Spreads do not fall by speech; they fall by structure. They attacked the price and forgot the design.

The Brazilian spread contains excess profit, but it also contains the cost of our bad institutions — and mixing the two protects both: the bank uses complexity to justify margin, the politician uses margin to avoid complex reforms.

The president has decided to attack the banking spread. Public banks were summoned to cut rates, private banks were pressured to follow, and the country received the message that high rates are a choice that can be corrected by political will. There is some truth in that. There is also the usual confusion between price and structure. Spreads do not fall sustainably by speech; they fall when the creditor needs less compensation for capital, cost, loss, illiquidity, and uncertainty. They attacked the price and forgot the design.

Spread is a useful word because it mixes everything. Inside it fit taxes, reserve requirements, administrative expenses, defaults, slow judicial recovery, cost of capital, banking concentration, margin, and plain market power. By treating the whole as abuse, the government obtains a visible enemy. By decomposing the whole, it would find hard problems that do not fit on a stage: fragmented registries, collateral of uncertain enforcement, poor financial information, limited competition, and companies that mix the business's cash with the owner's wealth.

Public banks can cut rates immediately because they do not answer to the same incentives as a private bank. They can accept smaller margins, use favored funding, take direction from the state controller, and treat portfolio expansion as an instrument of economic policy. That forces competition and can trim fat. But it can also replace price with fiscal risk. When the rate falls without the expected loss or the recovery cost falling, someone keeps paying the difference. It just does not appear in the debtor's contract.

The government will observe a reduction and declare victory. The private bank will respond in selected products, for selected clients, protecting margin through fees, insurance, reciprocities, and relationship requirements. The entrepreneur will see a lower rate in the advertising and discover it does not apply to his company. The system will become cheaper on average and remain arbitrary in the concrete case. Averages are excellent communication instruments because no borrower lives inside them.

The rate on a corporate loan is the end of a chain. Before it comes the question about the source of payment. Does the bank finance the company's general cash or a specific receivable? Is there linked collateral? Does the obligation mature when the asset generates cash? Is there monthly information? Does the creditor control the account? Does the debtor offer economic subordination through his own capital? Without answers, the bank charges for the unknown. With market power, it charges even more. Reducing only the abusive share requires knowing the necessary share.

Brazilian shelf credit works because it is cheap for the bank to produce. A standard contract, a score, the owners' personal guarantee, and blanket collateral allow fast decisions. The rate is high, but the cost of individual analysis is low. A designed operation could cost the debtor less and offer the creditor better risk, but it demands upfront work. As long as that work costs more than the margin saved, the small business will keep buying the bad product. The problem is industrial, not merely moral.

There are four durable ways to reduce spread. The first is better information, distinguishing good and bad risks before everyone pays for the average. The second is better collateral, reducing loss when the forecast fails. The third is broader competition, letting different sources of capital dispute the asset. The fourth is matching structure to flow, reducing the chance that a good business breaks under a mismatched liability. A presidential pronouncement can accelerate the third through public banks. It does not replace the others.

The government could use its influence to standardize registries, integrate data, accelerate enforcement, and support receivables markets. It could demand transparency of the effective cost, lower entry barriers, and encourage corporate portability. It could make a guarantee offered to one creditor visible and comparable to another. That would reduce the rent of opacity. It is slower than announcing a new rate, therefore less politically attractive.

The obsession with price produces a perverse effect: companies start hunting for the cheapest line without asking whether it is adequate. Subsidized working capital can be used for long-term expansion. A lower rate justifies higher leverage. The borrower reads the offer as a certification of solvency. When the policy changes, the liability remains. The cheap product can be the most expensive debt if it creates dependence on renewal.

I would not refuse a cheap public line on principle. Moralism is a bad investment strategy. If the capital is available and the risk of policy change is manageable, the shareholder should consider it. But I would treat the benefit as temporary and value the company without it. If the return on capital disappears when the rate returns to market, the business is not productive; it is subsidized. A valuation based on political favor should use political duration, not perpetuity.

As a creditor, I would distrust accelerated growth in portfolios pressured to charge less. Margin cuts tend to stimulate volume, and volume can hide deterioration for a while. Public banks are not exempt from adverse selection. When they offer the best price, they also attract the borrowers private banks refused for the right reasons. The public mission may justify accepting more risk, but the risk does not cease to exist because it received a social name.

The public-versus-private dispute also obscures the absence of a third way: a capital market accessible to the mid-sized company. The entrepreneur alternates between banks that look identical and concludes there is no competition. There is little competition because everyone finances the same generic object, the company's balance sheet, with similar contracts and cumulative guarantees. If receivables, inventory, equipment, and contracts were financed separately, different creditors could compete for different parts of the risk.

That requires the debtor to reveal more and accept discipline. The branch charges dearly, but it allows a form of operational privacy: the bank looks just enough to protect itself and demands a personal guarantee for the rest. A cheaper structure may require an escrow account, portfolio audits, concentration limits, and amortization triggers. Some entrepreneurs will reject those conditions and blame the market for the spread. In truth, they prefer to pay for the freedom to remain disorganized.

Technology will reduce that cost. Banking, tax, and commercial data will be processed continuously. Receivables will be verified instead of declared. Collateral will be registered with less friction. The creditor will price current behavior, not just an old balance sheet. When that happens, the good debtor will stop subsidizing part of the bad one. The reduction will come from distinction, not forced solidarity.

The policy of 2012 may manage to demonstrate that banking margins are not immutable. That is useful. The mistake would be concluding that pressure solved the problem. Banks respond to incentives; if a product stops paying, they reduce supply, change composition, or charge elsewhere. The State can sustain an artificial price for a while, but it cannot force private capital to ignore returns forever. Credit migrates like water: it goes around the dam.

A serious agenda would measure total cost and availability by profile, tenor, and collateral, not just an average rate. It would ask how many companies remain excluded, how long recovery takes, what share of the spread stems from concentration, and how much could be reduced by better documentation. Without that, each side picks a number: the government shows the decline, the banks show the defaults, and the entrepreneur shows the contract that remains expensive. Everyone is right within his own frame and nobody changes the machine.

My position in bank stocks would be cautious, not apocalyptic. Political pressure can compress margins and redistribute share toward public banks, but private institutions retain advantages in selection, services, deposits, and repricing capacity. The bigger risk lies in banks or finance houses that grow to defend market share while loosening credit. Lower margin with stable loss is competition. Lower margin with future loss is accounting postponement.

On the corporate side, I would favor companies able to finance specific assets and reduce dependence on the branch. Companies with good receivables, long contracts, unencumbered properties, and financial governance can create alternatives. The rest will keep negotiating a few points while handing guarantee, wealth, and future to the same creditor. The advantage will not be finding the friendliest manager. It will be arriving at the conversation with organized risk.

The Brazilian spread contains excess profit, but it also contains the cost of our bad institutions. Mixing the two protects both. The bank uses complexity to justify margin; the politician uses margin to avoid complex reforms. The entrepreneur pays for the involuntary alliance. The way out requires removing the layers one by one, until the remaining price can be genuinely contested.

They attacked the price because it is visible. They forgot the design because it distributes responsibility among registries, courts, regulators, banks, companies, and investors. There is no single villain, therefore there is no simple speech. But credit does not respect speeches. It respects precedence, information, and recovery. The rate can fall by decree for a season. Fear only falls when it knows exactly what it receives if it is right and what it recovers when it is wrong.

Leo Bentier