Plaid for Financial Advisors

Plaid is the plumbing behind most consumer finance apps, it is preparing an IPO, and it keeps coming up in advisor conversations about held-away visibility. Here is what it actually is for an advisory practice, without the vendor gloss.

What Plaid is, in one paragraph

Plaid is connection infrastructure, not an advisor product. An app built on Plaid lets a person link their financial accounts: the client clicks through to their institution's own login (OAuth where supported), approves access, and the institution issues a revocable token. The app then receives read-only data on an ongoing basis: balances, transactions, and for investment accounts, holdings and positions. More than 12,000 institutions connect. Nobody in the chain gains the ability to trade or move money, and where OAuth is supported nobody stores a password.

Advisors encounter Plaid through products built on it. You do not buy Plaid at an RIA; you buy software whose data layer is Plaid.

The safety question, answered by mechanism

The 2024-2025 custodian crackdown, most visible in the Pontera fight, targeted a specific practice: third parties storing client login credentials with trading capability attached. Fidelity's stated objection named credential storage and third-party trading power explicitly.

Token-based read-only connections are the practice custodians left standing, and the direction open banking regulation pushes. The distinction is worth internalizing because clients will ask, and "the institution issues a revocable token through its own login page, nothing can trade, and you can kill the link anytime" is both true and reassuring in a way "we handle it securely" never is.

For the custody rule, the posture matters the same way: visibility without any power to move assets is a categorically different thing from held credentials. Not legal advice; run your specific tooling past compliance counsel.

What it covers, verified

We checked Plaid's production institution catalog directly in late August 2026 (we build on it, so this is first-hand rather than quoted from marketing). Present and connectable: Schwab (OAuth), Fidelity including Wealthscape-side access, Pershing NetXInvestor, Altruist (OAuth), Raymond James, LPL, Edward Jones, and Interactive Brokers, plus the workplace layer that matters for held-away work: Empower, Vanguard retirement plans, and Morgan Stanley Shareworks / StockPlan Connect.

That list is the interesting part for advisors: the custodial logins and the retirement-plan logins, the two account families that decide whether whole-book visibility is real, are reachable. Coverage changes continuously, so re-verify any institution that matters to your book at evaluation time.

What it does not cover

The honest gaps, which are exactly where the enterprise aggregators earn their pricing:

• Annuity carriers and insurance products: thin coverage, and the data that exists is often balance-only.
• Small TAMPs, trust platforms, and odd recordkeepers: the long tail ByAllAccounts and Yodlee spent two decades wiring.
• Granularity variance: some investment connections return full holdings and transaction history; others return less. Test with real accounts, not the institution count.

The full landscape comparison, including the ByAllAccounts situation and Akoya's institution-owned model, is in our aggregation field guide.

The practical read for an RIA

If your clients' held-away assets are 401(k)s, equity comp, and mainstream brokerage accounts, which is the balance sheet of most working professionals, Plaid-based tooling covers the accounts that matter, on the connection mechanism that survives custodian policy. If your book is heavy in annuities and trust accounts at small institutions, enterprise aggregation still earns its quote. Many firms will land on both: Plaid rails for the mainstream accounts, patched with statements for the corners.

And either way, the data layer is only half the problem. A connected view shows the client's accounts; it does not tell you whose concentration crossed a line or what changed before the 2pm meeting. That analysis layer is the part the stack still lacks.

Where Helm fits

Helm Terminal runs on Plaid's read-only rails today for individual investors, which is why we could check the catalog first-hand. The advisor research we are running asks what those same rails look like as a practice tool: the whole book, custodied and held-away clearly separated, with intelligence on top and a read-only MCP layer for the AI tools a firm already uses. Research phase, built with practicing advisors, nothing for sale. If held-away coverage questions are live at your firm, we want twenty minutes of your experience.