The Big Idea

Framing the risk of bank sales in MBS

Steven Abrahams | March 2, 2024

This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors. This material does not constitute research.

When Truist Financial Corp. recently sketched plans to sell $23 billion of market value MBS, it became the latest and largest of nearly a dozen banks planning to sell in recent months. Spreads on securities likely to get sold by Truist softened in the days after the announcement, but the possible sale raises a bigger issue for the market: further MBS sales from banks, especially if interest rates eventually fall.

The Truist announcement centered on the sale of Truist Insurance Holdings, a subsidiary and insurance brokerage, to an investor group led by several private equity firms. Truist outlined a number of benefits from the sale including improved capital ratios, a bump in book value and an opportunity to reinvest more than $10 billion in cash proceeds. It also gives Truist a chance to use the new capital generated by the sale to reposition the balance sheet. And in footnoting the projected impact of repositioning, it assumed a sale of $23 billion in market value securities—$29 billion in book value—and reinvestment of half in cash and half in securities.

On the investor call announcing the sale, Truist Chairman and CEO Bill Rogers highlighted the opportunity to improve the bank’s interest rate profile by reducing the duration of the balance sheet. Although reducing interest rate risk was not the only thing motivating Truist, it arguably has become a much more important goal for all banks in the aftermath of the collapse last year of Silicon Valley Bank. Banks have many ways to reduce rate risk, but selling securities with long duration and crystalizing a loss could be painful. Truist’s gain from the sale of the insurance broker may have offset a lot of that pain. Absent an offsetting gain, Truist may have hesitated to rebalance.

The Truist investment portfolio at the end of 2023 held more than $67 billion in securities dominated by MBS. Conventional pass-throughs made up 42% of the portfolio, agency CMOs made up 17% and Ginnie Mae pass-throughs made up another 17% (Exhibit 1). The mortgage holdings had a book yield of 2.5%, suggesting, as a crude proxy, the equivalent of a pass-through with a 2.5% coupon. A TBA UMBS 2.5% pass-through currently trades around $82-06. If Truist had bought UMBS 2.5% pass-throughs when they traded at par, current pricing would explain most of the more than 20% difference between the book value and market value of the modeled sale.

Exhibit 1: MBS dominated the Truist portfolio at the end of 2023

Note: Market value of Truist AFS portfolio holdings as of 4Q 2023.
Source: S&P Capital IQ, Santander US Capital Markets

Truist follows at least 10 other smaller banks that have announced portfolio rebalancing—Auburn National, Bank of Oklahoma, Cadence, Equity Bancshares, FVC Bankorp, First Busey, Hancock Whitney, Heartland Financial, Pacific Premier and Synovus. Most of these involve a sale of securities trading below book value, crystalizing a mark-to-market loss and offsetting it by taking a gain on a business or other asset outside the investment portfolio.

Proceeds from these smaller transactions anecdotally get reinvested in assets with much shorter duration than the securities sold. For Truist, a UMBS 2.5% pass-through currently trades with roughly a 7.9-year duration. As my colleague Tom O’Hara outlined in January, a bank could sell that long duration, reinvest in shorter assets and still pick up net interest income. Those reinvestments could include cash, CMO floaters, CMO sequential classes or pass-throughs around par. Of course, banks could also reinvest in agency CMBS and swap it to floating or take other approaches. Shortening duration and picking up book income could be very attractive for a bank.

The news from Truist and other banks points to a clear interest in reducing duration and possibly the hard-to-manage negative convexity of MBS. MBS holdings at US commercial banks current stand at $2.6 trillion, down $167 billion or 6% since SVB collapsed. And a significant share of incremental buying of MBS by banks has come in CMO floaters, CMO sequential classes or other short, structured cash flows. Notably, demand for pass-throughs seems much lighter. Declining MBS exposure and a pivot toward shorter duration in MBS also signals broad interest in reducing duration and negative convexity.

For now, significant selling of MBS by banks is constrained by an unrealized mark-to-market loss of nearly $200 billion in available-for-sale securities, most of it in MBS. Banks also have significant holdings of MBS with unrealized losses of $271 billion effectively trapped in held-to-maturity accounts. Without offsetting gains somewhere else on the balance sheet, sale of MBS out of AFS accounts seems unlikely.

But if interest rates decline and reduce the mark-to-market penalty for selling MBS, bank behavior could change. Take an extreme case to illustrate the issue. If rates declined to the point where bank pass-throughs started to return to book value, banks could sell and reinvest in shorter assets with no mark-to-market penalty. If one of the big motivations for the recent rounds of bank portfolio rebalancing has been to reduce balance sheet negative convexity, a drop in rates would reduce the cost of doing that and potentially encourage more bank sales.

Of course, it is possible that a drop in rates and a return of MBS to book value would just allow management teams to conclude that danger had passed and business could continue as it had before SVB. But regulators, bank boards and senior management have now seen the potential impact of negative convexity. Business as usual could be a hard sell.

Even a small uptick in bank sales of MBS could make a big difference for the MBS market. Banks historically rarely sell except to take opportunistic gains into quarterly income. A drop in interest rates and a rise in MBS value could ease the frictions of portfolio rebalancing.

To be clear, the argument here is not that banks would sell MBS never to return. Rather, banks have new and clear incentives to reduce interest rate risk. That could entail selling long and negatively convex MBS and reinvesting in shorter and more convex MBS or other securities. Banks still likely will need the yield, capital and liquidity advantages of MBS. It’s just the form in which banks hold it that looks likely to change, especially if rates grease the wheels for portfolio rotation.

* * *

The view in rates

The market for now has completely converged with the Fed’s view of 75 bp in likely cuts this year. The market priced for 125 bp of cuts before the January 31 FOMC, closed at 100 bp a week and a half later and now stands at 72 bp. The market prices a 68% chance of a first cut in June, although Stephen Stanley sees the Fed likely on hold through the November US elections. Implied rate volatility remains elevated. And Treasury liquidity remains relatively poor.

Other key market levels:

• Fed RRP balances closed Friday at $441 billion, down $79 billion week-over-week but part of a steady trend down since April. Treasury bills and Treasury and MBS repo almost all trade at yields above the RRP’s 5.30% rate, giving money market funds good reason to move cash out of the RRP and into these higher-yielding alternatives.
• Setting on 3-month term SOFR traded Friday at 533 bp, unchanged week-over-week but up a few basis points since the end of January.
• Further out the curve, the 2-year note closed Friday near 4.53%, down 16 bp in the last week. The 10-year note closed at 4.18%, down 7 bp in the last week.
• The Treasury yield curve closed Friday afternoon with 2s10s at -35, steeper by 9 bp In the last two weeks. The 5s30s closed Friday at 18 bp, steeper by 8 bp over the same period.
• Breakeven 10-year inflation traded Friday at 232 bp, higher by 3 bp over the last week. The 10-year real rate finished the week at 166 bp, lower by 10 bp in the last two weeks.

The view in spreads

Credit has momentum and a strong bid from insurers and mutual funds, the former often funded with annuities and the later getting strong inflows. Liquidity is likely to trend lower through 2024 as QT and relatively high rates in the front end of the curve continue. Lower liquidity is likely to keep volatility elevated, keeping pressure on spreads, leaving them a tad wider than otherwise. Nevertheless, spreads should trade stable to tighter for the foreseeable future. The fact that markets have now priced to the Fed dots also looks bullish for credit spreads.

The Bloomberg US investment grade corporate bond index OAS closed Friday at 97 bp, wider by 8 bp over the last week. Nominal par 30-year MBS spreads to the blend of 5- and 10-year Treasury yields traded Friday at 155 bp, unchanged in the last week. Par 30-year MBS TOAS closed Friday at 47 bp, also unchanged over the last week. Both nominal and option-adjusted spreads on MBS look rich. Fair value in MBS is likely closer to 70 bp, so a widening toward 70 bp looks reasonable.

The view in credit

Most investment grade corporate and most consumer sheets look relatively well protected against higher interest rates, and eventual Fed easing—even late this year—should relieve pressure from interest rate expense and falling liquidity. Fixed-rate funding has large blunted the impact of higher rates on both those corporate and consumer balance sheets, and healthy stocks of cash and liquid assets allow these balance sheets to absorb a moderate squeeze on income. Consumer balance sheets also benefit from record levels of home equity and steady gains in real income. Consumer delinquencies show no clear signs of stress. Less than 7% of investment grade debt matures in 2024, so those balance sheets have some time. But other parts of the market funded with floating debt continue to look vulnerable. Leveraged and middle market balance sheets are vulnerable. At this point, mainly ‘B-‘ loans show clear signs of cash burn. Commercial office real estate looks weak along with its mortgage debt. Credit backing public securities is showing more stress than comparable credit on bank balance sheets. As for the consumer, subprime auto borrowers and younger households borrowing on credit cards, among others, are starting to show some cracks with delinquencies rising quickly. The resumption of payments on government student loans should add to consumer credit pressure.