The upper floor received a subsidy to design. The lower floor stayed on the rate table.

The incentivized debenture shows that price and tenor change when risk is presented differently — and that the correct question precedes the price: who is the natural creditor of this asset?

Brazil has decided that certain infrastructure projects deserve a tax advantage to attract investors. Law 12,431 offers an incentive for debentures tied to priority investments. The idea is reasonable: long tenors demand patient capital, and patient capital does not appear merely because a minister wants better bridges. The revealing detail is who manages to turn the idea into money. The upper floor received a subsidy to design. The lower floor kept receiving a rate table.

An incentivized debenture is not merely a cheaper loan. It is a sequence of decisions: which project receives the funds, which company issues, which guarantees protect the investor, how the cash is segregated, which covenants limit leverage, who supervises execution, and what happens if the construction is delayed. The tax incentive improves the buyer's net return, but it replaces none of those choices. It reduces the price of an architecture that must already exist.

The government seems to understand that infrastructure cannot be financed like working capital. A highway takes years to build and decades to generate revenue. A transmission system, a port, or a railway cannot depend on the quarterly renewal of a bank line. The liability must breathe at the same rhythm as the asset. That obvious truth, recognized for billion-real projects, disappears when the borrower is a mid-sized company financing equipment, inventory, or expansion.

At the branch, products are still defined by the creditor's shelf, not by the debtor's cycle. The company buys a machine with a ten-year useful life and receives a twenty-four-month tenor. It finances seasonal inventory with a line that matures before the sale. It uses property as collateral for short credit and rolls it over repeatedly, paying fees every time the bank pretends to reconsider a risk it already knows. The problem is not just the rate. It is the violence of forcing different clocks together because the creditor dominates the negotiation.

The incentivized debentures show, by contrast, that price and tenor can change when the risk is presented differently. The individual investor accepts a lower return because he receives a tax benefit. Funds and institutions buy because the obligation is standardized, held in custody, and tradable. The project can offer specific guarantees and periodic information. The issuer reaches a base different from the banking one. No single element explains the cost. The combination does.

That should destroy a Brazilian superstition: that the spread is merely greed. Greed exists in every market, including where credit costs less. The spread incorporates competition, taxes, reserve requirements, cost of capital, defaults, recovery, information, and bargaining power. Some components are excessive, others are real. Attacking the number without decomposing the machine produces strong speeches and fragile results.

The tax incentive is an indirect acknowledgment that the current architecture cannot carry private savings to long projects in sufficient volume. Instead of forcing the bank to lend cheaply, the State alters the investor's return and creates an alternative path. It is a smarter solution than price control, but it carries distributive risk: the benefit tends to be captured by issuers able to pay arranging banks, lawyers, audits, ratings, and fiduciary agents.

The fixed cost of structuring is a barrier. For a multi-billion concession, it is small. For a company that needs five million, it is prohibitive. Thus the paradox is born: whoever already has scale gains access to better instruments; whoever suffers most from the branch cannot pay for the exit. The market calls that efficiency. It is merely economies of scale applied to financial information.

There is an evident business opportunity in reducing that fixed cost. Standardized documents, automated receivables analysis, continuous monitoring, electronic collateral registration, and digital distribution can make a five-million operation as analyzable as a five-hundred-million one. It is not necessary to simplify the risk until it becomes false. It is necessary to industrialize repetitive tasks so that human work concentrates on the exceptions.

The investor, in turn, must resist the charm of the exemption. An asset does not become safe because the tax disappeared. The absence of taxation can turn a mediocre return into a competitive one, but it does not turn an uncertain flow into a certain one. The tax benefit is known; the loss of principal is asymmetric. A debenture with an attractive net return and poor recovery remains an elegant way of losing money without paying tax on what was never earned.

I would first examine the project's capacity to pay the debt without refinancing. Contracted revenue, predictable demand, adjustment mechanisms, operating costs, construction risk, and regulatory obligations. Then the structure: escrow accounts, reserves, distribution limits, acceleration clauses, guarantees, and priority. Only then the rate. The market usually starts with the rate because it fits on a screen. The risk occupies the annexes.

I would also distinguish project risk from corporate risk. A solid company can issue bad debt if the proceeds finance an uncertain asset and the guarantee is vague. A lesser-known company can issue acceptable debt if the project has segregated revenue, low leverage, and robust protection. A brand reduces the analyst's work, not necessarily the creditor's loss. The best credit is the one that remains comprehensible after the controller's name is erased from the cover.

My position would be selective in the incentivized debentures and more interested in the infrastructure the new market will demand. Competent fiduciary agents, data platforms, construction monitoring systems, independent technical appraisals, and distributors able to explain risk without selling the exemption as a guarantee. The instrument will grow if trust in the process grows. And institutional trust is produced by verifiable repetition, not by advertising.

I would avoid issues where the tax incentive serves to mask insufficient compensation. When demand is strong because investors compute only the net return, the issuer gains the power to compress the premium beyond reason. The exemption, created to enable infrastructure, can become shareholder rent if the price does not reflect risk. The State forgoes revenue, the investor accepts less, and the controller captures the difference. Public policy without additionality measurement is generosity with an unknown recipient.

The law also exposes an incoherence. The country accepts subsidizing the investor to finance priority projects, yet lets smaller productive companies pay rates that make equally necessary investment unviable. A regional factory, a network of clinics, or a distributor may not fit the official infrastructure label, though they generate jobs and capacity. The political criterion defines what deserves design. The rest is sent to the branch and blamed for not growing.

I do not propose extending the exemption to all debt. That would merely turn the tax system into a catalog of exceptions. I propose learning from the mechanism. The incentive works because it changes who finances, how, and for how long. Corporate credit should pursue the same principle without depending on the benefit: bringing investors closer to the asset, separating risks, creating standardization, and making the obligation tradable.

The transformation will be slow because banks will not abandon margins and companies will not learn structure spontaneously. The bank sells simplicity: sign here, offer the personal guarantee, and receive tomorrow. Architecture charges preparation: organize contracts, reveal data, accept covenants, and allow monitoring. Many entrepreneurs will prefer to pay dearly to remain opaque. They will call it agility. In truth, they are buying the right not to know themselves.

The capital market is not morally superior to the bank. It can be crueler, more procyclical, and more impersonal. Its advantage is different: it allows decomposing and distributing risks instead of keeping them inside a single institution. When well designed, it connects long tenors to savers who can bear them. When badly designed, it connects narrative sellers to yield buyers. Structure determines which version prevails.

Law 12,431 will be remembered as an infrastructure incentive. Perhaps its broader lesson is that the cost of capital changes when the channel changes. The Brazilian entrepreneur was taught to negotiate the rate within the same channel, like someone haggling over the price of water without looking for another source. The debenture shows that the correct question precedes the price: who is the natural creditor of this asset?

An insurer can accept a tenor a bank will not. A fund can buy a granular portfolio an individual investor cannot analyze. A family may prefer long-term tax-exempt income. A supplier can finance inventory if he controls the sale. Each risk has a more suitable buyer. The structurer's job is to find him and remove the unnecessary reasons for charging additional fear.

The upper floor received a subsidy to design because it already spoke the language of design. The lower floor stayed on the rate table because nobody translated its operation. The next great Brazilian credit market will not be created by forcing everyone to pay the same rate. It will be created by making thousands of small risks legible enough to stop depending on the same branch.

Leo Bentier