backtest-overfitting

canli-mcp: 345 free finance tools for Claude, Cursor and any MCP client, behind five (SEC filings, insider trades, 13F and 13D/G holders, options, FOMC, macro data, quant analytics, backtest validation). MIT, no API key, runs locally. Plus Canli Capital's open quant research, where every number links to its source.

Is your backtest real? Deflated Sharpe Ratio, backtest-overfitting (PBO) and lookahead checks with measured false-positive rates. Works with vectorbt, backtesting.py and TradingView.

Open-source quantitative portfolio optimization, risk analytics, and anti-overfitting strategy certification in Python. The research core of the CPZAI systematic trading operating system.

Backtest overfitting audit for factor research: probability of backtest overfitting (PBO), deflated Sharpe ratio, point-in-time data, purged walk-forward. Searches published factor libraries and reports what actually survived costs.

MCP server that catches fake backtests before you trust them: deflated Sharpe ratio, backtest overfitting (CSCV/PBO), data-snooping tests (White's Reality Check, Hansen's SPA), lookahead checks and pipeline placebos. 20 tools for Claude, Cursor and any MCP client; on npm, the MCP Registry and hosted at canlicapital.com/mcp.

The Null Zoo: Monte Carlo study of the size and power of backtest-overfitting corrections (deflated Sharpe ratio, haircut Sharpe, bootstrap) on synthetic strategy searches with known ground truth.

MSc dissertation: how automated strategy search inflates backtested Sharpe ratios, and why it gets worse for smaller stocks. 136,080 trials across 30 US equities.

Audits a backtest on two separate axes: selection bias (Deflated Sharpe Ratio, PSR, MinTRL, PBO via CSCV) and look-ahead leaks in the code (8 deterministic AST rules + an LLM semantic pass). Outputs one offline HTML report. 292 tests.

Controlled experiments on adaptive backtest overfitting, with auditable agent traces, frozen artifacts and reproducibility checks.

GitHub Action that checks a trading backtest in CI with the deflated Sharpe ratio, adjusted for how many strategies were tried, through canlicapital.com's free validation API, and writes the receipt to the job summary.

Did the winner of a search beat luck? Deflated Sharpe ratio, CSCV probability of backtest overfitting and a pass/fail checklist.

A backtest overfitting toolkit: probabilistic and deflated Sharpe ratios, probability of backtest overfitting, purged cross-validation and multiple-testing haircuts.

A gate that judges a backtest against the best its own search would find on noise, with an audited harness for LLM agents

코스피·코스닥 알파 심사 프레임워크 — 개별 트레이드 분포로 판정하고 랜덤 음성대조·purged CV·Deflated Sharpe 를 CI 가드레일로 강제. 기각 판정문까지 공개한다 · Alpha validation framework for KOSPI/KOSDAQ

Prove a trading backtest is real without revealing the strategy. Zero-knowledge (STARK) verification of backtest honesty: enforce the Deflated Sharpe Ratio, committed trial ledger, in-circuit. Reproduce a live credential to the digit.

Pre-registered, adversarially audited case study: why past-performance selection does not transfer out of sample (crypto perps, 2018-2026)

A backtested strategy walked from paper Sharpe to a funded book: selection bias, execution cost, tail risk and financing, one command and one table.