As a quant trader/quant reseracher/intelligent trader, you don't want to trade with a strategy that is extremely overfitting, right?.Let's join CorrAI's community and learn more about quant. We grow we learn we become stronger tgt!.#CorrAI #quantcommunity #quanttrading #trading.

After the change, the result is completely reversed. What about 2yrs, 3yrs and 5yrs?.What do you think about this strategy now?.Still a good alpha?

Here is the answer. This strategy might be overfitting because the backtesting period is too short, the strategy might be only working well during that specific period of time. Let's see what is going to happen if we change the period to one year.

Here are the stats:.Sharpe: 4.32.Sortino: 3.46.Calmar: 33.56.MDD: 7.14%.However this strategy has a problem. Do you see it?.

#CorrAI #QuantTrading #QuantLearningCurve #BTCUSDT #Backtesting .Sometimes good numbers and good stats do not necessarily represent a good strategy. Here is an example : What do you think about this strategy? From the numbers, it looks very decent.

Try that on CorrAI. You will have more thoughts about this!.

The trading fee is 0 in this case. What if we set the trading fee to 0.08%? To stimulate the actual trading environment and slipage. The table is turned. A small mistake can ruin everything.

This strategy seems pretty nice and it has a lot of trades to prove it works, but do you see where is the problem?

#CorrAI #QuantTrading #BTC #ETH #SOL #Backtest #quantlearningcurveThe importance of considering tradingfee. A lot of people forget or ignore trading fee when they are making strategies, but this is not appropriate. In CorrAI, I tried a lot of different factors and TA indicators.

RT @TyHaliburton22: Man. Don’t know how to explain it other than shock. Words cannot express the pain of this letdown. The frustration is u….

My new thread about tackling future factors in CorrAI.

#CorrAI #Quant #BTC #Backtesting #futurefactor.

By carefully identifying future factors and utilizing CorrAI's layer and lag functions, traders can design robust strategies that avoid reliance on unavailable data. This approach enhances the accuracy of backtesting and ensures logical consistency in trade execution.

Validate Strategies for Logical Consistency: Design strategies with clear temporal logic, ensuring that trade signals are based on data available at the decision point. This involves structuring conditions to avoid scenarios where future data influences trade entries.

Leverage Layer Transformations: Utilize mathematical transformations (via CorrAI’s layer function) to refine source factors, ensuring they are compatible with the primary trading timeframe and free of future data dependencies.

Use Lag Functions for Temporal Alignment: Apply lag functions (e.g., lag 1) to shift data references to prior periods, ensuring that only known data is used. For example, comparing the current 1-hour opening price to the previous day’s closing price avoids future data issues.

Approaches.Identify and Eliminate Future Factors: Carefully analyze strategy inputs to detect future factors, such as using a 30-day SMA that includes the current day’s closing price. Replace or adjust these inputs to rely solely on historical data.

Data Alignment is Critical for Robust Strategies: To ensure a trading strategy is practical, all data inputs must align temporally with the decision-making moment, using only historically available information.

Temporal Misalignment Creates Logical Flaws: Combining data points from different time points (e.g., opening and closing prices within the same timeframe) can inadvertently incorporate future data, resulting in strategies that are infeasible in real-world execution.

Future Factors Undermine Backtesting Validity: Future factors—data not available within the primary trading timeframe—introduce errors in trading strategies. Using such data leads to unrealistic trade signals, as it assumes knowledge of events that have not yet occurred.