Every company that sells in instalments finances its own customer. The goods have gone out, the service has been delivered, the invoice has been issued — and the money arrives in thirty, sixty, a hundred and twenty days. In that interval someone has to fund the operation, and that someone charges for it. For decades the question "who finances" had an automatic answer: whoever processed the sale. Not any more. The receivable stopped being a promise kept in the drawer of whoever sold and became a registered asset, with an owner, a term and an identity — and a registered asset is an asset that gets competed for. This text is the map of the four doors, with the rule behind each one and the traps that only appear once it is signed.
Start with the most used and least negotiated door. When the customer pays by card, a right to receive from the acquirer is created. Until recently that right lived inside the acquirer, and only the acquirer could buy it back before the due date — which is what an advance is. Being the only possible buyer, it was also the only price setter.
BCB Resolution 264/2022 ended that arrangement. Payment-arrangement receivables came to be registered with registrars authorised by the Banco Central, Brazil's central bank, and registration creates the Unidade de Recebíveis (receivables unit) — the asset, with an identity of its own, that can be traded or pledged as collateral with any institution, not only with whoever processed the sale.
The practical effect is simple and almost nobody exercises it: the advance rate became an offer, not a price list. The merchant can look up their own units at the registrar, take them to market and receive competing bids. Carrying on accepting the first one is a choice — except that most do not know they are choosing.
What to do this week: ask your acquirer for the statement of your receivables units and check, at the registrar, whether there is any lien or lock you do not remember granting. It is free, and it usually reveals more than expected.
This is the most expensive trap in the previous block, and it is recent. Many companies contract a pre-contracted advance — the kind in which everything sold is born already advanced, automatically, at the contract rate. It is convenient, and it is expensive: the convenience costs exactly the difference between the rate accepted without a contest and the rate that would be obtained by contesting it.
When the businessowner finds this out, they cancel. And here is the detail that changes the arithmetic: BCB Resolution 562/2026, in force since 11 May 2026, established that the effects of cancellation reach only the receivables created from transactions carried out after the cancellation. Everything already sold stays advanced on the old terms.
In other words: cancelling in December, after the season, does not free the season's revenue — it is already committed. The time to review the advance is before the sales peak, not during it. Whoever reviews in January saves the year; whoever reviews in November saves December.
Two contracts that sound alike in conversation and are opposite on the balance sheet. In an advance, the receivable is assigned: it leaves assets, cash comes in, and the risk goes with it — if the final buyer does not pay, the problem has stopped being yours. It is not debt, it does not use up a credit line, it does not appear as a liability.
In the trava, the receivable is pledged as collateral for a loan. The money that comes in is debt, with interest running, and the receivable is still yours — only blocked. For as long as it lasts, it cannot be advanced, nor offered to another creditor, nor used for anything. Companies find this out at the worst moment: when they need cash and find their own revenue handcuffed to a deal from two years ago.
Neither of the two is better in the abstract. The trava usually carries lower interest precisely because the creditor keeps the collateral; the assignment costs more and gives freedom back. The mistake is not choosing one: it is not knowing which one was signed.
The duplicata is the instrument of whoever sells to another company, and it has crossed eighty years on paper, with a counterfoil, a signature and a stamp. Law 13,775/2018 created the book-entry duplicata; CMN Resolutions 4,815/2020 and BCB 339/2023 gave it its frame; and the Banco Central put the ecosystem into operation on 30 June 2026.
The calendar of the mandate is phased and already published: assisted production in the second half of 2026, large companies from June 2027, medium-sized ones in December 2027 and small ones in June 2028. This is not a reform to be followed from a distance: it is a date in the diary of whoever sells on credit.
Three practical effects, and the third is the one rarely mentioned. First, the instrument now carries registration with an authorised entity, leaving a trail. Second, the drawee's acceptance becomes verifiable — and verifiable risk is cheaper risk, which shows up directly in the discount off face value. Third, the law bars any contractual clause that prevents the supplier from trading the instrument. It is worth rereading supply contracts with that eye: the clause forbidding assignment of the credit is common, it was signed without discussion, and it is exactly the one that turns a liquid receivable into dead paper.
Whoever gets ready before the mandate does not earn a seal: they earn price. The receivables buyer pays more for a portfolio already registered, with traceable acceptance and a history — because they stop pricing the uncertainty about your paperwork.
It is good form to declare the cheque extinct, and whoever does tends not to sell in the interior of the country. The clearing system's own figures say otherwise: in 2025, 112.5 million cheques were cleared, moving R$ 472.7 billion. That is down 18.2% on 2024 and 96.6% since 1995 — and even so it is almost half a trillion reais. The detail that explains everything: the average ticket rose to R$ 4,199.77. The cheque did not shrink, it concentrated: it vanished from the supermarket and stayed where the amounts are high and the relationship is old — agribusiness, construction, regional distribution.
It is still governed by Law 7,357/1985, and it has one feature that changes the price: the risk belongs to the issuer, and the issuer is not a regulated institution — it is the customer. That is why the discount on cheque discounting is the steepest of the four doors. It is not abuse; it is the price of a risk nobody can verify from outside.
The trap here has a name: coobrigação (co-obligation, discounting with recourse). By endorsing the cheque for discounting, the assignor usually remains jointly and severally liable. If the cheque bounces, the bank collects from whoever discounted it. The asset was sold and the risk was kept — the worst of both worlds, signed without anyone having read the line.
Every advance proposal arrives with an agreeable monthly figure. The real cost is almost never that one. Into the arithmetic go: the per-transaction fee, the registration fee, the portfolio's real average term (which is not the contract term), the days of float between settlement and credit, the minimum amount per transaction, and the treatment of the instalments that get advanced on their own.
There is only one honest comparison: how much landed in the account, divided by how much would have been received, over the term in which it would have been received. Done that way, the same portfolio taken to three buyers usually comes back with three answers that look nothing alike — and that distance is the work.
There is no such thing as "the market" for receivables; there are buyers with different appetites. Banks pay more for a large portfolio with impeccable paperwork, and they are slow. FIDCs — fundos de direitos creditórios, credit-rights investment funds, today under CVM Resolution 175 — buy risk that a bank will not, charge for it and demand structure. Securitisation companies make sense when the portfolio is recurring and spread thin. Factoring firms solve the urgent and the small, at the price of the urgent and the small.
The choice is not one of brand, it is one of fit: the same portfolio is worth different prices at each door, and the difference rarely comes from how likeable the manager is. It comes from how much work the buyer will have to understand what they are buying.
First, the inventory: every door at once — card, duplicata, cheque, progress billing — because almost every company negotiates one and forgets the other three. Then, the locks: what is already pledged, with whom, until when, and what can be released. Next, the tidying up: registration, traceable acceptance, contracts without the clause forbidding assignment — this is the stage that changes the price, and it is the one nobody wants to do. Only then the contest: the same portfolio, on the same day, on more than one buyer's desk.
And, finally, the part that is not operation but management: the yardstick. Advancing every month on autopilot is expensive; advancing only when money is tight is more expensive still, because whoever is in a hurry takes whatever comes. The good account is the one that decides before it needs to.
What we do not promise: a rate. Whoever turns up with a figure before looking at your portfolio is selling, not analysing. What we do promise is that your portfolio will be contested by more than one buyer, with you knowing exactly what you signed.
Official sources, at the exact point: the article of law, the service or the search you can use today. None replaces analysis of the specific case, which is our work.
Each note starts from a sourced figure, explains the mechanism behind it and ends with a test you can run yourself, today, without calling us. They are deliberately short pieces: what does not fit in them fits in a conversation.
A note does not replace the review — it shows the mechanism and gives the reader back the ability to check alone. That is deliberate: whoever understands the mechanism negotiates better, including with us.
If the reading raises a question the note does not answer, it is probably the right question. Worth sending: answering a new question is the cheapest way to find out what we have not explained properly yet.
We do not promise a percentage before looking at the data, we do not promise a homologation date — the procedure belongs to the State: and we do not chase speculative theses to fatten an opinion. Nor do we take commission for referring anyone: whoever executes is a dedicated partner of the house, contracted and paid by us, within our price.
A conversation, with no materials and no proposal, to understand what is blocked. If there is matter to work on, the next step is the letter of authorisation and secrecy, a document that limits our own access before anything else. If there is not, we say so, and the conversation ends there.
Every statement on this page comes from a public norm, and the links above lead to the full text: not to the agency's portal. That is deliberate: whoever wants to check can check alone, without asking anyone for anything, which is the only kind of trust that does not depend on whoever asks for it.
No thesis moves forward without its classification written beside it. That is what separates a survey from a promise.
Say in two lines what you need to resolve. An account manager replies personally, in business hours, and the conversation starts where it makes a difference.
This pagepoint the cameraIt reached the desk of whoever answers. The reply comes in writing, during business hours.
