Time is pressing, isn't it? Validity is a luxury that is becoming scarce. It is absolutely necessary to consider the quality of input data before blindly trusting statistical tests. The risk is that systematic biases in training data completely distort the results, making any statistical similarity misleading, especially in emerging markets where access to reliable historical data is often limited.