How to decide inside a company when money costs 14% a year

Summary: Brazil’s policy rate sits at 14.25% and the market projects 12% for the end of 2027, with 4.20% inflation and 1.65% growth. With double-digit real interest, idle cash earns too much for any internal project to be approved without a return calculation. This piece breaks down what changes in management, marketing and operations at a Brazilian company over the next eighteen months.

Brazil’s investment rate closed 2025 at 16.8% of GDP, almost exactly where it was in 2023. These four numbers explain most of what a Brazilian company has to decide right now. It is worth breaking down what each one does in practice.

Management and investment: the cost of money reorders the decision queue

Idle cash started earning too much

With the interbank rate near 13% a year, every internal project competes with a risk-free deposit. An investment promising an 18% return, over two years, with a real chance of failing, competes with a certificate paying 13% and asking nothing of anyone.

The effect is a new bar. Projects that used to clear it comfortably now need justification, and projects with a payback beyond 24 months join a queue, because the cost of waiting became lower than the cost of getting it wrong.

The same mechanism explains a figure that looks contradictory: Brazil received US$ 77 billion in foreign investment in 2025, the third largest volume in the world, and gross fixed capital formation still grew only 2.9%. Foreign capital comes in. Domestic capital stays in fixed income.

Anyone selling services to a Brazilian company is benchmarked against the policy rate

Most companies in the country export nothing. They sell to another Brazilian company: software, consulting, logistics, maintenance, engineering, accounting, communication, training.

For them, high interest arrives by an indirect and harsher route. The client assessing a service proposal uses the same yardstick they use for machinery. If their cash earns 13% sitting still, the service has to show a return above that, with a defined timeframe and the math done.

That changes what a proposal contains. Detailed scope and technical quality stopped being enough. The buyer wants to know how long it takes to pay for itself, which cost it removes and which revenue it sustains. In practice the decision cycle stretches, more people join the approval and procurement shows up earlier. A proposal without a return calculation tends to sit still until December.

Why cost-cutting capex beats expansion capex

Adding capacity in a year of 1.65% growth usually produces idle machinery. An investment that cuts energy use, rework or freight, on the other hand, returns regardless of what demand does. The difference is conditionality. Expansion needs the market to cooperate. Cost reduction works on its own.

Cash stopped being leftover

In high-rate cycles, undercapitalised companies show up on the market at low multiples. Whoever holds cash in 2027 will buy a competitor, a client portfolio or a ready-made team at a price that will not come back any time soon. That makes cash management a strategic decision, not a treasury task.

The November-to-February window

The first fiscal announcement from the next administration, whoever it is, sets the exchange rate, the yield curve and the risk appetite for the whole year. The inheritance is already on the table: a nominal deficit near 8.5% of GDP, gross debt around 81% and more than 90% of spending mandated by law.

Between November 2026 and February 2027, irreversible decisions are expensive. Permanent hiring, buying property, a long contract with no exit clause. A variable structure in that window costs more per unit and is worth the difference. For service sellers the window has another effect: the client freezes too. It pays to pull contract renewals forward to October instead of fighting for signatures in January.

Concentration is the silent risk

If one client accounts for 30% of revenue, that is the company’s most urgent problem. It ranks above any growth target. In services, concentration is even more common, because a large contract tends to renew once a year and nobody wants to look at that date. It is worth counting how many clients would have to leave to make payroll unviable. If the answer is one or two, the commercial plan for the year is already defined.

For exporters, the recent memory is useful. Since 22 July 2026, Brazilian products pay an additional 25% to enter the United States, plus 12.5% under the forced labour rule, with steel and aluminium still at 50%. Brazil became the country facing the highest average US tariff in South America.

Currency exposure exists even without exporting

Imported inputs embedded in a domestic supplier, software licences billed in dollars, international freight, spare parts. Plenty of companies that consider themselves fully domestic discover 15% of dollarised cost when they open the spreadsheet. Service companies usually have the most invisible version of this: the tool stack. CRM, cloud, automation, databases, media platforms. Almost all billed in dollars, almost never reviewed. With the currency projected at R$ 5.28 for the end of 2027, that is already in someone’s price.

The supplier queue is open now

BYD has approved 106 Brazilian companies for its Camaçari plant. Stellantis attracted eight new suppliers to the Porto Real hub alone, where it opened a second shift with 800 direct hires. Toyota opens its second Sorocaba plant in November, taking national capacity to 268,000 vehicles a year.

Supplier here does not mean parts only. It means engineering, maintenance, IT, logistics, training, technical communication, occupational health. Every new plant contracts dozens of services nobody associates with cars. Approval takes twelve to eighteen months. Whoever starts in 2027 arrives after the slots are handed out.

Marketing and brand: where the budget shrinks first and should shrink last

Cutting brand first is the most common mistake

When the economy slows, the communication budget is usually the first to shrink. Since everyone does the same, the media auction empties and cost per impression falls. Staying present while others leave is the cheapest way to buy market share. The cost is immediate and the return shows up later. Which is why almost nobody does it.

Pure performance gets expensive precisely now

A squeezed consumer researches more, compares more and takes longer to decide. Every extra step in the funnel raises acquisition cost. Brand is what lowers that cost, because it shortens the consideration step. People who already know the name skip part of the comparison. Moving part of the conversion budget into brand building protects margin in the medium term, even at the cost of volume in the short term.

Brand is what lets you pass inflation through

With inflation projected at 4.20% for 2027 and household income squeezed, every price increase becomes a test. Companies without perceived differentiation can only compete on discount, and discounting under high interest erodes cash twice. Pricing is a consequence of positioning. There is no shortcut here.

The made-in-Brazil argument gained substance

With tariffs multiplying and global chains reorganising, producing in Brazil became relatively more competitive in several segments. That gave economic grounding to a pitch that used to be pure sentiment. It works especially well in B2B and industrial, where the buyer can measure delivery time, logistics cost and supply risk.

Brazilian B2B brands have historically invested little in communication. The competitive space there is larger and cheaper than in consumer markets.

Frequency beats expensive production

On a tight budget, the choice between one big film per quarter and a constant monthly presence has a clear answer in consumer memory. Repetition builds recall. Sophisticated production builds agency portfolios.

Proof over promise

A sceptical consumer and a buyer under pressure respond to evidence: numbers, demonstrations, real cases, side-by-side comparisons. A manifesto without substance burns trust fast in this environment.

The wrong metric kills brand investment

Demanding ROAS from an institutional campaign is the fastest way to end the investment before it works. Brand is measured by brand indicators: branded search, direct traffic, share of search, unaided recall. Mixing the two usually destroys both.

Operational efficiency: the levers that work without depending on demand

Across-the-board cuts destroy the ability to react

Cutting 15% everywhere looks fair and is inefficient. Areas that generate future revenue lose the same as redundant areas, and the company reaches the end of the cycle unable to capitalise on the turn. Surgical cuts take more work and preserve the engine.

Energy is the cost with the highest risk of structural increase

Electricity demand from data centres in Brazil may go from 2.5 GW in 2025 to 13.2 GW by 2035. Grid connection requests to the Ministry of Mines and Energy went from 12 in May 2024 to 52 in June 2025. More demand on the same grid pushes tariffs up. Migrating to the free market and distributed generation stopped being a sustainability topic and became a margin line.

Inventory costs 13% a year

Every additional thirty days of inventory costs roughly 1% of the immobilised value, counting only the opportunity cost of capital. Under high interest, turning fast on a thinner margin beats holding a fat margin on slow turns. That calculation changes the entire commercial policy.

Selling on credit terms is extending credit

A 60-day receivable in a 13% rate environment means financing the client at a rate the company never calculated. Before touching list prices, it is worth looking at payment terms, early-payment discounts and credit policy. Service companies live the worse version of this problem. They pay payroll on the 5th and collect from clients in 45 or 60 days, every month, with no collateral. It tends to be the fastest lever available, and the least used.

Automation changed its argument

Applying artificial intelligence to internal processes is no longer an innovation pitch. It became fixed-cost substitution, which is why it competes with hiring. Where it works best: high volume, high repetition, expensive errors. Support, reconciliation, back office, operational content production.

The contract index matters more than it seems

Different inflation indices can diverge by several percentage points in a year of currency pressure. A long contract indexed to the wrong one erodes margin silently, year after year. In recurring service contracts, the annual adjustment is usually the only chance to correct price all year. Missing that date costs twelve months.

Tax transition calls for simulation, not reading

Brazil’s new consumption tax rollout shifts the burden very unevenly across sectors, and services tend to feel it more. Anyone who has not simulated their own model will discover the effect through cash flow.

A single supplier became geopolitical risk

Depending on one imported supplier, especially from the United States or China, stopped being an operational risk and became a political one. A second source costs margin today and keeps the operation standing when tariffs change overnight.

The decision calendar through 2027

Three traps

Confusing prudence with paralysis. A company that spends 2026 and 2027 only cutting arrives in 2028 with no brand, no team and no new product.

Reading foreign investment headlines as a hot economy. Capital arrives at record volume and fixed capital formation does not rise. They are different things, and the second is what generates demand.

Waiting for clarity. It does not arrive before 2028. The decision will be made with incomplete information, and what you can control is how reversible each choice is.

One observation that fits none of the three parts: Brazilian consumers may see cheaper coffee and beef in 2027 precisely because the product blocked in the United States will pile up in the domestic market. It is relief in the wallet born of a trade balance problem. Food sellers will celebrate the volume and puzzle over the margin.

Frequently asked questions

What is Brazil’s projected policy rate for 2027?

The market projects Brazil’s policy rate at 12% a year for the end of 2027, down from 14.25% in July 2026. With inflation projected at 4.20%, the real rate stays near 7.5% a year, still a double-digit real cost of capital for anyone investing with a risk premium.

Why does Brazil receive record foreign investment and still not grow?

Because foreign direct investment and gross fixed capital formation are different things. Brazil received US$ 77 billion in 2025, the third largest volume in the world, while gross fixed capital formation grew only 2.9% and the investment rate stayed at 16.8% of GDP. Much of the foreign capital buys existing assets instead of creating new capacity.

What is the US tariff on Brazilian products today?

Since 22 July 2026, Brazilian products pay an additional 25% to enter the United States under Section 301, plus 12.5% under the forced labour rule, totalling 37.5%. Steel and aluminium remain at 50% under Section 232. It is the highest average US tariff applied to a South American country.

How much does idle inventory cost under high interest?

Counting only the opportunity cost of capital at a 13% annual rate, each 30 days of inventory costs about 1.07% of the immobilised value. That is 3.21% at 90 days and 6.41% at 180 days, before storage, insurance and loss.

Should you cut the marketing budget in a downturn?

Cutting communication first is the most common mistake. When everyone cuts, the media auction empties and cost per impression falls, which makes staying present the cheapest way to gain share. Brand also lowers acquisition cost by shortening the consideration step, precisely when a squeezed consumer researches and compares more.

This text is informative analysis. It is not an investment recommendation and does not replace a specific assessment of your sector, cost structure and exposure.

Sources: Focus Bulletin (Central Bank of Brazil, 13/07/2026), IBGE, Central Bank of Brazil, UNCTAD, World Investment Report, BNDES, USTR and Ministry of Mines and Energy.