explaingit

cmatthwilkes-debug/trading-strategy-graveyard

Analysis updated 2026-05-18

2Audience · researcher

TLDR

A public log of 26 pre-registered trading strategy backtests, most of which failed once real trading costs were included.

Mindmap

mindmap
  root((repo))
    What it does
      Logs strategy backtests
      Pre registers pass fail rules
      Tracks kills and passes
    Method
      Honest fills
      Cost ladder
      Disjoint holdout
    Strategies tested
      Momentum scalping
      Mean reversion
      Options premium selling
      Pairs trading
    Audience
      Researchers
      Curious traders
      Skeptics of fintwit claims
    Outputs
      Kill table
      Companion website
      Build request system

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

What do people build with it?

USE CASE 1

Study which popular retail trading strategies actually survive realistic costs and fills.

USE CASE 2

See a worked example of pre-registering pass or fail criteria before running a backtest.

USE CASE 3

Submit a build request for a new strategy to be tested publicly.

USE CASE 4

Learn common backtesting mistakes like survivorship bias and unfair fill assumptions.

How does it compare?

cmatthwilkes-debug/trading-strategy-graveyard0-bingwu-0/live-interpreter010zx00x1/faresnipe
Stars222
LanguagePythonPython
Setup difficultymoderateeasy
Complexity2/52/5
Audienceresearchergeneralgeneral

Figures from each repo's GitHub metadata at analysis time.

In plain English

This project is a public record of 26 retail trading strategies that were tested with strict rules to see if any of them actually make money once real world costs are included. The author built a testing pipeline and committed to writing pass or fail thresholds down before running each test, then never changing them afterward. Out of 26 strategies, 22 failed and were retired, while a few showed promise before later checks knocked most of them down too. The repo documents strategies like momentum scalping, gap trading, opening range breakouts, mean reversion, trend following, options premium selling, and pairs trading. Each one is scored using a profit factor, a number that shows whether a strategy makes more than it loses. Most strategies looked fine on paper until the author added real world frictions: trading costs, honest order fills instead of perfect ones, and a broad universe of stocks instead of a cherry picked list of winners. Once those frictions were added, almost every strategy fell apart, including a pairs trading strategy that had the strongest raw edge in the whole record but could not survive the cost of crossing the bid ask spread four times per trade. The project includes eight rules the author follows to avoid fooling themselves, such as never trading un adjusted stock price data, always running a holdout test only once, and always checking whether a metric improves suspiciously smoothly as a warning sign of a data artifact rather than a real edge. A companion website hosts charts and a public request system so readers can suggest new strategies to test or new tools to build for the testing kit. No live trading account is used to generate any of these results, so nothing described here risks real money. The README does not name a specific programming language or list a tech stack, so it is unclear what tools are used to run the backtests. This is a research and transparency project rather than a ready to use trading bot, aimed at people curious about whether popular trading ideas hold up under honest scrutiny.

Copy-paste prompts

Prompt 1
Explain why a strategy with a strong gross profit factor can still fail after trading costs are added.
Prompt 2
Summarize the eight rules this project uses to avoid fooling itself with backtest artifacts.
Prompt 3
Walk me through how pairs trading in this repo went from the best raw edge to a loss after costs.
Prompt 4
What is survivorship bias and how did it affect the trend following test in this repo.
Open on GitHub → Explain another repo

This repo across BitVibe Labs

Verify against the repo before relying on details.