Programs may combine founder preparation, product development, commercial support, and possible investment. Setting resources aside preserves the ability to decide later, when more is known. At the outset, an investor cannot know which projects will demonstrate demand, which will need a different product, and which should end an experiment. Spending all available cash immediately removes that choice just as better evidence about teams, customers, and execution becomes available.
A reserve should protect a decision, not simply extend an operation
Funding an additional opportunity differs from absorbing an unexpected event. The first supports a proposition that has gained evidence; the second seeks to contain a loss or meet an obligation. Combining both under the broad label of portfolio support can allow emergencies to consume money intended for stronger projects. Unless that change is identified and approved, the program may change its investment strategy without acknowledging it.
The central question is which decision each reserve allocation makes possible, under what conditions, and within what limit. Money reserved for an already negotiated integration serves a different purpose from money held for an undefined future round. Ownership matters too: cash belonging to an investment vehicle, the program operator, and a supported company is subject to different decisions and obligations.
The numerical examples here are hypothetical illustrations of those choices. They are not market rates, achieved results, or personal allocation advice. The practical issue is how to support later investments and respond to incidents while keeping investment, service purchases, the operator's running costs, and protection of committed assets distinct. Reserves can improve flexibility, but holding cash also has costs and does not remove uncertainty about whether further spending is worthwhile.
Separate the cash pools before deciding how much to reserve
The first control is a map of ownership and purpose: who owns the money, who may move it, and which obligations can draw on it. An accelerator may provide services without investing. A venture builder may develop a product and also own an equity interest. Neither arrangement automatically makes the service provider's cash, the investor's cash, and the companies' cash interchangeable.
Within the investment budget, distinguish initial investments, follow-on investments, contingencies, and administrative expenses. These categories should appear in agreements and records, not only in a commercial presentation. Contributions that participants have promised but have not paid must be shown separately from available funds. An intention to contribute cannot settle an obligation falling due before the money arrives.
A second map should follow the money inside each supported company. Investment proceeds may pay suppliers, staff, Brazilian taxes, services purchased from the program itself, and other operating needs. This reveals how much remains usable to develop the business proposition. If a substantial portion returns to the operator or an affiliated provider, the investor needs to see that route to understand the arrangement's full cost.
Separation does not require pretending that connected parties are independent. It requires disclosure and prevents undisclosed offsets. An operator's expense should not become an investee's emergency merely because they share managers. Likewise, expected service revenue is not an available reserve before the work has been contracted, delivered, and paid for. Each entity's ability to use cash must follow its own rights and obligations.
Value services without artificially counting them as invested capital
Team preparation, customer research, programming, and commercial support cost money and may create value. That does not justify adding any assigned price for those activities to invested capital. Assessing services requires a defined scope, hours or delivery units, necessary expertise, external expenses, the provider's margin, and comparable alternatives. A long activity list does not establish that the company receives an equivalent economic benefit.
Comparison should reflect what the company would actually purchase outside the program. Unlimited mentoring might be worth less to it than a specific integration, even if the advertised mentoring price is higher. Shared services also need scrutiny. Charging every project the full cost of the same research can overstate the resources used unless the allocation method and the separate benefit to each company are explained.
The agreement should state whether services are paid for in cash, compensated with an ownership interest, or included in another obligation. Where equity is consideration for services, its source and conditions need separate documentation and assessment. Providing a service does not automatically constitute a capital investment. An internally assigned price does not prove that an asset of that value entered the company.
This distinction changes the reserve calculation. Services paid from investment proceeds reduce the company's operating cash. Services still to be delivered create dependence on the provider's capacity. If the plan assumes that the product can be completed without additional payments, check exactly which deliverables are covered. Otherwise, a follow-on investment may merely pay for work that appeared to be included in the initial arrangement.
Deliverables and intellectual property determine the cost of continuing
A reserve is meaningful only if the investor knows what the company can use after each stage. Completing a presentation or making a demonstration available is not equivalent to delivering usable code, documentation, and infrastructure access. Continued operation requires both legal rights and practical control, including when the relationship with the venture builder ends. A working demonstration alone may reveal little about those requirements.
Contracts should distinguish the provider's pre-existing materials, third-party components, and work created specifically for the project. For each category, clarify ownership, licensing, modification and transfer rights, and use restrictions. These questions require appropriate legal analysis. Their economic effect is equally important: a restrictive license can make changing suppliers much more expensive than the development budget initially suggests.
Acceptance criteria should describe verifiable behavior. An integration may require authentication, failure handling, and documentation as well as a functioning screen. Stage payments become more informative when tied to those criteria and provisions for defect correction. Vague acceptance can shift the cost of finishing already contracted work into the contingency reserve, obscuring whether the program is buying something new or paying again for an existing obligation.
Transition arrangements should cover company access to repositories, credential management, a dependency inventory, recoverable backups, and an agreed period of exit assistance. These controls do not require publishing secrets or private documents. They reduce reliance on a single person or provider and help estimate maintenance, replacement, or closure costs before a crisis. Any limitations should be visible when the reserve is approved, rather than discovered during an urgent handover.
A complete example of the budget and initial exposure
Consider a hypothetical program with R$1,200,000 already available and three companies, A, B, and C. Its approved budget allocates R$600,000 to initial investments, R$360,000 to follow-on investments, R$120,000 to contingencies, and R$120,000 to program administration. These four allocations add up exactly to the existing cash. No future fundraising is needed to make this starting budget balance.
Each company initially receives R$200,000. Planned follow-on limits are R$160,000 for A, R$120,000 for B, and R$80,000 for C. Together they total R$360,000, but they are limits, not promises to disburse. Each release depends on approved criteria. If B fails its milestone, its planned R$120,000 remains in the program pending a formal decision under the applicable agreements.
Suppose each company buys R$60,000 of services from a provider affiliated with the operator, using its initial investment. Each then has R$140,000 left for other purposes. The program has still invested R$200,000 in cash in each company. The services are a use of that cash, not an additional contribution taking the individual investment to R$260,000. Reporting both as separate investment amounts would count the same money twice.
The combined R$180,000 paid by the companies is the provider's revenue, subject to its own costs and obligations. It does not automatically replenish the program's reserve. The separate R$120,000 administration budget covers the investment arrangement's own expenses. Contract review should identify any overlap between administration charges and service fees, establish what each payment purchases, and disclose who benefits from it.
The cash timetable shows when a reserve can fail
In this example, the initial R$600,000 leaves at entry, so R$600,000 remains in the program. Administration is paid in four monthly installments of R$30,000. At the end of the first month, with no other payments, the balance is R$570,000. That balance includes cash assigned to different purposes. A bank balance is therefore not the same as an amount freely available for new decisions.
In the second month, assume A meets its requirements and receives R$160,000. Another R$30,000 administration payment brings the balance to R$380,000. In the third month, B receives R$120,000, administration uses R$30,000, and an approved contingency requires R$90,000. The remaining balance is R$140,000. The incident is examined below; here, its position in the payment sequence is what matters.
In the fourth month, C receives R$80,000 and the final R$30,000 administration payment is made. The closing cash balance is R$30,000, representing unused contingency funds. Reconciliation confirms that R$600,000 initially invested, R$360,000 in follow-ons, R$120,000 in administration, and R$90,000 for the emergency total R$1,170,000. This defined example assumes no other approved obligations beyond those amounts.
The same overall totals could conceal a temporary cash shortfall if some funding arrived only after payments became due. Investors should examine due dates, accessible bank funds, and commitments alongside the total budget. Here, the money already exists. An arrangement relying on capital calls would also need rules for payment periods and participant default. Expected receipts must not silently fill a liquidity gap in the analysis.
Release criteria should test the proposition and the additional use of funds
A follow-on reserve is more useful when every release answers two questions: what has been learned, and what will the additional money allow the company to test? A's R$160,000 might fund an integration requested by customers already using the product. The criteria should address actual use, delivery capability, and servicing costs consistent with the commercial hypothesis, rather than simply completion of a feature.
B's R$120,000 might depend on turning trials into contracts with clear obligations and understood payment terms. An expression of interest is weaker evidence than a contract, while a signed contract without collection does not eliminate cash risk. Evidence should suit the company's stage. Forecast sales should remain forecasts, and customer enthusiasm should not be presented as revenue already earned or money already available.
C's R$80,000 might support a narrower experiment with a defined duration and maximum loss. The committee should know which result supports continuation and which supports closure. Different projects can reasonably have different limits. The concern is a difference driven by personal closeness or bargaining power without documented justification. A smaller experiment can be appropriate if it answers a useful question at a bounded cost.
Criteria should still leave room to learn. If a material discovery invalidates the original milestone, changing it may be rational. The revision should identify the new hypothesis, cost, expected evidence, and effect on the reserve, while preserving the earlier criterion in the record. Rewriting past requirements until every outcome appears approved destroys the process's informational value and makes later decisions harder to assess.
A follow-on investment has its own price and rights
Reserving R$160,000 for A does not determine the ownership interest the investor will later receive. Suppose the initial R$200,000 was an equity investment at a pre-money company valuation of R$1,800,000. The post-money valuation would be R$2,000,000, giving the investor 10%. This simplified example assumes no other interests or instruments. The percentage follows from the investment divided by the post-money value.
For the follow-on round, assume a pre-money valuation of R$3,040,000 and only the reserved R$160,000 of new money. The post-money valuation becomes R$3,200,000, and the new issuance represents 5%. The existing 10% interest is diluted to 9.5%. Adding the new 5% gives the investor 14.5%, while the other existing owners collectively hold 85.5%. These figures describe ownership arithmetic, not investment performance.
The calculation excludes options, other investors, conversions, and special rights. Its purpose is to show that a reserve does not freeze valuation or ownership. If the original instrument includes preemptive or proportional participation rights in later rounds, their scope must be read in the agreement. Available cash alone creates no entitlement to invest, and a planned amount does not establish the terms another party must accept.
Conflicts become more sensitive when the operator also influences the round's price. A lower valuation may increase ownership for an investor connected to the program; a higher valuation may improve indicators presented by the manager. The process should justify assumptions, allow challenges, and record who negotiated for each side. A reserve preserves negotiating capacity, but does not authorize either side to impose any terms it chooses.
Equity, convertibles, debt, receivables, and donations have different consequences
The ownership calculations above apply only to the stated equity scenario. Equity capital remains exposed to business performance and an eventual investment exit, without an automatic repayment timetable. A reserve may support development and avoid a rushed round, but it does not make an ownership interest liquid. Even a business showing operational improvement may have no buyer for its shares and no distribution of profits.
A convertible instrument needs its own terms: conversion events, pricing rules, any discount or valuation cap, maturity, and treatment if the anticipated round never happens. The label alone resolves none of these questions. A planned additional investment may also affect negotiations over existing instruments. Investors should model the resulting ownership structure and consider the rights and incentives of other holders before treating the reserve as straightforward support.
With debt, due dates and repayments may compete directly with product development and sales spending. A reserve intended for new investment is not a guaranteed source for repaying an earlier loan. Receivables-based financing requires a further distinction between existing eligible claims and hoped-for sales. Expected billing is not an established payment right; eligibility, documentation, collection risk, and the financing terms must be examined for the actual receivables.
A donation to a nonprofit has a different purpose: it is neither an equity investment nor a promise of financial return. Brazilian entity form also matters. A MEI cannot admit equity partners while remaining a MEI. Appropriate advice should address these differences before funds are committed. Renaming a budget line or describing every transfer as support cannot correct incompatible rights, obligations, or business structures.
A contingency should address a defined event
Return to the R$90,000 contingency payment in the third month. Suppose an external supplier stops supporting a component used by A and C. A technical assessment identifies migration, data recovery, and validation work. The approved budget assigns R$50,000 to migration, R$25,000 to recovery, and R$15,000 to testing. These amounts add up to the R$90,000 disbursement and identify what completion should involve.
Before paying, the program should establish who contractually bore the risk and whether the original provider has a duty to correct the problem. A contingency should not automatically replace work already paid for. In this hypothetical scenario, review finds the work outside the contracted scope and asset preservation less costly than rebuilding. That conclusion requires evidence, including a credible comparison of alternatives, rather than reliance on the word emergency.
With appropriate authorization, payment might go directly to suppliers, with records explaining how the benefit is attributed to A and C. Its legal and financial nature must also be defined: a program expense, additional investment, loan, or another permissible arrangement. If it creates rights against either company or obligations for them, the agreements should reflect those consequences. Technical urgency does not remove the need for classification.
The decision needs a boundary. The reserve should not indefinitely support a model with no demonstrated demand. Preserving useful data through migration may justify spending; extending operations for months without a testable proposition requires a separate assessment. Give the incident an owner, deadline, spending cap, and completion criterion. If costs exceed approval, the difference returns for a decision rather than receiving an automatic extension.
Adverse scenarios reveal hidden concentration
The base budget ends with R$30,000, but that does not establish adequacy under adverse conditions. If the same incident cost R$150,000, it would exceed the entire R$120,000 contingency allocation by R$30,000. With all other payments unchanged, the program would be short by that amount. It would need to reduce still-discretionary commitments, secure genuinely available resources, or limit its response to the incident.
Now suppose B fails its requirement and does not receive the planned R$120,000. Keeping all other base-case payments unchanged, the final balance would be R$150,000: R$120,000 of unused follow-on funds plus R$30,000 of remaining contingency. The money does not change purpose automatically. Reallocation requires an authorized decision and compliance with existing agreements. A larger balance is not evidence that every use of it is permitted.
Concentration may exist beyond ownership interests. Three companies in different sectors can rely on the same technical team, infrastructure, or sales channel. A common failure may affect several companies simultaneously, precisely when the operator's ability to assist is reduced. Testing isolated incidents is therefore insufficient. Review shared dependencies, competing recovery needs, and whether the same limited specialists are expected to solve multiple problems at once.
Adverse scenarios should expose uncomfortable choices rather than produce an impressive reserve figure. Which payments are already mandatory? Which investments may wait? Which assets merit preservation, and when should an operation close? A smaller reserve with explicit priorities can be easier to assess than a larger one burdened by hidden commitments. Neither guarantees that resources will be sufficient, and neither removes the possibility of partial or total loss.
Conflicts of interest belong in the negotiation
In the example, the provider affiliated with the operator receives R$180,000 from the companies. That creates an identifiable incentive: maintaining programs and services can generate revenue even when investment progress is weaker. The relationship does not prove misconduct, but it calls for disclosure, price assessment, and a meaningful opportunity to discuss scope. Review should also cover commissions, cross-holdings, and paid referrals.
A common conflict arises when additional investment is conditional on purchasing more services from the same group. The proposal should show investors how much reaches the company and how much returns to related parties. Founders need clarity about alternatives and the consequences of declining. Combining everything into one ownership percentage makes it harder to assess service prices, dependence, financing terms, and each party's economic interest.
Disagreement over the reserve may be legitimate. Founders may prefer immediate funding, investors may require milestones, and the operator may want predictable staffing. One negotiable approach distinguishes a contractually committed portion, a conditional portion, and a wholly discretionary portion, provided cash supports every binding obligation. Agreements must use these terms precisely. A possible future contribution should never be presented as an unconditional promise.
Related-party decisions should be reviewed by people without the specific economic benefit, with recusals and comparable reasons recorded. Some structures require additional procedures. Oversight should continue after signature because price changes, scope extensions, and contingency spending can also shift value. A service that was reasonably priced initially may become expensive through repeated extensions, particularly if changing providers is difficult or ownership of deliverables is unclear.
Assign operating responsibilities so the reserve does not become an informal negotiation
A reserve policy can be brief, but it must assign concrete roles. The monitoring team gathers evidence; the financial officer checks availability and commitments; specialists assess material deliverables; the authorized body decides; and another person verifies payment execution. Separation can be proportionate to program size while preserving independent review at points of greater conflict. A written process is useful only if people can actually carry out these functions.
The investee's responsible representative should submit the purpose, amount, timetable, and consequences of not receiving the requested funds. The operator should explain whether it can deliver the promised support. The reserve administrator should show the effect on the portfolio as a whole. This turns a request that feels urgent to its sponsor into one that can be compared with the program's other needs.
Routine reconciliation should connect bank balances, approved expenses, signed contracts, and amounts still subject to discretion. When several legal entities are involved, consolidated reporting must retain each owner's identity. Cash in one company is not freely available to another. Reports should also disclose material conditional commitments without adding them together as though all were debts already due. Availability and possible future exposure answer different questions.
Absences and replacements need simple rules. If only one person can authorize payments or access documents, an operational interruption can become a financial crisis. Powers of attorney, access rights, and spending limits should respect applicable authorizations, with changes recorded. The aim is controlled continuity: resolve the event while preserving a record from which others can reconstruct the payment's purpose, approval, recipient, and treatment.
Use meaningful evidence and correct errors transparently
Monitoring should link each disbursement to the approved hypothesis. Product development evidence may include acceptance tests, actual use, and maintenance costs. For commercial support, contracts, collections, and reasons for lost business are more useful than a meeting count alone. Core reserve records include requests, decisions, statements, obligations, and related deliverables. Access should match a person's legitimate need, rather than exposing all company information indiscriminately.
A recurring error is counting the same value twice. Recording R$200,000 of investment plus R$60,000 of services as total investment, when the services were paid from that investment, overstates reported exposure. Correcting it requires reconciling sources and uses, revising reports, and explaining which indicators changed. Renaming a line while retaining the incorrect total does not repair the information used by investors and founders.
Another error is marking a delayed milestone complete to unlock funding. If discovered, record the discrepancy, pause further releases dependent on that information, and reassess completion costs. The treatment of money already paid depends on the agreements and facts. Keeping the history helps distinguish execution failure, poorly designed criteria, and improperly presented information. Those problems may require different remedies even if the immediate cash consequence looks similar.
Excessive rigidity can also cause mistakes, such as rejecting a low-cost adaptation supported by new evidence merely because it was absent from the original plan. The remedy is a documented revision with a limit and an accountable person, rather than removal of controls. Indicators should inform judgment: many recorded activities can produce little benefit, while a few well-substantiated outcomes may support a more useful decision.
Intermediation and documentation support diligence without replacing it
KEZS connects entrepreneurial projects and potential funders through technology and intermediation. In this context, organizing analysis by project can help distinguish funding needs, the intended instrument, deliverables, and supporting documents. The potential benefit lies in clearer information and comparison of assumptions. Investors perform their own due diligence and make their own decisions; KEZS does not assign a credit score.
A transparent reserve description should distinguish available money, conditional commitments, and intentions, and identify who is responsible for each statement. This is a documentation method, not a claim that a particular operational feature is available. A project's presence in a technology environment does not establish its quality, the legality of an offering, financing approval, or its ability to perform obligations. Organized records remain subject to verification.
Public distribution of securities and regulated financial activities in Brazil require structures and providers appropriate to the actual activity. Technology and a CNAE economic activity classification are not authorization. Materials from Banco Central do Brasil, the CVM, and other authorities can provide reference entry points for questions. They do not prove authorization for KEZS or any participant. The specific arrangement must be examined before fundraising or financial intermediation proceeds.
Documentation must also protect privacy and confidentiality through data minimization, access controls, and incident handling; ANPD guidance can be a starting point. Private documents should not be placed on public blockchains. Blockchain integrity does not establish that recorded content is true, that someone has legal title to an asset, that a reserve is solvent, or that payment is guaranteed. Technical record integrity cannot replace legal, financial, and factual verification.
Turn the analysis into a decision policy that can be revised
Before committing funds, an investor should be able to reconstruct the reserve policy without relying on the manager's presentation. The document should explain purpose, ownership, funding sources, permitted uses, payment timing, and authority to make changes. It should identify existing obligations and criteria for approving or declining further investments. Differences between contracts and spreadsheets must be resolved because a balanced spreadsheet cannot override a binding obligation.
The opening position is straightforward: R$1,200,000 available, three initial investments of R$200,000, R$360,000 for later decisions, R$120,000 for incidents, and R$120,000 for administration. Economic analysis adds what those totals conceal: R$60,000 of services per company, rights over deliverables, shared dependencies, and decision limits. The timetable then shows how the balance reaches R$30,000. Each view is necessary to understand the same arrangement.
Periodic review should ask what has changed in commercial evidence, delivery costs, obligations, shared risks, and continuation alternatives. Depending on the structure and agreements, unused funds may remain reserved, be reassigned, or be returned. Spending simply to exhaust a budget is problematic, as is leaving money idle while a known obligation worsens. Neither action should be justified solely by the existence of an approved budget category.
A sound decision cannot be judged only by its eventual outcome, which also reflects uncertainty. It should be defensible using information available when it was made, with clear costs and managed conflicts. For accelerator and venture builder investors, reserves are useful when they preserve real choices, enable corrections, and limit poorly understood commitments. They guarantee neither approval nor repayment, profitability, liquidity, business continuity, or recovery of invested capital.
Frequently asked questions.
Should the program reserve the same amount for every company?
Equal amounts are easy to communicate but can ignore differences in cost, stage, and dependence. One company may need a specific integration; another may need a longer commercial validation process. Different limits can be reasonable when the criteria are explainable and conflicts are addressed. Review shared exposure to suppliers and teams as well: individually balanced reserves may still be insufficient for a common failure. Equality of amounts does not establish equality of risk or usefulness.
Can a large reserve reduce pressure on a company to improve?
Yes, if participants expect every deficit to be covered. A defined purpose, release criteria, and a real possibility of refusal reduce that incentive. The company should understand which funding obligations are contractual and which require a new decision. Assess whether additional money buys useful learning or merely postpones a difficult choice. Available cash does not replace an assessment of the additional benefit expected from the proposed spending.
How can services be assessed when few comparable offers exist?
Break the work into required skills, estimated effort, dependencies, deliverables, and replacement costs. Partial proposals may provide useful comparisons even when no supplier offers the whole package. Pricing remains uncertain and negotiable. Independent technical review can help without turning an estimate into a fact. The greater the provider's influence over valuation, the more important objective acceptance, margin transparency, and workable exit alternatives become. Those controls help assess both initial price and future dependence.
Can an investor describe the entire reserve as capital already invested?
The categories should remain distinct. Cash available for a possible investment, a contractual commitment subject to conditions, and capital actually transferred describe different situations. Reports should identify each and explain material conditions. In the example, the reserved R$360,000 was not initially invested in the companies. Publishing a combined amount without explanation may overstate program execution and the investees' available cash. A future spending limit should not be reported as a completed transfer.
What if a new investor requests changes to rights associated with the reserve?
Start with existing agreements and the proposal's economic effect. Negotiations may involve preemptive rights, priority, valuation, governance, or information access. A discretionary reserve does not create rights that were never contracted. Before accepting changes, model ownership and obligations, consider alternatives, and address conflicts. A revision may be reasonable, but its cost should be explicit, including any reduction in flexibility or additional obligation that the reserve holder would assume.
Who bears the loss if a company closes before using its reserve?
That depends on who owns each amount and which obligations already exist. Money still held by the program follows its agreements. Cash already transferred to a company belongs to a different financial and legal situation involving expenses, creditors, and specific rights. Closure does not automatically make the balance repayable to the investor. Separate uncommitted resources, closure obligations, and potential recovery rights, without promising reimbursement or a priority that does not exist.
How can decision quality be preserved when information becomes outdated?
The policy can require updates before each release covering cash, obligations, use of earlier funding, deliverables, and material team or contract changes. The level of updating should match the decision's risk. If critical information is missing, the request may wait or, where legally permissible, be narrowed to a verifiable stage. Record the gap and its consequences. An estimate may help explore a scenario, but it should not be presented as established knowledge.
References and context
Institutional reference points for checking applicable rules and guidance. Calculations and scenarios are original examples, not quotes, historical results or individualized recommendations.
- Banco Central — correspondentes no País ↗
- CVM — Resolução 88, texto e alterações ↗
- CVM — Resolução 175, fundos de investimento ↗
- Portal do Empreendedor — perguntas sobre MEI ↗
- ANPD — orientações e perguntas frequentes ↗
AI-assisted educational content. Decisions require verification of the information, instruments and providers in each actual case.
