backtest
Testing an investment strategy by applying it retroactively to historical data to evaluate what returns it would have generated in the past; a way to validate trading rules before using them with real money
As defined in Cliff Asness: A Brief and Biased Survey of Quantitative Investing
Other definitions
A testing methodology that applies a trading strategy or investment model to historical data to evaluate how it would have performed in past periods
Unlocking the Secrets of Systematic Investing: Maximizing Your Wealth-Meb Faber
Where it comes up · 6
- Machine Intelligence and Risk Management6 claims
- Cliff Asness: A Brief and Biased Survey of Quantitative Investing5 claims
- Surviving the Meme Stock Bubble | Cliff Asness2 claims
- Quant in Private Markets1 claim
- Quant vs. Traditional Investors and How Alphas Become Betas1 claim
- Compound Your Face Off1 claim
Most important claims · 10 of 16Search all claims for “backtest” →
- Overfitting arises both from the model and from the person who believes a strong backtest or forward test is meaningful rather than random, so false confidence in the output is itself a form of overfitting.Machine Intelligence and Risk Management
- Machine learning tools have become so accessible that a researcher can type two lines of code to run a seven-layer deep neural network and obtain an amazing backtest, which makes overfitting easy to fall into.Machine Intelligence and Risk Management
- A strong quantitative investment strategy should be persistent (work over long time periods), pervasive (work across multiple markets and asset classes), and dynamic (not just backtest well but handle changing conditions).Cliff Asness: A Brief and Biased Survey of Quantitative Investing
- Backtesting is essential to quantitative investing because it provides 45 years of data showing base rates, worst-case drawdowns, and probability distributions; while backtests can be overengineered and data-mined, comparing a backtest to making investment decisions based on gut feeling and anecdotes ('I talked to my buddy') is not a fair tradeoff.Surviving the Meme Stock Bubble | Cliff Asness
- Out-of-sample testing—validating a strategy on data not used to develop it, across multiple time periods and geographies—is essential to distinguish genuine investment insights from data-mining artifacts; a strategy tested on 1926–1988 data and again on 1990 onward (25+ years), or across multiple countries, has stronger evidence of real effect than in-sample backtest alone.Cliff Asness: A Brief and Biased Survey of Quantitative Investing
- The critique that quantitative investing amounts to 'driving with only the rear view mirror' through backtesting is only fair when a model is based purely on historical pattern-matching without a prior investment thesis or theoretical justification; most serious quants have an investment thesis they believe in, and the backtest serves to examine how that thesis has performed in different periods rather than the basis for discovering it.Cliff Asness: A Brief and Biased Survey of Quantitative Investing
- Quantitative strategies historically required roughly a 50/50 balance between empirical backtest evidence and theoretical/intuitive reasoning about why a factor should work; machine learning shifts this balance because modern ML techniques can improve performance in ways that are theoretically opaque, forcing quants to accept approximately 2/3 evidence and 1/3 story rather than demanding equal weight.Surviving the Meme Stock Bubble | Cliff Asness
- Monthly returns are insufficient for judging a strategy; one must reconstruct the distribution that generated the returns, including the shape and tails, and test whether it is stationary and whether it changes over time.Machine Intelligence and Risk Management
- Private-market quant strategies are harder to validate for LPs than public ones because backtests cover fewer companies and shorter periods, and feedback loops are long, so proving the model to LPs is a major challenge.Quant in Private Markets
- Asness does not believe in pure 'quant style' without investment thesis; rather, quant investing requires an explicit thesis that patterns reflect behavioral or economic principles, validated through historical backtests and out-of-sample geographies, avoiding mere pattern-matching.Cliff Asness: A Brief and Biased Survey of Quantitative Investing