Cash instruments

Overview

quotes is the whole cash tape. These endpoints are the same rows split by instrument type, so you can ask for one class without filtering client-side.

Endpoint	Instrument type	What it holds
`GET /equities`	`equity`	Ordinary and preferred shares (`especi` `ON` / `PN`)
`GET /bdrs`	`bdr`	Brazilian Depositary Receipts (`especi` `DR*`)
`GET /units`	`unit`	Units — bundled share packages (`especi` `UNT*`)
`GET /fund_quotas`	`fund_quota`	Listed fund quotas: ETFs, FIIs and other `CI*` paper
`GET /cash_securities`	`cash_security`	The residual. Subscription rights, bonus rights and index lines were split out in catalog v12 and are no longer here
`GET /auctions`	`auction`	Auction prints (`tpmerc` `017`), ~210 on the whole 2019–2026 tape. Same columns as `quotes`, minus `lot` / `share_class` / `fund_type`. Events, not a series — never compute returns over them

The type comes from vw_b3_instrument_typed, which classifies only from B3’s published TPMERC and ESPECI. Nothing is inferred from a ticker’s shape. fund_quotas rows carry fund_type — etf | fii | fidc | fiagro — from B3’s published CODBDI (14 → etf, 05/12 → fii, 13 → fiagro), validated against the CVM ETF registry. NULL on boards that carry no family signal, never guessed from the ticker:

/fund_quotas?fund_type=eq.etf&lot=eq.standard

equities rows carry share_class (ON | PN | PNA | PNB | PNC | PND) and governance_segment (NM | N1 | N2 | MA | M2 | MB), parsed from ESPECI and cross-checked against the ISIN class code — never inferred from the ticker’s 3/4 suffix:

/equities?share_class=eq.ON&governance_segment=eq.NM&lot=eq.standard

Lot size is part of the grain

These endpoints carry both lot sizes, with lot derived from tpmerc:

standard — tpmerc 010, the round lot
odd — tpmerc 020 / 021, the fractional market

The grain is (ticker, trade_date, board, term_days, lot) — one column wider than quotes. A query that ignores lot sees what look like duplicate dates.

Odd lot is not a rounding error. Measured 2026-08-27, equities carry more odd-lot rows than standard-lot — 153,072 against 140,227, across 496 codnegs versus 476. Summing volume without pinning lot double-counts. For round lots only: ?lot=eq.standard.

api.quotes is unchanged — still tpmerc = 010 only. These endpoints are additive.

Columns

Identical to quotes plus lot (and share_class + governance_segment on equities, fund_type on fund_quotas), minus asset_class — the endpoint is the type. Everything is unadjusted.

curl -s -G "$SILO_URL/equities" \
  -H "apikey: $SILO_ANON_KEY" \
  --data-urlencode "select=ticker,trade_date,lot,close,volume" \
  --data-urlencode "lot=eq.standard" \
  --data-urlencode "trade_date=eq.2026-08-25" \
  --data-urlencode "order=volume.desc" \
  --data-urlencode "limit=5"

params = {
    "select": "ticker,trade_date,lot,close,volume",
    "lot": "eq.standard",
    "trade_date": "eq.2026-08-25",
    "order": "volume.desc",
    "limit": 5,
}
rows = requests.get(f"{URL}/equities", headers=H, params=params).json()

const { data } = await supabase
  .from('equities')
  .select('ticker, trade_date, lot, close, volume')
  .eq('lot', 'standard')
  .eq('trade_date', '2026-08-25')
  .order('volume', { ascending: false })
  .limit(5)

Price series stay unified

There is deliberately no equity_history or bdr_history. A codneg belongs to exactly one instrument type, so quote_history('PETR4') is already unambiguous — and a typed history would force you to determine the type before you could ask for a price. Split the cross-section; keep whatever is keyed by id. The same rule is why options and termo do get their own history endpoints: those carry columns the cash shape has no room for.

Empty and error behavior

No rows matching your filter → 200 []. Never a fabricated price.
A ticker on the wrong endpoint → 200 []. BOVA11 is a fund_quota, not an equity. Use /quotes, which is untyped, or lookup, which reports the type it resolved.
These views have no built-in cap — always send a limit.

Overview

Data

Get started

Guides

API reference

Endpoints

Cash instruments

Overview

Lot size is part of the grain

Columns

Price series stay unified

Empty and error behavior

​Overview

​Lot size is part of the grain

​Columns

​Price series stay unified

​Empty and error behavior

Overview

Lot size is part of the grain

Columns

Price series stay unified

Empty and error behavior