BINARY
FINTECH

BXF is the proprietary trading and backtesting platform of BINARY FINTECH s.r.o. — one engine for backtesting, genetic optimization, walk-forward, Monte Carlo simulation and live deployment. The same strategy code that passes the tests runs live unchanged.

From individuals with modest capital all the way to institutional players — companies, banks, prop firms and asset managers. There are no limits on client size: the same engine, the same testing protocol and the same honesty of results apply to everyone.

With a multi-phase protocol: from the first backtest through searching the whole settings landscape (exact grid or genetic optimization), walk-forward analysis and Monte Carlo distributions, to tests over instrument universes and concurrent strategies in a portfolio. A strategy that fails any phase does not move on — details on the blog.

Order book, kline (candles) and tick data. Backtests run on an event-driven engine with accurate spreads, slippage and fill-probability simulation — results as precise as they can be. The details are in our event-driven backtesting article.

Stocks, futures, options, forex, crypto and prediction markets — ten integrated exchanges and brokers including Interactive Brokers, Binance, Kraken and Coinbase. All under one engine, on one equity curve.

For registered users the platform provides backtesting, optimization and analysis. Actual trade execution is always performed by the exchange or broker — the regulated entity where the trader holds an account. Zero-delta deployment — the same code going live in one click — is used internally.

Only as much as you are willing to lose. Not even our platform can guarantee that what passes a backtest will carry into the future without a hitch — we push robustness as far as it can go, but not every consequence can be foreseen. Money put into risky trading must be money whose loss you can bear; we believe we can substantially reduce that risk, but we will not promise it. How much to stake, and how, is covered in our position-sizing article.

A naive backtest tests a friendlier market: it fills you at the price on the chart, with no fees, spread or slippage, and it doesn't see the fill queue — the price touching your limit doesn't mean you were filled. The deepest trap is that your order isn't in the historical data at all, so the market never reacted to you stepping in. Demos are also simplified and sometimes deliberately flattering. The full breakdown — what a backtest hides and how to get close to reality — is in the article Why Backtests Lie.

Yes — deliberately. Negative results and failed tests are part of the testing life cycle: every shining equity curve has a counterpart of dreadful ones, and every quality edge strategy comes with those that didn't make it. The platform never hides negative results — they are what puts the positive ones in perspective. It's part of how we see ethics and marketing: we show the markets as they are.

An off-the-shelf bot is a black box: it promises profit, but you can't look inside. With us the opposite holds where it matters — in the results. Our internal strategies are not public, but every one of them comes with complete results for the user: every trade, every order, the equity curve and the distributions of all tested variants. Robustness is proven by thousands of tests — the settings landscape, walk-forward windows and Monte Carlo distributions — not by one pretty equity screenshot.

Yes. Custom software is half of our craft: high-performance backends, integrations, specialized systems — PHP, Python, Rust, Laravel. We build software for distributed data centers and their monitoring, deep-tech projects, applications with AI features and agentic systems, and prediction and testing models of all kinds — as well as websites, presentations and customer demos, web-based CRMs, and much more. Trading technology is our specialty, not our limit.

Cultivating the algotrading space — its strategies and portfolios. We hold ourselves to an ethical standard in how we inform clients and the public: about the ceilings and limits of algorithmic trading, about realistic returns, and about the approaches that work from where we stand. We don't know everything; we've found a path that works for us. And above all: this is not about just switching on automated trading. It's about knowing what is happening, why and how, what to expect — and how to get your head right so that panic over market swings doesn't drive the bad decisions that end up costing money.

Write to us via the contact form or at [email protected]. You will discuss your project directly with the people who build the platform.