Radar Perene / Archive / science

What the American finance literature assumes that Brazil does not have

◦ Index methodology v2.2 (working papers with DOI). See the methodology.

Science

Every paper carries two layers: what it claims and what it assumes. The first is debated at conferences, replicated, contested. The second is almost never read — it lives between the lines of the sampling design, in the silent choice of universe, in the "we use all listed stocks" that sounds neutral and is not. When a method crosses borders, the first layer travels for free. The second gets held at customs, and no one tells you.

The empirical finance literature — mostly written about the American market — assumes, without declaring it, a set of structural conditions: thousands of listed assets, deep liquidity in nearly all of them, decades of stable institutional history, populated sectors, and a financing system in which the capital market plays the lead. None of these conditions describes Brazil. Importing the formula without auditing the premises is applying conclusions to a world that is not theirs.

This article maps the broken premises — the diagnosis that precedes any adaptation. How the house adapts is the subject of the next piece in the trail; here, the inventory.

Premise 1: assets in abundance

The founding papers of American financial empiricism operate over universes of thousands of stocks. Portfolios are formed in dozens of buckets, each bucket holding hundreds of names; the individual asset is statistically disposable. The Brazilian universe counts its liquid names in dozens — and it was not always even that: in the early 2000s, the set of stocks with usable liquidity for this kind of research was about a dozen; today it approaches eight dozen. In numbers: from 12 to 82 stocks since 2000, by the count the house uses in its own breadth work — real growth, and still two orders of magnitude below the universe the original formula assumes. Every portfolio, bucket and average technique loses its legs at that scale.

Premise 2: liquidity as landscape

In the source literature, liquidity is landscape: it exists in sufficient quantity for the closing price to be fresh information in practically every name. In Brazil, outside the most traded core, the closing price can be an echo from days ago — and techniques that treat every price as a valid observation import, along with the formula, a fiction that each day brings new data. The problem reaches collective measures: sector series computed over barely trading components look stable exactly where nothing is being measured.

Premise 3: a long history in one piece

American empiricism rests on series that cross a century under the same essential monetary regime and continuous institutions. The useful Brazilian series, for many purposes, begins after monetary stabilization — and even that short stretch is sliced by structural changes that make the local "long run" fit into a few barely comparable regimes. Patterns that need many decades to manifest simply have nowhere to manifest here; tests that need many observed cycles operate with two or three.

Premise 4: sectors as populations

The imported sector taxonomy assumes that each label — energy, consumer, healthcare — contains a population of companies. In Brazil, entire labels are occupied by two or three liquid names, and some by one. The consequences for baskets and rotation were examined in the previous articles of this trail; the broken premise is the same: the American sector is a collective, the Brazilian one is often a club too small to have statistics of its own.

Premise 5: the capital market at the center of financing

The least visible premise is institutional. Much of the literature assumes that companies finance themselves primarily through the market — equities and traded corporate debt — and that the prices of those instruments therefore carry the central information about the economy's cost of capital. In Brazil, bank credit occupies a space the American capital market does not know, with a system concentrated in a few institutions and spreads that are a public policy topic in their own right — Track 1 covers that structure in credit and bank spread. Formulas that read the cost of capital from market prices see, here, only a slice of the film.

The inventory before the tool

Nothing in this map says the American literature is useless for Brazil — the questions it formulates and the rigor with which it attacks them remain the discipline's standard. What the map says is that every imported method arrives with a list of undeclared dependencies, and that the first task of anyone researching a small market is to write that list before running any regression. The house learned this the empirical way: its own breadth and sector-basket studies were born from formulas that failed locally before being rebuilt. The diagnosis of the premises is public by nature — it is above. The rebuilding is craft.

Frequently asked questions

Do these broken premises apply to other emerging markets?

To varying degrees, yes — few assets, thin liquidity, short history and underpopulated sectors are the normal condition outside deep markets. Brazil has particularities (the weight of bank credit among them), but the inventory serves as a template for any small market.

Does validating on American data and then applying to Brazil solve it?

No — it validates the formula in the world where it already worked. Transferring the result still depends on exactly the premises Brazil does not meet; the relevant validation is local, with the sample limitations stated plainly.

Doesn't Brazilian academia correct this?

Part of it does, and there is good national research sensitive to these constraints. But the dominant literature's sheer volume means the imported standard tends to win by inertia — including in applied material, where the audit of premises almost never appears.

Won't the growth of the Brazilian market retire the problem?

It softens the first premise and none of the other four. Liquidity, history, sector population and financing structure change on a scale of decades — and research is done with the market that exists, not the one it may become.

---

Continue the trail: Adapt or import: the decision that precedes any Brazilian sector index →

House reading: today's reading is in the Diário; the episodes of the market that exists, in the Atlas.