The Return Nobody Announced
Collateralized Mortgage Obligations carry a reputation so damaged by the 2008 financial crisis that simply mentioning them in a portfolio meeting can clear the room. Yet a quiet reallocation is underway. Institutional allocators – pension funds, insurance companies, and a growing number of family offices – are revisiting CMO tranches not out of nostalgia, but because the yield math has started to win arguments that reputational caution used to shut down.
This is not a story about reckless risk-taking. The CMOs drawing interest today are largely agency-backed structures, meaning the underlying mortgage pools carry the implicit or explicit guarantee of government-sponsored enterprises. The appeal is straightforward: in a rate environment where investment-grade corporate bonds have compressed spreads and cash equivalents are beginning to roll off peak yields, CMO tranches are offering duration-adjusted returns that are genuinely difficult to replicate in other corners of the fixed-income market.
What CMOs Actually Are – And Why They Disappeared
A CMO pools individual mortgage loans and slices the resulting cash flows into sequential tranches, each with a different priority claim on principal and interest payments. The earliest tranches receive principal repayments first and carry shorter effective durations. Later tranches absorb prepayment risk and volatility in exchange for higher yields. The structure was originally designed to give investors precise control over duration exposure – a genuinely useful tool that got buried under the wreckage of non-agency, subprime-laden variants that collapsed spectacularly in 2007 and 2008.
The post-crisis regulatory response – Basel III capital requirements, stricter underwriting standards under Dodd-Frank, and heightened scrutiny of structured products – effectively pushed CMOs off the approved lists of most retail-facing advisors and many institutional risk committees. What remained were plain-vanilla mortgage-backed securities and the occasional agency CMO purchased by the Federal Reserve as part of its quantitative easing programs. Retail investors largely forgot the product existed. Institutional memory, though, is longer.
The distinction between then and now matters. The structures that collapsed carried credit risk layered on top of structural complexity – low-documentation loans, inflated appraisals, and originator incentives pointed entirely in the wrong direction. Agency CMOs carry none of that credit exposure. Fannie Mae, Freddie Mac, and Ginnie Mae guarantee the underlying principal and interest payments, which means the risk an investor is actually taking is prepayment risk and interest rate risk, both of which are measurable, modelable, and – depending on which tranche you hold – deliberately bounded.
The Yield Argument in Detail
The case for CMOs right now rests on a specific set of conditions that do not arrive together very often. Mortgage rates have stayed elevated relative to historical norms, which means the underlying pools backing new CMO structures contain loans originated at higher coupons. That creates real income. At the same time, elevated rates have dramatically slowed prepayment speeds – borrowers with 3% mortgages are not refinancing into 7% mortgages – which means the prepayment risk that traditionally complicated CMO modeling is substantially reduced. A tranche that might have seen its duration collapse in a low-rate environment is now behaving closer to a standard bond.
For insurance companies in particular, the asset-liability matching logic is hard to argue with. A planned amortization class tranche – one of the more predictable CMO structures, designed to maintain a defined principal payment schedule within a range of prepayment speeds – can match liability durations with a degree of precision that a bullet corporate bond cannot always offer. The spread pickup over comparable-duration Treasuries is not enormous, but it is consistent, and for an insurer trying to close a funding gap without taking on meaningful credit risk, consistent matters more than exciting.
Family offices present a slightly different picture. The allocators showing up in this market are not looking for insurance-style precision – they are looking for yield that survives a rate cut cycle. The concern is straightforward: if the Federal Reserve begins easing, short-duration instruments reprice lower almost immediately, while longer-duration corporate bonds may see spreads tighten enough to offset the rate benefit. A CMO tranche with stable prepayment characteristics sits in an interesting middle position – it benefits from lower rates in theory, but because of the embedded prepayment optionality baked into the mortgage pool, it does not extend duration the way a standard long bond does. That convexity profile, while technically negative, is actually tolerable when you are already pricing it in through a yield premium.
There is also a liquidity consideration that often goes unmentioned. Agency CMOs trade in a deep, reasonably liquid secondary market anchored by the same dealer networks that handle agency MBS. This is not the illiquidity premium story you find in interval funds absorbing demand from RIAs – CMO positions can be unwound without the gating mechanisms and redemption queues that define private credit or alternative fund structures. For allocators who need to demonstrate liquidity to their investment committees or beneficiaries, that matters.
Where the Risk Actually Lives
The honest accounting of CMO risk starts with extension risk. When rates rise unexpectedly or stay high longer than projected, prepayments slow further – and a tranche that was priced assuming a certain average life can extend significantly, leaving the holder with a longer duration instrument than intended at a moment when that is exactly the wrong thing to have. Planned amortization class tranches are specifically engineered to resist this, but support tranches – the structures that absorb prepayment volatility so PACs can maintain their schedules – can extend dramatically. Holding a support tranche without fully modeling the extension scenario is where allocators historically got into trouble, and not only in 2008.
Interest rate modeling is the other live wire. CMO pricing depends heavily on prepayment assumptions, and prepayment models are built on behavioral data – specifically, how quickly borrowers historically refinanced at various spread levels. When market conditions fall outside historical ranges, those models lose predictive power. The current environment, where mortgage rates have stayed high while home prices have also remained elevated, creates a borrower population with less refinancing incentive but also less ability to move, which suppresses prepayments through a mechanism that standard models do not always capture cleanly.
None of this means the product is unsuitable. It means the analysis has to be done correctly, which requires either internal structured products expertise or a manager with a genuine track record in agency CMOs – not just MBS broadly. The distinction between a shop that buys agency pass-throughs and one that actively manages CMO tranche selection is significant, and allocators who blur that line tend to find out why it matters during the next rate cycle.
The real question sitting underneath all of this is whether the current window – elevated coupons in the pool, muted prepayment speeds, spread pickup over Treasuries – holds long enough to justify the operational and analytical investment required to get the position right. If rates drop sharply and quickly, the prepayment picture changes, the duration math shifts, and the yield advantage compresses at the exact moment allocators are congratulating themselves on the trade.
