Detailed descriptions may be reached in the Methodology section.
Backtesting simulates a trading strategy using historical market data, in order to evaluate the past performance and risk characteristics of such strategy. Traders use backtesting to simulate how a particular trading strategy would have performed in the past, given specific rules, risk management parameters, and other trading variables.
Backtesting is a crucial step in reviewing any trading strategies displayed in the ProPick Strategies. By applying the rules of the different ProPick Strategies to historical data, we attempt to assess each strategy's profitability, risk-adjusted returns and other essential metrics, and obtain valuable insights into each strategy’s strengths and weaknesses, in order to refine and optimize each such strategy.
As detailed above, using our AI model, ProPicks Strategies systematically analyze historical stock data with the intent to identify stocks that may have potential to be profitable in the future. These predictive ratings are the basis for developing our ProPicks Strategies. After a ProPicks Strategy is created, it is subsequently audited through our proprietary backtesting module. This step allows us to assess the performance of each ProPicks Strategy according to the historical data.
Subsequently, the strategies that pass our rigorous assessment are selected, based on our sole discretion, for publication on Investing.com.
Please note that past performance does not guarantee future performance. See additional information in the disclaimer specified below and in the question below regarding past performance.
No, the backtested results do not represent actual trades made in past conditions, but only represent the results of hypothetical trades based on historical data. The backtested results only represent a hindsight analysis of the historical data, as we understand it today.
Please see additional information in the disclaimer specified below and in the question below regarding past performance.
- Total Return: The cumulative return across the backtest period.
- Annualized Return: The geometric average return across the backtest period, annualized for comparability.
- Beta: A measure of the strategy's volatility in relation to the market, providing insight into its systematic risk.
- Sharpe Ratio: A risk-adjusted measure, calculated by dividing the excess return of the portfolio (minus the risk-free rate) by the portfolio's standard deviation. A higher Sharpe Ratio suggests superior risk-adjusted performance.
- Sortino Ratio: Adjusting for downside risk, this ratio divides the excess return of the portfolio (minus target return) by the portfolio's downside deviation, highlighting performance with respect to adverse price movements.
- Max Drawdown: Representing the largest percentage drop in the portfolio from a peak to a subsequent trough, providing insight into the strategy's historical risk and volatility.
These metrics collectively enable an analysis of a strategy's historical performance and inherent risks, with a view to give investors information that they may use in order to make well-informed investment decisions.
No, past performance does not guarantee future performance. Although the ProPicks Strategies are based on backtested results, they are for general informational purposes only, should not be considered as investment, legal, tax or accounting advice, and do not represent the personal views of Investing.com or its employees. Investing.com is not a licensed securities dealer, broker, investment adviser or investment bank. Furthermore, any stock or trading strategy displayed on Investing.com does not constitute a recommendation that a particular security, strategy, or action is suitable for you. We urge you to consult your broker or investment advisor before making any investment.
Backtested results are based on certain assumptions related to the market (including for example, the assumption that an investor would have been able to acquire the stocks suggested by the ProPicks Strategies, and that there would have been consistent market liquidity at the time of the hypothetical trade). Deviations from these assumptions can substantially affect the backtested returns, and the real world results stemming from such results would be different.
Additionally, these data points, formed in hindsight, do not account for actual trading influences or unforeseen economic and market events, and therefore may not reflect the potential impact of various economic conditions. The ProPicks Strategies were tested against the historical data available and therefore against the economic conditions present at the time of such historical conditions. If the market conditions change in a way not observed during the historical data, the effectiveness of such strategy would be substantially different. Backtesting can be adjusted until past returns seem optimal, so real-world results may differ.
Backtesting, while valuable, may be prone to several biases which may affect the results, which stem from the fact that the results are based on a hindsight analysis of historical data. Such biases include, for example:
- Survivorship Bias -- survivorship bias occurs when the dataset does not include data regarding companies that did not “survive” (i.e., companies that were delisted, went bankrupt or were otherwise dissolved), causing us to only analyze the “surviving” companies and leading to overperformance. In order to combat this bias, we strive to include in our dataset companies that were later delisted or bankrupt, in order to ensure the data is not biased towards companies still operating today.
- Look Ahead and Restatement Bias -- look ahead and restatement bias occurs when information used in the backtest would not have been available at the time of the backtested trade, due to the date on which the statement was published or the fact that a company restated its previous financial statements, leading to unrealistic results. We try to ensure that our data points are in point-in-time format, meaning they only reflect publicly available information, which would have publicly known to market participants at the date of the trade.
- Corporate Actions -- companies may perform certain corporate actions, such as dividends or stock splits, which can skew the numbers and confuse the analysis. Therefore, we adjust our data to take into account such corporate actions.
Please note that despite our ongoing efforts to optimize our datasets in a way that reduces the aforementioned biases, it is impossible to ensure that all biases have been eliminated, and this may affect real world results.
Comments
0 comments
Article is closed for comments.