US CLO issuance set a record for the second year running in 2025, with nearly $200 billion in new broadly syndicated and middle-market volume, and Deutsche Bank Research expects 2026 to keep pace. Secondary CLO trading followed the same trajectory, reaching $219 billion, up 19 percent year over year. More managers are pricing more deals, refinancing more liabilities, and holding more names across more strategies than at any point in the asset class’s history.
The operating models behind most of these desks were not built for this. Indenture compliance, hypothetical analysis, and allocation are still handled in spreadsheets designed for CLO programs a fraction of today’s size. That gap between deal flow and operational capacity is structural, and it’s the reason so many operations teams feel like they’re sprinting just to stay in place.
Where The Strain Actually Shows Up
Talk to enough heads of operations at CLO managers and the same five pressure points come up, in roughly this order:
Deal sourcing is fragmented across market data feeds, banker emails, and PDF term sheets, with no single system pulling that pipeline together. Hypothetical analysis moves too slowly to be useful, since every trade needs to clear WARF, WAS, WAC, and Diversity Score tests before, during, and after execution, and spreadsheet models rarely return an answer before the market has moved on. Allocation outcomes go untracked across funds and SMAs, which means desks often cannot say with confidence which lead banks reliably fill their orders versus which ones just circulate deal flow. There is no portfolio-level view of how trading activity is actually affecting credit quality, spread, and diversification, so building one means manually reconciling trustee reports on top of everything else. And the compliance burden keeps compounding, because every new CLO brings its own OC and IC tests, concentration limits, and CCC bucket rules that someone has to track by hand.
None of these are new problems. What’s different now is the volume running through them, enough to turn a workaround that used to hold up into a real operational risk.
What the Firms Handling This Well Are Doing Differently
A pattern shows up consistently among managers who have scaled without their operations function buckling under the weight of it.
They consolidate their data workflows onto a single platform before layering on new strategies, so they aren’t stitching together five systems’ worth of mess after the fact. Specialized tools, like expense allocation or wire management, get added in a deliberate sequence tied to real triggers in the business. And the internal case for investment gets built early enough that it’s a considered decision, made with time to think it through.
The technology side of this matters, but it is not the whole story. Getting stakeholders aligned, from portfolio management to compliance to the desk itself, is usually the harder part of a change this size, and it is often the part that gets the least planning time.
Questions Worth Asking Before You Build the Business Case
What actually triggers the need for a unified data platform in a CLO business? Usually it’s an accumulation, not one dramatic event: a hypothetical that took too long to run before a bid expired, a reconciliation break that took days to trace, an allocation dispute nobody could fully document. By the time those add up, the cost of staying manual is already higher than the cost of switching.
How do managers decide when to add specialized modules like expense allocation or wire manager? The clearest signal is when a workflow that used to be occasional becomes constant. A firm running one or two CLOs can often manage expense allocation by hand. A firm running a dozen, across multiple strategies, usually cannot, and trying to force it tends to show up first in slower month-end closes and then in audit findings.
What does ROI look like for CLO operations technology, in practice? The most defensible numbers tend to come from reconciliation and compliance testing, since both are measurable before and after. Firms that automate this work typically see reconciliation runtime drop from hours to minutes, and reconciliation accuracy improve by as much as 90 percent. Those are the kinds of figures that hold up well in a business case review, since finance and audit can trace them directly to before-and-after data.
Does AI change any of this in the near term? It’s changing the compliance and document side faster than most operations teams expected. Covenant extraction from credit documents, automatic exception flagging, and parsing a credit agreement to create a security record on upload, capabilities IVP’s own semantic layer and AI agents already handle inside client environments, are moving from pilot projects into production use this year.
Where This Conversation Continues
This is the same architecture behind IVP’s own CLO solution: what-if analysis, indenture compliance, trustee reconciliation, and data management, brought into one workflow instead of four separate ones.
These are the exact questions we will be working through on October 1 in New York, at the next IVP Buy-Side Dinner. Dwayne Weston, Head of Operations at AGL Credit Management, will walk a small group of senior buy-side operations and technology leaders through how his firm actually made these decisions: what triggered the investment, what the implementation looked like in practice, and what he would tell his past self to do differently.
It’s an invitation-only evening, built around a candid, practitioner-level exchange with a small group of peers. If scaling your own operating model is on your desk right now, this is worth being in the room for.
Invitation only. Request An Invite: Building a CLO Operating Model That Scales: A Practitioner’s Perspective
Related IVP resources
