The Plumbing Is the Product: BNY Says Alpha Is Leaving the Back Office

What's New

Gurjit Jagpal, Brian Ruane and Carolyn Weinberg of BNY argue that private markets have reached a structural inflection point where the binding constraint is no longer capital formation but the operational machinery underneath it, in a paper published August 6 titled A New Architecture: Accelerating Private Market Growth. Their case is that AUM has expanded while the systems supporting it have not, leaving asset servicing, cash movement, collateral pledging and ownership records fragmented across intermediaries and disconnected platforms. The proposed remedy is a four pillar architecture built on persistent identifiers, structured terms and programmable cash flow logic, settlement certainty, and collateral utility. What separates this from the standard tokenization pitch is the concession buried in the foreword: BNY states that in the next phase, the source of alpha will come from origination, underwriting and investment insight rather than operational complexity. That is a custodian telling its clients that the operational moat many of them quietly rely on is scheduled for demolition.

Why It Matters

The paper is a statement about convergence more than about digitization. BNY observes that investors have begun treating private and public assets as a single portfolio, which collapses the conventional divide and forces the infrastructure question into the open. If private credit is going to sit beside public fixed income in the same allocation, the same book and eventually the same collateral pool, it needs identifiers, evaluated pricing and settlement certainty that public markets have had for forty years. For GPs and LPs that turns operational capability from a cost line into a distribution enabler, and for technology platforms it means the durable position is not workflow automation but ownership of the asset identity and terms layer that every downstream process depends on.

Big Picture Drivers

• The wealth channel is the forcing function, not tokenization. Evergreen and semiliquid structures require continuous subscription, redemption and valuation cycles that closed end operating models were never designed to run, and the volume arriving through advisor channels is what makes the manual approach fail rather than any technological inevitability.

• Private credit has become a category rather than a strategy. BNY describes the shift from a narrow alternative allocation into a financing category spanning corporate loans, asset backed lending, real estate debt and infrastructure debt, and breadth of that kind makes document driven bespoke servicing untenable at scale.

• The real data problem is legal, not numerical. The economics of a private credit asset live inside customized agreements, side letters, amendments and servicing records, which is why consistent monitoring, valuation and mobilization remain hard even for firms with sophisticated data functions. BNY's answer is extraction of legal and economic terms into structured data that can drive automated behavior.

• Settlement latency operates as a liquidity tax. Loan transfers can take up to 30 business days to complete, with cash wires that are difficult to match and reconcile, and every secondary market ambition and semiliquid redemption window is priced against that number whether or not anyone says so.

• Collateral utility is where the commercial prize sits. The fourth pillar depends entirely on the first three, because an asset can only be pledged once it can be identified and its cash flows understood, and BNY frames private assets entering the collateral ecosystem as a driver of secondary activity and asset class growth in its own right.

• Interoperability determines whether any of this compounds. The paper is unusually candid that without connectivity across blockchains, custody platforms, fund administration systems, triparty venues and analytics tools, digital infrastructure simply reproduces today's fragmentation in newer silos.

By The Numbers

• $500 billion in evergreen and semiliquid AUM as of September 2025, more than double the 2022 level, which is the wedge that breaks the closed end operating model.

• 30 business days as the upper bound on loan transfer completion today, against near instant delivery versus payment as the stated target, which is the entire gap the architecture is meant to close.

• More than $2 trillion in private credit AUM in 2026 approaching $4 trillion by 2030 on Moody's projections, growth that cannot be absorbed by adding operations headcount.

• More than $200 billion in nontraded perpetual life BDCs, up from effectively zero in 2021, a vehicle class that did not exist five years ago and now carries daily operational obligations.

• Four pillars in one dependency chain, since identifiers enable structured terms, structured terms enable settlement certainty, and settlement certainty enables collateral utility, which means firms sequencing these out of order will build expensive dead ends.

• Two named technologies carry the entire thesis, tokenization for transfer and settlement and generative AI for document extraction, and the paper is explicit that neither delivers much without the other.

Key Trends to Watch

• The credit agreement becomes the golden source. BNY says it is exploring AI powered extraction of unstructured credit agreement data to drive smart contracts encoding covenant hierarchies, financing dependencies and payment behavior, which relocates the competitive question from who stores the position to who parses the document.

• Payment form stops being a footnote. The paper explicitly lists payments in kind and revolver activity among the behaviors that must be translated into rules governing how a tokenized loan operates, and once PIK status is machine readable rather than buried in servicing records, the visibility problem in private credit changes character entirely.

• Whole portfolio views become the commercial battleground. Identifiers, reliable settlement and evaluated pricing for private instruments are framed as the combination that lets an allocator see private and public exposure with the same precision, which is the capability every platform serving this market will claim within eighteen months and few will actually hold.

• Custodians move up the stack and collide with everyone. A custodian defining the asset identity and terms layer is competing simultaneously with fund administrators, data vendors and portfolio management platforms, so the partnership announcements that follow should be read as territory negotiations.

• Tokenization gets judged on financing, not pilots. The measure that matters over the next 12 to 24 months is whether tokenized private assets can be pledged and financed against tokenized cash in production, because pilot volume has never been the constraint.

The Wrap

The most revealing thing in this paper is what BNY gives away. A firm that earns fees on operational complexity is arguing that operational complexity should stop conferring advantage, which suggests the incumbents now believe the scale curve is bending past what manual processes can absorb. The framework itself is sound and the sequencing is right, but nothing in it is achievable by any single participant, which is why interoperability gets its own section and why the honest read is that this is a coordination problem wearing an architecture diagram. For technology providers serving this market, the exposed seam is the asset identity and terms layer, because valuation, servicing, settlement and collateral are all downstream of knowing precisely what the instrument is and how it behaves. Platforms that treat private markets as a reporting problem will end up integrating into someone else's architecture.