The Fed wants to let a dangerous asset loose on bank balance sheets

• Key insight: The Federal Reserve recently proposed three changes to the way mortgage loans and mortgage servicing rights would be treated under bank capital requirements. Two of the three proposed changes are a step in the right direction.
• What’s at stake: Removing the deduction of mortgage servicing rights over 10% of a bank’s common equity Tier 1 invites one of the most notoriously volatile assets back onto banks’ balance sheets.
• Forward look: A potential solution lies in maintaining a cap on MSRs at a level between that set currently and by Ginnie Mae, permitting some offset for hedging and aligning that regulatory treatment between banks and nonbanks to maintain a level playing field in mortgage banking.

To further prop up an ailing mortgage market, the Federal Reserve recently proposed three changes to the way mortgage loans and mortgage servicing rights, or MSRs, would be treated under bank capital requirements. Two of the three proposed changes are a step in the right direction but the other, removing the deduction of MSRs over 10% of a bank’s common equity Tier 1, or CET1, capital, invites one of the most notoriously volatile assets back onto the balance sheets of some banks in a big way.

Processing Content

To understand the risk of MSRs we must look back to a period before the 2008 financial crisis when banks were dominant players in both mortgage origination and servicing. Back then large banks maintained sizable mortgage servicing portfolios. Mortgage servicing is an inherent economy of scale business where volume and automation greatly affect financial performance. Unlike a mortgage loan, a secured asset that a bank might place in its portfolio, MSRs are an intangible asset reflecting the discounted present value of servicing cash flows the company earns from each mortgage loan it services. If the loan remains in the portfolio, the bank continues to receive cash flows. However, mortgage loans, unlike standard bonds, provide borrowers with a free option to refinance and prepay their loans when market rates fall below their mortgage note rate. In a falling rate environment, prepayments rise and that causes MSR values to decline. The scale aspect of the servicing business combined with prepayment significantly increases the risk profile of a bank’s balance sheet over time.

Compounding these twin features, increased asset concentration in an inherently volatile asset is the trifecta of significant model risk. Unlike other assets on a bank’s balance sheet, such as Treasuries and corporate bonds, MSRs do not actively trade and so are treated under fair value accounting rules as a Level 3 asset where valuations are determined by models and not market prices. MSR valuations hence are driven by critical assumptions regarding the trajectory of interest rates, discount rates and prepayments, among other factors. Accurate valuations can be a tricky business as a result and have caused headaches for any number of banks over the years.

Another potential area of concern in managing MSRs centers on hedging. Given their inherent volatility a bank needs to hedge this asset closely and as the adage goes, the only perfect hedge is an English garden. I’ve been at two of the largest depositories at the time that experienced billion-dollar MSR valuation problems due to inaccurate MSR modeling and hedging issues. MSRs are not an asset to be taken lightly by banks and regulators.

MSRs are also touted as providing a natural hedge for banks against accumulated other comprehensive income, or AOCI, an adjustment to equity for gains and losses to various bond investments. Under a rising rate scenario, bond losses increase which would flow into AOCI and thus cause bank capital to fall. At the same time, MSR values rise as prepayments slow. Following the failure of Silicon Valley Bank in 2023, the loophole allowing large banks to opt out of counting AOCI in capital was closed. This effectively eliminates the potential by banks like Silicon Valley Bank in the future to mask large bond losses right up to the end.

The Fed’s proposal to eliminate the MSR capital deduction provides a counterbalance to AOCI, as any hit to bank capital from bond losses would be offset to some extent by higher MSR valuation that would flow into retained earnings. A problem with this result is that it creates an incentive for banks to tinker with their MSR valuation models to enhance their overall capital position. Despite strong regulatory guardrails in place on model risk, models can be greatly influenced by assumptions and other inputs and MSRs are in a league of their own in that regard.

The Fed’s proposed adjustments to mortgage and MSR bank capital are well-intended as regulatory requirements on these assets went too far after 2008. The proposal to vary regulatory risk weights on mortgages by loan-to-value, or LTV, ratio buckets is consistent with credit risk principles as well as Basel capital standards elsewhere. And taking a second look at the 250% risk weight on MSRs is also sensible considering risk weights assigned to other risk assets such as high volatility commercial real estate that are set at 150%.

There is no doubt that an unintended consequence of significant regulatory tightening after 2008 led to the prominence of nonbank financial institutions in mortgage origination and servicing, institutions that are not regulated at the federal level for safety and soundness and depend on sources of less stable funding compared with banks. Ginnie Mae recognized this risk and imposed similar but less restrictive treatment of MSRs on nonbank capital in 2024. Reducing the potential systemic risk from nonbanks in mortgage banking by eliminating the MSR cap on bank capital is certainly laudable, but not at the expense of incentivizing the buildup of one of the riskiest assets known on bank balance sheets. A potential solution lies in maintaining a cap on MSRs at a level between that set currently and by Ginnie Mae, permitting some offset for hedging and aligning that regulatory treatment between banks and nonbanks to maintain a level playing field in mortgage banking.