Finance Function Reference

This document is generated from the extension registration surface in src/ and the executable SQL coverage in test/sql/. All functions live in the fin_ namespace and use standard DuckDB types such as DOUBLE, VARCHAR, DATE, TIMESTAMP, LIST, STRUCT, and table results.

Usage Conventions

• Load the extension before use. After community publication, use INSTALL finance FROM community; LOAD finance;; for current source builds, use the repository Makefile targets to build and load the unsigned local extension during validation.
• Scalar functions are called in ordinary SELECT expressions.
• Aggregate macros are called over grouped rows, for example SELECT fin_total_return(r) FROM returns;.
• Table functions and bind-replace functions appear in FROM, for example SELECT * FROM fin_calendar('weekday', DATE '2026-05-04', DATE '2026-05-08');.
• Several broad compatibility entries are v1 aliases or placeholders. Those entries are documented explicitly as placeholders and are covered by tests so future implementations can be swapped in behind stable names.
• make check verifies this reference against every public registered fin_* function and verifies that every registered function appears in the gold SQL tests. The checks scan split .cpp and .inc source units under src/.

Compatibility Aliases

These entries preserve familiar helper names where a SQL shape is practical. Most are aliases over the canonical fin_* implementation already documented below; entries marked as v1 placeholders intentionally keep a stable callable surface while local SQL-native semantics mature.

GS Quant Portfolio Measure Compatibility

These names mirror gs_quant.timeseries.measures_portfolios. They operate on local SQL series supplied by the caller instead of fetching entitled Marquee portfolio reports.

GS Quant Report Measure Compatibility

These names mirror gs_quant.timeseries.measures_reports and use local SQL series supplied by the caller.

GS Quant Market Measure Compatibility

These entries represent smaller GS Quant timeseries market-data modules whose Python implementations fetch entitled Marquee datasets. The DuckDB forms operate on local SQL series supplied by the caller and keep exact source names where possible; display-name aliases are included when GS Quant exposes a different plot name.

GS Quant Rates Measure Compatibility

These names mirror additional decorated functions in gs_quant.timeseries.measures_rates. The Python source retrieves rates, swaption, curve, OIS, and policy-rate datasets. The DuckDB forms operate on local SQL series and preserve the market-data parameters as optional metadata arguments.

The first argument is always the local numeric series to summarize. Metadata arguments such as swap_tenor, benchmark_type, expiration_tenor, termination_tenor, source, and real_time are accepted for query compatibility and filtering conventions; they do not fetch Marquee data.

WITH curve AS (
  SELECT * FROM (VALUES
    (DATE '2026-01-02', 'USD-SOFR', '5y', 0.0361),
    (DATE '2026-01-03', 'USD-SOFR', '5y', 0.0364)
  ) AS t(as_of_date, benchmark_type, swap_tenor, rate)
)
SELECT
  fin_swap_rate_calc(rate, swap_tenor := swap_tenor, benchmark_type := benchmark_type) AS latest_swap_rate,
  fin_swap_term_structure(rate, benchmark_type := benchmark_type, tenor := swap_tenor) AS curve_point
FROM curve;

GS Quant Data Vendor And Risk Model Compatibility

These names mirror decorated GS Quant functions for cognitive credit, countries, FactSet/GIR, and risk-model datasets. The SQL forms operate on local series supplied by the caller.

GS Quant General Measure Compatibility

These names mirror the remaining decorated functions in gs_quant.timeseries.measures. They are local-data equivalents for GS Quant market, fundamentals, ESG, thematic, forecast, and commodity dataset wrappers.

Function Index

Numerical And Money Helpers

Returns, Risk, And Statistics

Fixed Income, Rates, And Cash Flows

Options And Volatility Models

Technical Indicators And Microstructure

Portfolio, Matrix, And Factor Analytics

Validation, Parsers, Specs, And Calendars

Table Functions And Time-Series Builders

General Helpers

Testing

The reference surface is exercised by make test, which builds the extension, runs smoke SQL, loads test/sql/gold_dataset.sql, and evaluates test/sql/gold_tests.sql. The gold dataset is intentionally small and deterministic so expected values are easy to audit.

For model and unit conventions, see Quant Developer Guide. For workflow-oriented examples, see Finance SQL Playbooks.

GS Quant-Inspired Pricing And Risk Helpers

These helpers port the parts of GS Quant-shaped pricing-and-risk workflows that fit a local SQL extension: instrument descriptors, risk-measure descriptors, pricing context descriptors, scenario transforms, and portfolio aggregation. They are independent DuckDB implementations and do not call Goldman Sachs APIs. See Compatibility for the supported boundary.

Compatibility Examples

Price and risk an equity option descriptor:

WITH option AS (
  SELECT fin_gsq_eq_option('call', 'SPX', 100.0, 100.0, 1.0, 0.05, 0.20) AS inst
)
SELECT
  fin_gsq_eq_option_price(inst) AS price,
  fin_gsq_eq_delta(inst) AS delta,
  fin_gsq_eq_gamma(inst) AS gamma,
  fin_gsq_eq_vega(inst) AS vega
FROM option;

Use a risk-measure descriptor for local dispatch:

SELECT fin_gsq_calc_eq_option(
  fin_gsq_eq_option('call', 'SPX', 100.0, 100.0, 1.0, 0.05, 0.20),
  fin_gsq_measure_price('USD')
) AS price;

Apply a GS Quant-style market-data shock:

SELECT fin_gsq_apply_shock(
  0.20,
  fin_gsq_market_data_shock('Absolute', 0.0001)
) AS shocked_vol;

Aggregate a local portfolio result set:

SELECT
  portfolio_id,
  fin_gsq_portfolio_value(value, quantity) AS portfolio_value,
  fin_gsq_portfolio_risk(risk, quantity) AS portfolio_risk
FROM gsq_goldman_portfolio_cases
GROUP BY portfolio_id;

GS Quant-Inspired Golden Dataset

The deterministic fixtures in test/sql/gold_dataset.sql use the prefix gsq_goldman_. They cover equity, FX, rates, inflation, credit, scenario, and portfolio cases. The expected values are constants, and test/sql/gold_tests.sql compares public fin_gsq_* functions against those constants with explicit tolerances.