Introduction

What is a backtest?

A backtest is a simulation that applies a trading strategy to historical market data and records what would have happened — fills, positions, equity changes, commissions, slippage and drawdowns — as if the strategy had been running in real time.

Instead of risking real money to see whether an idea works, you replay the past and measure the result. A single backtest answers questions like:

Would buying every time the 50-day moving average crosses above the 200-day average and selling on the opposite cross have made money on the S&P 500 over the last 10 years?
What would my maximum drawdown have been?
How does this strategy compare to simply holding the index?

Past performance never guarantees future returns, but it's the closest thing retail investors have to a laboratory for trading ideas.

Why should you backtest?

Most retail investors rely on intuition, tips, or a handful of cherry-picked chart examples to decide whether a strategy is worth following. Backtesting replaces anecdotes with data:

Benefit	What it means in practice
Objective evaluation	Numbers replace gut feelings. You see exact returns, drawdowns, win rates and risk-adjusted ratios before committing any capital.
Risk awareness	A strategy that returned 40% but had a 60% drawdown on the way is very different from one that returned 30% with a 10% drawdown. Backtesting surfaces those differences.
Parameter sensitivity	Changing a single look-back period from 14 to 20 days can flip a strategy from profitable to losing. Running multiple experiments reveals which parameters are fragile.
Benchmark comparison	It's easy to feel good about a 15% return — until you notice the benchmark returned 25%. Backtesting against a benchmark keeps expectations honest.
Emotional detachment	Real-time trading invites fear and greed. A backtest is dispassionate: it follows the rules exactly, every single bar.
Faster learning	You can compress years of market experience into minutes of simulation time and iterate on ideas much faster than trading live.

Warning

Backtesting has well-known pitfalls: overfitting (tuning parameters until they match the past perfectly), look-ahead bias (using future data that wouldn't have been available in real time), and survivorship bias (testing only on assets that still exist). Backtide's engine guards against look-ahead bias by default, but overfitting and data quality are the user's responsibility.

How Backtide compares to other Python backtesters

The Python ecosystem has several well-known backtesting libraries. Each one makes different tradeoffs between speed, flexibility, ease of use and feature scope. The table below summarizes how Backtide sits relative to the most popular alternatives.

Feature	Backtide	Backtrader	Zipline	vectorbt	bt
Engine language	Rust	Python	Python / Cython	Python / NumPy	Python
Event-driven loop	✅	✅	✅	❌ (vectorized)	❌ (vectorized)
Built-in strategies	20+ ready to use	❌	❌	❌	❌
Built-in indicators	12+ (Rust)	100+ (Python)	~10 (via ta-lib)	Via pandas-ta	❌
Built-in position sizers	7 (Rust)	1	1	❌	❌
Multi-asset support	✅	✅	✅	✅	✅
Currency conversion	Automatic FX table	Manual	❌	❌	❌
Interactive UI	Streamlit app	❌	❌	❌	❌
Data download	Yahoo, Binance, Kraken, Coinbase	Manual	Quandl bundle	Manual	Manual
Local storage	DuckDB	❌	HDF5 bundle	❌	❌
Analysis plots	20+ built-in	Manual via matplotlib	Via pyfolio	Built-in	Built-in
Configuration	TOML + Python API	Python only	Python + YAML	Python only	Python only
Custom strategies	Python class	Python class	Python class	NumPy functions	Python tree
Actively maintained	✅	⚠️ Stale	⚠️ Archived	✅	✅

Where Backtide stands out

Speed without complexity. The core simulation loop — bar alignment, order matching, portfolio accounting, equity tracking, FX conversion, indicator computation and all built-in strategies — runs in compiled Rust. You write plain Python; the engine does the heavy lifting at native speed.

Batteries included. Most backtesters give you an engine and leave the rest to you: data download, storage, indicators, position sizing, result analysis, plotting. Backtide ships all of these out of the box so you can go from an idea to a result without gluing together five different libraries.

Interactive application. The built-in Streamlit app lets you configure experiments, run strategies and inspect results visually — no code required. When you do need code, the same API powers the app and your scripts, so there is zero translation overhead.

Reproducibility by default. Every experiment config, equity curve, order log and trade log is persisted to local storage. Returning to an old experiment months later is as easy as querying it.

Why Backtide works better for beginner and retail investors

Institutional backtesting platforms (and even some open-source ones) assume you already know what you're doing: you're expected to write your own strategy from scratch, hook up a data pipeline, handle position sizing, implement commissions and slippage, and build your own analysis dashboard. That's a lot of work before you even get to the question you started with: "Would this idea have worked?"

Backtide flips the script:

Start without code. Launch the app, pick a few symbols, select a built-in strategy, hit Run. You have a full backtest result — equity curve, trade log, benchmark comparison and 20+ analysis plots — in under a minute. No Python required.
Learn incrementally. When you're ready to go deeper, switch to the Python API. The same concepts you learned in the app (experiments, strategies, indicators, sizers) map one-to-one to Python classes and functions.
Sensible defaults everywhere. Commission, slippage, warmup periods, empty-bar handling, currency conversion — all have reasonable defaults that match what a retail investor trading through a typical online broker would experience. You only need to change what you want to change.
Built-in strategies as learning tools. The 20+ built-in strategies (SMA Crossover, RSI, Bollinger Mean Reversion, Turtle Trading, portfolio rotations, etc.) are not just trade-ready — they are readable examples of how to structure and evaluate trading ideas. Compare their performance side-by-side to build intuition before writing your own.
Guard rails against common mistakes. The engine prevents look-ahead bias by only feeding data up to the current bar to each strategy call. Auto-injected indicators are deduplicated so you don't accidentally compute conflicting values. Orders that would exceed available cash or violate margin limits are rejected with a clear reason string.
Affordable infrastructure. Everything runs locally on your machine. There are no API keys, cloud subscriptions, or rate-limited data feeds to manage. Market data from Yahoo Finance, Binance, Kraken and Coinbase is downloaded and stored in a local database for fast, offline access.

In short, Backtide is designed so that your first backtest takes minutes, not days — and the same tool scales with you as your skills and ambitions grow.

Tip

If you are completely new to backtesting, start with the Getting started guide, then run a simple Buy-and-Hold experiment to see the full workflow end to end.