Buried in more than 1,500 pages of proposed capital rules is a policy goal that regulators have stated almost in passing: bringing mortgage origination and servicing back within the agencies' regulatory perimeter.
That phrase deserves more attention than it has received, because it describes an attempt to reverse one of the most significant structural changes in American finance since 2008.
What actually changed on March 19
The Federal Reserve, OCC, and FDIC jointly proposed a package that formally rescinds the 2023 Basel III Endgame framework and replaces it with something considerably more permissive. The agencies estimated the standardized approach proposal would decrease CET1 capital by approximately 8.6% for covered depository institutions and 6.8% for covered holding companies. The agencies acknowledge that overall system capital would modestly decrease, while arguing levels remain substantially above pre-crisis.
Three mortgage-relevant changes stand out.
First, the proposal would remove the threshold-based deduction for mortgage servicing assets for all banking organizations, including those using the community bank leverage ratio framework. The agencies state this is intended to promote mortgage origination and servicing by banking organizations in a risk-appropriate manner.
Second, the proposal introduces more granular, loan-to-value-based risk weights for residential mortgages, so capital scales with measured leverage rather than applying broad categories.
Third, and cutting the other way, Category III and IV banks would be required to include most elements of accumulated other comprehensive income in CET1, reflecting unrealized gains and losses on certain securities in regulatory capital.
Why the MSA deduction change is the big one
The mortgage servicing asset deduction is technical enough that it rarely gets explained, and it is the single provision most likely to move business between institutions.
When a bank originates a mortgage and retains the right to service it, collecting payments, managing escrow, handling delinquency, that servicing right is an asset with real economic value. But under existing capital rules, MSAs above a threshold must be deducted from capital, on the theory that their value is volatile and hard to realize in stress. Servicing rights swing violently with interest rates: when rates fall and borrowers refinance, servicing portfolios evaporate.
The practical effect of that deduction has been to make servicing expensive for banks to hold. And that is a substantial part of why servicing migrated to nonbanks, which face no equivalent capital charge. Independent mortgage banks now dominate origination and hold enormous servicing portfolios precisely because the capital framework made it costly for banks and free for everyone else.
Removing the deduction erases that asymmetry. Analysis suggests banks more heavily engaged in mortgage activities are expected to see the greatest capital relief from the combination of granular LTV risk weights and MSA relief. Whether it pulls business back is an empirical question, but the incentive it removed was real and the direction is unambiguous.
The regulatory-perimeter argument, and the objection to it
There is a serious case for this, and it is not simply deregulation dressed up.
Since 2008, mortgage lending shifted heavily from banks to nonbanks. Nonbanks do not take deposits, so they are not subject to bank capital rules, examinations, or the same supervisory oversight. They fund themselves with warehouse lines and securitization, which can vanish in a liquidity crunch. If a large nonbank servicer fails, there is no FDIC resolution framework and no clear backstop, and the borrowers whose loans it services still need someone to collect their payments and manage their escrow.
So the argument runs: pushing mortgage activity outside the regulated banking system did not eliminate the risk, it relocated it somewhere regulators can see less and control less. Recalibrating capital so banks can compete brings the activity back where supervision reaches. The agencies' own framing, returning mortgage to the regulatory perimeter, is exactly this argument.
The objection is equally straightforward. Capital requirements exist because banks are leveraged institutions holding insured deposits, and the discipline they impose is the point, not a bug. Reducing capital by 8.6% system-wide to make banks competitive with unregulated rivals accepts more risk inside the insured system in exchange for visibility. That is a genuine trade, and reasonable people weigh it differently. It is worth noting that the same logic, if applied broadly, would justify loosening any rule that pushes activity to less-regulated competitors.
The provisions that quietly tighten
The proposal is not uniformly permissive, and the conservative elements deserve mention because they complicate the deregulation narrative.
The AOCI requirement for Category III and IV banks is a direct response to the 2023 regional bank failures. Silicon Valley Bank held large unrealized losses on securities that did not appear in its regulatory capital ratios until it was forced to sell them and crystallize the loss. Requiring mid-size banks to reflect those unrealized losses in CET1 is the fix for exactly that failure mode, and it is a meaningful tightening for banks that had been excluded.
Two mortgage-specific details are also notably conservative. In calculating LTV, the property value would generally be maintained at the value measured at origination, rather than updated as prices rise, which prevents banks from claiming capital relief from paper appreciation. And banks would not be permitted to recognize private mortgage insurance when calculating the LTV ratio, meaning a high-LTV loan carries its full capital weight regardless of insurance. Both choices assume less protection than a bank might claim, which is the correct direction for a capital rule.
The timing question worth asking
The most useful skeptical question is not whether the framework is well designed but whether now is the moment to reduce mortgage capital.
The current credit picture is uneven. FHA delinquency has climbed to levels not seen since 2012 outside the pandemic, negative equity is concentrated among 2022 to 2024 vintage borrowers in Sun Belt markets, and escrow costs from insurance and property taxes are squeezing homeowners in ways mortgage rates do not capture. None of that is a systemic crisis, and prime bank mortgage portfolios are performing well. But it is a moment of visible stress in specific pockets of housing credit.
Reducing capital held against mortgage exposures is a bet that the loss experience will stay benign. The granular LTV weights partly mitigate this, since riskier loans attract more capital under the new framework than under the current one. Still, an 8.6% aggregate CET1 reduction is a real decrease in loss-absorbing capacity, arriving alongside deliberate encouragement of more bank mortgage lending, at a time when parts of the mortgage market are deteriorating. That combination deserves scrutiny even from people who think the perimeter argument is sound.
What to watch
Comments closed June 18, 2026, and the proposals are not final. Three things will indicate how this actually lands.
Whether banks measurably reclaim mortgage origination and servicing share is the test of the entire rationale. Large banks posted strong mortgage volume gains recently, and it has been genuinely unclear whether that reflected a growing market or share taken from nonbanks. These rules are designed to make the second outcome more likely, which means the perimeter thesis will be observable in the data within a few years.
Whether the AOCI requirement survives industry pressure will show whether the post-2023 lesson holds. It is the most expensive provision for affected banks and the one most directly tied to a recent failure.
And whether the final rule addresses combined leverage across liens is a live question. Capital rules keyed to loan-to-value depend entirely on measuring the borrower's actual leverage, and a framework that weights a first lien without visibility into subsequent junior liens is measuring something narrower than the risk it intends to capture. That gap has been raised in the comment process, and how the agencies resolve it will say a great deal about whether the granular LTV approach delivers the risk sensitivity it promises.