Skip to main content
Skip to content
Agentic Finance

7 Questions About Who Earns on an Agent’s Idle Budget

Classify agent budgets and compare documented interest rights, withdrawal conditions and same-currency opportunity costs without inventing provider yields.

A budget display resolves into different forms of authority, service credit and funded money.
Reading options: audio, text size & sections

Decentralised News · Agentic Finance · Decision research

By Heath Muchena · Source review 2026-11-11 · Instrument version 1.0.0

What Matters

An agent budget may be an unfunded spending limit, prepaid service credit or funded monetary balance. Only the last supports a direct cash-return comparison. Establish ownership, earnings entitlement, withdrawal conditions and the necessary operating buffer before estimating opportunity cost. DN’s tool preserves those categories and treats every return input as a scenario, without claiming to measure a provider’s profit.

A 10,000 budget is not always 10,000 sitting in someone else’s account. The first task is to identify what was funded and what rights came with it. Only then can a business ask who receives any earnings and whether holding the balance still makes commercial sense.

DN Evidence Block

Sources reviewed: 2026-11-11 · Evidence type: Primary-source documentation review and reproducible hypothetical arithmetic

Scope: Contract-specific balance classification and hypothetical cash economics. PayPal example is US personal Balance Account only; cited USDC terms apply outside the EEA.

Not established by this work: No account balances, actual reserve returns, live withdrawals or investment products were measured or recommended.

Original DN instrument · version 1.0.0

DN Idle-Budget Economics Table

Classify the balance before comparing a same-currency customer return scenario. This tool never estimates a provider’s actual earnings.

Enter your assumptions or load a labelled example. Required blank numbers stay missing; unknown evidence remains unknown. Use nonpersonal labels only. This instrument makes no network requests and uses no persistent browser storage.

Use your own inputs

Decision inputs

No current result. Complete inputs or load an illustrative example.

No investment recommendation, current market yield, provider profitability, insurance determination, foreign-exchange model or account verification.

1. Is the budget a limit, a credit balance or funded money?

The first question is what the displayed number represents. A spending limit authorises activity up to a bound. It does not necessarily move money anywhere. A purchased service credit represents a right to consume an eligible service under its terms. A funded account contains a different kind of claim. Treating all three as idle cash produces misleading yield calculations.

A hypothetical agent has a 10,000 monthly limit but has spent nothing and deposited nothing. There is no 10,000 provider-held balance to analyse. The same dashboard number could instead represent 10,000 of prepaid service credits or 10,000 in a funded payment account. Each requires a separate record.

OpenAI’s Service Credit Terms explicitly distinguish its credit balance from a bank account or payment device. That example is useful because a currency-like display can encourage the wrong mental model. The record must retain the applicable service and rights, not merely the numerical balance.

For each agent, record the funding event, account owner, balance category and evidence. If no money moved, investigate spending authority. If credits were purchased, investigate use, transfer and recovery conditions. If money was funded, investigate interest entitlement and withdrawal. The DN tool stops the cash-return calculation for an unclassified balance, service credit or unfunded limit rather than assigning a false zero yield to a different product.

2. Who owns the claim against the account provider?

The second question is the owner of the claim against the account provider. The answer should come from the applicable agreement and account record, not the identity of the agent operating the interface.

Match the legal account holder, the source of the funding and the party entitled to request withdrawal. If an agent operates a customer account on a company’s behalf, the agent’s access does not by itself establish that the company owns the balance. Record any mismatch for resolution rather than folding several parties into a generic “our cash” field.

The practical check is a joined record: account identifier, named holder, funding event and relevant agreement. This establishes the evidence to review; it does not resolve ownership disputes or grant withdrawal authority. Keep control of the interface separate from the documented entitlement to the funds.

3. Who receives the interest or investment earnings?

The third question is the recipient of any interest or investment earnings. Read the account-specific allocation; do not infer it from the person who funded the balance or the platform’s description of its reserves.

The US PayPal Balance terms, updated 4 November 2026, state that relevant earnings belong to PayPal and that balance holders do not receive interest. Those terms cover a specified US personal Balance Account, not every PayPal product, business account or country. Their conditional program-bank arrangements also distinguish protection against a program bank’s failure from protection against PayPal’s failure.

Circle’s current USDC Terms for holders outside the EEA state that holders are not entitled to reserve interest. Direct redemption with Circle also depends on the applicable account eligibility and conditions. These terms should not be transplanted into an EEA account or treated as the terms of every third-party platform holding the token.

These examples establish particular contractual allocations. They do not measure the actual return a provider earns from a particular customer’s balance. To make that latter claim would require evidence about the provider’s holdings, allocation and costs. A buyer can make a useful decision without inventing it: identify what the customer receives, what the customer gives up and which party has the documented entitlement.

4. What is the relevant holding period?

The fourth question is how long cash is unavailable for another use. The correct interval follows actual funding and release events. A monthly budget can sit empty for most of the month, be funded only before a task, or remain funded continuously. Applying one month of interest to the headline limit confuses those cases.

Take a hypothetical 10,000 funded balance held for 30 days. At an assumed 4% annual simple rate, its gross time value is about 32.88. This is an opportunity-cost illustration. It is not the provider’s measured profit, a current offered yield or a promise that a comparable alternative will earn that amount.

The calculation becomes more useful when balance changes are recorded. Five days at 10,000 and 25 days at 2,000 create a different cash-time exposure from 30 days at 10,000. A detailed model sums each balance interval. The interactive instrument below deliberately uses one constant balance so its assumption remains visible.

Identify why the money stays. A required reserve, expected purchase, withdrawal delay or forgotten funding rule calls for a different response. An agent may have no authority to withdraw money simply because it appears idle. The person who owns the cash policy should approve any changed operating buffer and preserve enough liquidity for agreed obligations.

Recurring underuse is a signal to revisit the funding rule, not proof that the provider has behaved improperly. The service may provide availability, access or operational convenience worth paying for. The question is whether that benefit still matches the amount and timing of the funded balance.

5. Can the balance actually be withdrawn and redeployed?

The fifth question is exit. Record the permitted withdrawal destination, currency, account owner, processing window, limits, fees and any pending obligations. A theoretical alternative return is irrelevant if the funds cannot arrive before the next payment must be made.

Suppose the hypothetical 32.88 gross time-value difference requires 3.00 in total transfer costs. The modelled benefit falls to 29.88 before tax, uncertainty or operational effort. If the money must return within a day, a 30-day comparison no longer describes the feasible choice. If the move creates an unmet operating requirement, the arithmetic does not make it acceptable.

Compare equivalent claims. A withdrawable balance with a particular protection and settlement arrangement is not directly interchangeable with a volatile investment, a locked product or a loan to a different counterparty. A higher quoted return can compensate for changed risk or access. This article does not recommend a yield destination.

Record an evidenced current customer return and a scenario alternative separately. If the current entitlement is unknown, do not infer it from a marketing statement about what reserves earn. Keep the uncertainty in the comparison and obtain the account-specific terms before taking action.

Withdrawal can also leave residual commitments behind. Automatic top-ups, pending purchases and refund obligations should be reconciled before reducing a balance. The cash calendar should show both the withdrawal and the replenishment needed if another authorised task arrives.

6. What operating buffer must remain available?

The sixth question is how much funding the service needs to operate dependably. An idle-looking balance can be a deliberate reserve for a known obligation. Identify that obligation and the funding policy before treating the entire balance as discretionary surplus.

Separate minimum operational cash from discretionary surplus. A hypothetical 12,000 account with a documented 2,000 floor leaves 10,000 for the scenario comparison. If that floor is only a guess, run a larger-buffer case and resolve the requirement. Avoid counting the same surplus as both a withdrawable amount and the resource for a pending payment.

Recheck the floor against the next agreed payment, pending purchases and the time needed to replenish the account. A lower balance is useful only if the business can still complete authorised work. Keep the person responsible for that operating decision in the record; an opportunity-cost calculation does not approve a change in the funding policy.

7. Does currency change the comparison?

The seventh question is whether the return comparison changes currency or purchasing power. An apparently higher return can introduce conversion exposure. Identify the currency of the obligation and the currency in which the alternative would return funds, then preserve the conversion assumptions and movement costs separately.

The DN Idle-Budget Economics Table calculates only a constant-balance, same-currency scenario. It subtracts the entered current customer rate from an assumed alternative rate, applies the interval and then deducts movement costs. It does not calculate the provider’s reserve earnings or approve a transfer.

For a hypothetical obligation denominated in the original currency, a return earned elsewhere does not establish how much of that obligation can be paid after conversion. The DN instrument intentionally models one currency. If conversion is required, use a separate dated cash comparison rather than presenting the same-currency result as a complete answer.

Methodology, limitations and corrections

Methodology. All amounts and rates are hypothetical. For a funded balance only, gross return difference equals constant balance ×(alternative annual customer rate−current annual customer rate) ×days ÷365. Net difference subtracts total movement costs. A 10,000 balance,30 days and 4 percentage points produces 32.876712 gross units;3.00 of costs leaves 29.876712. This is not a provider earnings estimate. Noncash credits and unfunded limits receive no cash-return result. Equal currency, feasible access and comparable risk require separate evidence.

Maintain the balance-rights record

Methodology and evidence comparison
Record fieldWhat it resolvesWhat it does not prove
Balance categoryWhether cash-return arithmetic appliesLegal ownership by itself
Interest recipientDocumented entitlementProvider’s actual earnings
Available withdrawal dateFeasible holding periodFuture route performance
Operating floorAmount retained for the defined scenarioA universally sufficient reserve
Currency and comparable alternativeBasis of the return comparisonEquivalent risk without further evidence

The next useful action is to classify the largest balances across the agent estate. A clean inventory often changes the question from “Who gets the yield?” to a more actionable one: why is this amount funded here, what rights does it carry, and what would have to be true before the business could reduce it?

Keep the review date and the source version. A change in terms, funding method or account eligibility can alter the answer while the interface continues to show the same number. The maintained record is the asset; a single yield percentage is only one input.

For corrections, identify the claim, source and date when contacting Decentralised News. A new record-creation timestamp does not refresh its underlying evidence.

Change log: version1.0.0, 2026-11-11. Initial documentation review, original instrument and labelled examples.

Frequently asked questions

Does every agent budget earn interest for a provider?

No. A budget can be an unfunded control or a service entitlement. Even with funded money, actual earnings require evidence beyond the displayed balance.

Are service credits the same as cash?

No. Their use and recovery depend on the applicable service terms. The tool keeps them separate.

Does the alternative rate recommend an investment?

No. It is an entered scenario. Liquidity, risk, currency, protections and withdrawal conditions must be compared separately.

Can the agent withdraw every idle balance?

A numerical surplus does not grant authority. Required reserves, pending obligations, ownership and permitted access determine what can move.

Primary sources

  1. PayPal Balance Terms and Conditions: United States. US-specific interest ownership and conditional program-bank arrangements. Reviewed 2026-11-11.
  2. USDC Terms: outside the EEA. Mint eligibility, direct redemption and reserve-interest entitlements. Reviewed 2026-11-11.
  3. Service credit terms. Applicable-service restrictions, nontransferability, expiry and legal refund exceptions. Reviewed 2026-11-11.

Turn the finding into a decision

Commercial next step. A DN Research Desk brief can classify one agent funding arrangement and document the rights that change its holding cost. Agree the scope and written quote before sharing redacted terms and balance categories.

Related DN research

Disclosure: Analysis and instrument results are editorial. Where labelled affiliate links appear, DN may receive compensation when you sign up or buy through them. Commercial relationships do not determine findings. No tested integration or guaranteed outcome is implied.

Get the most talked about stories directly in your inbox

Join the Decentralised News briefing for independent crypto, DeFi and AI analysis. No spam, unsubscribe anytime.