an entity that is less directly affected by the measured constraint. Whether that shift violates a rule depends on the applicable law and the facts; the evidence at issue does not establish unlawful behavior. Nor does it show that the parent escapes supervision. Consolidated oversight, reporting requirements, funding dependence and the parent’s own incentives can still connect the affiliate to the regulated bank.
How the parent reroutes the relationship
The most consequential resource in this mechanism is not simply the affiliate’s legal status. It is access to the group’s internal capital market. A bank-owned NBFI may be able to obtain funding, liquidity support or favorable terms through the parent’s wider relationships, even though the NBFI itself cannot collect deposits like a bank. Those group resources can make the affiliate a practical substitute rather than a theoretical one.
The IMF study does not directly observe each internal funding flow. That boundary matters. Its lending estimates show that affiliated NBFIs expand credit when bank subsidiaries contract after the relevant tightening, which is consistent with a parent shifting capacity toward the NBFI. Related evidence cited by the paper finds favorable credit conditions for some nonbank affiliates, and the overall pattern fits the operation of internal capital markets. But a statistical reallocation does not reveal every transfer agreement, the exact price of funding or the parent’s full decision process.
Several motives may therefore operate at once. Regulatory pressure can make the affiliate route more valuable. A parent may also use an NBFI to diversify funding channels, preserve a client relationship, manage liquidity risk or maintain its position in a market. These motives are not mutually exclusive. Treating the mechanism as pure regulatory arbitrage would discard the organizational reasons that make a group own multiple financial businesses in the first place.
The relationship with the borrower is another part of the mechanism. A bank’s reduction in new lending does not necessarily mean that the parent has abandoned the borrower. The group may preserve corporate credit by moving the exposure to the affiliated NBFI. From the borrower’s perspective, the source of funds may remain within the same wider group even though the legal lender and the directly applicable prudential constraints have changed.
That continuity can benefit the borrower in the short run. If the affiliate has room to lend, the company may avoid a sharper interruption in corporate financing. It can also benefit the parent by protecting its position in a lending market. But the continuity makes the policy question harder, not easier. A measure designed to cool credit growth may have a smaller effect on the group’s total lending when the parent can route credit around the most constrained entity.
The mechanism is especially revealing when the parent is financially weak. The IMF finds stronger reallocation among groups with indicators such as higher distress probability or nonperforming loans, lower Tier 1 capital ratios or smaller size. A weaker group has more reason to preserve income and lending relationships when a rule raises the cost of using its bank subsidiary. It may also face a more urgent need to avoid a contraction in activity. The affiliate margin becomes more valuable precisely because the constraint is more painful.
This finding is about relative behavior, not a claim that weak groups are better at lending. It says that the difference between the bank subsidiary’s response and the NBFI subsidiary’s response is larger for groups with fewer comfortable options. The parent’s condition changes the value of flexibility. A strong group might absorb a regulatory cost in the bank or choose not to move a marginal loan. A weaker group has greater incentive to preserve the group’s lending footprint through the entity that remains less directly constrained. The same observation also raises a verification question: to understand the full effect, one would want direct evidence on internal funding and on whether the affiliate’s expansion changes the group’s consolidated risk.
That incentive has a double edge. If the parent is strong enough to support its affiliates, the group’s internal market may help maintain lending through a difficult period. If the parent is weak, however, the same dependence can concentrate risk. The affiliate may rely on the parent for funding or confidence, while the parent relies on the affiliate to sustain business. A decision that protects current origination can leave the group more exposed to a later shock in either entity. The evidence supports the stronger response by weaker groups; the broader fragility implication follows from the dependence described by the two institutional sources and must be treated as a risk channel, not as a measured failure rate.
This also explains why an independent NBFI is not an interchangeable alternative. It may be subject to lighter regulation, but it does not share the parent’s internal funding, information and distribution network. When credit is scarce, those connections can matter more than the label “nonbank.” The IMF’s comparison finds that independent NBFIs reduce lending relative to banks after tightening, while bank-owned NBFIs expand relative to them. The difference is consistent with an ownership advantage, although the study does not directly measure every funding transfer or implicit guarantee.
The same logic helps separate a group-level adjustment from a system-wide conclusion. If a bank subsidiary cuts lending and an affiliated NBFI expands it, the group may cushion the borrower-level effect. But the loan has not become risk-free, and the bank–NBFI relationship has not disappeared. The parent may still finance the affiliate or remain exposed to its performance. The formal location of the loan has changed; the economic network around the loan may have become more important.
The legal and economic distinctions should therefore be kept together. Legal separation tells us which rules and supervisors may apply directly. Economic affiliation tells us how funding and risk can travel. Neither category alone answers whether the parent has truly reduced its exposure. A nonbank affiliate may be outside a particular bank-level capital calculation while remaining deeply connected to the parent’s funding and strategy.
The mechanism also has a stopping condition. If the relevant rule is imposed on the group’s consolidated capital position, shifting the loan to an affiliate may not relieve the constraint. If the affiliate lacks lending capacity, market access or funding, it cannot serve as a substitute. If the parent is too weak to support it, the internal capital market may transmit stress rather than provide a buffer. “Sidestepping” is therefore conditional on the rule, the organization and the available resources.
The central conclusion is narrower and more useful than the claim that regulation simply pushes lending into shadow banking. A parent banking group can sometimes change the legal vehicle through which it supplies corporate credit because its bank and NBFI subsidiaries face different immediate constraints and share organizational resources. The empirical question is whether that pattern appears systematically after tightening and how large it is. The next chapter turns to that evidence, showing what the study compares, how it controls for borrower demand and where its inference stops.