How Big Wall Street Banks, From JPMorgan to Citi, Are Using AI - Business  Insider

On Thursday, the Federal Reserve released a proposal that would reduce capital requirements for the largest U.S. banks by 4.8%, for large banks with $100 billion to $700 billion in assets by 5.2%, and for banks under $100 billion by 7.8%. After implementation, the biggest banks would still hold more than $800 billion in capital to guard against downturns, roughly twice what they held before the 2008 financial crisis.

The immediate reaction from most observers was predictable: loosening capital requirements means more risk, right? Less cushion, more fragility, more potential for another 2008.

We think the opposite is more likely to be true.

Here is the logic. For nearly two decades, post-2008 capital requirements have been ratcheted upward on traditional banks. At the same time, those rules do not apply to nonbank lenders: mortgage companies, private credit funds, fintech platforms, and business development companies. The result has been a massive migration of lending activity out of the regulated banking system and into less transparent, less supervised corners of the financial system.

The numbers tell the story clearly. In the mortgage market, nonbank lenders have gone from originating roughly a third of home loans in 2010 to more than half today. Two-thirds of mortgages for home purchases in 2024 were made by nonbank lenders. Bank market share has declined from 42.5% in 2018 to just 30.1% in 2024. Of the ten largest mortgage originators in the U.S., only four are traditional banks. The top two, United Wholesale Mortgage and Rocket Mortgage, are both nonbanks.

This trend is not limited to mortgages. The Financial Stability Board reported that the nonbank financial intermediation sector grew 9.4% in 2024, double the 4.7% growth rate of the traditional banking system. Nonbank institutions now hold roughly 51% of total global financial assets, at $256.8 trillion. A narrower subset of nonbanks whose activities carry bank-like stability risks grew even faster at 12.7%, reaching $76.3 trillion.

Private credit alone has ballooned from $46 billion in 2000 to roughly $2.5 trillion in assets under management today. That is not a gradual shift. That is a structural remaking of the American credit system.

The problem is straightforward. Banks are regulated, stress-tested, required to hold capital against losses, required to disclose their loan portfolios, and backed by FDIC insurance. Nonbank lenders are subject to far fewer of these constraints. They operate with less transparency, lighter regulatory oversight, and no deposit insurance. When a bank makes a bad mortgage, regulators can see it coming and force the bank to set aside reserves. When a nonbank lender makes the same bad mortgage, the visibility is dramatically lower.

Fed Vice Chair for Supervision Michelle Bowman said it directly: overly aggressive requirements “produced unintended consequences, including constraining credit availability, pushing banking activities into the less-regulated non-bank sector, while adding complexity and costs without meaningfully enhancing safety and soundness.” The proposals, she said, aim to bring traditional lending back into the regulated banking system.

This is the core of the argument. The question is not whether banks should be tightly regulated. They should. The question is whether regulations that are so punitive that they drive lending activity out of the regulated system entirely are actually making the financial system safer, or whether they are doing the opposite.

Think of it through a simple analogy. If you set the speed limit on the highway to 25 miles per hour, you will not eliminate fast driving. You will push drivers onto back roads that have no speed limits, no guardrails, and no traffic enforcement. The total amount of risk in the system does not decline. It just moves to where you can no longer see it or control it.

That is what has happened in American credit markets. Post-crisis capital requirements made traditional bank lending more expensive. Banks responded rationally: they pulled back from mortgage origination, middle-market business lending, and other credit activities. Nonbanks filled the gap. Academic research from Buchak et al. (2018) and multiple Federal Reserve studies confirm that increased regulatory burdens on banks, including stress tests and higher capital requirements, directly contributed to the rise of nonbank lending.

The IMF’s October 2025 Global Financial Stability Report flagged that U.S. and European banks now have $4.5 trillion in exposure to nonbank financial institutions, including hedge funds and private credit groups. The interconnections have grown large enough that a stress event in the nonbank sector would transmit directly to the banking system anyway, through credit lines, co-lending arrangements, and shared counterparty risk. The regulated system is exposed to the unregulated system whether regulators like it or not.

The Fed’s proposal does not eliminate the post-crisis framework. Capital levels would still be roughly twice what they were before 2008. What it does is recalibrate: easing the pressure that has been pushing lending activity into the shadows for the better part of a decade, and creating incentives for that activity to migrate back into a system where regulators can actually see it.

There are several direct implications worth tracking.

Traditional bank stocks benefit. Lower capital requirements mean higher return on equity, all else being equal. JPMorgan, Bank of America, and other large banks will be able to deploy more of their balance sheet toward lending rather than holding sterile capital buffers. TD Cowen’s Jaret Seiberg noted that these proposals are likely “floors on capital relief rather than ceilings,” suggesting the final rules could be even more favorable after the comment period.

Nonbank lenders face competitive pressure. If banks can re-enter mortgage origination and middle-market lending more aggressively, nonbank lenders like Rocket Mortgage, United Wholesale, and private credit funds lose some of the structural advantage they have enjoyed for a decade. That does not mean they disappear, but their growth trajectory may moderate.

Mortgage availability could improve. More bank participation in mortgage lending typically means more competition, which tends to benefit borrowers through lower rates and better terms. Banks also tend to be subject to the Community Reinvestment Act, which requires them to serve low- and moderate-income communities. That is something nonbank lenders largely sidestep.

Credit transparency increases. Every dollar of lending that moves from a private credit fund back onto a bank balance sheet is a dollar that becomes visible to regulators, stress tests, and public disclosure requirements. That is a net positive for financial stability, even if the headline reads “Fed loosens capital rules.”

The Bottom Line

The financial system is not made safer by rules that look tough on paper but push risk into places regulators cannot reach. For fifteen years, escalating capital requirements on banks have driven a structural shift in American lending toward less regulated, less transparent institutions. The Fed’s new proposal is an acknowledgment that the pendulum swung too far, and that recalibration can improve stability by bringing lending back into the light.

Lower capital requirements sound risky. In this case, they may be the safer path.

See ya later folks & stay curious.

—J&E

Sources

Federal Reserve; HMDA/CFPB via NCRC; Financial Stability Board Global Monitoring Report 2025; FDIC Risk Review 2025; Federal Reserve Bank of Boston; Federal Reserve Bank of Kansas City; IMF Global Financial Stability Report (October 2025); ABA Banking Journal; Yahoo Finance; Buchak et al. (2018); ProSight Financial Association.

This article is for informational purposes only and does not constitute investment advice.