Zooming In and Out: Operating ABF Portfolios at Scale
April 27, 2026
Read Time: 4 minutes
Authored by: Juhi Ghosh
Operations & Growth
Private Markets
Authored By
Juhi Ghosh
Juhi Ghosh is a senior product and technology leader with 20 years of experience building and scaling financial platforms. As Senior Vice President of Product at Arcesium, she founded and led the development of the firm’s UBOR platform — Arcesium’s core investment and portfolio accounting engine — taking it from inception to a mission-critical system processing millions of positions and trades per day.
Juhi’s expertise spans product strategy, enterprise platform architecture, complex financial domain modeling, and large-scale engineering execution in highly regulated environments. She specializes in translating complex financial infrastructure into durable, scalable platforms that drive long-term business value.
Juhi holds a Master of Science in Management Information Systems from Temple University and a Bachelor’s degree in Computer Engineering from Pune University.
As asset-based finance (ABF) portfolios grow, complexity increases. More loans, more counterparties, more lifecycle events, and more data from various sources raise the bar. It’s growing as institutional asset managers and owners ramp up their participation in ABF via consumer loans, credit cards, buy now, pay later (BNPL) products, and other forms of credit. These can dramatically increase loan volume and operational workloads.
Much of this complexity is framed as a debate over loan-level versus aggregated data. Clients often ask us questions about which one is right, which one is safer, and which one scales. Those questions miss the larger point, however, because there’s no single correct level for running an ABF portfolio. The real requirement is the ability to choose the right level for the situation and to change that choice as the portfolio evolves. And the answer depends on who is using the data to answer what question.
A good analogy is planning a cross-country road trip. Sometimes you want the big picture: highways, major cities, the whole itinerary. Other times, you need street-level details because you hit traffic or missed an exit. You don’t choose one map and stick with it. You zoom in and out based on what you need to know in the moment.
Use loan-level detail when it affects decisions
In many situations, loan-level management is the right choice. There’s no reason and a lot of risk in abstracting loan details when individual loans are economically significant or when their behavior materially impacts portfolio outcomes. That is commonly the case with large debt, from commercial loans to jumbo mortgages.
For example, if delinquency ticks up or cash collections fall short, loan-level precision lets you see whether the issue is concentrated in a particular originator pool, geography, vintage, or collateral type. If outcomes often come from discrete events, you need clarity on modifications, extensions, charge-offs, repurchases, covenant triggers, or collateral substitutions. Loan-level detail lets you trace the impact to the loans where the event occurred. Many ABF portfolios should, and do, operate this way.
Loan-level visibility also feels rigorous. Auditors like it. Investors find it reassuring. The problem is that operational costs scale with the portfolio. More loans mean more data to review, more exceptions to chase, and more attention spread thin. At some point, clarity becomes noise. Teams chase minor tape discrepancies and one-off exceptions, while bigger problems with a specific originator pool or segment go unnoticed.
Loan-level operation stops being the right default when you want to scale from 5,000 to 50,000 loans or expand into consumer assets.
Aggregate when volume makes loan-level work unusable
Aggregation groups loans together so teams can monitor and act on the portfolio at the proper level. That becomes essential in high-volume ABF strategies, where a single portfolio may include tens or hundreds of thousands of small consumer receivables, such as credit card balances, auto loans, or BNPL accounts. In those settings, loan-by-loan oversight quickly turns into noise. These books generate constant lifecycle activity: payments, delinquencies, modifications, charge-offs, and repurchases, often delivered through daily tape updates and servicing files from multiple counterparties.
Firms that are used to loan-level detail can sometimes see aggregation as a loss of control or a shortcut, but that’s the wrong way to think about it. It’s simply a different operating view of the same underlying data.
The benefit of aggregation comes from allowing portfolio managers to run the book by focusing on segments that drive behavior, such as originator pools, vintages, or collateral types. Choosing aggregation does not mean loan-level data disappears. It means day-to-day oversight happens at the cohort level, while loan-level traceability remains available when exceptions, lifecycle events, or reconciliation breaks demand a closer look.
Focus on the dimensions that drive behavior
Aggregation is only useful when the grouping reflects how loans behave and how the portfolio is managed. The most informative lenses tend to be tied to economics and operations.
For example, collateral behavior can show where risk is building, including advance-rate volatility, turnover, and liquidation patterns. Structural features can reveal where cash flows and downside protection differ, including amortization type, covenants, triggers, and waterfall mechanics. Servicer patterns can expose operational risk, including reporting timeliness, variance versus expectations, and exception rates. Sponsor behavior matters too, especially under stress. These views help portfolio managers see what is changing.
Keep diagnostic access intact
Aggregation helps run the book day-to-day at scale. Diagnostic access is a different topic. You can't just aggregate and walk away. Even when a portfolio runs in aggregate, operations teams need the ability to drill into detail to prove what sits underneath the rollup when something breaks.
These situations happen constantly. A servicer might send a file with a few hundred loans marked as delinquent. It causes a sudden hit to your NAV, but you can’t immediately determine if the issue comes from processing or the underlying loans. Which do you trust? If you can't drill from the rollup into those loans and explain whether the picture is correct, you’re managing a black box.
What’s needed is the ability to trace an aggregated metric back to the loans and lifecycle events that drive it, isolate what changed, and explain why. It also includes controls that make the investigation repeatable, so breaks get resolved quickly and counterparties stay confident in the numbers.
Build infrastructure that preserves trust
Many data platforms push firms towards loan-level workflows that collapse under volume or aggregated reporting that looks clean until the first break. Their design forces different teams to compensate with shadow spreadsheets, one-off reconciliations, and manual workarounds to bridge the gap between rollups and reality.
That is why infrastructure matters more than features. It must provide a data foundation that can take daily loan tapes and servicing activity, standardize them, and keep rollups tied to the underlying loans as the book changes. Achieving that flexibility in zooming in and out means being able to do five things:
Normalize inputs into a consistent loan schema across counterparties
Validate files and loans so schema breaks, missing fields, duplicates, and out-of-tolerance values don’t flow downstream
Process lifecycle updates so balances, statuses, and cash expectations stay current
Route exceptions into clear queues with ownership and resolution tracking
Preserve lineage so every rollup can be traced back to source files and loan-level drivers
That is what separates scalable ABF operations from portfolios held together by spreadsheets and after-the-fact explanations. Operational maturity shows up when you can explain the rollup when it stops matching the outside world.
Design for the exceptions you haven’t seen yet
If current issuance and private credit forecasts are any guide, the volume of loans flowing through ABF will keep rising, and the operational risk of losing diagnostic access will rise as the market approaches a potential $10 trillion.
More firms are moving into consumer debt. Portfolios are getting larger. Loan populations are getting more heterogeneous. More counterparties sit in the chain, and more daily files and lifecycle events flow through the system.
In that environment, aggregation becomes unavoidable. Without preserving data lineage and visibility for diagnostic access, the risk is you limit your ability to participate in the phenomenal growth of ABF.
“Today, it is estimated that non-bank ABF lending activities combine into a $6 trillion+ global market, exceeding the size of the direct lending ($1.2 trillion) and private equity ($6.1 trillion) markets. Furthermore, we expect the private ABF market will continue to grow to nearly $10 trillion by 2028, spurred by further bank regulation in both the US and Europe.”— AB CarVal, Asset-Based Financei
