The Beacon | Banking On Nonbanks
New NY Fed research reveals that liquidity risk hasn’t disappeared, just gone dark.
The conventional narrative about financial intermediation is fundamentally wrong. For decades, analysts have watched bank-led finance decline while nonbank financial institutions (NBFIs) surge, interpreting this as a structural shift toward a nonbank-centric system. But a new paper from the Federal Reserve Bank of New York reveals that this story obscures a critical reality: much of the NBFI growth has occurred inside banking holding companies (BHCs) themselves, and regulatory policy is now pushing banks to unwind these internal liquidity arrangements, potentially concentrating risk outside the regulatory perimeter.
This is not merely an organizational footnote. It goes to the heart of how monetary policy transmits, where systemic risk accumulates, and whether post-crisis regulation has displaced problems rather than solved them. For macro practitioners focused on liquidity mechanics, this work from Nicola Cetorelli and Saketh Prazad deserves immediate attention.
Banks Are Not What You Think They Are
Begin with a simple but overlooked fact: the textbook definition of a bank—an entity that takes deposits and makes loans—has been obsolete for decades. The transformation began in the mid-1980s, when regulatory interpretations allowed BHCs to integrate nonbank subsidiaries spanning investment funds, securities dealers, specialty lenders, insurers, and broker-dealers .
The scale of this integration is staggering. Nonbank subsidiaries grew from ~10% of consolidated BHC assets in the mid-1990s to over 30% just before the 2008 financial crisis. While that share declined post-GFC, it has stabilized around 20% since then.
The scope of activity is equally revealing. By 2020, 64% of the top 200 BHCs had specialty lenders, 69% held securities brokers, 66% held insurance subsidiaries, and 74% had investment funds. Before the GFC, the average BHC was engaged in about 37 unique nonbank business lines—double the count from the 1980s .
In essence, the modern banking firm is a diversified financial conglomerate—not a deposit-loan institution.
The Liquidity Synergy Logic
Why did banks build these complex financial ecosystems? The answer lies in imperfect correlations in liquidity demands across business lines.
During market stress, a traditional bank might face deposit outflows while a nonbank affiliate—like an investment fund—experiences redemption pressure. But these events often occur at different times. By combining them under one roof, BHCs can economize on total liquidity needs and rely on internal reallocation, rather than external borrowing .
Between 1995 and 2022, intra-company funding between bank and nonbank affiliates averaged around 5% of bank subsidiary assets. These transfers weren’t incidental—they were the internal liquidity plumbing.
The real stress test came in 2007. Banks with high exposure to ABCP conduits drew significantly from their nonbank subsidiaries while avoiding reliance on the Fed’s emergency liquidity. According to Cetorelli and Prazad, this internal backstopping absorbed ~$176 billion in liquidity pressure—25% of the Fed’s total emergency lending during the GFC .
Regulation Breaks the Synergy
After the crisis, the Dodd-Frank Act’s “living wills” requirement disrupted these internal liquidity mechanisms. Regulators discouraged funding interdependence because complex internal ‘plumbing’ makes a bank impossible to unwind in bankruptcy. To make BHCs resolvable, they demanded these entities be severable, effectively forcing banks to dismantle their own shock absorbers.
Fed researchers tested the effect of this shift by comparing BHCs subject to the living wills requirement with those that weren’t. The result? Post-regulation, BHCs reduced their nonbank asset share, number of subsidiaries, business line diversity, and intracompany funding flows .
The Paradox: Where Does the Risk Go?
Here’s the paradox: regulations meant to reduce risk inside banks may have merely displaced it to the unregulated periphery.
When BHCs divest nonbank subsidiaries, the intermediation function doesn’t disappear, it migrates to independent NBFIs. But banks still provide liquidity to these entities through credit lines, repos, and deposit flows. So instead of monitoring internal funding relationships, regulators now face a sprawling, fragmented, and opaque system of external dependencies.
This externalization has grown. According to 2025 Fed research, bank credit lines to NBFIs now exceed 3% of GDP, becoming the dominant form of bank-NBFI connectivity.
What You Should Watch
For allocators and macro practitioners, this reshaped landscape highlights three critical areas of focus:
1. NBFI Credit Line Exposure: These commitments, now ~3% of GDP, are likely to be drawn simultaneously in a systemic liquidity event. Monitoring these exposures and the quality of underlying collateral is critical.
2. Intra-BHC Liquidity Network Erosion: Living wills policies have reduced internal shock absorption. This makes BHCs more reliant on Fed facilities and less able to self-insure during stress events.
3. The Regulatory Boundary Paradox: Risk hasn’t vanished; it has gone dark. It now lives in the blind spots between banks and unaffiliated nonbanks.
The Larger Frame
This research leads to a deeper insight: macroeconomic stability today depends not just on interest rates, but on liquidity mechanics.
The 2020 COVID crash illustrated this vividly. The stress wasn’t from deposit runs, it was from margin calls, investment fund redemptions, and drawdowns on bank credit lines to NBFIs. These vulnerabilities weren’t visible in traditional banking metrics, they were buried in the interstitial space between affiliated and unaffiliated liquidity flows.
As Cetorelli and Prazad’s work shows, banks haven’t retreated from intermediation. They’ve simply been forced to reorganize—shifting risk from visible internal channels to opaque external webs .
Implications
Portfolio managers should monitor FR Y-9C filings for intracompany funding levels… BHCs now operate with diminished internal liquidity flexibility.
Macro allocators should prepare for the next crisis to emerge from the NBFI sector, with faster transmission and less buffer capacity than in 2008.
Policymakers must revisit whether regulatory “ring-fencing” has enhanced resilience or simply fragmented the system into something harder to track, understand, and support in crisis.
That’s our view from The Watch. Until next time… we’ll be sure to keep the lights on.