The best tools in investment management today will not be the best tools in three years. Portfolio managers see that evolution firsthand. They craft the strategies, move the needle, generate alpha and they do it with a toolkit they would happily swap for something better when something better appears.
The problem is what happens once you find a tool you like: the cost, the dependencies, the months of integration work and the ultimate decision to stick with what you have because changing anything means tearing half your current operating model apart.
Interoperability is the word our industry has used for that problem for well over a decade.
But today, interoperability can finally mean what it was always supposed to mean: the freedom to choose the best tool for each part of your operating model, replace it when something better comes along and introduce new technologies, including AI, without rebuilding everything around it.
Such freedom depends on having one part of the architecture that does not move: a trusted, real-time investment accounting foundation that every other system can connect to.
What is Interoperability in Investment Operations?
Interoperability is how easily different systems can work together. For example, in investment operations interoperability could define whether an order management system, a risk platform, a custodian feed, a reporting tool and an accounting engine can exchange data and trigger work in each other without custom development for every pairing.
For most of its history, this workflow has been achieved through file movements and bespoke integration projects, which is a large part of why the operating models built on it have stayed so fragmented.
But connecting more systems is not the same as creating an interoperable operating model. The real goal is to give firms the freedom to choose the best technology for each function, and replace those technologies as their needs change, while keeping the underlying portfolio record consistent.
Multiple Systems, Multiple Versions of the Truth
This concept only works if the tools being connected are working from the same underlying truth.
Connecting systems becomes difficult when each calculates its own positions, cash and valuations. Moving files and API payloads between them may make the systems connected, but it does not make the answers consistent. A best-in-class risk system, for example, is only as useful as the portfolio data feeding it. If that data comes from separate accounting environments that do not reconcile, the firm has connected best-in-class technology to conflicting versions of the truth.
This is where investment accounting becomes fundamental to interoperability. A unified accounting system of record can provide the consistent postions, cash, valuations and transaction-level data that the tools around it consume. Instead of every new application having to reconcile with multiple environments, it connects to one trusted foundation.
The Integration Tax
Anyone who has worked in operations knows the routine of a legacy integration. Onboarding a new custodian turns into weeks of custom interface work before the first position even arrives and every one of the projects you work on is a direct tax on adopting new software, with firms ending up paying for integration costs or deciding not to bother.
The monolithic all-in-one platform was sold as the way out and it does remove some of this integration tax by keeping everything inside one vendor, but it also introduces a different cost. No single vendor is the best at every layer of the investment lifecycle. A platform with an excellent OMS may have a weaker investment book of record while one with deep accounting may offer a poor compliance experience. Buying the whole stack from one place means accepting the weakest parts along with the strongest and being unable to swap them out when something better comes along.
Interoperability resolves this. When systems connect through APIs and open protocols to one accounting foundation, a firm can run the best available tool at each layer and change any of them without the integration tax.
Accounting as the Connective Foundation
It helps to trace where a portfolio manager’s numbers come from. Morning cash and position views are the baseline for every trading decision a portfolio manager makes, but relying on an overnight batch run to generate those figures leaves a massive blind spot once the market opens.
Portfolio managers increasingly want that figure during the day as well. They have been trading since the open and they want to know their cash balance now, which means every one of those hops has to run again, in the middle of the session, without waiting for an overnight cycle. On a batch platform that is a request the operating model cannot meet. A real-time accounting engine handles this automatically, updating positions, cash, prices and corporate actions the moment they arrive.
This is what makes investment accounting the connective layer of the operating model. It sits at the intersection of everything:
- Trades coming in
- Custodians confirming
- Middle office reconciliation
- Front office decision-making
- Client reporting going out
A unified, real-time accounting system of record gives all of those functions the same context at the same moment. Every tool connected is working from one set of positions and one accounting state.
Open Architecture and Composability
Modernization does not happen all at once. Firms carry existing systems that work, contracts that run for years and teams who know the current tools. Any credible modernization path has to let a firm introduce new capabilities while the existing systems keep running and change one component without touching the others.
With open APIs and modular architecture, data flows in and out of the platform through documented interfaces, so a firm can choose its own data platform, whether that is Snowflake, Databricks or something else, its own reporting vendor and its own front office tools and connect each through a standard adapter. Partnerships extend this further. FundGuard’s work with Gresham, which integrates with nearly every custodian in the market, means adding a new custodian is a configuration request and the data is there the next day, in place of a prolonged interface build.
Supporting multiple asset classes inside a single accounting system of record means adding private credit or digital assets is just a configuration change. You do not have to add on another specialist platform or break your operating model every time the investment strategy evolves.
What Clients Are Asking to Connect Today
Most implementations start at the same foundation: custodian and administrator feeds to reconcile cash and positions. Once those core balances match, market data, reference feeds and existing reporting tools plug in downstream. Middle-office settlement workflows through CTM and DTCC follow the same logic, relying on direct APIs instead of legacy file drops.
But now, firms are asking how their own AI tools reach accounting data. That changes what it means to interoperate, because an internal chat interface over portfolio data or an agent handling reconciliation triage needs governed, current access to the accounting record.
AI x Interoperability
A firm experimenting with a frontier model in the front office still needs the data story right for the rest of the organization, or the model is reasoning about positions and cash that may not be current. That is the first way AI changes interoperability.
The second is that firms are building their own AI environments, with multiple models, multiple agents and multiple technology partners, and they expect their accounting platform to support that without requiring AI activity to stay inside it.
MCP and A2A
Standards like Model Context Protocol and Agent-to-Agent communication are how that gets done technically. MCP means a client’s own AI tools can reach accounting data through a universal interface, under the same entitlements as the client’s people, while A2A means a client’s agents can work with a provider’s agents, so a reconciliation query or an audit request can run agent to agent with governance intact. Together they make the accounting platform something a firm’s AI strategy runs on, with no requirement that the strategy run inside it.
Data Access and Insight
Teams can now query accounting data directly through MCP and automated oversight tools can explain the specific drivers behind a NAV shift. While broader agent-to-agent workflows are still early in adoption, the core requirement is already clear: an AI agent is only as good as its underlying data. Give an agent outdated or fragmented feeds, and it will automate bad decisions. Give it a real-time, fully versioned accounting engine, and it has a reliable foundation to act on.
Choice, Modernization and the Ability to Grow
Picture where this is heading for the portfolio manager: they come in to work and their system already knows their book, their clients, their usual strategies and what happened in the world overnight. It presents a view of the day and a set of recommendations and the portfolio manager decides.
This way of working cannot be built on a monolithic tool, because the data has to be correct and current across every asset class and the tools around it have to be swappable as the models improve.
Interoperability is what makes that operating model possible, and it comes down to three things a firm should be able to do:
- Change components of the architecture when something better appears
- Introduce new technologies, including AI, on its own terms
- Do both without rebuilding your accounting foundation each time
To do this, your accounting system of record needs to be able to stay fixed while the rest of your operating model shifts. Build it on a clean, unified foundation, and its value grows exponentially as you bring on new asset classes, third-party software and automated agents.
Book a FundGuard Demo
If you are assessing how your operating model connects, or planning how your own AI tools and agents will reach your accounting data, book a FundGuard demo today to see all this in action.
Frequently Asked Questions
What does interoperability mean in investment operations?
Interoperability is how easily systems across the investment lifecycle exchange data and trigger work in each other without custom development for every pairing. Historically it was achieved through file movements and bespoke integrations. In a modern operating model it means systems, data, workflows and AI agents working against one trusted accounting foundation through APIs and open protocols.
Why has interoperability not solved fragmentation in the past?
Connecting systems that each calculate their own positions, cash and valuations moves data between several versions of the truth without making them agree. Interoperability reduces fragmentation only when the systems being connected operate against a single accounting system of record.
What is the integration tax?
The cost, in money, time and dependencies, of connecting a new tool to an existing operating model. Legacy integrations involve flat file generation, encryption, SFTP delivery and weeks of interface work per counterparty. That cost discourages firms from adopting better tools and locks them into what they have.
Why is investment accounting the connective foundation of the operating model?
Accounting sits at the intersection of trading, settlement, custody, middle office, front office and client reporting. A unified, real-time accounting system of record gives every function and every connected tool the same positions, cash and accounting state at the same moment.
How does AI change the interoperability requirement?
Firms are building their own AI environments with multiple models, agents and partners. Those tools need governed, current access to accounting data and increasingly need to coordinate with a provider’s agents. Standards like MCP and A2A enable that, and they depend on a real-time, versioned accounting foundation to be useful.
Does FundGuard require AI activity to stay inside its platform?
No. FundGuard supports clients’ own AI environments with governed access to trusted, real-time accounting data through MCP and A2A, so a firm can pursue its own AI strategy with its own models and partners while working against one accounting foundation.
How does FundGuard support open architecture?
Through documented APIs, standard adapters to data platforms such as Snowflake and Databricks, client choice of reporting vendor and partnerships such as Gresham for custodian connectivity. New asset classes are configured within the same system of record, so the operating model stays stable while the portfolio changes.
