The capital call nobody can verify: making drawdown notices cryptographically real
Private capital moves more money per instruction than almost any other business process, and authenticates those instructions worse than a consumer bank authenticates a card payment. The instrument is a PDF and the deadline is built into it.
How does a limited partner verify a capital call notice?
By comparing the wire instructions to a previous notice and, if something looks odd, calling a number in the footer. A capital call arrives as a PDF attached to an email instructing a very large transfer by a deadline, and neither check tests whether the notice came from the general partner.
- Capital calls are documents, not transactions. There is no shared infrastructure between GP, administrator and LP, so every LP verifies independently through channels the attacker can also reach.
- Deadline pressure is a property of the instrument itself, which is what makes it an unusually effective fraud vehicle.
- A GP-signed call receipt distributed with the notice removes the callback from the critical path entirely.
Part of Private capital and fund identity
The transaction, described as a protocol
A general partner needs capital. An administrator computes each limited partner's share under the partnership agreement. Notices are generated and emailed. Each LP's treasury team reads the notice, verifies it by some means, and wires.
Described that way, the weaknesses are obvious to anyone who has designed a payment protocol. The instruction is transmitted in-band with the identity claim. The verification channel is supplied by the instruction. There is no shared state between the parties. And the protocol has a built-in deadline that penalises verification latency — under most partnership agreements, failing to fund a call carries severe consequences including forfeiture provisions.
Why LPs cannot verify well
An institutional LP may hold commitments across sixty funds. Each fund has its own administrator, its own notice format, its own contacts, and its own portal if it has one at all.
| Verification method | Why it fails or is unavailable |
|---|---|
| Compare wire instructions to a prior call | Defeated by an attacker who has the prior notice, which came from the same compromised inbox |
| Call the administrator | The number may be from the forged packet; the correct number may not be current |
| Log into the LP portal | Available for some funds; the notice may be the only artefact for others |
| Confirm with the GP directly | Relationship-dependent and does not scale to sixty funds and quarterly calls |
What 2026 demonstrated
Private equity firms experienced documented wire fraud losses from email compromise, with fraudulent capital-call notices and spoofed wire instructions identified among the most damaging attack types against funds. Separately, vishing campaigns in 2026 funnelled staff at large private equity firms into dozens of spoofed authentication sites, harvesting credentials in real time.
Those two facts belong together. Credential harvesting produces the inbox access; inbox access produces the convincing capital call.
The Capital Call Receipt
One artefact, generated with the notice, verifiable by every LP independently.
{
"type": "manav-stmt/1",
"action": "capital_call",
"render": [
"Fund: [name] Vehicle: [entity]",
"Limited partner: [name] Commitment ref: [id]",
"Call number: [n] Call amount: [amount] [currency]",
"Cumulative drawn: [amount] of [commitment]",
"Due date: [date]",
"Beneficiary: [full account name]",
"Account: [full account number] Routing: [full code]",
"Intermediary: [details, where applicable]"
],
"issuer": "[GP authorised signatory, credential assertion]"
}Three properties matter. The account details are rendered in full and unmasked, because masking hides the characters an attacker alters. The cumulative position is included, so a call that does not reconcile against the LP's own records is visible. And verification is offline against a published key, so no LP has to reach anyone.
Where the signature belongs
A design question worth resolving explicitly: the administrator generates the notice, but the GP is the party whose authority is being asserted.
The defensible arrangement is that the GP's authorised signatory signs, under a delegation the GP issued, with the administrator operating the mechanics. That way an administrator compromise does not produce valid calls, which is the failure mode most worth designing against — administrators serve many funds and are correspondingly attractive targets.
Getting it into the documents
This works bilaterally before it works industry-wide, and the place it enters is the side letter. Language along these lines, adapted by counsel:
Each drawdown notice delivered to the Limited Partner shall be accompanied by a cryptographic receipt issued under the General Partner's published verification key, binding the fund, the Limited Partner, the call amount, the due date and the full beneficiary account details. The Limited Partner shall be under no obligation to fund a drawdown notice for which a valid receipt is not provided.
That last clause is the operative one. It converts the receipt from a courtesy into a condition, and it gives the LP's treasury team a defensible basis for not wiring against an unverified notice — which today they lack, because the partnership agreement obliges them to fund.
What this does not address
It does not help with a genuine call whose funds are stolen after arrival, which is a banking problem. It does not prevent a compromised GP from issuing a legitimate-looking call with altered details, which requires the signing credential to be compromised as well. And it does nothing for funds that decline to participate, which is why the side letter matters more than the technology.
Why the deadline is part of the attack
Partnership agreements attach consequences to late funding, up to forfeiture of interest. That pressure is designed to make LPs pay promptly, and it works equally well on a fraudulent notice. An attacker who knows a fund's call cadence does not need urgency to be plausible — the document supplies it.
| Check | Strength |
|---|---|
| Notice formatting and letterhead | Weak — prior notices are the template |
| Comparison to previous wire instructions | Moderate — and legitimate changes happen |
| Call the number on the notice | Weak — the number is in the document |
| Call a number from the subscription documents | Strong, and rarely done under deadline |
| Verify a GP signature against a published key | Strong, and instant |
Objections and honest limits
“Our LPs know us well.” Which is what makes a convincing forgery effective. Familiarity supplies the credibility the attacker would otherwise have to manufacture.
“This needs every LP to adopt something.” No. A GP can sign notices unilaterally. LPs who can verify do; the rest receive an unchanged PDF.
What a GP can do alone
- Publish a key at a stable location. Referenced in the subscription documents, so LPs know where to look.
- Sign the notice contents. Fund, LP, amount, due date and the full remittance details.
- Flag banking changes explicitly. A changed account between calls is the single highest-signal event.
- Tell LPs the rule. No valid signature, no wire — regardless of the deadline.
Terms used here
- Capital call
- A notice requiring a limited partner to fund part of their committed capital by a stated date.
- Drawdown notice
- The document carrying the call, including the amount and remittance instructions.
- Default provision
- The partnership agreement term penalising late funding, which supplies the time pressure an attacker relies on.
Frequently asked questions
What if the GP will not issue receipts? Then the LP knows that, and can price it. A large institutional LP raising this at the side letter stage has more leverage than it typically uses.
Who should hold the signing credential? An authorised signatory at the GP, under an explicit delegation. Administrator-held credentials reintroduce the failure mode the design targets.
Does this slow down calls? It removes verification calls and reduces funding delays. LPs currently delay wires while verifying; a verifiable notice funds faster.
What about non-institutional LPs? Verification is a signature check and requires no infrastructure. A family office verifies the same way a pension fund does.
Why can't an LP just phone the GP? They can, and under deadline pressure most compare documents instead. The number on the notice is also in the attacker's control.
What is the highest-signal event? A change in remittance details between calls. Rendering the delta makes it visible rather than something to notice.
Does this need industry adoption? No. A GP can sign unilaterally, and LPs who can verify benefit immediately.
Where this fits in Manav
Manav binds the authorising individual to the exact instruction being given — the amount, the destination and the entity — and produces a receipt an administrator, a custodian or an LP can verify without calling the issuer.
Sources and further reading
- Reported private equity wire fraud losses arising from email compromise, 2025–2026.
- FCC — protecting consumers from SIM swap and port-out fraud
- FBI IC3 2025 Internet Crime Report
- ILPA — model documents and private markets guidance