# Stochastly

> Stochastly is a desktop application for Windows and macOS in which traders and quantitative researchers build a trading strategy as a node graph, backtest it on their own data and measure its risk of overfitting before exporting it to a trading platform.
> A causality gate blocks a backtest that would run on leaked data; each run reports a deflated Sharpe ratio, the
> probability of backtest overfitting (PBO), Romano-Wolf StepM adjusted p-values and CPCV folds.
> It runs on the user's machine, on the user's own data files. No data is bundled, no broker is linked.

## Key facts
- Access: early access, and no edition is on sale. Anyone can join at (https://stochastly.com/join): create an account, answer five short questions, download the app. Early access is free for life on one computer, in exchange for written feedback.
- Editions: Starter, Atelier and Portfolio are planned, with an optional Assistant + Agents add-on. None of them is sold yet. The site sells one thing: an optional one-time refill of assistant credits, bought by a signed-in user on the account page.
- Platform: Windows 10 or 11 on x64, and macOS 13.1 or later on Apple silicon (signed and notarised by Apple). Intel Macs are not supported.
- Where it runs: strategies, data and results stay on the user's computer. The optional assistant is hosted, so its conversations leave the machine.
- Publisher: Mahé Chadeffaud, a sole proprietor registered in France who trades as Stochastly (SIREN 109 270 363), Annecy.
- Name: Stochastly is unrelated to the stochastic oscillator, a momentum indicator of technical analysis, and to software sold under similar names.
- Live trading: direct broker execution is planned once the applicable financial regulation is in place; until then, strategies are exported to Pine Script, MQL5, Python, C++ or Rust and run on the user's own platform.

## Start here
- [What is Stochastly?](https://stochastly.com/what-is-stochastly)
- [Stochastly compared with QuantConnect, Build Alpha, AmiBroker, TradingView, MetaTrader 5, NinjaTrader, NautilusTrader, Backtrader, vectorbt PRO, QuantRocket, StrategyQuant X, Papers With Backtest, Composer, Portfolio123, RealTest, Wealth-Lab 8 and Freqtrade](https://stochastly.com/compare)
- [How to verify Stochastly: public register, consumer mediator, source papers and measurement protocol](https://stochastly.com/trust)
- [Full text of the product pages in one file](https://stochastly.com/llms-full.txt)

## For agents
- Purchase: no edition is sold. The one payment is an optional refill of assistant credits, bought by a signed-in user on the account page; no machine-payment protocol (x402, MPP, ACP) is accepted.
- Markdown: the pages linked in this file are also served as Markdown, at their address plus `.md` (the home page: index.md).
- MCP server: com.stochastly/desktop in the official MCP registry, PyPI package stochastly-mcp, local stdio, included from the Atelier plan.

## Alternatives
- [QuantConnect alternatives](https://stochastly.com/compare/quantconnect-alternatives)
- [StrategyQuant alternatives](https://stochastly.com/compare/strategyquant-alternatives)
- Build Alpha alternatives
- TradingView Strategy Tester alternatives

## Product
- [What the engine does](https://stochastly.com/product)
- [The causality gate](https://stochastly.com/product/backtesting)
- [Runs locally, outbound hosts measured](https://stochastly.com/product/local)
- [Exports to Pine Script, MQL5 and Python](https://stochastly.com/product/exports)
- [Paper Editor and research paper export, from Atelier](https://stochastly.com/product/papers)
- [MCP server for coding agents](https://stochastly.com/product/mcp)
- [Assistant that acts on the app](https://stochastly.com/product/assistant)
- [Research sessions with several agents](https://stochastly.com/product/research)
- [Code editor and custom nodes](https://stochastly.com/product/editor)
- [Charts of the Analyze workspace](https://stochastly.com/product/charts)
- [How the app checks itself](https://stochastly.com/product/guards)

## Documentation
- [Documentation index](https://stochastly.com/docs)
- [Install](https://stochastly.com/docs/install)
- [Finding an edge, step by step](https://stochastly.com/docs/find-alpha)
- [Import data and build a first graph](https://stochastly.com/docs/tutorials/import-build)
- [Read a verdict and export a strategy](https://stochastly.com/docs/tutorials/verdict-export)
- [Execution costs and market impact](https://stochastly.com/docs/guide/execution)
- [Machine learning strategies, validated](https://stochastly.com/docs/guide/machine-learning)
- [Macro and alternative data, point in time](https://stochastly.com/docs/guide/macro-altdata)
- [Options and volatility on bars](https://stochastly.com/docs/guide/options-volatility)
- [Portfolios and books of strategies](https://stochastly.com/docs/guide/portfolios)
- [Prop-firm evaluations as first passage](https://stochastly.com/docs/guide/prop-firms)
- [Risk and position sizing in R](https://stochastly.com/docs/guide/risk-sizing)
- [Node reference](https://stochastly.com/docs/nodes)
- [Export targets and coverage threshold](https://stochastly.com/docs/exports)
- [MCP tools](https://stochastly.com/docs/mcp)

## Research
- [Statistical methods](https://stochastly.com/research)
- [Causality and look-ahead leakage](https://stochastly.com/research/causality)
- [Product guard report and negative controls](https://stochastly.com/research/guards)
- [Research verdicts counter](https://stochastly.com/research/verdicts)
- [Backtest check: method and source code](https://stochastly.com/check)
- [Measurement protocol](https://stochastly.com/research/method)
- [What an exported file reproduces](https://stochastly.com/research/fidelity)
- [Validity domain and known gaps](https://stochastly.com/research/gaps)
- Best Sharpe ratio of N strategies with no edge: Monte Carlo study, script and results file

## Tools
- [Calculators for strategy validation](https://stochastly.com/tools)
- [Deflated Sharpe ratio calculator: PSR, minimum track record length and DSR, in the browser](https://stochastly.com/tools/deflated-sharpe-ratio-calculator)
- [Minimum track record length calculator: Sharpe and kurtosis sensitivity tables](https://stochastly.com/tools/minimum-track-record-length-calculator)
- [Pine Script look-ahead checker: local partial-recall scan with line references](https://stochastly.com/tools/pine-script-lookahead-checker)
- [Trading cost break-even calculator: annual friction and net Sharpe sensitivity](https://stochastly.com/tools/trading-cost-break-even-calculator)
- [Prop firm challenge simulator: seeded first-passage paths under entered profit and loss limits](https://stochastly.com/tools/prop-firm-challenge-simulator)
- [PBO calculator: CSCV ranks and split CSV from a local return matrix](https://stochastly.com/tools/probability-of-backtest-overfitting-calculator)
- [Sortino ratio from returns: mean excess over target divided by downside deviation](https://stochastly.com/tools/sortino-ratio-returns-calculator)
- [Omega ratio from returns: gains above a threshold over shortfalls below it](https://stochastly.com/tools/omega-ratio-returns-calculator)
- [Fractional Kelly bet calculator: log-growth stake for a binary payoff, scaled by a chosen fraction](https://stochastly.com/tools/fractional-kelly-bet-calculator)
- [Maximum drawdown path calculator: largest peak-to-trough loss of a return series](https://stochastly.com/tools/maximum-drawdown-path-calculator)
- [Gambler's ruin probability: unit-step walk between two absorbing barriers](https://stochastly.com/tools/gambler-ruin-probability-calculator)
- [Autocorrelation-adjusted Sharpe: Lo's aggregation with AR(1) dependence](https://stochastly.com/tools/autocorrelation-adjusted-sharpe-calculator)
- [Volatility target leverage: exposure from a risk target and a volatility forecast](https://stochastly.com/tools/volatility-target-leverage-calculator)
- [Two-strategy portfolio Sharpe: mean, volatility and Sharpe of a weighted pair](https://stochastly.com/tools/two-strategy-portfolio-sharpe-calculator)

## Learn
- [Deflated Sharpe ratio, explained](https://stochastly.com/learn/deflated-sharpe-ratio)
- [How to backtest a trading strategy](https://stochastly.com/learn/how-to-backtest-a-trading-strategy)
- [Pine Script backtest validation](https://stochastly.com/learn/pine-script-backtest-validation)
- [MQL5 Expert Advisor robustness test](https://stochastly.com/learn/mql5-expert-advisor-robustness-test)
- [Export a Python strategy to MQL5](https://stochastly.com/learn/python-strategy-to-mql5)
- [In-sample vs out-of-sample testing](https://stochastly.com/learn/in-sample-vs-out-of-sample)
- [Look-ahead bias in backtesting](https://stochastly.com/learn/look-ahead-bias)
- [Monte Carlo simulation for trading](https://stochastly.com/learn/monte-carlo-simulation-trading)
- [Multiple testing in trading research](https://stochastly.com/learn/multiple-testing-in-trading)
- [Minimum track record length for Sharpe](https://stochastly.com/learn/minimum-track-record-length)
- [Monte Carlo permutation test for signals](https://stochastly.com/learn/monte-carlo-permutation-test)
- [Regime change in backtesting](https://stochastly.com/learn/regime-change-in-backtesting)
- [Overfitting in trading, explained](https://stochastly.com/learn/overfitting-in-trading)
- [Parameter robustness in backtesting](https://stochastly.com/learn/parameter-robustness-in-backtesting)
- [Point-in-time data and vintages](https://stochastly.com/learn/point-in-time-data-and-vintages)
- [Probability of backtest overfitting (PBO)](https://stochastly.com/learn/probability-of-backtest-overfitting)
- [Prop firm challenge pass probability](https://stochastly.com/learn/prop-firm-challenge-pass-probability)
- [Purged k-fold and CPCV for finance](https://stochastly.com/learn/purged-k-fold-and-cpcv)
- [Sharpe vs Sortino ratio in backtests](https://stochastly.com/learn/sharpe-vs-sortino-ratio)
- [Signal timestamps and execution delay](https://stochastly.com/learn/signal-timestamp-and-execution-delay)
- [Survivorship bias in backtesting](https://stochastly.com/learn/survivorship-bias-in-backtesting)
- [Transaction costs and slippage in backtests](https://stochastly.com/learn/transaction-costs-and-slippage-in-backtesting)
- [Walk-forward analysis for trading](https://stochastly.com/learn/walk-forward-analysis)

## Library
- [Quantitative research library, sourced notes by subject](https://stochastly.com/library)
- Glossary of quantitative research terms, each linked to its library note or guide
- [Backtest overfitting: the research library](https://stochastly.com/library/backtest-overfitting)
- [Market data practices for backtesting](https://stochastly.com/library/data-practices)
- [Econophysics: the research library](https://stochastly.com/library/econophysics)
- [Machine learning in finance: research notes](https://stochastly.com/library/machine-learning)
- [Market practice: evidence on how trading works](https://stochastly.com/library/market-practice)
- [Market microstructure: the research library](https://stochastly.com/library/microstructure)
- People who shaped quantitative trading
- [Research method: rules for tests that can fail](https://stochastly.com/library/research-method)
- [Risk and portfolio construction: the library](https://stochastly.com/library/risk-and-portfolio)
- [Statistics for traders: the research library](https://stochastly.com/library/statistics)
- [Trading strategies: what the research found](https://stochastly.com/library/strategies)
- [Engine notes: bars, custom code and tests](https://stochastly.com/library/using-stochastly)