

Straight-through processing for private markets
Author:

Straight-through processing is within reach for private markets: a record that reflects every document on the day it arrives, and decisions made on trusted data. Intelligent document processing (IDP) for private markets documents is becoming routine, so the difference lies in the governance around it: validation of every value, a traceable source for each one, control of model change, and security. This leaves a decision: who maintains the record, and who handles the documents along the way. In our view, the allocator keeps the record and the platform that holds it carries the processing.
Your private markets record, current the day documents arrive
Sovereign wealth funds, pension schemes, insurers, endowments, and foundations now commit a third to half of their portfolios to private markets. The reporting behind those books has scarcely changed in decades: documents still arrive as PDFs, each manager, lender, and operator in its own format. An analyst works out which fund and entity a capital call notice concerns, keys the amount, the date, and the fee split into the book of record, and someone else checks the keying. The book is marked on data that is late, partial, and held in documents instead of a database. That is now solvable, so the record can be current the day each document arrives. A current record provides three things: it keeps exposure, liquidity, and look-through up to date, gives risk and portfolio analytics structured data, and replaces a spreadsheet per team with one shared record. 1
- Exposure, liquidity, and look-through that update as documents arrive
Managers still report quarterly, and the book then updates the day each document arrives: the adjusted net asset value (NAV) carried forward between reporting dates, the underlying companies, loans, and assets visible rather than inferred, and every cash flow in the liquidity forecast. For an allocator pursuing a Total Portfolio Approach (TPA), that brings the private book inside the sector, geography, and factor views applied to everything else, with exposures, liquidity, and risk budgets managed as one book. - Structured data for risk and portfolio analytics
Structured data is what risk, exposure, and portfolio-construction tooling for private markets needs, and it is what lets an analytics layer form as it has in public markets. Techniques that carry a fund's reported NAV forward between reporting dates, adjusted for cash flows and currency, can strike an adjusted NAV every day. They are only as current as the last document someone keyed, so they often sit unused. - One shared record instead of a spreadsheet per team
The documents are read, but the knowledge rarely lasts. A deal team will mine a quarterly report closely for one decision, capture what it needs in a spreadsheet, and move on; the knowledge evaporates when the spreadsheet is closed. The same document is then read again by other teams for other purposes, and none of it persists in a shared record. Capturing it once replaces several readings.
Large language models make any format routine to handle
A current record is achievable more efficiently as large language models (LLMs) can extract data from any format a manager sends. The gains above have been out of reach for years, because rules-based code could not process documents written for human eyes. Manager reporting carries much of its meaning in narrative, layout, and context rather than in fields a machine can address directly. Decades of Optical Character Recognition (OCR), templates, and rules broke on the next template, and the machine-learning extractors that followed broke on the next format because each needed new training data. LLMs change those economics: IDP has moved from building rules to building training data to building context. Comprehension arrives pre-trained and the allocator's own schema directs extraction. An unseen format becomes routine to handle, and the cost of extraction is falling toward the cost of the computation. A commercial LLM can already extract the data from a capital call notice. The task now is to give it the right context, such as fund structures, the investor register, and the institution's own data model, inside the controls described above. The next question is whether the record can be trusted.
Governance is what lets an allocator trust every value
With any document now able to become data, what decides the quality of the record is whether every value can be trusted, and governance is what delivers that. Governance covers what happens to each value between the document and the book of record. Once a machine extracts a fee split, someone has to show that it is right, trace it to its source, explain it to an auditor, and keep it right as the models change. That accountability sits with the allocator, because the number sits in the allocator's book.
In practice, the system validates every value, because confidence scores alone cannot catch a misread fee split. A value is flagged when the components do not sum, when the amount does not reconcile to the commitment, when a field disagrees with the prior notice, or when the model's own confidence is low, and a flagged value goes to a reviewer. The reviewer sees the exact span the model read, corrects it, and the correction joins the evaluation set against which the system's quality is measured over time.
Five controls decide whether the record can be trusted: validation of every value, an evaluation set that measures quality over time, a traceable source for every value, change control as models update every few months, and security and compliance controls over the documents. Together, these controls let the allocator show an auditor or a risk committee that each number in the book is right and where it came from.
Governance applies whether or not a standard arrives. Standards are how public markets reached straight-through processing, and their history shows why processing documents at the allocator's end is the more direct route for private markets. Public markets standardized because the law required it, and the law followed a crisis. Between 1968 and 1970 paper overwhelmed Wall Street 2, and the response was a depository, universal identifiers, and book-entry settlement, which Congress made compulsory in 1975 3. Every later standard came through those same central institutions, because that is where a regulator can impose one. Equal access to information forces agreement on what it looks like. Private markets rest on the opposite premise.
Why you can’t wait for standards in private markets
Private markets are bespoke by design and private by nature. Their returns come in large part from privileged insight and negotiated access, and investors who pay for an informational edge have every reason to keep what is shared with everyone to a minimum. That same nature keeps standards from taking hold, for two reasons: the managers who would have to adopt them have little incentive to, and new strategies keep arriving before any standard can describe them.
The private nature does not incentivize standards
Funds are sold under exemptions that keep information within a closed circle of sophisticated investors, and the partnership agreement and its side letters set what a manager must report. The one regulatory attempt to compel standard reporting, the SEC's rules of 2023, survived 10 months 4. Without regulatory compulsion, and none is in sight, private markets will keep reporting on bespoke terms.
The Institutional Limited Partners Association (ILPA) updated its Reporting and Performance Templates in January 2025, the first revision since 2016 5, to align with the SEC's Private Fund Adviser rules. The Fifth Circuit, a US federal appeals court, had already vacated those rules in June 2024, so adoption depends on voluntary endorsement. That is the first reason standards trail: adoption fades as the regulatory leverage behind it fades, and managers bear the cost of standardization while allocators receive the benefit.
Standards cannot keep up with new strategies
The second reason is that even with stable leverage, standards would trail strategies, because every novel strategy arrives before the reporting standard that describes it. The first European guidelines, issued in 2000 by what is now Invest Europe, were written for private equity, and the later growth of private debt forced far more granular reporting 6. The pattern repeats: exceptions are handled ad hoc, a few leading investors impose their own requirements, and a standard forms roughly a decade later. Standards have raised the floor and will likely keep raising it. Bargaining power and innovation make it unlikely they close the gap on their own, so allocators need controls that work whatever format arrives. On the flip side, there's no need to wait for a shared template, it’s possible to use the data whatever format it arrives in and keep a governed, current record.
Three routes to a governed private markets record
With standards trailing for private markets, there's a decision to be made on who processes the documents and who keeps the record. Three routes put IDP to work, and the five controls above sort them. A pooled service processes documents for many allocators at once, an allocator builds its own tooling on top of its platform, or IDP is embedded in the platform that holds the record.
- The pooled model
The pooled model suited the previous generation of technology, when the largest training corpus won, and it retains real strengths in portal collection, exception handling, and benchmark data. But LLMs make format coverage nearly free, so what remains scarce is context, and context belongs to the allocator, along with the ground truth for evaluation, the confidentiality of its managers' reporting, and the audit trail of every value. Smaller allocators have long outsourced private markets operations altogether, and those without a book-of-record platform of their own may reasonably keep using pooled services for collection, exceptions, and staffing, so providers keep a role. - Building tooling on top of the platform
Building on top keeps the context in-house, and the lifecycle with it: models change every few months, the evaluation framework has to follow them, and the controls that satisfy compliance, risk, and security requirements become a standing responsibility for the allocator's own teams. - The platform holds the record for the allocator to maintain
That leaves IDP embedded in the platform that holds the record as the one route that keeps the context with the allocator without handing it the lifecycle. The allocator can maintain the record in-house without building the tooling in-house. The context is the allocator's and cannot be bought; the machinery around it can be embedded in the platform that maintains the record. Model updates, the evaluation framework, and the controls that support compliance, risk, and security requirements then ship with the platform, so the allocator keeps the record and the documents without carrying that lifecycle itself. The platform processes each allocator’s documents against that allocator’s own parameters (schema, validation rules, thresholds, and entity data), and client data is segregated.
The platform that holds the record also holds the rest of the portfolio. When private markets documents are processed where the listed book already lives, the controls draw on the same entity data, Security Master, and ground truth as every other asset. Validated values land in one investment book of record with no reconciliation to a separate system, so the private book joins the total portfolio view on the day each document arrives.
A current, governed record brings private markets into the total portfolio view
With the platform carrying the processing, keeping the record pays off in two places. For CIOs, a record maintained by the allocator means a private markets portfolio managed on a current record instead of last quarter's reconciled view. That is what lets the Total Portfolio Approach include the part of the portfolio that has so far resisted it. For COOs and CFOs, it means an operations budget that holds steady as commitments grow, and a confidential document estate governed by the institution's own parameters and policies.
"You decide where the record is maintained and who processes the documents on the way. To start, trace one manager's capital call notices from arrival to your book of record and see where automation fails."
FAQs
References
1 Scope: this article concerns the reporting that flows from managers to their investors: positions, cash flows, and valuations of private funds, and borrower reporting in direct lending. Valuation methodology (International Private Equity and Venture Capital Valuation (IPEV) Guidelines), sector benchmarks and standards (European Association for Investors in Non-Listed Real Estate Vehicles (INREV), National Council of Real Estate Investment Fiduciaries (NCREIF)), regulatory reporting (Solvency II), and environmental, social, and governance (ESG) reporting each have their own standardization history and are outside its scope, as are hedge funds.
2 The Paperwork Crisis, Brown Brothers Harriman, 200 Years of Partnership. https://www.bbh.com/us/en/bbh-who-we-are/our-story/200-years-of-partnership/the-paperwork-crisis.html
3 Regulatory Impact on Efficiency on the U.S. Clearance and Settlement System, U.S. Securities and Exchange Commission. The paper describes the 1975 amendments to the Securities Exchange Act and the national clearance and settlement system they authorized. https://www.sec.gov/news/speech/spch021004leb.htm
4 National Association of Private Fund Managers v. SEC, No. 23-60471, US Court of Appeals for the Fifth Circuit, June 5, 2024. The court vacated the SEC’s Private Fund Adviser Rules (adopted August 2023) in full. See Fifth Circuit Vacates the SEC’s Private Funds Rule, Sullivan & Cromwell, June 2024. https://www.sullcrom.com/insights/memo/2024/June/Fifth-Circuit-Vacates-SECs-Private-Funds-Rule
5 ILPA Releases Updated Reporting Template and New Performance Template for Industry Adoption, Institutional Limited Partners Association, January 2025. This was the first update to the Reporting Template since 2016; adoption is by voluntary endorsement. The updated Reporting Template applies from Q1 2026 to funds still in their investment period and to funds commencing operations on or after January 1, 2026. https://ilpa.org/news/ilpa-releases-updated-reporting-template-and-new-performance-template-for-industry-adoption/
6 Investor Reporting Guidelines (2024 edition), Invest Europe, 2024. The edition builds on the 2018 and 2015 editions and the original EVCA Reporting Guidelines first issued in 2000. The 2018 update introduced tailored sections for real estate, private debt, infrastructure, venture capital, fund of funds, and secondaries, with the standard catching up with the strategies roughly a decade behind each. https://www.investeurope.eu/investor-reporting-guidelines/
Related content

