tradefloor

Formerly pretium. Versions through 0.4.3 were published under that name and remain installable forever; results recorded against them replay under those exact versions. The rename changed no behaviour: 0.5.0 reproduces the same known-answer digest on every platform.

A market simulator you can run a strategy against, with a Rust core and a Python API.

Give it a seed and a list of companies. It runs a market forward: prices, a limit order book, fills, and an economy that moves each day. Your orders match against real depth, so your trades move the price.

Real market data cannot tell you what happens if you trade differently, or what caused a move. This can, because it computed every price.

Documentation

The full documentation lives at https://tradefloor.dev: install, core concepts, the realism envelope, presets, the API surface, and the notebooks, whose source is maintained in a separate repository.

Install

pip install tradefloor

Wheels for Linux, macOS and Windows on CPython 3.11+. No dependencies. The same engine is a Rust crate: cargo add tradefloor.

The API can still change before 1.0. A published result cannot: new coefficients ship as a new preset, so a run you cited last month replays this month.

The demo

python examples/rate-shock/counterfactual.py

Run an agent in a controlled market, checkpoint the world and fork it, then raise rates by 200bps in one branch and compare what the same agent does next.

One changed variable, in the same world and with the same agent. The run takes two seconds and needs no keys and no network. It prints the nine checks that prove the two branches started identical, the step at which the agent's behaviour changed, and the side-by-side. The walkthrough is Your first counterfactual experiment.

First run

import tradefloor as tf

universe = tf.Universe.random(40, seed=111)

spec = tf.StrategySpec.momentum(lookback_days=1.0, top_k=5)
scores = tf.evaluate({"mine": spec}, seed=7, universe=universe, days=10)

scores["mine"].return_pct            # what it made
scores["mine"].impact_bps            # what its own footprint cost
scores["mine"].strategy_fingerprint  # sha256, cite this

That is one market draw. It tells you as much about the seed as about the strategy. tf.rank runs many seeds and compares them with a paired sign test.

Contents

No historical dataset carries the labels truth() gives you. Real data shows you that a stock fell. It never shows you that 60% of the fall was order flow.

To drive it from an agent:

pip install "tradefloor[mcp]"
claude mcp add tradefloor -- tradefloor-mcp

Strategies, universes and scenarios are data, so a tool argument cannot reach code. Each result carries its own caveats. See the MCP page.

Controlled scenarios

scenario = tf.Scenario.load("liquidity_crisis")   # ships with the package

control, stress = tf.branch(engine, 2)
for day in range(80):
    scenario.apply(stress, day)
    ...                                  # run both branches

A scenario is an explicit collection of market interventions and assumptions, for controlled experiments rather than for forecasting. It names targets from a registry of fields the engine actually reads, and it keeps what it asserts happened apart from what it assumes happened next:

tradefloor scenario show scenarios/oil_price_spike.yml

Exogenous shocks
  day 50+            commodity.oil            x1.4

Assumed transmission
  day 55..74 ramp    macro.inflation          +1.50pp
  day 55+            macro.corporate_yield    +0.50pp

tradefloor does not claim what a war, an election, an oil shock or a recession will do to markets. It lets you state those assumptions and measure how an agent behaves under them.

Six scenarios ship inside the package, so Scenario.load works on a plain pip install -- each one recording what it was measured to be worth. Their source is here. tradefloor scenario list names them, and tradefloor scenario targets lists every target and what it actually reaches.

Realism

tf.facts.measure() scores fourteen statistics against real-market bands. The default preset, pt-v16, holds all fourteen at one year and at two years on each of twenty-six seed blocks, thirteen of which no search ever touched, with no out-of-band row anywhere.

Five of the fourteen were calibration targets, and the bands both tuned the model and graded it. So this is a stated envelope, not a test against market data that was held back.

Five limits are measured and written down:

tf.envelope.check() refuses a question that falls outside a limit, and the realism envelope says what each one forbids.

Good results here do not predict real returns. The prices come from a known model, so a strategy shaped like that model looks excellent and teaches you nothing. A strategy that fails here tells you more. There is one venue, no latency, and no counterparty that adapts to you.

Seed determinism

Each release builds five targets, runs one fixed simulation in each, and compares digests. A disagreement stops the release. The crate ships its own exp, log, sin and cos, so the platform libm cannot change a result.

pt-v16 became the default at 0.6.0, taking it from pt-v14. Naming your preset explicitly makes a run replay exactly, and every preset from pt-v1 on is still selectable.

eng = tf.Engine(seed=42, universe=u, model="pt-v10")

Examples

Twelve numbered examples/ in reading order, run by the test suite:

Beside them, one directory per self-contained study. Each asks one question, and they come in no particular order.

FinRobot integration

Evaluate a FinRobot financial AI agent inside a controlled Tradefloor market. Run shared history, checkpoint the world, fork it, raise rates by 200bps in one branch, and compare how the same agent responds. The canonical rate-shock experiment, with the agent swapped and nothing else moved.

pip install "tradefloor[finrobot]"
python examples/finrobot/rate_shock.py            # replays a real recorded run
python examples/finrobot/rate_shock.py --live     # calls FinRobot

The default run replays a genuine recorded FinRobot run and needs no API key, no network and no FinRobot install.

• examples/finrobot/rate_shock.py
• examples/finrobot/rate_shock.ipynb

FinRobot is a project of the AI4Finance Foundation, licensed Apache-2.0. The notebooks above are a Tradefloor integration for FinRobot, maintained in this repository without endorsement from AI4Finance, and they form no part of FinRobot's own interface.

More

Full docs: simoncoombes.github.io/tradefloor.

To contribute, see CONTRIBUTING.md and RELEASING.md. One rule shapes the rest: a change to the simulated trajectory is a breaking change, whatever its size.

To cite the software, see CITATION.cff. To cite a result, use its RunManifest.

Licence: MIT OR Apache-2.0, at your option. See LICENSE-MIT and LICENSE-APACHE.