Skip to main content
Pre-release

This release is a pre-release and may not be stable for production use.

XXFIN

About

The xxfin package of the 10x project serves a dual purpose:

  • to illustrate the capabilities of the 10x system by implementing a sufficiently rich example solving an important problem in an important subject domain
  • to provide a complete working implementation of the most foundational valuation problem in the financial domain: calculating today's price of a fixed amount of a future cash flow denominated in any currency.

To achieve this goal, one needs to determine the relevant currency/foreign exchange (FX) rates and discount factors in a manner consistent with the current market data. Hence, we define all the necessary data structures for market data, including:

  • an ingestion/translation facility for collecting market data from any chosen data providers
  • generation/population/storage objects for standard quoted market data
  • models for calibrating complex/"synthetic" objects (forward curves), allowing calculation of rates/values that are not directly observable but are consistent with all the relevant markets

The sequence of the valuation steps is as follows:

  • We create market data objects ("quotables") that contain the market quotes and are stored in a user-specified market data store.
  • The user specifies an external market data provider.
  • An adapter is needed to translate the market data from the provider's format and populate our quotables.
    • We may provide such an adapter for a commonly used provider (e.g., Bloomberg), and, at a later date, add more adapters if demand warrants it.
    • We create a sample set of quotables which allow us to illustrate the use of market data.
  • Each piece of market data is quoted as a number (or a collection of numbers) representing values/parameters of certain standard financial contracts, and the set of specifications sufficient to describe and value such standardized contracts is saved in "market convention" objects.

The subset of all market quotables used to build/calibrate synthetic market data (forward curves) is called a market assembly. It usually consists of the most liquid contracts that provide the most precise view of market participants on the market (e.g., the most liquid swap or FX forward quotes). Once we ingest the market quotes from the data provider(s) into the quotables, we use a market assembly to build/calibrate the forward curve using a chosen calibration procedure. The subsequent use of the curves built this way is completely agnostic to the curve calibration choices. Hence, the market assembly, as well as the calibration analytics, can be replaced by the user's preferred choices.

The discount curve for a discount rate in a given currency is built using that currency-denominated standard cash deposit and fixed-floating swap market quotes, with that rate as the underlying/reference floating rate for the instruments listed in a market assembly (e.g., for the SOFR curve we use USD cash deposits and SOFR standard annual fixed/floating swap quotes).

For FX forward curve construction, we currently use FX spot and FX forward (point) quotes for rates funded/collateralized in the same currency (e.g., USD). We build an FX curve for a market assembly-specified set of FX spot and FX USD-collateralized forwards consistent with the funding currency's discount curve.

Such an FX forward curve allows us to calculate the present value, in any given currency, of a cash flow denominated in any other currency.

Optional JIT acceleration via AADC

xxfin/jit_aadc/ optionally records curve/valuation code into a JIT-compiled adjoint differentiation kernel using MatLogica's AADC (Adjoint Algorithmic Differentiation Compiler), enabled via the use_cxxfin/aadc_license settings in xxfin_env_vars.py. aadc installs freely from PyPI and is a required dependency on non-macOS platforms; it's excluded on macOS because MatLogica doesn't publish an aadc build for that platform. It ships with a time-limited trial license out of the box, but continued or production use requires a license key obtained from MatLogica. xxfin itself works fully without it — AADC only accelerates the JIT-compiled code path.

An easy setup to start playing with xxfin package

  • install mongodb on your machine and run without authentication (default)
  • set environment variables:
    • XX_MAIN_TS_STORE_URI=mongodb://localhost/xx_fin (or any other MongoDb)
    • XXFIN_DEFAULT_PRICING_CONTEXT_NAME=Abu Dhabi 20251010
  • run xxfin/dev_data_helpers/RUN_ME.py (it will store a number of objects)
  • you are ready to go!

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

py10x_fin_base-0.1.0rc4.tar.gz (83.1 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

py10x_fin_base-0.1.0rc4-py3-none-any.whl (85.1 kB view details)

Uploaded Python 3

File details

Details for the file py10x_fin_base-0.1.0rc4.tar.gz.

File metadata

  • Download URL: py10x_fin_base-0.1.0rc4.tar.gz
  • Upload date:
  • Size: 83.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for py10x_fin_base-0.1.0rc4.tar.gz
Algorithm Hash digest
SHA256 cd15d7a652f1ab896f1f1b83dc696df021d08e8d9fa1719ecae38cd5403838e4
MD5 dee5903e6e74ccbc05d09eec1f566596
BLAKE2b-256 c7fe5930f591756ea9a2608eae399b12c0d37e1ea717c28a1f4eabdadb87c992

See more details on using hashes here.

Provenance

The following attestation bundles were made for py10x_fin_base-0.1.0rc4.tar.gz:

Publisher: finbase_wheel.yml on 10x-software/py10x

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file py10x_fin_base-0.1.0rc4-py3-none-any.whl.

File metadata

File hashes

Hashes for py10x_fin_base-0.1.0rc4-py3-none-any.whl
Algorithm Hash digest
SHA256 955fbe0d418aada1198a14e6a1fe60a144b54ab278f08840154f942f7949323f
MD5 f1ead744a58e2f8723a021bceaf1796c
BLAKE2b-256 08cb6e2c635b63c1b3d8b9e868c16f5b4077cd38669f6bb12d8c204a494eaac2

See more details on using hashes here.

Provenance

The following attestation bundles were made for py10x_fin_base-0.1.0rc4-py3-none-any.whl:

Publisher: finbase_wheel.yml on 10x-software/py10x

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page