Shelf debt hides; designed debt answers

R$ 20 billion was the alarm; R$ 18.4 billion in supplier finance was the anatomy — and the Collateral Framework Law arrived in the same year: Brazil delivered the disease and part of the treatment within twelve months.

In January, Americanas announced "accounting inconsistencies" on the order of R$ 20 billion. Months later, the anatomy became clearer: an enormous share was tied to supplier finance operations that did not appear as bank debt in the way a reasonable reader would expect. In October, Brazil passed the Collateral Framework Law, trying to make collateral more usable. The same year delivered the disease and part of the treatment. Shelf debt hides; designed debt answers.

I wrote in January that debt that changes its name does not change its nature; it changes its date. October delivered the anatomy and, by a coincidence of the calendar, the beginning of the antidote.

Supplier finance is not fraud by nature. A supplier sells merchandise, a bank prepays the receivable, and the buying company pays at maturity. The operation can improve the chain's liquidity. The problem begins when the economic substance is financing and the accounting presentation lets the market see supplier terms. The obligation exists, the bank exists, the financial cost exists; only the word "debt" disappears from the place where the investor looks for it.

That distinction changes everything. Suppliers finance commercial operations and tend to follow the inventory cycle. Banks finance balance sheets, charge interest, and can react jointly to covenants, ratings, and confidence. Treating a financial liability as a commercial account alters leverage, working capital, and cash generation indicators. The investor believes he is observing operational efficiency when he may be observing financing hidden inside the suppliers line.

The scandal is not merely accounting. It is architectural. A company with a known brand, celebrated shareholders, an auditor, a board, and market access managed to build obligations whose economic design was not legible on the surface. That should humble anyone who believes governance is a collection of prestigious names. Reputation reduces questions until the day the questions become losses.

The market will hunt for individual culprits, as it should. But limiting the story to people allows the mechanism to survive. Who negotiated the operations? How were they recorded? What confirmations reached the banks? How did auditors reconcile suppliers, financial institutions, and cash flow? What incentive did executives have to maintain the appearance of working capital? Fraud, when it exists, is a human act; scale requires a system.

The first lesson for the creditor is that debt should not be identified by the name of the account, but by the economic origin of the obligation. If there is prepayment by a financial institution, compensation for time, and a payment commitment from the company, there is financing in substance. The analyst must reconstruct the balance sheet from the flows. Accounting is a useful map, not sovereign territory.

The second lesson is that concentration of trust is risk. Banks may run their own analyses and still depend on the same information produced by the company. Auditors may review procedures and miss collusion. Investors may diversify funds and keep exposure to the same issuer through different channels. When everyone trusts the central reputation, apparent independence becomes hidden correlation.

The third lesson is about suppliers. They look like operational creditors, but they are the first to discover that selling and financing are inseparable. A company can show large revenue because suppliers grant terms, banks prepay, and the cycle expands. When trust breaks, the supplier cuts limits, demands payment, and halts delivery. The operating asset enters crisis because the financial architecture that sustained the shelf disappears.

That is why every supplier should treat payment terms as a credit decision. It is not enough to check whether the client has a bad record. It is necessary to decide how much to sell, for how long, at what margin, with what security, and what signal ends the exposure. Americanas is not just a story about banks and shareholders; it is a warning to thousands of companies that deliver merchandise before getting paid and call it commercial policy.

Selling on credit contains a free option granted to the buyer: using the supplier's capital until maturity and, in deterioration, delaying while negotiating. The supplier charges a commercial margin, not an explicit credit premium. With large clients, he accepts worse conditions because he fears losing volume. Thus purchasing power turns suppliers into subordinated, dispersed, uncoordinated financiers. The retailer's balance sheet grows supported by creditors who do not perceive themselves as creditors.

The Collateral Framework Law arrives in the same year promising to improve the use, registration, and enforcement of collateral, including mechanisms that may allow better economic use of certain assets. The direction is correct: dead collateral protects no one. A property can be worth a lot and finance little if pledging it is expensive, slow, and legally uncertain. Making collateral more divisible and reusable can reduce the fear concentrated in the contract.

But a law does not automatically turn wealth into credit. It takes conservative appraisal, clear priority, reliable registration, monitoring, and an enforcement process that survives conflict. An asset pledged to several obligations requires knowing who gets paid first and up to what limit. Reuse without transparency can multiply credit over the same collateral, reproducing in the registry the problem Americanas exposed in the liability: several people believing they hold protection that cannot fit simultaneously.

The thesis common to both events is legibility. Americanas shows the cost of an obligation whose substance does not appear. The framework tries to increase the legibility and usefulness of collateral. Credit works when the creditor understands both the flow that pays and the asset that recovers. Hiding the first or overestimating the second produces the wrong rate, the wrong limit, and the wrong confidence.

My position in January would have been to avoid any heroic attempt to buy the stock merely because the market value fell. When the balance sheet stops being trustworthy, there is no multiple. The investor does not know the debt, the working capital, the profit, or the cash need. A stock at a ninety percent discount to false numbers is not cheap; it has no denominator. I would wait for independent reconstruction and treat the equity as the residual of a process whose liability was still being discovered.

In credit, I would analyze documentation, collateral, and position in the restructuring before the coupon. Unsecured bonds of an operationally relevant company can recover value, but they depend on negotiation, asset sales, and continuity. Banks with specific rights, essential suppliers, and secured creditors have different incentives. "Creditors" is not a homogeneous class; it is a line in dispute. Structure determines who arrives with an argument and who arrives with nothing but hope.

I would also look for risk-infrastructure companies: reconciliation of receivables and suppliers, independent balance confirmation, collateral registration, exposure monitoring, and systems connecting procurement, treasury, and accounting. The scandal reveals that large organizations may not possess a unified view of their obligations. The solution is not another dashboard. It is preventing an economic transaction from changing its nature as it crosses departments.

For investors in other companies, the checklist would change. Does supplier growth track inventory and sales? Does operating cash flow depend on stretching payment terms? Are there supplier finance, forfait, confirming, or reverse factoring operations? Where are they classified? What do they cost? What share can the banks cancel? A company can look like a cash generator because it pushed payments forward with financing it did not call financing.

For suppliers, the event should destroy the belief that size reduces risk. Large companies have more assets and access, but they can also accumulate larger obligations and negotiate more aggressive terms. The absolute loss can be fatal to a small supplier. A credit limit must consider the creditor's capacity to survive the default, not only the client's probability. A "safe" exposure can be irresponsible if its loss breaks whoever sold.

The Collateral Framework Law offers an opportunity for better products, but also for worse marketing. Banks will advertise "secured" credit without necessarily passing the full risk reduction to the price. Fintechs will promise to monetize properties in minutes. The entrepreneur must compare not only the rate, but how much of the asset is committed, which future obligations get blocked, and how enforcement works. Collateral that is cheap for the creditor can be a very expensive option sold by the debtor.

Brazil holds enormous real estate and productive wealth, but much of it is poorly documented, indivisible, or mixed with personal assets. The new law can expand leverage capacity. That is benefit and danger. Making debt easy does not make investment productive. An unencumbered property offers resilience; a fully monetized property offers cash and fragility. Design should preserve margin, not extract the last available real.

This year's letter should not say that I "predicted Americanas." Retroactive prediction is cheap vanity. It should say that the mechanism was already visible to anyone who refused to accept accounting names as substance. The point is not to call one company; it is to build questions that work before the next case. Where is the debt that is not called debt? Where is the guarantee that does not guarantee? Who retains risk and who merely collects a commission?

R$ 20 billion was the initial alarm, not the final classification. About R$ 18.4 billion in supplier finance later appeared as the central part of the explanation. That precision matters because authority is not born from rounding scandals to fit the headline. It is born from showing how the number was produced, what obligation existed, and why so many readers looked without seeing.

Shelf debt hides because it sells a standardized name and lets the company adapt reality to the label. Designed debt answers because it starts from substance and defines flow, priority, collateral, and information. Americanas showed what happens when the liability is organized for appearance. The Collateral Framework tries to organize collateral for enforcement. Between the two lies the future of Brazilian credit: making it visible before making it cheap.

Leo Bentier