Horizon AI launches no‑code backtester
- Horizon Trade opened early access for a no-code trading platform that turns plain-English ideas into backtests and live automations across brokers and exchanges. - The core pitch is speed: users describe a setup, then Horizon returns strategy rules, historical results, and metrics like win rate, drawdown, and Sharpe ratio. - It matters because retail traders usually bounce between code, charts, and brokers; Horizon tries to collapse that workflow into one interface.
Trading backtesting is usually a plumbing problem before it becomes an investing problem. You have an idea, then you need code, data, a charting platform, and some way to check whether the thing actually worked in the past. Horizon Trade is trying to flatten that whole stack into one step. Its new pitch is simple — type a strategy in plain English, get a structured ruleset, run a backtest, and, if you want, wire it straight into live execution. ### What did Horizon actually launch? The product on Horizon Trade’s site is an early-access platform for building, testing, and automating trading strategies without coding. The company says users can describe a setup in natural language, have Horizon turn that into strategy logic, test it against historical data, and then connect brokers for live deployment. The site frames this as one workflow rather than separate tools stitched together. ### What makes the backtester the interesting part? (horizon.trade) Because backtesting is the bottleneck for most non-technical traders. Plenty of people can describe a rule like “buy when RSI drops below 30 but price stays above the 200 EMA.” Fewer can translate that into working code, debug it, and make sure the test is using the assumptions they think it is. Horizon is selling the jump from idea to simulation — not just idea to code. Its preview shows the system mapping entry and exit conditions, evaluating signals, and running historical tests almost immediately. ### Is this code generation or a real no-code tool? Basically both, depending on which Horizon property you look at. Horizon Trade emphasizes a visual no-code builder plus instant backtests and broker connections. A related HorizonAI product leans harder into generating Pine Script for TradingView and MQL5 for MetaTrader 5 from prompts, then helping users iterate and fix errors. The shared idea is the same — natural language becomes executable strategy logic — but the Trade product is broader and more workflow-driven. (horizon.trade) ### Why are people reacting to this now? Because the pitch lands on a real pain point. Retail systematic traders usually live in fragments — TradingView for charts, Python or Pine for logic, spreadsheets for analysis, then a broker for execution. Institutional desks have integrated research tooling, but solo traders usually do not. Horizon’s appeal is that it promises “research desk” convenience without requiring engineering overhead. That does not make the strategies better by itself, but it does make iteration faster. (horizon.trade) ### Does faster backtesting actually help? Yes — but only in one narrow sense. It helps you reject bad ideas quickly and refine decent ones faster. If a trader can test ten variations in an afternoon instead of wrestling with syntax for two days, that changes who can participate in systematic trading. But the catch is that easy backtesting can also make overfitting easier. A beautiful equity curve is cheap if the assumptions are sloppy or the rules were tuned too tightly to old data. Horizon includes familiar summary stats like win rate, drawdown, and Sharpe ratio, which are useful, but none of them rescue a weak strategy design. (horizon.trade) ### What’s the business angle here? Horizon is not just selling a backtest button. It is trying to own the full lifecycle — idea creation, validation, deployment, and eventually strategy sharing. Its blog describes a marketplace where traders can publish, sell, and follow strategies from others, which points to a platform model rather than a single utility feature. If that works, the no-code backtester is the wedge product that gets users in. (horizon.trade) ### So what should readers take from this? The real news is not that AI can write trading rules — that part is already crowded. The interesting move is packaging strategy design, backtesting, and execution into one plain-English workflow. If Horizon’s tooling is reliable, it could lower the barrier for solo quants and discretionary traders who want to test ideas systematically. If not, it becomes one more slick interface sitting on top of the same old backtesting traps. (horizon.trade 1) (horizon.trade 2)
