Why are my backtest results showing unrealistic returns?

Often, backtest engines fail to deduct capital upon entry or ignore unrealized P&L, leading to inflated 'ghost' capital. This skill fixes that logic to ensure every dollar is accounted for.

How does it handle commissions and slippage?

It applies slippage directly to the execution price and deducts commissions from capital on both entry and exit to simulate real-market friction.

What is the 'silent bug' mentioned in the documentation?

A silent bug in backtesting is when code runs without errors but uses incorrect math—like adding P&L only on close while ignoring market fluctuations—rendering metrics like Max Drawdown useless.

Does this skill support short selling?

Yes, it includes specific formulas for short position market value calculations (2*entry - current price) to ensure P&L is calculated correctly when prices fall.

Trading Backtest Capital Accounting

Name: Trading Backtest Capital Accounting
Author: smith6jt-cop

bysmith6jt-cop

0•

数据科学与机器学习

Corrects financial logic in trading simulators to prevent inflated equity curves and inaccurate drawdown metrics.

This skill provides a specialized framework for handling capital accounting in trading backtests, ensuring realistic financial modeling. It addresses critical silent bugs where backtest engines fail to deduct capital during position entry or incorrectly calculate equity values for long and short positions. By implementing standardized logic for slippage, commissions, and market value calculations, it enables developers to generate reliable performance metrics like Max Drawdown and Sharpe Ratio that accurately reflect real-world trading constraints.

主要功能

01Real-time equity tracking (Cash + Market Value)

02Built-in slippage and commission impact modeling

03Correct position-entry capital deduction logic

04Accurate market value calculations for long and short positions

050 GitHub stars

06Robust drawdown and buying power guardrails

使用场景

01Auditing custom-built trading bots for accounting accuracy

02Implementing realistic short-selling logic in financial simulators

03Fixing 'too good to be true' backtest results in Python trading engines

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add smith6jt-cop/skills_registry backtest-capital-accounting

For use in Claude.ai and ChatGPT

Download Skill