Strategy Metrics Out Of Sample / In The sample Ratios
StrategyQuantPlugins
Measures the ratio of Out-of-Sample to In-Sample performance to assess strategy robustness and detect overfitting. Evaluates how well a strategy generalizes to unseen market data.
The OOS/IS ratio expresses the degree of degradation of the strategy in the Out-of-Sample (OOS) against In-Sample (IS)
Strategy degradation refers to the decline in a trading strategy’s performance when it is applied to new, unseen data. This degradation often occurs when a strategy that has been optimized and fine-tuned on In-Sample (IS) data is tested on Out-of-Sample (OOS) data. The degradation can be a result of overfitting the strategy to the IS data or due to changing market conditions that the strategy is not able to adapt to.
Out-of-Sample (OOS) and In-Sample (IS) ratios are crucial concepts in the development, testing, and validation of trading strategies. These ratios help assess the robustness of strategies and ensure that they are not overfit to a specific data set. The following is an extensive explanation of OOS/IS ratios:
Purpose of OOS/IS Ratios: The primary goal of using OOS/IS ratios is to assess the robustness and generalizability of trading strategies. By evaluating the strategy’s performance on both IS and OOS data, users can identify potential overfitting and ensure that the strategy is not overly optimized for a specific data set. A strategy that performs well on both IS and OOS data is more likely to be resilient and adaptable to changing market conditions.
You know me as clonex. I’m Ivan Hudec — I built the free blocks you’re using here, and I write about the research behind them at algotrading.space. Need something built? Work with me.
