# Build Reliable Strategy Backtests

Trading backtests often overstate performance because they ignore bias, costs, and uncertainty. This skill provides practical structures for realistic simulation and validation.

## Install

```bash
npx skillstore add sickn33/backtesting-frameworks
```

## Metadata

- Status: approved
- Slug: sickn33-backtesting-frameworks
- Skillstore revision: r2
- Version status: missing
- Tree hash: 0e8d728a21a1652be4c5f3e0b3b6163a213b97ceb723e0391eec12c045215f93
- Author: sickn33
- GitHub username: sickn33
- License: MIT
- Repository: https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/backtesting-frameworks
- Ref: 5a26d1d61d694db29af9b138c661c1981076d9df
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: scripts, external\_commands
- Quality score: 78
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/sickn33-backtesting-frameworks
- Manifest: https://skillstore.pages.dev/api/skills/sickn33-backtesting-frameworks/manifest

## Capabilities

- Designs event-driven simulations with orders, fills, positions, slippage, and commissions.
- Provides a vectorized backtesting pattern for simple signal-based strategies.
- Structures rolling or anchored walk-forward optimization with separate test periods.
- Applies bootstrap Monte Carlo analysis to drawdowns, loss probability, and return intervals.
- Calculates return, volatility, Sharpe, Sortino, Calmar, drawdown, win rate, and profit factor metrics.

## Use Cases

- Prototype a Research Backtest: Create an event-driven or vectorized design with realistic transaction costs and delayed signal execution.
- Review Validation Methodology: Identify look-ahead, survivorship, selection, and overfitting risks before accepting reported strategy performance.
- Plan a Backtesting Platform: Define portfolio, execution, optimization, and reporting components for a maintainable internal research system.

## Prompt Templates

### Define a Basic Backtest

```
Design a backtest for [strategy] using [market] data from [period]. Include assumptions, signal timing, transaction costs, and evaluation metrics.
```

### Choose a Simulation Architecture

```
Compare event-driven and vectorized designs for [strategy]. Recommend one based on order complexity, data volume, execution realism, and maintenance needs.
```

### Create a Walk-Forward Plan

```
Create a walk-forward validation plan for [strategy]. Specify training, validation, and test windows. Explain parameter selection and leakage controls.
```

### Audit an Advanced Methodology

```
Audit this backtesting methodology: [details]. Assess bias, cost modeling, capacity, overfitting, regime sensitivity, Monte Carlo uncertainty, and reporting gaps. Prioritize corrections.
```

## Limitations

- It does not execute live trades or connect to brokers.
- It does not acquire, clean, or verify historical market data.
- The examples require adaptation and testing for each strategy and market.
- Backtest results do not guarantee future performance and are not investment advice.

## Best Practices

- Use point-in-time data and delay signal execution to prevent look-ahead bias.
- Model commissions, slippage, market impact, liquidity, and portfolio constraints.
- Keep an untouched test period and report uncertainty alongside average performance.

## Anti Patterns

- Optimizing parameters across the full history and reporting the same period.
- Testing only surviving securities or using revised data unavailable at the decision time.
- Presenting gross returns without costs, capacity limits, drawdowns, or uncertainty.

## Security Audit

- Audited at: 2026-08-14T09:47:28.977\+00:00
- Summary: All 11 static findings are false positives caused by Python type annotations, dictionary methods, ordinary prose, and Markdown code formatting. No executable scripts, shell commands, credential access, reconnaissance behavior, or prompt injection were found.

## Stats

- Views: 136
- Downloads: 13
- Favorites: 0
- Popularity score: 0