The thesis

The widest spread in Brazilian credit right now isn't compensation for risk. It's compensation for complexity: for the structuring work most desks won't do. Brazil's mid-market is full of companies that rate institutional but don't fit a standard template, and the firm that can analyze and structure those deals at speed captures a premium without taking on more credit risk. That is the trade. It is under-priced because almost nobody is set up to run it.

Where the idea comes from

Marc Rowan built Apollo on a single principle inherited from the Drexel/Milken school: "find those areas where you're not compromising on credit risk but you're willing to do something that may have a little more complexity in it or that has a little less liquidity, yet still maintains the same investment-grade rating." His phrase for it is excess return per unit of risk.

The mechanism is plain once you see it. Public markets pay a premium for things that trade daily and fit a template. The mirror image of that premium is a discount on anything complex or illiquid. A firm that can out-analyze the rating agency and out-structure the market collects that discount as spread, at the same credit quality. John Zito, Apollo's co-president, has a line for the upside: the Wikipedia page for french fries runs about 4,000 words; the page for private credit runs about 500. Credit can be as varied as food, once you stop forcing it into standard products.

Why Brazil's mid-market is the pile

Every credit desk has a "too-hard pile": deals that are creditworthy but don't fit the assembly line. In Brazil that pile is enormous. The corporate and structured credit market is a fraction of what an economy this size should support, and the missing portion isn't missing capital. It's deals that never get structured.

A mid-market company in Brazil frequently rates institutional on the fundamentals and still can't reach the market, because the deal needs a covenant package built from scratch, or a real read on a sector the agencies cover thinly. The work to bring it to market doesn't scale with headcount, so the traditional desk passes. The company stays unfunded, or pays bank spreads that don't reflect its actual risk. That is the complexity premium, and it sits uncollected.

What the track record shows

The clearest evidence that the trade works is who shows up in the book. Bamboo has structured R$900M+ across 25+ deals, and about 60% were first-time institutional issuers.

That last number is the thesis in one statistic. First-time issuers are, almost by definition, the companies that didn't fit a template, the ones the market hadn't bothered to structure for. Bringing them to institutional investors at institutional ratings is the complexity premium expressed as a track record rather than a theory. It is market expansion, not redistribution: these are deals that wouldn't have existed in standardized form.

How the edge actually works

The edge is not cheaper money, and it is not more leverage. It is analytical capacity. The constraint on this trade has always been that out-analyzing the rating and structuring a bespoke deal is senior, expensive, slow work. The thing that changes the math is an intelligence layer: a proprietary data and decision system that lets a small senior team do that work at a velocity a conventional desk can't reach.

This also answers the pricing question that confuses most people about applying AI to finance. You don't price the software. You price the spread the analysis unlocks. A per-seat license captures none of the value; a take on the originated spread captures the thing that was actually scarce. The discipline cuts both ways: roughly nine of every ten deals that come in get turned down. The premium isn't paid for saying yes to complexity. It's paid for knowing which complexity is worth underwriting.

The honest limits

None of this is free alpha, and it's worth being precise about where it breaks.

Apollo runs this at US scale, built over thirty years. Bamboo runs it at Brazilian mid-market scale today. The premium only translates where there's enough deal volume and enough investor sophistication to reward structured analysis with real spread, and not every Brazilian institution prices that way. Many still price off the rating agency's opinion. The thesis gets proven by track record, not asserted.

"Complexity" can also be a polite word for hidden risk. The premium is compensation for work that, done badly, produces correlated losses under stress. The track record is the asset that makes the whole argument credible, and it's the asset most easily destroyed by scaling faster than the analysis can support.

And "industrialized origination," Apollo's manufacturing of credit at volume, is the endgame, not the current state. Near-term, the leverage is the analytical engine compressing the time it takes to underwrite one complex deal, not a factory. The model is proven; the scale is the work ahead.

Regulated activities are conducted by Bamboo Securitizadora S.A. (CNPJ 48.343.871/0001-34), which acts as coordinator of public offerings under its CVM Resolution 161 coordinator license and issues and services CRI, CRA and debentures under CVM Resolution 60. This content is informational and is not an offer, recommendation or promise of returns.