The Limits of Backtesting: A Conceptual Perspective on Quantitative Research

Backtesting has become one of the most widely used methodologies in quantitative finance. By evaluating investment strategies against historical market data, researchers and practitioners attempt to estimate how a strategy might perform under future market conditions. The approach has played a central role in the development of algorithmic trading, factor investing and portfolio management systems. Despite its widespread adoption, backtesting remains subject to important limitations that may reduce its reliability as a predictor of future performance. Financial markets are adaptive environments influenced by changing economic conditions, technological developments and evolving participant behaviour. Consequently, relationships observed within historical datasets may not persist indefinitely. This conceptual preprint examines the limitations of backtesting from a systems-oriented perspective. The manuscript argues that excessive confidence in historical simulations may encourage false assumptions regarding strategy robustness and market predictability. Particular attention is given to issues including overfitting, survivorship bias, regime changes, data mining and market adaptation. The paper further explores how reliance on historical optimisation can create a disconnect between simulated success and real-world implementation outcomes. Rather than rejecting backtesting as a research tool, this work proposes a more balanced interpretation of its role within quantitative finance. Historical simulations should be viewed as exploratory instruments for hypothesis evaluation rather than definitive indicators of future profitability. By emphasising uncertainty, contextual awareness and market evolution, the manuscript seeks to contribute to ongoing discussions concerning the future direction of quantitative research and investment strategy development. Keywords: quantitative finance; backtesting; algorithmic trading; market regimes; overfitting; quantitative research; investment strategies; financial markets